with my o wn researche s in 1 90 6 - Forgotten Books

470

Transcript of with my o wn researche s in 1 90 6 - Forgotten Books

D ISCOVER IES IN FORMER TIMES COMPAREDWITH MY OWN RESEARCHE S IN 1906

— 1908

VOL. I

LAKE MANASAROVAR AND THE SOURCE S OF THE GREATI NDIAN RIVERS . FROM THE REMOTEST ANTIQUITYTO THE END OF THE E IGHTEENTH CENTURY

S T O C KHO LMLITHOGRAPH IC INSTITUTE OF TH E GENE RAL STAFF OF TH E SWE DISH ARMY

STOCKH bL iviKUN GL. BOKTRYCKE RI ET. P . A. NORSTE DT 56m

i s l s87

C O N T E N T S.

I ND I A N , GRE EK, ROM A N AN D M O H AMM ED A NG E OGRA PH ERS

CHAPTE R I . South -wes tern T ibet in Ind ian antiqu ity

I I . The Greeks and RomansI I I . PtolemyIV . The Arab ian geographersV . Later Mohammedan writers

C H I N E S E AN D T I B E TA N G E OGRA PH ER S

CHAPTE R VI . Ch inese works on the h ydrograph y of South—wes tern T i betV II . Wei -tsang-t

u-ch ih

VII I . The Ch inese geographers on the source of the BrahmaputraIX . The northern tributaries to the Ts angpo

X . The s ource of the Brahmaputra on the Ta-ch ’ ing map

XI . Some descri pt ions of the Manasarov ar and its su rround ingsXII . The S hu i-tao-t i-kang on the source of the Brahmapu tra and the Satlej

XIII . Manasarov ar and surround ing regi ons in T i betan work s .

E ARLY EURO P E A N KN OWLEDGE OF T IB E T

CHAPTE R XIV . E urope ’s knowledge of T i bet in the middle ages

XV . Friar Odoric . S ir John Mandev illeXVI . E as t Ind ian trave llers in the s i x teenth and s eventeenth centuriesXVI I . When d id the name Manasarov ar become known to E urope ?XVII I . Bened ict Goes and Antonio de Andrade

T I BE T AN D T H E SOURC E S OF THE GRE AT I N D I A NR I V ER S O N A N C I EN T M A P S

P e

CHA PTE R Maps of the fi fteenth century . 33Maps of the s ixteenth century 1 78

Maps of the s eventeenth century 1 88

XXI I . T ibet in E uropean books and narrat ives of the s eventeenth century 206

XXI I I . Verb iest and Gerb i llon 2 1 0

Maps fromthe firs t th ird of the e ighteenth century 2 1 4

XXV . Lag o de Ch iamay 2 30

XXVI . The Maps of Strah lenberg and Renat 246

XXVI I . The Jesu it cartographers in Ch ina,and d

Anv ilIe’

s map of the lakesand the ir surround ings

XXVI I I . Ippolito Des ideriXXIX . Fath er Tieffenthaler, Father Gaub iI and Anquet i l du Perron

ILLUSTRATION.

Mount Kailas , looking N 340

E fromKhaleh

MAPS .

The Indus of Ptolemy,Codex Constantinopolitanus , 1 5th century

The Ganges of Ptolemy . Codex Constantinopolitanus , 1 sth centuryCentral and Southern A s ia Ptolemaeus Romae 1 400

Map illustrating >>Histoire de T imur Bec» Transl by Petis de la Croix , Paris 1 7 2 2

Map of the Western Regions and IndiaChinese Map of T ibet

,published by Klaproth in 1 8 3 0

The region of the Manasarov ar and the Sources of the great Ind ian R ivers , as re

presented on the Ta-ch ’ ing Map . (A fter Dutreu il de Rhins) .

Route Map of Ulterior T ibet. Taken fromH s i-chao-t’u -lueh . (After Rockhill)Map of T ibet

,fromthe work ed ited in 1 88 6 by Houang T ’

ci-k ’ iaoThomas Herbert : Some Years Trave ls into A frica A s ia the Great. London 1 63 8

The Map of A s ia in K ircher’s China i llustrata , Amsterdam1 667

The Genoese Map of the World , of 1 447

Fra Mauro’s Map of the World

N icolas Desliens Map of the World , 1 54 1

A s ia on the World-map , published at Ven ice , 1 554

Gastaldi'

s Maps of the Ind ian peninsulas , in Ramus io, 1 554

Tertia Pars As iae . Jacopo Gastaldi , 1 56 1

The Wall-map of As ia in Sala de llo Scudo, Ven ice , restored in 1 76 2

Part of Diogo Homem’

s Map of the WorldPart of Mercator’s Map of the World, 1 569

Ortel ius : Tartarize s ive Magn i Chami Regni typus , 1 570

Orte l ius : Indiae Orientalis , Insularumque -adiacientiumtypus , 1 570

Ortel ius : Pers ic i s ive Sophorumregn i typus , 1 570

Ludouicus Georgius , 1 584

India and Central As ia on Jodocus H ond ius'

Map 1 6 1 1

Part of the Map of Franciscus Hoeius , about 1 640

India in Blaeu ’

s Atlas , 1 640

Iohannes Ianssonius , Amsterdam 1 64 1

VIII

XX IX . Sanson d’

Abb ev iIIe , 1 654XXX . Sanson d’

Abb ev ille , 1 654XXXI . Imperii S inarumnova descriptio,1 655XXXI I . N icolaus V isscher's M ap , about 1 680

XXX III . Cantell i ’s Map of 1 68 3XXX IV . N icolaes Wits en ’

s Map of 1 68 7

F . de Witt’s Map , middle of 1 7th centuryXXXVI . Coronelli s Map , 1 695XXXVII . Sir Thomas Roe s Map . FromOg

ilby

s work . London 1 673XXXVIII . The Map of Isbrants Ides , 1 704 .

XXXIX . The Map of Isbrants Ides illustrating R ubruck s journey, in Bergeron ’

s »Voyages » ,I 73 SDe li s le ’s Map 01 1 705

XLI . Carte de Tartaric. Delis le 1 706XLI I . De lisle ’

s Map of 1 7 23XLIII . >>Carte Nouvelle de l ’As ie Septentrionale» , 1 7 2 6XLIV . The second Map in H istoire Genéalogique des Tartars , Leyde 1 7 26

XLV . Part of a Map published in >>Journal du S ieur Lange» , 1 7 2 7

XLVI . Martini ’s Map , as reproduced by Thevenot. Showing that the Ganges was supposedto come fromthe Lake Chiamay (Kia 1 666 '

XLVI I . The Map in Wicquefort’

s edition of Mandelslo’

s narrative , 1 7 2 7XLVII I . Homann ’

s Map , about 1 73 9XLIX . Southern part of S trahlenberg’

s Map of As ia, Stockholm 1 73 0

L . Renat’

s Map of Central A s ia, brought to Sweden in 1 73 3

LI . D’

Anv ille’

s Map of the Lakes and of the Sources of the three great R ivers , 1 73 3

LII . T iefi'

enthaler’

s Map of the Lakes , 1 766

LI II . Gaub il’

s Map of the Lakes and the Sources of the Ganges (i . e . Satlej), 1 7 29

E R RATA.

Page 24 , line 20 fromabove stands coast for court.

2 7 . 1 43 tw ice t/zrz

64 , 1 >f s ana P arana

99, 5 Sa under'

s Saunders’

.

I 43 1 I i t I '

1 90 , 1 9 XXV

2 8 7 , 8 below

THE SURV EY O F IN D I A

AS A TOKEN OF ADM IRATION

BY T H E AUT H O R

PRE FACE .

I have fe lt it to b e my firs t du ty to extend to the Swed ish D ie t of 1 9 1 0 mys ince re thanks for the s tate s ubvention granted towards the publ ishing of this work .

With special grati tude I remembe r the e fficient way in which the then Min is te r ofS tate , Admiral Arvid Lindman , was pleas ed to support to a succe s s fu l is s ue my re

que s t to have the expens es cove red out of public funds . And the favourable re

ception ofmy demand was , in no smal l meas ure , due to the sympathe tic atten tiongivento the que s tion on the part of the Roya l Swed ish Academy of S cience .

I have ded icated this work to the Survey of India as a mode s t s ign of the

great es teem in which I have always he ld this high Admin is tration for its bri ll iantachievements , during a centu ry, in the s e rvice of geographica l re s earch . E arlyin the summe r 1 9 1 4, I wrote to the Surveyor Gene ral submitting that the dedicationmight b e accepted by the Survey of Ind ia . By le tte r, dated S imla the 1 3 th Ju lyof the s ame year, this offe r was , in ve ry cordial te rms , agre ed to by S ir S . G .

Burrard. No one be tte r than he was in s ituation to appreciate how much I cons idered mys e lf in debt to the gene rations of surveying officers and Pundits who ,

s tep by s tep , had, th roughou t decades pas t, unde r d ifficul ties unheard of, forced theHimalaya wi th i ts adjoin ing tracts of land in the North to yie ld the ir s ecre ts . Theyhad made my ways even , and carried our know ledge of Southe rn Tibe t so far thatmy exploring journey only became a natural cons equence of the ir pers is tent andundaunted labou rs . From them I had learnt what rema ined to b e done , and

,by

the aid of the i r exce llent maps , I was e nabled to de tect the wants s ti ll exis ting.

The me thod of treating the s ubject in >>Southe rn Tibe t» i s qu ite an othe rthan in >>Scientific Re s u lts of a Jou rney in Cen tra l As ia , 1 899 In the

las t mentioned work , I re lated all the obs e rvation s , day by day, during the wholecours e of the voyage , while I withhe ld from the ve ry t ime—was ting task of as s emb

XII

l ing, arranging and analys ing the re su lts brought about by othe r trave l lers in the

s ame regions . I was giving too much of my own and too l i ttle of what was gainedby my prede ce s sors . One got bew i lde red in the deta i ls , los ing all ge neral views ,and one became unable , withou t s e rious , preparatory s tudie s , to as s ign , to my s cien

tihe journey, its right place in the cha in of human progres s . In the pres en t ins tanceI have the re fore , in deal ing w ith the matte r, fol lowed qu ite an Oppos ite sys tem. To

the his toric point has be en given the preponde rance in >>Southern Tibet» . As

far as it has be e n pos s ible to command the whole geographica l l ite rature in thisbranch , I have , w i thin i ts precincts , sought out eve ry treatis e , ancient or mode rn ,contributing to the knowledge of the coun try that had be come the object of myinves tigation. From the las t contemporary trave ls of exploration I have pas sed on

to my own journey, the reby enabl ing the reade r to decide by hims e lf in whatmeas ure the latte r has brought about new real resu lts . My own journey has beendes cribed as succ inctly as pos s ib le , w i th indication only of its chie f characte ris tictrai ts . Ce rtain parts have not even been ment ioned at all. Thus , for ins tance , Ihave altogethe r left out the voyage round Kai las for the reas on that I had gi ven a

qu i te sufficient account of its features already in my popu lar workThe voyages down the upper Indus fromGartok to Drugub , and down the uppe rSatlej through Himalaya, we re affecting regions so we ll known , that I needed not

now dwe l l upon them. Al l unes s ential de tai ls have been omitted, and on ly the res u ltss e t forth . In this way, a clear gene ral view may b e taken of the matter, al low ingit to b e more eas i ly compared with the harves t reaped by othe r explore rs .

In cons equence of the modified plan, my new work has become cons iderablyreduced as to its dimens ions , when compared w ith >>Scientific Resu lts ». It is truethat also this time I amappearing with four volume s of geographical text, b ut boththe s ize and the numbe r of page s are les s than previous ly. The

n

fact that i t hasnot been poss ible to proceed with the publishing earl ie r than e ight or n ine yearsafter my re turn home , is of no account. The work has in no manner grown an

t iquated by the laps e of t ime .

I As far as I know , no one has vis ited the te rri toryin ques tion late r than mys e lf, and no new l ight has b een thrown ove r Transhimalaya s ince my caravan las t time wande red ove r its mountains .

The geographical text is divided into fou r volumes . The firs t is re lating to

the lake Manasarov ar and the Sources of the great Indian R ive rs fromthe remotes tantiqu i ty to the end of the e ighteenth century . The s econd forms its continuation

V ol. I was printed fromJuly 1 9 1 4 till Fe bruary 1 9 1 6, V ol . II during February and March1 9 1 6, V ol. III fromMarch to May 1 9 1 6 and V0] . V fromDecember 1 9 1 5 to February 1 9 1 6 .

XI II

from the e ightee nth centu ry to my own jou rney 1 906 to 1 908 . The third volumeis devoted to the problemof the Transhimalaya . The fou rth one , which is unde rpreparati on , is applying to the Karakorummounta ins , chiefly regarding the part scompris ed in my rou tes . This volume w i ll b e published during the cours e of nextyear, and b e accompan ied by an index of names and a table of matters re fe rringto all four volumes .

I hope the public w i l l not b e too critical in re s pect of the treatment of thelanguage . Contrary to what was forme rly the cas e when I had my Swedish manus cript trans lated , I have now written out the te x t di rect in E nglish , a procedu rewhich , i t is true , has been faci l itated ow ing to the reason that mos t of the s ourcesof its his torical chapte rs are of E nglish origin . Ye t the treatment of the languagecannot under s uch circums tance s b e d is tinguished by any lofty flight or beauty. On

the contrary , i t i s ve ry plain and bearing s tra ightway on facts . In orde r to get thetext cleans ed from dimor e rroneous expres s ions , I have had the manuscript goneove r and corrected by two E nglishmen l iving in S tockholm. I amals o indeb ted tothes e gen tlemen for the i r, I hope , care fu l revis ing of the proofs .

The his torical and geographical text of the three firs t volumes is i l lus tratedby 1 08 maps , s e lected fromold and new atlas es , explorative works and treati s es on

Tibe t . The i r number might ha v e be en increas ed w ithout cons ide rable d ifficu l ty, b ut

to no great advantage . The maps appended are s erving the end I have had inview in thi s respect . They make i t pos s ible for the reade r to fol low the deve lopmen t of our know ledge of Tibe t and, in part icu lar, of the sou the rn part of thatcountry . The fou rth volume is accompanied by an addi tional numbe r of his torica lmaps , the s e be ing mos tly of later date , which is qu i te natural con s ide ring that Karakorumwas as unknown to the cartographe rs of antiqu ity a s Transhimalaya .

The i l lu s trative mate rial of the geographical volume s cons is ts , moreover, of

photographs , reproduced in phototypes or auto types , a few photographic panoramas ,and a lso of sketches and waterco lours repres en ti ng Tibe tan lands capes .

The map of my lates t journey in Tibe t cons is ts of 25 she e ts in the s cale of

1 My fa ithfu l and cons cientious as s is tant, Lieutenant C . J . O tto Kjells trom,

had be fore his death , which occu rred on Decembe r a8 th 1 9 1 3 , comple ted e ight ofthes e .

I He had fu rther drawn up my rou te on the fol low ing shee ts and, afte r my

On P l . 7 and PI. 8 , the spe lling of names i s in s ome cases not quite the same as in the

text : for instance Lug-la, Tugdan, Rokso, A jang, inste ad of Luk-la , Tugd ii n ,

Roksdo, Ayang. I always us ed to write down the name s ph onetical ly . The spe l l ing in the text is more correct than on

Pl . 7 and PI. 8 .

XIV

panoramas , reproduced a great part of the featu re s of the country in mos t of them.

At his death , my expe rienced friend Colone l H . Bys tromtook up this work , and themap mate rial cou ld s carce ly have falle n in to be tter hands . Colone l Bys tromhas

at pres ent comple ted the draw ing and topographical arrangement of all the remain ing1 7 shee ts .

In my previous work , >>S cient ific R esu lts » , Vol. IV , pages 542 and 547 , I speakof my intention to publish a gene ra l map of Tibe t in the s cale of 1 :

Preparations for this work we re then , in 1 905, already made . On page 534 of the

s ame vo lume was , during my abs ence , ins e rted an explanation of the reasons why

the promis ed mi ll ion s cale map cou ld not then b e published . The volume in ques tionwas publ ished in 1 907 , and my cartographers qu ite correctly realized that a gene ralmap, which did not conta in my dis cove ries made in 1 906 and 1 907 , shou ld have beenou t of date already at my re turn . The comple ting of this map was in cons equencepos tponed .

The time for i ts publ ication has now arrived. The ne t of coord inates hadalready been cons tructed by Kjells trom, who had als o ins e rted the rou te ofmy jou rney of 1 899

—1 902 . The rema in ing part, s til l unde r preparat ion, is the work of

Colone l Bys trom. The forming of this map has taken s eve ral years of d i l igen tlabou r , and enta iled so great expens e s that the s tate s ubvention was not s ufficientto cove r more than half the cos ts of the entire work .

I The mi l l ion s cale map isintended to give a gene ral , clear and un iformView of E as te rn Turkes tan and Tibe t,and to contain the ma in features of all that we know at pres ent of the geographyof the s e coun tries . It will fill an es s e ntia l blank in Penck ’

s inte rnational world mapin the s ca le of 1 It con ta ins all known trave ll ing rou tes of which wepos s e s s re l iable surveys . The col lection and combination of the exis tingmaterial hasbee n ve ry arduous and d ifficu lt, and often has a smal l part of the map, as for in

s tance the region around S e l l ing-tso , requ ired we eks of dis cus s ion , of shifti ng and

adjus tmen t . At times i t has been n igh to impos s ible to make the routes of diffe re nt trave l lers agre e . Occas iona l ly a route has proved too short or too long be tween

Not even a gift, placed at my d isposal froma private source , has been sufficient to coverthe expenses . In add ition to this , my budget has to support the cost of a work concern ing my journey through E astern Pers ia , for which a map of 9 sheets is

already completed , and of Profes sorDr A . Conrady

s ed ition of the collection of manuscripts on paper and wood , compos ing about 1 50

numbers , wh ich I found during my d iggings at Lou -Ian . The work of Prof. Conrady was to a largeextent ready in the spring of 1 9 1 4 , b ut i ts publication was prevented by the outbreak of the war.

Lastly, it is my hope that the s ale of »Southern T ibet» will make it pos s ible for me al s o to publish ina fitting manner the deta iled survey of km. of the middle course of the E uphrates wh ich I prepared in the spring of 1 9 1 6.

XV

two points , the s ituation of which for certain reasons it was pos s ib le to fi x . Itoften happened that , in the ende avour to bring abs o lu te ly e s s entia l points to agree ,encroachments we re caus ed on adjoin ing te rritorie s and rou te s , and in such cas es

conformity could on ly be atta ined by d ivid ing the fau lt ove r cons ide rable fie lds of

the map . But on the other hand, it is only natura l that a country , comparative lyso l ittle inve s tigated as Tibe t, shou ld offe r both the trave lle r hims e lf and , late r onhis cartographe rs , the greate s t di fficu lties . The enormou s mounta inland be twe enTranshimalaya and Kwen-lun is cle ft into an intricate mosa ic of bas ins without outle t ,

~

and an inextricable confus ion of cha ins of moun tains and ridges s e emingly w i thout connection . It lacks the s imple large and clear l ine s of E as te rn Turke s tan w ithits gu iding hydrographica l ske le ton .

In the s e conditions , our mi l l ion s cale map no doubt is encumbe red w ithmanyfau lts , caus ed by the de fective mate rial and the frequently ve ry vagu e natu re of the

rou tes . Colone l Bys tromhas , however, tried to make the be s t of what was to b e

obta ined .

The mi l lion s cale map mu s t the re fore only b e cons ide red as a firs t edition ,

which I hope i t w i l l b e pos s ible to improve by degre es in the fu tu re . All correc

tions nece s s itated by new d iscove ries shall b e introduced . After a certain numbe rof ye ars , i t wil l b e poss ible to publish a new improved edition . It is my w ish and

intent that this map shal l b e pe rmanent and even in fu ture , during d i fferent epochs ,b e cons ide red to give the be s t obtainable cartographical repre s entation of E as te rnTurkestan and Tibe t .

Profe s sor Dr K . V . Ze tters téen has control le d the spe ll ing of name s on the

mi ll ion sca le map .

Colone l Bys trom has drawn a smal l s cale map show ing the divis ion of the

mi ll ion scale map into 1 5 she e ts . The whole of the topographic mate rial , as we l las the text, i s to b e published during next . year. For the atlas , Colone l Bys tromwill wri te a pre face in which he explains his working me thods .

During my late s t jou rney, I made it a ru le to draw a panorama of the wholeregion , within my horizon , fromeach camp and fromeach dominating pas s . Theyamount to 552 in numbe r, and are printed in numerica l order on 1 05 she ets , forminga special volume of the same s ize as the atlas . For the comple tion of the topo

graphy of the trave ll ing route s , thes e horizontal views of the lands cape have beena valuable s upport to the cartographe rs , not leas t when they we re to ge t an ideaof the re lative alti tudes . They s hou ld also offe r the reade r a comparative ly clearview of the gene ral fiaéz’z‘zcs of the Tibe tan landscape and of its mos t s tri king charac

XV I

teris tics . More dis tinctly than fromlong descriptions or even qu ite detai led maps ,he w i l l fromthes e landscape s b e able to dis cern the sharply chis e l led outlines w i ths teep s lope s and deeply cut valleys of the pe riphe ric country, and he w i l l rece ive as

vivid an impre s smn of the endles s s tre tche s of table land with no outle t, and the i rcomparative ly great un iformi ty . He will , as I did yonde r in the fie ld, fee l as i f hewe re s tanding on a rock in the mids t of a troubled s ea, whe re the ridge of one

wave b ut rare ly ris es above the othe rs .

The bearings for each panorama have be en found by me on the spot, bymeans of a hand compas s , after the drawings we re completed . They do not therefore pre tend to any greate r accu racy than what is obtained by practice . E ach panorama is provided with a note contain ing a defin i tion of the poin t fromwhich the

view is drawn , and also s tating on which shee t of the map its proje ction is to b e

found . Thu s i t will b e pos s ible , in each s eparate cas e,to compare the horizontal

vi ew w ith the ve rtical one . I was not in pos i tion to bring w i th me the thousandsof photographic plates which wou ld have been neces sary for a s urvey of the countryby mean s of the came ra . For s uch a photogramme tric work

,much greate r res ources

than those at my command wou ld have been needed. My panoramas are mean tin s ome meas ure to compens ate for the wan t of photogramme tric mate rial .

In volume V of the text, the collection of petrografic s pecimens which I havebrought home , is worked out and examined by Profes s or Dr Anders Hennig. In

his introduction , he indicates the principles on which the collection has been made ,as we l l as the me thod he has followed in examin ing the material . With great s kil land patience he has made the be s t of the opportun i tie s given . If the pains I havetaken in collecting the spe cimens of rocks , have rendered it pos s ible for the hand ofthe special is t to draw up the main l ine s of the geology of Sou the rn and Wes ternTibe t

,they have not been in vain . I need not point ou t that the geologica l picture

we have produced,mus t in the fu ture unde rgo many and great alte rations . The

de ta i led work of the special is t on the fie ld s til l remains to b e done .

Now , as we l l as forme rly , I have for the working out of the me teorologicalobs ervations , had the advantage of the experienced and valuable as s is tance of Profe s sorDr N i ls E kholm.

Dr K . G . O ls son also remains one of my faithfu l collaborators , and he has

charged hims e lf w i th the working ou t of the as tronomical obs ervations . Thoughthe obs e rvat ions are , th is time , les s accurate and comple te than thos e res u lting fromthe journey of 1 899

—1 902 , Dr O ls son has made us e of themas far as i t has been

XV II

pos s ible . The works of Profes s or E kholmand Dr Ols son are ready in manus criptand under printi ng.

To my above named collaborators I b eg to expre s s my warmand s ince rethanks for the i r never fai l ing patience and inte re s t .It has been of invaluable he lp and advantage to me to b e al lowed to apply

to the S tate Librarian Dr E . W . Dahlgren , whos e profound know ledge of the

his tory of cartography , and of the geographical dis coveries , is appreciated and

acknow ledged ove r the whole geographical world . Dr Dahlgren has cal led my at

tention to many of the older maps , reproduced in chapters XIX—XXIV, and he hasobtained for me a great number of the origi nal geographical documents referred to

in the same chapters . He has als o been kind e nough to pe rus e this part of mywork in orde r to guard me fromeventual e rrors . By so doing, he has ins piredmepersonal ly w i th a fee l ing of s ecurity , when I now venture to submi t also the s e cartographically his torical chapters to the judgment of the learned world .

FromInd ia Office , I have as a loan got s ome pre cious unpublishedmanu s crip tsquoted in the firs t chapter, and also a couple ofmaps reproduced in the s econd

vo lume . In connection he rew i th , the Librarian in the said office,Dr F . W . Thomas ,

has given me s ome good advice s and informati ons .

The we l lknown German s inologu e , Dr Albert He rrmann , has had the amabilityto perus e the proofs of the Chine s e chapters VI—XII in the firs t volume and, at the

s ame time , he has es tablished a un i formand cons is tent s pe ll ing. H e a lso took the

opportun ity , on this occas ion , to give me s eve ral good hints and new ideas .

Chapter XIII , in which Tibe tan works are treated, has be en read in proof byDr A lbe rt Grunw edel, who a lso has drawn my attention to a couple of works uh

known to me .

Dr G ilbe rt T . Walke r has supplied me w ith the tables indicating the quantityof rain for India , reproduced in chapte r XXVI of the s econd vo lume , be fore the i rbe ing printed .

I am indebted to the di rector of the Roya l Library at Berl in , Profes sor DrH . Me is ne r, for s eve ral of the Ge rman maps reproduced in the third volume , whichare kept in the cartographical ins titute of the same l ibrary, and some of which are

un ique hand drawings of Klaproth and R itte r.Genera l 0. E . v on S tub endorff of S t . Pe te rsburg, and Commande r Roncagl i

of Rome , have given me some valuable sugge s t ions regarding Rus s ian and Ital iantrave lle rs .

Las tly , Profe s sor Chavannes of Paris has place d at my d isposal the Chines eIII—1 3 1 3 8 7 . I .

XV III

map of Tibe t, so d ifficul t to obtain , which is reproduced on PI. IX in the firs tvolume .

To all thes e ge ntlemen I b eg to expres s my s ince re thanks . My thanks are

a lso due to my publishers , The Lithographic Ins ti tute of the General S taff of theSwed ish Army in S tockholm, e specially to its able chie f, Mr Axe l Lagrelius .

Now that I offe r this work to the geographical world of our time,I do it

w i th a fee l ing of insufficiency and imperfectnes s . The resu lts I have won by myjou rney , are in no adequate re lation to the grand w idene s s and enormous magn i tudeof the task i ts e lf. The forces of a s ingle man w e re not sufficient to embrace and

as s imi late all that me t himon his way du ring months and years . If, l ike mys e lf, hemakes the geographical discove ries his principal aim, and cons equently cons ide rs of

mos t importance to forma mode rate ly re l iable pre l iminary map, he has not muchtime le ft for minute and c ircums tancial inves tigations . Thes e deficiencies mus t needsb e vis ib le in his work .

And, notw i ths tanding, I hope that thes e volumes w i l l not b e looked upon as

an ins ign ificant addition to our knowledge of the globe . They s ti l l contain the

des cri p tion of great tracts of Tibe t, neve r before vis ited, even by Indian Pundits ,much les s by E uropeans . For my own part , I cons ide r this jou rney, e ffected unde runusually difficu l t condi tions , to b e a pioneer ente rpris e , which in s ome measu re hasopened the way for future conques ts in Tibe t on the different fie lds of s cientificres earch .

IN TRODUCTION .

For the reader’s in itiation and gu idance , I think i t e s s ential briefly to give an

indication of the plan and arrangemen t of the follow ing fou r geographica l volume s .

The object I have in view, is to show in what ways , during the cours e of the cen

turi e s , the knowledge of Tibe t has s low ly ga ined the ears of E u rope . When firs theard of, the news of its exis tence came in the shape of vague and mys terious ru

mours which , a l ready in the middle age s , we re rece iving a more fixed formand,

late r on , in times not too far remote , were s ucceeded by narrat ives of more or les sfantas tica l colou r. My intention is to expos e in how unw ie ldy lines , labourious lydrawn up, the notion of Tibe t appears on the maps , how uns teadi ly and vaci l latinglyi t hovers above the dark expanse s North of the Hima layas , and how long i t is ableto shie ld

,as in a fortress , i ts s ecre ts fromthe insatiable explorative spiri t of the

E uropeans . The s i tuation of Tibe t, among the highes t and mos t inacces s ible mountains of the globe , was by i ts e lf a guarantee to the e ffect that, of all countries on

the earth the polar regions excepted i t shou ld las t b e conquered by the E u ropean attacks for discove ry .

I go back as far as to the Indian an tiqu ity and endeavour to s earch out fromitsepic songs , legends and re ligious tales , the foundation of geographica l real ity on whichthey are bu i lt up , though even in a poeti c mou ld . The Indian cosmographers allowthe i r myths to soar around the legendary mounta in of Me ru , on whos e dizzy he ightthe great city of Brahma rais e s its shin ing bas tions . The i r hymns s eek out Himavant , the sacred, and love to dwe l l amid the dazz l ing splendou r enci rcl ing the homeof S iva on Kailasa , the king of mountains . They s ing the prais e s of Mana-sarov arawith a charmmighty of arous ing the s urpris e and admiration even of pres ent-dayoccidentals . It is true , the geographical know ledge which can b e dis entangled fromall this splendou r of ancient lore , is very unce rtain and vague , b ut s till i t i s sufficientto convince us that the Hindoos we re no s trangers to the country to the ir North ,

and , above all, that they were fami l iar w i th the regions around Kai las and Manas arov ar. They were indeed des cendants of the Aryans and had pos s ibly at one timecome down into India by this road . Pe rhaps in the i rminds the re remained a dor

mant recol lection at the wanderings of the ir ances tors acros s the mountain fas tnes se sof the wes te rn Himalayas .

India was known to Herodotu s , b ut , contrary to Megas thenes , he doe s not

mention the moun ta in s to its North , much les s Tibe t. And yet we hold fromhis

times the cu rious s tory of the golddigging ants . Alexander and his generals carriedthe know ledge of the wonde rfu l world of the E as t a giant s tep forward, and theycal led the Himalaya Emodus . E ratos thenes speaks of Paropamisus and Imaus , yetS trabo is the one , among the ancien t geographe rs , who create s out of the materialwhich he has succeeded in col le cting , the grandes t des cription of the countries of

Sou the rn As ia . Far be tte r than his predece s sors , he knew the mountains to the

North , and a long time had to pas s before E urope got as clear an idea of Ind iaand its rive rs as S trabo pos s e ss ed . The geography of Arriamis founded on that ofMegas thenes and E ratos thene s . The mounta ins among which lie the sources of the

Ganges , he calls Par0pamis us or Caucasu s , and w ith a sure hand he draws up the

orographical and hydrographical sys temof the great continen t. But as l i ttle as any

other clas s ic author, he had any idea of the exis tence of Tibe t . They knew on lythe sou thern parts of the Himalayas . Ptolemy s urpas s es the old Gre eks and Romans .

Much of his know ledge he drew fromMarinus of Tyrus . His picture of India and

its hydrography is mas te rly. Q u i te correctly he makes Satlej out to b e the larges tri ve r of the Panjab , and even shows on the map we re its sources are s ituated. But

no trace of Tibe t , or the rive rs of Tibe t, is to b e found in his pictu re of the coun tryNorth of India . On the con trary

,he pres s es Kwen-lun and the Himalayas toge the r

into one s ingle moun tain-range , s o that Tibet is effaced into nothingnes s be tweenthem. E ven years after his time , the E uropeans commi tted qu i te the samemis take in the i r repres en tations of Sou the rn As ia . But also the fundamental l inesof the As iatic world, drawn up by Ptolemy, put the ir s tamp inde l ibly on 1 5 cen

turi es . There in cons is ts his enormous s ign ificance for the his torica l deve lopment ofgeography . H e gave to h is own time a fund of collected and s olid knowledge , buthe acted as a check on later epochs who , in blind be l ie f in his authority, he ld fas tto his image of Southe rn As ia as to an unshakable dogma . In a time when Ga l ile imade his revolutionary d is cove ries in the heavens , Ptolomy s ti ll ru led a lmos t absolute ly on the earth .

Du ring the age of the Cal iphs , geographica l learn ing flou rished among the

XXI

Arabs,but they too bu i lt on the sys temof Ptolemy . They knew of Tibe t, b ut had

ve ry vague ideas in the matte r. In the year 851 , the name of the mys teriouscountry is mentioned by the me rchant Su le iman ,

b ut i t is the musk-trade thatawakens his greate s t inte re s t . Through the trade that the Cal iphs of Bagdad carriedon w ith India and China , the Arabs came into clos e r contact with the more dis tantcountries . Ibn Khordadhb eh speaks cursorily of Tibe t, so also does Al Ya ’kubiin the year 8 80. Mas ud i vis ited India in 9 1 2 and le ts us su spect that he had someno tion of the lake Manasarov ar. He endeavours to fix the s i tuation of the sourcesof the Panjab rive rs

,and touches as lightly as his predece s sors on the name of Tibe t.

Ib n Haukal and Alb eruni are not able to des ignate more dis tinctly the place of thisrea lm, b ut the latte r quote s the s tatements of the Ind ian Puranas abou t the land to

the North .

E dris i wrote in the year 1 1 54. For any one s eeking for information aboutthe hydrography of South-wes te rn Tibe t, he is the mos t va luable andmos t surpris ingamong the Arabian geographe rs . H is wri tings are , howeve r, of no easy interpre tation .

Also in his s tatements we are a llowed to get a glimps e of Manasarov ar, the holylake of the Hindoos . By Tibe t he meant ih ' rea l ity Ladak , as did many of his pre

deces sors . In 1 340 , Ib n Batuta trave lled in the s e parts . Accord ing to him, the

mounta ins of As s amare connected with the Tibe tan range s whe re the >>musk-gaze l le s »are to b e found .

Late r Mohammedan authors have not much to add above what has a lreadybe en told by the i r predeces sors . Mirza Ha idar, who trave l led in 1 53 3 , i s the firs tone able to give us anything pos i tive out of his own experience . But the re s u lt iss l ight, and the geographica l names di fficu lt to identify . The Empe ror Akbar gavehi s attention for both practica l and re ligious reasons to geographic inves tigation , ins ide as we l l as ou ts ide the borde rs of his empire . H e cau s ed a work to b e published,called Ain-i-Akbari , which is flow ing ove r w i th information abou t India . Of specia linte res t to us are the descriptions of the hydrographical sys temof the great Indianrive rs . But even to the author of Ain-i-Akbari , the concept ion of Tibet only cove redLadak and Baltis tan .

The Chines e forma splend id exception fromthe nebu lous obscurity in whichothe r Orienta l nations have floated w ith regard to the ir know ledge of Tibe t . The irpol itics , the ir trade and, not leas t, the i r s ens e of geographica l real i ties , brought themto this po int long be fore the name of Tibe t was known in E urope . A lso the des i refor re l igious re s earch has induced Chine s e buddhis ts , s uch as Fa-Hian and Hiuen

Tsang to admi rable journeys in the inte rior of As ia . I amcoming back to the ir

xxu

jou rneys in the fourth volume of thi s book . In the firs t volume , I have part icu larlyfol lowed the revis ions and trans lations made by Klaproth , Abe l Rémusat, Bushe l l ,Brets chne ider, Rockhill , de Rhins , He rrmann and othe rs . Klaproth cou ld alreadyby the a id of Chines e sources give a ve ry good de s cription of the hydrographicalcondi tions of Southe rn Tibet . In many ins tance s his conclus ions we re remarkablyaccurate , and they prove how w e l l informed we re the au thors of whos e writ ings hehad made us e . Gene ral ly, the picture of, for ins tance , the sources of the great Ind ian rive rs , left by the Chine s e , is much more re l iable and true than the concep tionE u ropeans s ti l l al ive have formed of the same regions . I try to analys e the Chines ege ographe rs ’ de scriptions of different extents of land in S outh-wes tern Tibe t and to

compare the ir resu lts w i th my own . In this po int I have found exce l lent gu idance ind’

Anv ille’

s map of Tibe t , founded on Chines e s ource s , and in Dutreu i l de Rhins ’ cons cientious explanat ions of the Ta—ch ’ ing map. The Chine se are gene ral ly much les saccurate in the i r orographic than in the ir hydrographic draw ings . The i r maps of

the sources of the S atlej and the Brahmaputra, and of the northe rn tributa ries to

the Ts angpo , w e re the bes t mate rials exis t ing be fore my journey.

Afte r this , I pas s on to speak of the know ledge E u rope had of Tibe t duringthe middle age s . Firs t among the trave l lers fromthis pe riod s tands Pian de Carpine , who began his journey in 1 245, and only incidentally te l ls some fantas tic s to

ries fromTibe t. Rubruck s tarted s even years later for the Far E as t , and he increas edto a ce rtain extent the mate rials of informati on le ft by Pian de Carpine . MarcoP010 is the firs t E u ropean who has given any re l iable de scriptions of the inacces s ibleland . Al l that he has to te l l , e special ly about the cus toms and usage s of the peoples ,carries the s tamp of the greate s t ve racity. He i s the foremos t of all the trave l le rsof the middle ages and the pionee r for As iatic exploration of all time s . Odoric is ,on the othe r hand, the firs t E uropean to have trave l led s tra ight through Tibe t inits prope r s ens e , and he has even vis i ted Lhasa , in the year of 1 3 30. But he has

s carce ly anything to te ll abou t the country . H is contemporary Mandeville was a

swindle r who had s tolen his informat ion fromOdoric and othe rs . To this epochalso be longs Benjamin of Tude la , who comple ted his jou rneys in 1 1 73 . The fol lowing is all that he has to say abou t Tibet : »The country of Tub oth , which is threemonths journey dis tant fromArabia» , and : >>In four dayes journey fromhence (Samarcheneth) you come to Tub ot a Me tropol i tan Citi e , in the Woods whereof swee tsme ll ing Mos s e is Afte r the time s of the great trave lle rs , the exis tence of

I Purchas His P ilgrims , V ol . V III page 559 and 576 . [ quote himhere , because I haveneglected to menti on his name in the following text.

XXIII

Tibet was forgotten among the E u ropeans , and the country had to b e rediscove red300 years afte r Odoric. In the 1 6th and i 7th centuries , innume rable journeys weremade to India. But it s e ems as i f thes e trave l lers were neve r able to s ee anythingwith the i r own eyes or make any new obs ervations . They repeat each othe r’s wordsinterminably , they quote the clas s ic authors and look upon Ptolemy as the principalauthority. All knowledge of thes e parts has , unti l the middle of the 1 7th century,taken a crys tal l is ed form afte r antiqu e patterns , and the clas s ic authors have often

gi ven to the world much be tter de s criptions than have the recorde rs of journeysof . this pe riod. Thus we re the Indus , the S atlej and the Ganges be tter knownduring the Gre ek and Roman antiqu ity , than by the ge ographe rs of the 1 6thand 1 7th centurie s . The Himalayas pas s unde r both ancient and modern namessuch as Imau s , Tau rus , Caucasus , Paropamisus , Nagrakot and the Us sonte -mountains ,and are cons ide red to s eparate India fromTartary or Scythia to the North . Kashmir and S rinagar are names that occu r qu ite often , and some times even the kingdom of Tibe t is mentioned , b ut vague ly and dimly, and is common ly looked upon

,

even by the renowned Thevenot, as a part of Tartary . Tave rn ier has a few thingsto s ay about Boutan , though mos tly out of inte res t for trade . In his des cription he ,almos t w i thou t know ing of i t, touche s on Tibe t . Van Twis t in 1 63 8 is the firs t oneto mention the name of Mas oroor i . e . Manasarov ar, so renowned among the H in

doos and, late r on , even in E u rope .

None of the s e trave lle rs had pene trated beyond the high mountains to the

North . They were E as t India trave l lers who only we re able to gathe r s econd and

third hand informations . Wide ly more important and eminen t we re the two Cathol icMis s ionarie s Bened ict Goe

'

s and An ton io de Andrade . Goes was one of the Jesu i tswho came to Lahore in 1 594 on the invitation of the Empe ror Akbar. At Goa , he

was commis s ioned by his frate rn ity to d is cove r Cathay , be l ieved to b e anothe r countrythan China . H is journey carried himacros s a lmos t the whole inte rior of As ia , ove rKabu l , Pami r, Khotan, Turfan , and Su-

chou . It is a mos t remarkable journey, b utonly touching on the borders of our te rri to ry, which he pas sed round. Andrade isthe firs t E u ropean to have cros s ed the Himalayas fromInd ia , which he did in 1 624 ,

on his renowned and important journey to Ts aparang on the uppe r S atlej . Afterhaving trave lled to this city a s econd time , he was fol lowed by a numbe r of M iss ionaries , b ut of the i r jou rneys we have got no information . The geographicalknow ledge , left us by Andrade , is a min imum

,b ut s ti l l of the greates t interes t. H e

s peaks of a smal l lake on the Mana pas s , be l ie ved by geographe rs s til l l iving to

b e Manasarov ar. I t was the learned Kircher who, w ithou t having any idea of the

XXIV

exis tence of this lake , e rroneous ly attributed to Andrade the meri t of having discove red the lake forming the source of the Gange s and the res t of the great rive rsof India . E ven in our days the re exis t geographe rs who , w i thou t any shade of

foundati on , s uppos e Andrade to have dis cove red Manas arov ar and to have pene tratedthrough Tibe t to China . In his own writings there is not a s ingle line to sucheffect. By others he has , again with as great injus tice , been cal led a sw indle r.

In 1 66 1 and 1 662 , a memorable journey was e ffected by Gruebe r and Dorvi l le fromChina through Tibe t to India . I have re lated this voyage in Vol. IV of

this work , to whe re it chiefly be longs . If we paus e for a moment at the las tmentioned epoch looking back ove r the pas t, we mus t admit that E u ropeans ’ knowledge conce rn ing Tibe t was s till ins ign ifican t , although Odoric, G ruebe r, Dorv ille and

Andrade had comple ted the ir journeys as we l l in the inte rior of that country as

through the whole of its exten t .The next part of the firs t volume of my work embraces the cartographical

his tory of Tibe t . Cons idering the nature of the subject , it is obvious that wemus t he re pas s on fromgene ral views to deta ils . On the oldes t maps of whichmention can b e made in this connection ,

it is as much as the name of Tibe t hasbe en bare ly put down . Late r on it is roaming abou t here and there through the

regi ons North of India , and i t is firs t in times far advanced that this rea lm, so

narrow ly engirded , begins to take shape also on the map . In orde r to b e enabledto follow the cours e of this deve lopment, and to s e e how Tibe t s low ly appears as

if grow ing forth on the map unti l i t final ly forms a sharply defined geographicalconception , we mus t go back as far as to the middle of the 1 5th century . But i tis firs t from the year 1 700 that, by application of the me thod of the S tateLibrarian Dr E . W . Dahlgren , we can try to dis engage ce rtain type s . For prior to thispe riod , Tibe t me re ly exi s ted as an acces sory de tai l , a name on the maps , playingno active part. From the year 1 700, the image on the map is more and moreconsol idated, deve loped and refined, unti l i t gains the appearance which is repre s entedon the mi ll ion-

s cale map be longing to this work . Within the range of this pe riodof more than 200 years , we can leave the ordinary i . e . the gene ral maps of the

whole of As ia , and pas s to the detailed maps of Tibe t and its special parts . The rewe can fo llow the cours e of deve lopment of the cartographical repre s entations of

the lakes Manasarov ar and Rakas -tal, of the s ou rce s of the Indus , the Satlej and

the Brahmapu tra , and of the mighty moun tain-mas s es to the North of the valley of

the Ts angpo .

I commence w i th the world map of the year 1 447 and Fra Mauro’

s of the

XXVI

altoge the r too long more incorrectly even than Ptolemy . The fantomlake Chiamay Lacus has been accepted s ince the days of Me rcator. According to Hoe ius ,

abou t 1 640, the Ganges has , i t is true , been given its correct place and i ts outflowinto the Bay of Bengal , b ut the le ngt h of the rive r is , as a lso the cou rs e of the

Indus , according to Hondius , enormous ly exagge rated, and here also Ptolemy , 1 5

centurie s e arl ie r, was more to b e depended upon . It is wonde rfu l to see how loos eand shifting i s the ground on which the cartographers of this pe riod are s tanding ,as soon as they leave Ptolemy . Whe re he is fo l lowed , as the case i s regardingIndia , the Oxus and the Jaxarte s , the maps are mos t accurate du ring this pe riod.

But a frightfu l confus ion take s place , when they throw out the mounta in-range running E as t and We s t to the North of India

,and replace i t by more longi tud inal moun

ta ins , as doe s Hoe iu s . The S atlej , which s ti l l rema ined in the maps of Gas taldi ,Me rcator , Orte l ius and Hondius , i s mis s ing in the ske tche s of Hoe ius , Blae u and

Ians sonius from 1 640 and 1 64 1 . This is due to the s lowly fad ing away of the au

thority of Ptolemy , which i s no advantage in a time whe n the re was nothing be tte rto offe r ins tead . Of the course of the Brahmapu tra the re is , if pos s ible , s ti l l les strace than on the map of the Alexandrian geographe r.

But the ecl ips e did not las t long. Al ready in 1 654 , S anson d’

Abb ev ille marksa huge progre s s . He force s the Indus and the Gange s back into more rea l proport ions , and le ts the bas in of the i r sou rce s b e borde red by Mont de Caucas e , whichcorre sponds to the Himalayas . To the North the reof, he places Tibe t. Of s t il lgreate r s ignificance is Martin i ’s brill iant repre s entation of China in 1 655. What heoffers is bu i lt on rea l obs ervations . H e locate s Tibe t Regnumqu i te correctly inthe regi on to the We s t of the uppe r cours e of the Ye l low R ive r. In 1 680,

V is s cher

is the firs t one who endeavou red to make u s e of the d iscove rie s brought home byGruebe r and Be rnie r from the i r trave ls . Cante ll i ’s map from 1 68 3 is of greatin te re s t . The Indus and the Gange s he trace s in the same manne r as do his neare s tpredece s sors , and the Brahmapu tra i s mis s ing. Like Sanson d’Abb ev ille , he has a

Ra ia Tibb on , which in all probabil ity can b e trans lated by Raja of Tibe t, or the

King of Ladak . In his map, a numbe r of the geographica l name s of Goes are

also recurring . But s ti ll , the mos t inte re s ting fact i s that , ove r a comparative lylarge te rritory , which , i t i s true , l ie s a l ittle too far to the E as t, he has ente red ind is tinct wri ting thre e d iffe rent name s for the hidden land of the Lamas , Tobat, Thibe tor Theb e t . Here we have aga in a fe e l ing that the ve i l is be ing ren t as unde r , andthat the unknown land is appearing in view .

VVi ts en , in 1 68 7 , has removed Tibe t yet a s tep to the E as t, which is no im

XXVII

provement on the map of Cante l li . H is picture of the inte rior of As ia was le s s of

a s ucce s s than the ske tch made by Gas taldi . We notice e specia l ly that he le ts theGanges take its origin froma lake which he cal ls S iba lacu s . This is a ls o to b e

found in the map of de Witt , who in i ts ne ighbou rhood has anothe rmys terious lake ,cal led Be rvan . To the s ame period be long the va luable de s criptions le ft by the

prominen t Mis s ionarie s Ve rb ie s t and Gerb i llion . The latte r in particu lar, who e ffecteda numbe r of jou rneys through Mongol ia betw ee n the years 1 688 and 1 698 , des cribesbe tte r than any earl ie r ge ographe r the roads to Tibe t , though he had neve r trave lledby themhims e lf.

In the beginn ing of the 1 8 th cen tury, appeared De lis le as a pione e r and re formerin cartography , which he treated in an entire ly s cientific manne r . H i s map of 1 705

show s that, wi th cri tical acu tene s s , he has made u se of all mate ria l exis t ing at thattime . N ow , afte r 1 600 years , Ptolemy i s played out and dis appears for eve r. De

l is le endeavours to inte rpre t in picture s the de s cript ion Andrade has made of hi s

jou rney , b ut , natural ly enough , he has misunde rs tood him. He has given Tibet itsright pos ition and le ts it b e s eparated fromIndia by the mountain—range of M . de

Purb e t ou de N augracut . He doe s not ye t su spect the exis tence of the Tsangpo

Brahmaputra , though this rive r was cros s ed by Odoric, Gruebe r and Dorv ille . The

des criptions more than lean of the i r journeys , are ove r sufficien t to expla in this de fect.Ye t it did not tarry any longe r than ti l l the year 1 7 2 3 , be fore De l i s le cou ld s ti l lmore improve his va luable map. One can read off, s tra ight fromthe map, that hehas made us e of informations , fe tched fromFrench Mis s ionaries in Peking . Herewe even find Las s a located on the banks of a great rive r, the Bramanpou tre , whichis the uppe r cours e of the Bengal rive r Laqu ia . On th is account , the re is no longe rany room for the legendary lake Chiamay Lacus , which now d isappears for eve rfrom the map of As ia . S til l w e mu s t obs e rve that the problemTsangpo-Brahmapu tra , so difficu lt of solution even in our times , was already qu ite correctly e lucidatedby De l is le . It was Klaproth who, more than a hundred years late r , enti re ly spoi ltand dis torted thi s hydrographical problemby his awkward be l ie f, that the Irrawaddiwas the lowe r cours e of the Tsangpo . The mas terly maps of De l is le s eemumfortunate ly to have made no de epe r impre s s ion on his time , and on the next fol lowing era after him. H is greate s t me rit was that he unde rs tood how to explain the

fundamental documents in an exact and crit ical manner. The s ame materia l hadbeen at the d isposal of h is predece s sors 30 or 40 years earlie r, b ut they had not

known how to make use of it .

XXVIII

Through nearly 1 60 years , the mons trou s apparition which had rece ived the

name of Lago de Chiamay, had haunted all the maps of As ia . In Ramus io ’

s Na

v igat ion i e t V iaggi , 1 550 , we find this lake firs t mentioned . It originates fromthe

des cription of G iov an de Barros . This lake was located to the dis tri ct N .N .E . or

NE . of the Bay of Benga l , and i t was be l ieved to give ris e to fou r or more largerive rs . In the year 1 649 , Vincent le Blanc even knows the dimens ions and othe rcharacte ris tics of this lake . The gre at cartographe rs Sanson d ’

Abb ev ille and Mar

tin i are blind be l ieve rs in its exis tence . Ptolemy had no idea of such a lake . Lagode Chiamay mus t the refore b e a new conception which had got a footing in the

maps . I have succe eded in proving that this s eemingly phenomenal lake neverthe le s shad a solid foundat ion in rea l ity. According to V is scher, Lago de Chiamay has already begun to wande r. Oh the map of Cante l l i , it has made so large a s tride tothe \Ves t, that Barantola and Las s a are s ituated to the E as t of i ts b ed. Isb rants

Ides in 1 704 remove s the lake far to the N W . b ut , to b e on the safe s ide , he le tsanothe r lake lie in the old place and formthe source of the great rivers . In 1 705,

De l is le retains the lake , b ut le ts on ly one rive r take ris e fromits wate rs . On his

map of 1 7 2 3 , he has , as mentioned, final ly s truck it out . But in the meantime the

lake has wande red towards the Wes t to the place whe re it is in real ity s ituated.

Lago de Chiamay is name ly noth ing e ls e than Manasarov ar, which has be en con

s idered , and by the people s of the E as t is , to this day, be l ieved, to give ris e to the

great Indian rivers .

In my re lation I then pas s on to two Swede s of great me rit, S trahlenbe rgand Remat who had s e rved in the army of Charles XII and occupied the ir timedu ring captivity to make peace fu l conque s ts in the interior of As ia . The map of

S trahlenberg, published at S tockholmin 1 730, touche s pri ncipal ly on the northe rnparts of the continent, b ut even extends so far to the South as to compris e the

source s of the Indus and the Gange s . In s eve ra l respects , it is supe rior to manyothe r contemporary maps . This can b e s a id in a s til l highe r degree of Re nat

smapof the Tarim bas in and its adjo in ing d is trict . He re , Renat ’s map of 1 7 3 3 is at

leas t as near to real ity as the map re fe rring to Central As ia in S tieler’s Hand-Atlasof 1 875.

I

In this connection I wish to mention another Swede , whose name i s not to b e found in myrelation . For l ingu istic purposes , Johan Otter made a journey in 1 73 8

—1 73 9 to Bagdad , Ispahan and

Basra , thereby taking the Opportun ity a l so to note down certain geographica l obs ervations . About theGange s he says : >>Le Gange , une de s grandes riviere s de l ’Inde , sort des montagnes du Tebet ou dela Tatari e . II coule du nord au sud et se décharge par deux principales bouches dans le golfe de

XXIX

In the beginn ing of the 1 8 th century, the great French Je su its in the Pekingof the Empe ror Kang Hi , comple ted the i r renowned map of China . A copy of i treached d ’Anv ille at Paris . On this map Tibe t is repres ented in accordance w ith theprovis ional s urvey carried ou t, at the command of the Empe ror, by specially ins tructedLamas . They commi t the e rror of le tting the Ganges take its origin from lakeManasarov ar. D

Anv ille caused both the uppe r Indus and Satlej to b e included inthe rive r-sys tem of the Gange s . O the rw ise , the powe rs of obs e rvation of the La

ma is tic cartographe rs we re extraordinarily accurate , and the ske tch of the bas in of

the , sources of the Satlej and the Ganges which they pres ented to d’

Anv ille , is

admi rable . The i r achie vments su rpas s all that cou ld b e produced by E uropeangeographe rs 1 70 years late r.

In connection w ith the re l iable des cription of the sacred lake and its sur

round ings , which was thus obtained by Lamais tic res earch , the re s u lts of the journey which was begun in the year 1 7 1 5 by the Je su i t Fathe r De s ide ri , are of ex

ceedingly great inte re s t. I amendeavouring to prove that his lake Re toa can b e

no othe r than Manasarov ar. He is the firs t E u ropean to have vis ited and describedthis lake , and he i s the d is cove re r of Kailas. He is one of the mos t remarkabletrave lle rs having eve r journeyed through the land of Dala i Lama .

Pate r Tieffenthaler is anothe r Je s u it whos e me rit howeve r le s s cons is ts in hisown obs ervations , than in hi s col lecting of geographica lmate rial . In the y ear 1 784,his maps were published by Anque til du Pe rron , who dis cus s es themw ith greatle arning . Fromthe narrative s of Indian pilgrims , the Fathe r had got the impre s s ionthat Manasarov ar gave rise to the Brahmapu tra . The s tatement that the Satlej hadits source fromthis lake , was , on the othe r hand , qu i te correct . Anque til du Pe rroncommitted the great mis take of re lying more on Tieffenthaler than on d

Anv ille ,

though the latte r bu i lt on be tte r mate ria l than did the forme r .Vol. II of my work forms the immediate cont inuat ion of Vol. I and

,l ike this

Bengale .» Concern in g »K ichemir», he says that it is a »canton» , »environné de montagnes qui dominent du cote’ du sud sur les frontieres des Royaumes de Dehl i e t de Labour, a l ’e s t sur ce lle du Tebet,au nord sur ce lle de Bedahchan, e t s ur une partie du Khorasan , e t a l 'ouest sur les cantons des Afgans .

Quant aux montagnes de K ichemir, on diroit qu’

e l les ont été place’ es expres comme une mura i lle pourfortifier cc pays , et pour le mettre hors d’ insulte . On ne peut y entrer que par trois dehles , savoir,ce lu i du Khorasan , qu i es t s i étroit e t s i d ifficile , que les hommes sont ob liges d ’y porter les chargessur leurs e

’ paules , n’

e’

tant pas pos s ible d’

y fa ire passer une bete chargée ; ce lu i de l ’Inde qu i n ’

e st pasmoins étroit et d ifficile que le premier, et ce lu i de Tebet. Quoique ce dern ier soit un peumoins rudeque les précédens , il n

est guere practicable pour les b éte s de charge , par ce qu ’

on n’

y trouve dansl’

espace de plus ieurs journées , que des herbes ven imeuses Voyage en Turqu ie et en Perse . Tomesecond , Paris MDCCXLVIII , p . 68 e t 1 2 1 .

XXX

part, treats the hydrographical problem of which Manas arov ar forms the centre ,and which has through centuries been the s ubject of a more or le s s we l l informedd is cus s ion . Bu t while V ol. I was chiefly devoted to Tibe t in gene ral , and, fromi tspe riphe ra l regions , s tep by s te p became concentrated in the d irection of the s acredlake and the s ource s of the gre at rive rs , Vol. I I is divided be twe en the diffe rentde tai ls of the problem in que s tion . We have thus gradual ly proce eded fromge

ne ra l views to particu lars , and the vo lume ends with my own obs e rvations , made at

the two lake s , and at the s ou rce s of the Indus , the Satlej and the Brahmaputra .

In his map of 1 78 2 , Major Renne l l le ts the Indus take its origin from the

moun tains We s t of Kashgar, and the Satlej fromthe southe rn s ide of the Himalayas . What is in real i ty the uppe r cours e of the s e rive rs , is by himchanged intosou rce -rive rs of the Ganges . He had ce rtainly nothing be tte r to do than to tru s tthe Lamas of Kang Hi , who traced the origin of the Satlej fromRakas -tal, b ut hadca lled the rive r the Gange s . A lso Wilford who , at the beginning of the 1 9th cen

tu ry, col lected all mate ria ls then exis ting regarding the geography of the se parts ,le ts the Ganges take its origin fromManasarov ar.

E lphins tone and Macartney tried in va in to solve the problemof the sourceof the Indu s , while Webb and Rape r in 1 808 succe eded in pene trat ing to the truesource of the Gange s on the sou the rn s ide of the Himalaya , withou t know ing thatthis feat a lready had be e n achieved 1 84 ye ars earlie r byAndrade .

In 1 8 1 2 , Moorcroft at las t pene trated to the two lakes , of which he gave an

exce ed ingly va luable de s cription . He confirmed that , at the s aid pe riod , no rive rflowed ou t of Manas arov ar, and h is re l iabili ty was que s tioned fromce rtain quarte rs ,when Henry S trachey in 1 846 and R ichard S trachey in 1 848 found a l ive ly flowcoming from the s acred lake . During the decade s which have pas sed s ince then ,d i ffe rent s tatements have apparently s tood agains t e ach othe r. Some explore rs havema inta ined that a cu rrent of wate r was runn ing out fromthe lake , while othe rshave pos i tive ly den ied the e x is tence of s uch an outflow. At the t ime of both myvis its , in 1 907 and 1 908 , I found the outgoing channe l dry, b ut I learnt froma

re l iable Mohammedan me rchant at Leh , whos e caravans every year take this roadto Lhasa , that in 1 909 to 1 9 1 1 an abundant s tream of wate r flowed fromManas arov ar to Rakas -tal, while , in 1 9 1 2 and 1 9 1 3 , the s e two lakes we re aga in cut off

fromeach other.With the s upport of all this mate rial , I prove that the appare ntly contrad ictory

s tateme nts in real ity are s tand ing in the gre ates t and mos t natu ra l harmony witheach othe r . The outflow fromlake Manas arov ar is a pe riodica l phenomenon, whos e

XXXI

pu ls ations s imply depend on the quantity of the annua l ra infal l in the Himalaya and

the Transhimalaya . Afte r dry s umme rs and au tumns , the lake has no surplus of

wate r to part w ith . After copiou s ra infall, its s urface ove rflows , and a portion of

its wate r runs out through the channe l to Rakas -tal. The same cl imat ica l change sthat characte ris e the pos t glacial pe riod , are a ls o to b e obs e rved he re . Withou tdoubt, the great curve of cl imatica l change has hithe rto proceeded towards des iccat ion , becaus e , as far as I have found, no wate r has flowed out of Rakas—tal s ince1 760 . Wha t the future can hold in s tore in this re spect, is unknown to us . Per

haps the curve of ra infa ll shal l again begin to ris e fromits drift to de s s icat ion . The

osci l lations in the leve l of wate r in lake Manasarov ar, be long to a curve of lowe rorde r than the oscil lations of Rakas -tal. The forme r depend on years and decade s

,

the latte r on centurie s .

With the s upport of all this mate rial , the re is then no difficu lty in proving thatthe pe riodici ty i s a secondary phenomenon , which doe s not in the leas t prevent thetwo lake s from be longing to the cours e of the Satlej and the rive r-sys temof the

Indus . This also goe s to es tablish that the sou rce of the Satlej shou ld not b e

s ought for in any of the place s forme rly de s ignated as such , b ut on ly and exclu s ive lyat the po int where the large s t tributary to Manasarov ar, or Tage—tsangpo , is bornat the front of the glaciers of Ganglung . This view even agre e s with that of theChine se and the Tibe tans . It will , moreove r, prove to b e the only correct one infuture , when the s e parts have been thoroughly mapped and explored.

In Vol. II I , I treat the Transhimalaya and analys e in his torica l orde r all theexpedit ions having touched on the pe riphe ry of the unknown centra l parts of thissys tem, whose w ings to the E as t and We s t we re comparative ly we l l known . D e

s ideri , Beligatt i and de l la Penna are the firs t E u ropeans who have at all be en in

contact with this great and attract ive orographica l problem. Afte r them fol lowedAbbé Huc, Thomson , Hooke r, Campbe l l , S chlagintwe it , Drew , S trachey , Cunn ingham and the Pundi ts of Montgomeri e and, fina l ly , a few mode rn trave lle rs , las t ofthem Ryde r , Raw l ing and Wood . But even more fe rtile is an analys is of the de

scriptions of the mounta inland to the North of the Tsangpo , given by prominentgeographe rs , and partly founded on Chines e documents , partly be ing of theore ticalcons truction . The foremos t of thes e geographe rs are d

Anv ille , Klaproth , R itte r,Humboldt , B . Hodgson ,

R ichthofen , Saunde rs , Markham, Dutreu il de Rhins , E . T .

Atkinson and, las tly , S . G . Burrard and Hayden . Afte r having given in turn an

account of each of the ir varying sys tems , of which not a s ingle one corresponds to

rea lity , I pas s on to a de scription of my own jou rneys in the Transhimalaya , and

XXXII

along the cha in of lakes , s i tuated at the i r northe rn bas e . In this introduct ionI shal l des is t fromgi ving a gene ra l s ta tement of the i r re su lts . The Transh imalayas tands to the We s t in clos e connection with the Karakorumsys tem, and on this account I have occas ion to come back in Vol . IV of this work , as be ing the rightplace , both to a gene ra l view of the d iffe rent cha ins of the Transhimalaya , des cribedin Vol. III, and to the ir orographica l re lation to the mounta ins of the Karakorum.

S tockho lm,October 1 7th , 1 9 1 7 .

S VE N H

CHA PTE R I.

SOUTH-WESTERN T IBET IN IND IAN ANT IQ U ITY.

To the Hindus and the i r ances tors Sou th-Wes tern Tibe t has be en known s incethousands of years , and indeed we are he re movi ng on ground which was clas s icwhen the foundation s tone of Rome was laid. The i r knowledge has , howeve r, be enve ry s upe rficial , and the few glimps es of actua l geography w e ob tain fromthe i rancient books almos t d isappear in inte rminable epic songs , in legends and in tales .

But there i s at any rate a ce rtain foundati on of real i ty , which proves that the peoplesof India knew the exis tence of the ve ry high and extens ive mountains to the

north , as we ll as Mount Ka i las,the lakes Manasarov ar and Rakas -tal, and the

regi ons in which the great Indian rivers have the i r sources . As an introduction to

the geographical inves tigation of Sou th-Wes te rn Tibet I wil l the re fore u s e some quotations fromold Indian books .

In the Buddhis t world-sys temMount Me ru or Sumeru , the R i-rab of the Tib etans , the great vene rable O lympus

,ris es fromthe axis of the earth , and forms

the centre and foundati on of the un ive rs e . Its four s ides cons is t of gold , crys tal ,s i lve r and sapphire . In concen tric ci rcles Mount Me ru is su rrounded by s even s eas

and s even rings of golden mounta ins . In the ocean around this King of Mountainsare the fou r great continen ts or world-is lands

wi th the i r s ate l l i tes . Three of thes econtinen ts are fabu lous , and only the fou rth , the one to the south

, Jambudv ipa , hasa real geographical foundation , and corres ponds to the world as far as i t was knownto the Indians

, name ly , the Indian pen insu la , with the Bodhi-tree of Budh Gaya in itscentre . »It is shaped l ike the shou lde r-blade of a she ep , this idea be ing evidentlysugges ted by the shape of the Indian pen insu la which was the prototype of Jambudv ipa , as Mount Ka i las in the Himalayas and NE . of India was that of MountMe ru .

»

Waddel l : The Buddhismof T ibet or Lamaism. London 1 895, p . 7 8 et seq . O f Jamb udv ipaKoppen says : »ih ihm erstreckt s ich der H imav ant mit demSee Anav atapta (imPal i Anotatta , derbuddhistische N ame des Manasa-Sees ) , aus we lchemdie vier gros sen Strome nach den vier Himme lsgegenden flies sen , der Ganges nach O sten , der Indus nach Suden , der V atsch (Oxus) nach Westen ,

SOUTH -WESTE RN TIBE T IN INDIAN ANTIQUITY .In the Hindu mythology Me ru is a fabu lous moun ta in , for on i t is s i tuated

Svarga , Indra’s heaven wi th the ci ties of the gods , and the habitations of ce les tia lsp i ri ts .

This mighty upheaval , the highland of Tibet , the Me ru of the Indian cosmography

,was pra is ed in the fol lowing words by the Ma/caélzara ta , the great epic poem

of the Hindus , and probably the greates t poemin the world ’

s l iterature : »There is amoun ta in named Meru of blaz ing appearance , and looking like a huge heap of effulgence . The rays of the sun fal l ing on its peaks of golden lus tre are dispers ed bythem. Abounding w i th gold and of variegated tints , that mounta in i s the haun t ofthe gods and the Gandharvas . It is immeasurable , and unapproachable by men of

man i fold s ins . Dreadfu l! beas ts of prey inhabit its breas t, and i t is i lluminated wi thd ivine herbs of healing vi rtue . It s tande th kis s ing the heavens by its he ight and isthe firs t of mounta ins . Ordinary people cannot so much as think of as cending i t

S tanding high for infin ite ages , upon i t once all the mighty ce les tials sat themdown and he ld a conclave» .

2

The s ame source te lls us that Meru is in the Himalaya between Malyav ant andGandhamadana . >>This gold-mounta in is the highe s t of all mountains . I t is roundas a ball , shines l ike the morn ing s un

,and is l ike a fire w i thou t smoke . I t is

Yojanas high and goes as far down in depth , and i t overshadows the worldsabove and be low and acros s I t is furn ished w ith heavenly flowe rs and fru i t ,and cove red everywhere w i th bright gold dwe ll ings The top of Me ru is

covered with fores ts that are b eautified w ith flowe rs and the wide-s tre tching branchesof Jambu tre es , and which res ounded with the me lodious voices of kinnari es .» 3

V ishnu ’

s dwe l l ing place is on the top of Mount Mandara to the eas t of Me ruand to the . north of the Milk-Sea . On the sou th of the Ni la mounta in and

the northe rn s ide of Me ru are the s acred Northern Kurus , the res idence of the

S iddhas .4

In the same ancient poemwe read : >>Fromthe H imav ant mountains which arethe mos t exce l lent in the world and which are extol led as divi ne , holy and lovedby the gods who s eek thes e regi ons , they us ing themas pleas ure-grounds ; fromthes e mounta ins of which Meru is the centre you ascend through the air to Svarga .

di e S i ta (der F lus s v on Yarkan) nach Norden.» Sun , moon, and stars turn round the Meru , and

above the K ing of Mountains are the heavens . K0ppen : Die Religi on des Buddha, Berl in ,1 857 , I,

2 3 2 et s eq .

1 Froma purely geographical point of v iew Dowson be l ieves it should b e identified with somemountain north of the H imalayas . It is a lso Hemadri , »golden mountain» , R atnasanu , »jewe l peak» ,Karnikachala , » lotus mountain» , and Amaradri and Deva-parv ata , amountain of the gods ». Dowson ,

A Clas s ica l D ictionary of Hindu Mythology and Re l igion , Geography, H istory, and L iterature . Second

Ed ition . London 1 8 8 8 p . 2 08 .

2 Indian Mythology according to the Mahabharata. In outl ine by V . Fausb oll. London 1 902 ,p . 20.

3 Ibidemp . 43 .

4 Ibidemp . 1 03 and 1 69.

MAHABHARATA AND RAMAYANA .J

FromMandara in this mountain-range Arjuna as cended to Indra ’

s heaven conveyedthere by Matali in Indra ’s chariot, afte r he had firs t sung this beau ti fu l hymn to the

mounta in Then follows the hymn glori fying H imav ant .

According to Fausb oll Civ a is mos t often in the Mahabhara ta called Mahadeva ,the great god. He dwe l ls on the holy H imav ant . He re on the ridge of Himav antthe mighty mas te r always s i ts , shin ing like the fire at the end of a

'

Yuga . In othe rplaces he is recorded to b e found on a horn of the moun tain Me ru , cal led Sav i tra ,or on Mount Mandara . »The H imav ant mounta ins hold the highes t place amongs tall the mounta ins in the world , they are prais ed as be ing divi ne , holy and loved bythe gods and they are protected by Rakshasa ’

s and Pigaca’

s . Fromthemthe as cen tis

'

made,through the air, up to Svarga—heaven with the Nandana fores t

,the home

of the inhabitants of heaven , the Deva ’s . The high Kuvera enjoys a fou rth partof Meru ’

s treasures and he give s a s ixteenth part to mankind . S ou th of Moun tNi la and on the north s ide of Me ru l ies the holy northe rly Ku ru-land whe re the

S iddhas dwe l l Among Himalaya ’s many great and smal l mountains Kuverachos e Kai lasa (also cal led Hemak i‘ i ta) and Gandhamadana as his favou ri te dwe llingplace The whitish-ye llow Kai lasa is 6 Yojana ’

s high ; and a gigantic jujubetree is found there . It is likewis e cove red w i th love ly woods , rive rs , lakes and

caves . Here Kuvera was ins tal led by Brahma hims e lf in domin ion ove r all richesand ove r Rakshasa ’

s , Yaksha’

s and Gandharva ’

s , and he rejoiced greatly. At the en

trance to Kai lasa the re is a golden gate H is (Kuve ra ’s) rive r i s the beau tifu lMandakini , the firs t of rive rs whos e waters are decked with golden lotu s es , thatres emble the s un . H is lotus-lake , Nalini or Jamb i

i nada-saras , the golden lake is

called A laka. It is fu l l of d ivine , fragran t, golden Saugandhika Iotus es and all sortsof aquatic birds , surrounded by love ly woods with thick tre es and cl imbing plantsIts wate r is clear and cool , and has an ambros ia l tas te H is ci ty is cal ledAlaka H is as s embly-hal l is as bright as the peaks of Kai lasa, and i ts

white sheen eclips es even the s plendour of the moon .»

In the lMahabharata Brahma is des cribed unde r the name of Manasa , and is

some time s a lso called Puru sa. The Gandharvas are s a id to res ide near the lakeManasa and on Moun t N isadha .

2

Only as glimps es or me re names do the e terna l mountains of Himalaya , Me ruand Ka i las , and the sacred lake Manasarov ar l ight up the interminable and ve rydu l l s tories of the R amayazza . I amgiving a few examples of this kind of ve ryvague geography. Raghav a is said to have »proceeded leaving on his left s ide the

junctions of fou r roads and reached afte r all the fami ly dwe l l ing of the king ,with its palatial tops pie rcing the sky, looking beau ti fu l , res embling amas s of clouds ,white as the ce les tia l cars and high as the hil l Kai lasa, and with sporting hous es

Ibidemp . 1 4 7 , 1 8 3 , et seq.

Ibidemp . 58 and 1 2 7 .

6 SOUTH -WE STE RN TIBET IN INDIAN ANTIQUITY.adorned with pearls . »

I E ve rywhere appears the deepes t veneration for the Kai las , .

>>that lord of mounta ins , Kai lasa .»2 Fu rthe r : >>The re i s a monarch of mountain s in

a mighty fores t, the great refuge of asceti sm, the worke r of the weal (of all), an

Asura , ce lebrated by the name of H imav ant contain ing great cas cades and furn ishedwith many founta ins and caves .> 3 In the monkey s tories in the same work bothHimalaya and Ka i las and othe r mountains are ment ioned . In Vol . V it is said of

a palace that i t was rhigh as the lord of mountains and white as the Kai las a hill»,

and in Vol. VI that the God of Death was l iving on mount Kai lasa , e nci rcled byYakshas . Of a certa in Kumbhakarna i t is said : »as he yawned

,his mou th res emb

l ing the subte rraneou s regions , looked l ike the sun s tat ioned at the s ummi t of Meru» .

A Brahmarshi , foremos t of as ce tics , repai ring to the asylumof Trinav indu at the s ideof the mighty mountain , Me ru ,

took up his abode there .4 The same volume te lls

of Ravana : »And leaving behind ci ties and s treams and hills and woods and groves ,he in a moment came to the moun ta in Kai lasa . » Fu rthe r : >medi tating for sometime th e lord of Lanka went to the region of the sun and spent the n ight uponthe pictu res que summi t of the moun t Me ru .» »When Ravana arrived at the banksof the wes te rn Ocean a person was s e en on the is land , a highly powe rfu l pe rsonamongs t men l ike the Me ru amongs t the mounta ins , with legs l ike the mountKai las a , and he was not more agitated than the Sumeru by the King of s er

pen ts ; and of Ravana is sa id : the mounts Himalaya,Hemakfita, Mandara , and Me ru

are his bones .» >>There i s a mountain named Me ru

,highly pictu resqu e

,made of

gold and greatly charming . The middle peak is much liked by the ce les tials , on

which is s ituate the beau ti fu l as s emblage hal l of Brahma extend ing over a hundredleagues . The fou r-mou thed de ity ,

sprung fromlotus , a lways res ideth there » .

The P uranas ce lebrate , as DOWSON s ays , the powe rs and works of pos i tivegods , and repres ent a later and more extravagan t deve lopmen t of Hindu ism, of whichthey are in fact the S cri ptures .

5 The bes t known is the Vz’

s/zmc P a ran a trans latedby Profes sor WIL SON . In the firs t chapte r of its s econd Book is given the des crip

tion of the earth , much as has be en re lated above , though s ti l l with some interes ting geographica l de tai ls making it worth a fu l ler quotation .

Jambu is one of the s even great insu lar con tinen ts , and there are s even grea ts eas . »Jambu-dv ipa is in the centre of all thes e : and in the cen tre of this continent is the golden mountain Meru . The he ight of Me ru is e ighty-four thousandYojanas ; and its depth b elow the s urface of the earth is s ixteen thousand. Its

d iame ter at the summi t is thi rty-two thousand Yojanas ; and at its bas e , s ixteenthou sand : so that this mountain is l ike the s eed-cup of the lotus of the earth .

The R amayana . Trans lated into E ngl ish Prose fromthe Original Sanscrit of Valmiki, Valakandam. Vol . I I . Publ ished by M anmatha Nath Dutt. Calcutta 1 8 89, p . 2 3 8 .

3 Op . c it. V ol . II I, p . 6 1 8 .

3 Op . c it. Vol. IV , p . 7 1 9.

4 Op . cit. V ol. VII , p . 1 557 .

5 Dowson 0p. c it . p . 245.

VISHNU PURANA .I

The boundary mountains (of the earth) are H imav ant , Hemak i‘ i ta , and N ishadha ,which lie south of Meru ; and Ni la , Cv e ta , and Cringi , which are s i tuated to the

north of it. The two centra l ranges (thos e next to Me ru , or N ishadha and Ni la)extend for a hundred thou sand (Yojanas , runn ing eas t and wes t). E ach of the

others d imin ishe s ten thousand Yojanas , as it l ie s more remote fromthe centre .

They are two thousand Yojanas in he ight, and as many in breadth . The Varshas

or countries between these ranges are Bharata (Ind ia), sou th of the H imav antmountains ; etc. The re are also (near Meru) fou r great lakes , the waters of whichare partaken of by the gods

,ca lled Arunoda

,Mahab hadra , S i to

da , and Manasa .

The principal mountain ridges which proj ect fromthe bas e of Me ru , l ike filamentsfromthe root of the lotus , are (here fol low s eve ra l names ).

On the summi t of Me ru is the vas t ci ty of Brahma, extend ing fourteen thousand leagues , and renowned in heaven ; and around i t, in the card ina l po in ts and the

inte rmed iate quarte rs , are s ituated the s tate ly cities of Indra and the othe r regentsof the sphe res . The capita l of Brahma is enclos ed by the rive r Ganges , which ,is su ing fromthe foot of Vishnu

,and washing the lunar orb , fal ls he re fromthe s kies

,

and , afte r encircl ing the city , divides into fou r mighty rivers , flow ing in oppos ited i rections . Thes e rivers are the S i ta, the Alakananda, the Chakshu , and the Bhadra.

The firs t , fal l ing upon the tops of the infe rior mounta ins,on the eas t s ide of Meru ,

flows over the ir cres ts , and pas s es through the country of Bhadragv a to the ocean

the Alakananda flows sou th , to the country of Bharata , and , dividing into s even rive rson the way, fal ls into the s ea : the Chakshu falls into the s ea , afte r trave rs ing all the

wes tern mountains , and pas s ing through the country of Ketumala : and the Bhadrawashe s the country of the U ttara Kurus

,and empties i ts e lf into the northern ocean .»

I

The V ishnu Purana goe s on : >>Me ru,then , is confined be tween the mounta ins

Ni la and N ishadha (on the north and sou th), and be twe en Malyav ant and Gandhamadana (on the wes t and eas t) ; it l ies between them l ike the pe ricarp of a lotus .

The countries of Bharata , Ketumala , Bhadragv a , and Uttarakuru lie , l ike leaves of

the lotu s of the world,exterior to the boundary mounta ins . Jathara and Dev aknta

The V ishnu Purana, a systemof H indu Mythology and Tradition, trans lated fromthe origina lSanscrit, and il lustrated by notes derived chiefly from other Puranas , by H . H . Wi lson . London

1 840, p . 1 66 e t seq . In a note to the passage quoted Wi lson says that the Vayu Purana has thesame account but also another which is found in the Matsya and Padma Puranas as we l l . In th is

,

as re lated above , the Ganges is sa id , after e s caping fromCiva to have formed the seven rivers : N al ini , Hladin i and Pavani going eastwards , Chakshu , Si ta, and S indhu westwards , and Bhagirathi or

Ganges to the south . Wilson finds some actual geography in the legend. The fol low ing suppos itionof Wi lson , however, seems somewhat audac ious ; he says of the legend that »it seems not un l ike ly tohave originated in some imperfect account of four great rivers flowing fromthe H imalaya, and thehigh lands north of that range , towards the card inal points : the Bhadra, to the north , representingthe Oby of S iberia ; and the S i ta, the river of China , or Hoangho. The Alakananda is we l l knownas a main branch of the Ganges , near i ts source ; and the Chakshu i s very poss ib ly, as Major Wilford supposed , the Oxus . » For how wou ld even the most vague account of any S iberian river everhave reached India ! On ly Alakananda is certain , Oxus very likely; by the north-go ing river Yarkanddarya may have been meant and by the S i ta Tsangpo-Brahmaputra .

8 SOUTH -WE STERN TIBET uv INDIAN ANTIQUITY .are t wo moun ta in ranges , runn ing north and sou th

,and connecting the two chains

of N ishadha and Ni la. Gandhamadana and Kai lasa extend, eas t and wes t, e ightyYojanas in breadth , from s ea to s ea . N ishadha and Pariyatra are the l imi tativemounta ins on the wes t , s tre tching, l ike thos e on the eas t, be tween the Ni la and

N ishadha ranges : and the mountains Trigringa and Jarudhi are the northe rn l imitsof Me ru , extending, eas t and wes t, be twe en the two s eas . Thus I have repeatedto you the mounta ins des cribed by great sages as the boundary moun tains , s i tuatedin pai rs , on each of the fou r s ides of Me ru .

»

In the B /zag a v a ta and Vayu P a rana Wilson finds the e ightmountains enumerated in a s imi lar way. The Bhagavata places the mountains Kai lasa and Karavi ra south of Meru . Wilson does not be l ieve in the pos s ib il ity of verifying the

pos i tion of thes e d i ffe ren t creati ons of ancient legendary geography of the Hindus ,b ut he thinks the s cheme was sugges ted by imperfect acquaintance with the actualcharacte r of the country. The refore , quoting Humboldt and R itter he indicates an

iden ti ty w ith the Al ta i , Mus tag or Tian-shan , Kwen-lun and Himalaya . And hereaga in we have good reason to doubt whe the r the ancien t sages of Ind ia cou ldposs es s even the fain tes t idea of the exis tence of such sys tems as Alta i and Tianshan , and even Kwen-lun . The ve ry natives of thes e mounta ins can even nowadaysnot b e s aid to have a clear concep tion of the great orographical sys tems . The

authors of the Pu ranas have cons tructed a ve ry regular and symme tri cal orographybordering the Me ru . Humboldt has cons tructed a geometri cal sys temof ranges inCen tra l As ia . How cou ld eve r thes e cons tructions agre e w ith each other ! The

exis tence of ranges was known , b ut how they w e re s i tuated in re lation to eachothe r was unknown . E ven so late as in Humboldt’s days ve ry l i ttle was known of

the orography be tw een Kwen-lun and Himalaya . To sati s fy the demands of the

Purana texts the wes te rn Himalaya and the Ka ilas are su fficient ; the res t aecomplishing the symme try round Me ru is fiction .

The Vayu Purana says there are s ix moun ta ins in Jamb udv ipa, »which lengthento the eas te rn s ide , have exce l lent summi ts

,are all re lated to each other, and have

th e s ea both to the eas t and wes t» . They are named as fol lows , name ly, HemasH imav ant

,H emakt' i ta , Hemav ant , Hairanya (which las t is) pos s es s ed of a

splendou r equal to the rays of the midday sun and N ishadha .

Of all thes e the mount S umeru , which i s made of gold IS the bes t.The Vayu Purana te lls us of four Sarov aras or lakes , »which res ound with the

nois e of large birds ; the i r wate r is clear, propiti ous , healthy The lakes are adorned

Chapter 3 2 , p . 3 2 8 of the manuscript trans lation of the Vayu Purana, which be longs to the

Library of the India Office , and wh ich is the same copy that was prepared for Profes sor Wi lson . Ithas been placed at my d isposal by the courte sy of the Ind ia O ffice . Of i ts existence in London

I rece ived information through the kindne s s of Dr. F . W. Thomas . I amgiv ing here a few extractsfrom this trans lation though they, in some respects , are nearly the same as in the V ishnu Purana.

The manus cript copy of the trans lation of the Matsya Purana whi ch was al so sent to me proved to

b e superfluous , as it d id not contain any geograph ical deta i l s of interest to us .

I O SOUTH -WE STE RN TIBE T IN INDIAN ANTIQUITY.even so late as 1 30 years ago, its upper cours e has been confounded with thos e of

the Indus and the Satlej that we cannot omi t paying some attention to the his toryof human know ledge abou t the origi n of this river. And as cou ld b e expected fromthe immens e importance of this rive r for north-eas tern India, the Ganges has beenan object of i l l imi ted vene ration of the Indians eve r s ince the remotes t antiqu i ty . Itis said to b e mentioned only twice in the s g

Ganga was H imav ant ’s e ldes t daughter, and conce rn ing her the following even ti s re lated both by the Mahabharata and the Ramayana : King Sagara in Ayodhyahad sons . Once while they were s earching for a hors e that had been s tolenfroma sacrifice of hors es , they met the sage Kapila . They accus ed himof havings tolen the miss ing hors e , and in cons equence , in his ange r, he trans formed themintoas hes . Only by the aid of Ganga’s holy wate rs cou ld they again come to li fe . I twas Sagara

s great-great-grandson , Bhagi ratha, who at las t s ucceeded in bringingthe s treamAkaga-Gaga, Vyoma-Gaga, (Ai r—Ganges , Heaven-Ganges ), down fromheaven , and its vi olence was on ly res trained by Civ a rece iving it in the cu rls of his

hai r. This ta le is often refe rred to in the Mahabharata , as : »Her, who is difficu l tto bear even by moun tains , bore afte r this the holde r of Pinaka (Civ a) on his headfor hundred thousand yea rs .» Or : »He re Mahadeva rece ived the down-pow ring

from heaven s ent Gaga, and gave her to the world of men , 0 thou who is fu l l ofBrahma knowledge .

» And : »That holy Bhagi rathi , who is worshipped by gods and

Gandh arvas shines far and wide over the heavens like a pennant banne r by the

wind , i t is pe rpe tually conducted downwards over the lowe r mountain t0ps , lying l ikea s erpent King’s mis tres s , trembling on the rocks , she , the dear que en of the s ea

overflows all the sou thern d is trict (nou rishing it) l ike a mothe r afte r having firs ts treamed fromCambhu ’

s (Civ a’

s) hai r» . And at another place in the Mahabharataw e read of a sacrifice pe rformed »on Himav ant , in that place whe re Gaga burs tsforth fromthe moun ta ins . »

2

According to Wilson the legend of the descent of the Ganga is told , in itsmos t ancient and authen tic shape , in the Ramayana. The Ganges , he says , iscalled Bhagi rathi , from King Bhagi ratha, and the s ea, which was formed by the

rive r, is te rmed S agara, afte r his great-grand s i re 3 WHE E LE R s ays that the legendhas no real value , excepting that i t is un ivers ally be l ieved by the Hindu s . Fromthe Ramayana Whee ler re lates a poe tical des cription of the d ivine Ganges , the rive rwhich Rama saw flow ing in three d irections , »the rive r which dashes agains t i tss trong banks with a terrifi c sound res embling a de ep laugh, or smi l ing w ith its purefoam the river which abounds with the pure lotos , and where the gods perform the ir ablu tions , the rive r which removing every load of impuri ty is i ts e lfclear and pel lucid, the river which fe l l fromthe feet of the divine Vishnu , and

Dowson , op. cit. p . 1 08 .

2 Fausb oll , op. cit. p . 1 60 et seq .

3 Works by H . H . Wi lson , Vol. I I, London 1 862 p . 1 68 .

TH E DIVINE GANGA. I I

from the matted hai r of the great god Civ a , through the influence of the sageBhagi ratha ; the river Ganga wife of Sumudra , pure and des troying s in .»

Again , the Ramayana te lls us how Civa »cas t Ganga off in the direction of

the Vindu lake . And as she was let off, s even s treams branched out fromher.

And the three s treams of the exce l len t Ganga of auspicious wate rs took an eas terlydi rection ; while the Suchakshu , the S i ta, and that mighty rive r the S indhu flowedon the auspicious wes t. And the s eventh fol lowed Bhagi ratha

s car. And that royals ain t, the exceedingly pu is sant Bhagiratha , mounted on a superb car, went before ;and Ganga followed him . 1

2

At anothe r place Ganga is descri bed as the daughte r of Himalaya : »And

having repai red to the north s ide of that mountain (Himalaya), Mahegv ara alongw i th the goddess became engaged in aus te riti es on the peak Himav atprabhav a . Ihave now re lated unto thee , O Rama , the spread of the Moun tain ’

s daughte rAnd furthe r : »When Ravana proceeded by the aerial way and again going up

ten thousand leagues he reached the e ighth aerial region where Ganga, known as

the Ganges of the sky, having s trong cu rren ts and s ending high roars , and uphe ldby air, is s ituated on the s un

s way.» 4

On accoun t of the Gange s ’ descent from heaven this river is also called Devab hi

i ti,»produced in heaven» , and Khapaga, »flow ing from heaven».

The Ganges des cended fromCiv a’

s brow in s even rive rs,the S apta

-s indhav as .

According to Dowson this expres s ion is often met with in the Vedas and was wide lyknown , even by the Romans in Augus tus ’ days , for V i rgil , E ne id , IX . 30, s ays :Ceu s eptem surgens s edatis amnibus altus per tacitumGanges . In Zend theyare called Hapta

-hendu . Dowson re lates the hymn in which the names of the

rivers have been given : »E ach s et of s even (s treams) has followed a threefold cours e .

The S indhu surpas s es the othe r rivers in impe tuos ity Rece ive favorably thismy hymn, O Ganga, Yamuna, S arasvati , Cutudri , Parushni ; hear, O Marud-vridha,wi th the As ikni and Vitas ta, and thou , Arj ikiya, wi th the Su shoma. Un ite firs t inthy cours e with the Trishtama, the Susartu , the Rasa, and the Cv etya; thoumee tes twith the Gomati , and the Krumu with the Kubha and the Mehatm‘

rl» According to

this the s even rive rs are 1 ) Ganga (Gange s) ; 2) Yamuna (Jumna) ; 3) S arasv atr(Sarsuti) ; 4) Cutudri (Satlej) ; 5) Parushni ; 6) Marud-vridha ; 7) Arjikiya (the V ipaga,Hyphas is , Byas) In the Mahabharata the s even rive rs are named in one placeVasokasara, Nalini , Pavani , Ganga, Srra, S indhu , and Jamb i

rnadi ; and in anothe r,

Ganga, Yamuna, Plakshaga, Rathastha, S arayu (Sarju), Gomati , and Gandaki (GanIn the Ramayana and the Puranas the s even s treams are the Nalini , Hladini ,

The History of India from the earl iest ages . By J . Talboys Whee ler, Vol. II,p . 45

3 The R amayana, Vol . I , p . 1 04 .

3 Ibidemp . 9 1 .

4 Ib idem, Vol . VII, Calcutta 1 894 , p . 1 642 .

1 2 SOUTH -WE STE RN TIBE T IN INDIAN ANTIQUITY.and Pavan i going eas t, the Chakshu , S i ta, and S indhu to the wes t, and the Gangesto the south .

The hydrography of the Vishnu Purana begins : »The Catadru , Chandrabhaga,and other rive rs , flow fromthe foot of Himalaya , e tc. Wilson te l ls us thatthe othe r Puranas enumerate many othe r rive rs beyond thos e of the Vishnu. Catadruis »the hundred-channe lled» , the Zaradrus of Ptolemy , the Hes idrus of Pliny

, or Satlej .The Chandrabhaga, Sandab alis , or Aces ines , is the Chinab . Thu s even the S atlejis regarded as coming fromthe foot of the Himalaya , not from beyond the mountains . This is the s ame view as that of Ptolemy who makes all the ri ve rs s tart fromthe southern s ide of the Himalaya .

According to Wilson the popu lar notion concern ing the origin of the Ganges isthat Civ a or Mahadeva rece ives the rive r on his head, though this is refe rred to the

des cent of the Alakananda, or Ganges of India , not to the ce les tial Ganges . In the

Vishnu Purana we read : 3 »From that thi rd region of the atmosphe re , or s eat ofVishnu , proceeds the s treamthat washes away all s in , the rive r Ganga, embrownedwith the unquents of the nymphs of heaven , who have sported in her waters . Havingher source in the na i l of the great toe of Vishnu ’

s le ft foot, Dhruva rece ives herand sus tains her day and n ight devou tly on his head ; and thence the s even Rishispractis e the exercis es of aus te rity in her waters , wreathing the i r braided locks withher waves . The orb of the moon , encompas s ed by her accumu lated cu rren t, de rivesaugmented lus tre from her contact. Fal l ing fromon high , as she is sues fromthe

moon , she al ights on the summit of Meru , and thence flows to the four quarters of

the earth , for its purification . The S i ta, Alakananda, Chakshu , and Bhadra are fou rbranches of b ut one rive r, d ivided according to the regi ons towards which it proceeds .

The branch that is known as the Alakananda was born affectionate ly by Mahadeva ,upon his head , for more than a hundred years , and was the rive r which rais ed to

heaven the s infu l sons of Sagara , by washing the i r as he s . This sacred s tream, heardof, des i red, s een , touched , bathed in , or hymn ed , day by day, s anctifies all be ings ;and thos e who, even at a dis tance of a hundred leagues , exclaim’

Ganga, Ganga’ ,attone for the s ins committed during three previous l ives . The place whence thisri ve r proceeds , for the purification of the thre e worlds , is the third d ivis ion of the

ce les tial region s , the s eat of V ishnu .» To which Wilson adds : »The s i tuation of the

source of the Ganges of heaven iden tifies it with the mi lky way.»

Dowson , op. cit. p . 2 8 1 . In the Vayu Purana the story of Bhagi rathi runs thus : »the riverran into seven courses , three towards the e ast, name ly N alin i , H ladini , and Pavan i, three towards

the west to wit, S i ta, Chakshu , and S indhu» , the seventh , Bhagi rathi , ran to the south . These sevenrivers are s aid to have sprung fromthe V indu lake , s ituated at the foot of a mountain to the northof Kailasa . They flow through the H imav ant, inundate severa l countries , and go to the place , whenceIndra rains , that is to the ocean . S i ta follows the S indhu to the western ocean . Chakshu fal ls intothe ocean . S indhu flows through Darada, Kagmira, Gandhara, etc. The course of the three easternrivers cannot b e fol lowed .

2 Op . cit. p . 1 75.

3 Op . cit. p . 2 2 8 .

THE GANGE S OF HE AVE N. LAKE MANASA . 1 3

In the Ramayana the follow ing piece of Himalayan hydrography is also to b e

found, in connection With the tale about the lord of monkeys , the s ucces s fu l Sugri v a ,who said to a leade r of he rds and lord of apes , named Vinata »there , inmounta in fas tness es , and fores ts and rive rs , do thou s earch Videha ’s daughte r, S i ta,as we l l as the abode of Ravana . And whi le on the s earch around for Rama ’

s

be loved wife , S i ta, thou shou lds t s earch the beautifu l Bhagi rathi , and the Sarayu ,and Kaugiki , the Kal indi , and the charming Yamuna, and the mighty hil l borde ringthereon ; and the S arasvati , and the S indhu , and the Cona with water res emblingruby .»

1

Finally I quote the follow ing pas sage abou t the Manasarov ar, which for us is

the mos t inte res ting in the whole Ramayana : 2 »And Bibhishana obtained for his

wife the righteous (dams e l) named Sarama daughte r un to the high-sou ledCailt

i sha sove re ign of the Gandharvas . (Sarama) was born on the shores of the

lake , Manasa . And while the lake , Manas a, was swollen with wate r on the arrivalof the rains , her mothe r, hearing her cries , affectionate ly said, Saromav ardhata

O lake , do not swe l l ’ , and fromthis ci rcums tance , her name became Sarama.»

As the Ramayana is the oldes t of the Sanscrit epic poems , suppos ed to have beencomposed abou t 500 years B C , and even i f i t rece ived its pres ent formone or two

hundred years late r, the quoted pas sage mu s t indeed b e sa id to b e the firs t placewhe re not on ly the s acred lake has been mentioned b ut a lso the fact of its swe l lingor th e ris e of its s urface on arriva l of the rains . Nay, i t wou ld requ i re some 2 3

cen turies be fore E u ropean geographers and explorers became aware of this mos tcharacteris tic fact of the pe riodical fluctu ations of the lake , which in late r ages gaveris e to s o many mis takes and misunders tandings , and which I will try to explain , sofar as pos s ible , in s ubs equent chapters .

Through S ir JOHN STRACHE Y E . T. ATKI NSON is able to give a paraphras eof a portion of the S éana

’a P a rana ,

known as the .fl/ci na s a-kaana’a .3 »It occurs in

the usual formof a dia logue be tween Sata , a pupi l of Vyasa , and Janamejaya , theson of Parikshit, the Pandava rule r of Has tinapu r, and profes s es to re late what wasforme rly communicated by Vyasa to Vas ishtha . In formand Often . in ve rbiage i tfollows the mode l of the Olde r Puranas and minute ly describes the coun try fromthe

lake Manasarowar in Tibe t to Nanda Devi The work i ts e lf is ve ry popu larand is deeply inte res ting as showing the form in which the actual l iving be lief ofthe people is exhib ited .» 4

Ramayana , op. cit. , Vol . IV, Calcutta 1 89 1 , p . 808 .

2 Op. c it. V ol . VII , p . 1 594 .

3 The H imalayan Districts of the North-Western Provinces of Ind ia. A l lahabad 1 8 8 2 . Vol.

I I , p . 297 et seq .

4 Suppos ing that the Library of the Ind ia Office should possess some trans lation of theManasa-khanda e ither in manu script or print, I wrote to Dr. F . W. Thomas , who, however, communicates me the fol lowing : »For the Manasa-khanda the passage which you cite fromA tkinson appearsto be the on ly authority, e ither E ngl ish or otherwise . I have found on ly one reference to a Manasa

I 4 SOUTH-WE STE RN TIBET IN INDIAN ANTIQUITY.In the in troduction of the Manasa-khanda Janamejaya address es the Suta des i r

ing to hear of the creation of the world and its s tate subs equent to that eventand the mahatmyas Of the grea t ti rthas or places of pilgrimage . We need only toq uote a passage which is of special inte res t to us in connection with the lake . The

Samte l ls the legend of H imachala to Janamejaya . Dattatreya Rishi , one of the

s even human incarnations of Vishnu , afte r his vis i ts to H imachala , went to Kagi(Benares) and proclaimed the glories of H imachala to the Raja Dhanvantari . Theytalk of the ti rthas , and the s age glorifies H imachala » that H imachala whe reCiv a l ived and whe re Ganga falls fromthe foot of Vishnu l ike the s lende r threado f a lotus flower and where the R ishis worship and whe re the Civ a l ingas are

nume rou s . I behold Mana-sarov ara and the re in the form of the raja-hansa(royal goos e) dwe l ls Civa . This lake was formed fromthe mind of Brahma, the refore was i t called ’Manasa-sarov ara ’ . There dwe ll also Mahadeva and the gods ,thence flow the S arayu (here probably Karnal i) and othe r (female) rivers and the

Catadru and o the r (male) rivers . When the earth of Mana-sarov ara touches any

ones body or when anyone bathes there in , he shal l go to the paradis e of Brahma, andhe who drinks its waters shal l go to the heaven Of Civ a and shal l b e re leas ed fromthe

s ins of a hundred births , and even the beas t who bears the name of Mama-s arov aras hal l go to the paradis e of Brahma. Its waters are l ike pearls . There is nomounta inl ike H imachala , for in i t are Kai las and Mana-sarov ara . As the dew is dried up by themorn ing sun , so are the s ins of mankind dried up at the s ight Of H imachala . AtMama-sarov ara , the king, Bhagi ratha , pe rformed the aus te ri ties by which the holyG anga was produced and Vas ishtha Obtained the Sarayu . The coun try around thisholy lake is » cal led Manas a-khanda .» Conce rn ing the creation of Mana-sarov ara we

read : »The sons of Brahma, Mari chi and Vas ishtha and the res t proceeded to the

north of H imachala and performed aus teri tie s on Kai lasa. There they s aw Civ a and

Parvati , and the re they remained for twe lve years absorbed in mortification and

praye r. There was then very little rain and l ittle water, and in the i r d is tres s theywent to Brahma and worshipped him. Then Brahma asked what the i r des ire mightb e . The R ishis answerd and sa id.

’We are engaged in devotion on Kai lasaand mus t a lways go thence to bathe in the Mandakini ; make a place for us to

bathe in ’

. Then Brahma by a menta l e ffort formed the holy lake of Manas a . The

Rishis re turned and rejo icing at the succe s s of the i r jou rney aga in engaged in mort ification and praye r on Kailasa and worshipped the golden l ing which rose fromthe

mids t of the waters of the lake .»

khanda, there exists in India a Ramagi lamahatmya , profes s ing (b ut such profess ions are very untrustworthy) to be long to the Manasa-khanda of the Skanda Purana . But no such Manasa-khandaappears to exist in manuscript or print. Atkinson ’

s vers ion must have been made especial ly for himfrom some unknown manuscript.» FromAtkin son ’

s reference to Sir John Strachey th is seems indeedto have been the case .

Atkinson adds the explanation that the Catadru i s Satlej , »which riscs in the Rakas lake ,wh ich i s itse l f connected with the Mana lake .»

THE SACRE D LAKE , MANA-SAROVARA. 1 5

Then follows the s tory of the Raja Mandhatri who married the earth in the

shape of a woman , who, afte r years of happines s refus ed to die wi th him. »ThenMandhatri was enraged and drew his sword to kil l the earth

,and she fled towards

H imachala and the raja fol lowed her, and she reached Mama-s arov ara . Then on the

banks of this lake the raja cut off her head, b ut the ea rth cou ld not pe ri sh and

van ishing in the wate rs went down to Patala, whe re she worshipped the gods whowere s eated on Kai lasa .

»

To the raja ’

s ques tion : »which is the road to the holy lake ?» Dattatreya gavethe principal s tations and the du ties of the pilgrim, amongs t which , afte r his des cending

. to Mana-s arov ara , was the fol lowing : »let himbath the re and give wate r to the

manes of his ances tors and worship Mahadeva in the name of the raja-hansa . Thenlet himmake the parikrama (ci rcum-ambu lation) of the holy lake Mana and lookon Kai las a and bathe in all the rive rs around.

»

The raja ’

s next ques tion conce rns the return jou rney, and then fol lows,unde r

the heading : M afza-s arov ara : »On the south of the Mana lake i s the mountainCambhu , fromwhich is sues the river which flows to the north into Mana

,

near which are mines of s i lve r and lead : near this the s ands are red and the waterswhite . To the north is the Nala mounta in , whence is su es the river Kapila , whichflows into Mana-s arov ara , while to the south is a cave and a gold mine . Fromthe

Nala mountain , a rive r called Pushpabhadra, flows to the eas t into Mama-s arov ara

and a ls o the Dev abhadra. He re,Ramachandra propitiated Mahadeva

,and fromthis

went to Svarga , leaving his hors es and e lephan ts , which s t i ll rema in the re . Nearth is flows the Chandrabhaga fromthe head of Civ a on Kai lasa . Fromthe mountain Gauri flows the Sarda into the Mana lake . FromKai lasa flows the Mandakini

or Bhadra to the s ou th into the lake . Wes t of this rive r are fiv e lakes , Kal i-b rad ,Kan-hrad, Padma—b rad, Kal i-hrad and Hari-hrad . To the left of Kai lasa is the

Kalapa peak , whe re are many caves and mines of gold and s i lve r ; fromit flowsthe rive r Comanda, of which the water i s the colour of gold ; this flows into the

Mana lake . Near Kalapa is moun t Meru ; this moun tain is blue and from i t fal lsthe Saras vati , and Suv arna-dhara, which a lso flow into Mana-sarov ara . Beyondthes e i s the Mahendra mounta in ,

fromwhich flows the rive r Mahendri into Manas arov ara ; from it also flows the rive r Baruni with ye l low waters into the lake and

the Swati .» I

1 To this Atkinson adds some explan ations in notes . The mountain Cambhu i s Gur-la . The

Nala mountain north of the lake he regards as a peak of the Ka i las range , fromwh ich two streamsflow into the lake near Sarniah-Uniah . The Mandakini or Bhadra s aid to flow fromthe Kai lasa to ,

the south into the lake , Atkinson suggests to b e the Som-chu (Samo-tsangpo) of the Pundits , comingfromthe Ka i las range and going to the lake , which obvious ly i s a mistake , as the Som-chu flows to.

the west and not to the south and as it does not rise fromthe Kai las . The Mandakin i may b e the

Gyuma-chu or any other river coming fromthe north , fromthe mountain s just east of the Kailas .

This enumeration on l y talks of rivers fal ling in to the lake . But in the above quoted passage it issaid that from the Mama-sarov ara flow the Sarayu and the Qatadru. The Sardjou or Gagra on.

T ieffenthaler’

s map flows out of Rakas-tal,but his Satloudj , or Catadru, flows out of Manasarov ar.

1 6 SOUTH -WE STE RN TIBE T IN INDIAN ANTIQUITY.D is regard ing the problematic value of the Manasa-khanda we mus t, at any

rate , confes s that, in the quoted pas sages , it conta ins a good deal of real geography ,which , even if i t has , as usual , be en mixed with mythologi ca l legends and tales ,s ti l l s eems to refe r to au thentic obs e rvation in the sacred region round the Manas arov ar.

The famous piece of poe try of Kal idasa , cal led M eg/za-a

'zzta or Cloud-me s s enge r,

i s too we ll known to b e ente red upon in th is connection . I will only quote two or

three vers es as examples Of its geography . In OUVRY’

s trans lation vers e 1 1 runsas fol low s : »The Rajhansas , hearing thy ear

-pleas ing thunde r , that has the powerto make the earth (in a s tate) cropping up with mushroomparasols , and longingto go to the lake Manas a , wil l become thy compan ions in the sky as far as Kai lasa,as soon as they have made provis ion for the jou rney of the fresh cu ttings of the

Lotus plant .» 2

Ve rs e LXIV in Wilson ’

s trans lat ion runs thusW here bright the mountain’

s cry s tal glories break ,E x p lore the golden lotus -covered lake ;Imb ibe the dews of Manasa, and s preadA friend ly ve il round Airav ata’s head

And vers e LXXV runs as fol lowsThere is the fountain

,emerald s teps denote ,

W here golden buds on s talks of coral float ;And for whose l imp id waves the swans fors ake ,P leas ed at th y s ight, the mount-encircled lake .

3

The latte r ve rs e Ouvry trans lates l iteral ly thus : »And fromthis , a fl ight ofs tai res formed of emerald s labs leads to a large oblong pond , covered with goldenlotus , with s talks glos sy like lapis lazu l i . The Hansas , which make the i r re s idencein its wate rs

,throw as ide regre t, for when they s ee thee

,they think no more of

the adjacent Manasa» . This means , in Ouv ry’

s Opin ion , that the Hansas , s ee ingthe cloud which wil l fil l the i r pond with rain , no longe r think Of taking the i r accu s tomed fl ight to Manasa .

The re is a good deal of othe r geography as we ll as descriptions of ve ry pict uresque s cenery in the Megha-data , as for ins tance the source of the Ganges , and

H e,

or rather his native informants , have been lead by re l igious prejud ice . Atkin son explains themistake of the origina l text, saying that the Catadru in rea lity goes out of Rakas -tal , b ut as th is lakei s in connection with the M anasarov ar, the river can b e sa id to start fromthere . H is pos itive statement of the connection between the Manasarov ar and Rakas-tal, gi ven in 1 884 , must b e foundedupon S trachey

s observations . Whether the channe l b e dry or not the old text is correct in maki ngthe Satlej flow out of the Manasarov ar, b ut the statement about the Sarju i s wrong.

2 The M egha Dt‘ i ta or Cloud M es senger by Kal idasa, trans lated into E ngl ish prose by ColonelH . A . Ouvry . London 1 8 68 .

3 The Megha-data (Cloud-messenger) by Kal idasa. Trans lated fromthe Sanscrit into E ngl ishv erse , by the late H . H . Wi lson . T hird E d ition . London 1 8 67 .

1 8 SOUTH -WE STE RN TIBE T IN INDIAN ANTIQUITY .But i t is not old and real Indian conception to let the Ganges take its origin

from the Rakas -tal. From this lake on ly one river takes i ts ris e , as Las s en s ays,

namely , the Catadru or S atlej , which , l ike the Indus , has i ts origin »on the Tibe tans ide of the Himalaya» .

Las s en regards the S atlej or Catadru , » in hundred branche s runn ing»,as tak

ing i ts origin fromRav anahrada , and the name i s expla ined by a legend,re lat ing

that the river, fromfear Of V as ishtha , divided i ts e lf into a hundred branches . In the

K a/z'

éa P a ran a i s anothe r legend s how ing that the ancien t Indians knew the originof the Brahmapu tra to b e s i tuated in the ne ighbourhood of Ka i las .

2

In our days Brahmakunda is regarded as s i tuated in Uppe r As sam, and

Las s en be l ieves that i t has been placed the re s ince the real origin Of the river wasno more known . As to the e tymology of the name Ka i las Las s en de rives i t fromKilasa , »dwe l ling place Of the peak» . The fu l l name of the holy lake is Manasarov ara , Manasa , the mos t beau tifu l of lakes ; Manasa means »created fromthe

sou l» , 3 and of his s ou l Brahma created the Manasarov ar. The Rakas-tal or Ravanahrada was called Anav atapta , i . e . not hot , thus cold , by the Buddhis ts .

4 Las s en

D ie he il igen Seen l iegen unter 3 0°

4o'

-

3 0°

n . Br. und 99°

0. v . F . auf e iner P lateau -Hohe v on

1 4000—1 5000 F . ; ihnen im Norden Kailasa, imSuden der ste i le Rand des H imalaja ; s ie werden

durch Bache v on be iden Gebirgen genahrt, haben klares Wasser und s ind die He imat der w ildenGans e , we lche hier, die Regenze it des S ii dens fl iehend , in ungestOrter Ruhe brii ten . Aus demN .W .

See , Rav anahrada Oder See des Ravana, auch Lanka genannt, s t IOmt e in Q ue llflus s der Catadruaus demzwe iten , Manasa , le itet die Ind is che Vorste l lung d ie Sarayu , aber mi t Unrecht ; s ie

entspringt jedoch nur e ine Tagesre ise sudwarts davon . D ie Quel lgebiete der Ganga und Yamuna l iegeny on hi er W. auf der innern Ind ischen Se ite des Hochgebirges . D er Indus entspringt auf der Nordse itedes Kai lasa und flies s t N .W . ab , hier Sanpu genannt ; der Brahmaputra , Tibetisch Dzang

-b o, auf demOstgehange de s Geb irges , we lches irn Osten den Manasa begrenzt

I»Durch se ine HOhe bewahrt der H imalaja stets e inen uners chOpflichen Vorrath an Schnee , der

geschmolzen e in so grossartiges Systemv on gros sen Fli i ss en nahri , wie nur wen ige in derWe lt vorkommen .»

2»E s gebar namlich d ie Amogha, die Frau des B ii ssers Cantanu , v on Brahma geschwangert

e inen See , in des sen M itte e in Sohn entstand,wie Brahma v ierkopfig und rOthlich ge l b . Dieser bil

dete s ich in der M itte der vier Berge : Kailas a imNorden , Gandhamadana imSuden , Jarudh i imWesten , Samv artaka imOsten , e inen Te ich (kunda) , in demer wuchs und zu letzt zu e inem5 yojanagros s en See wurde . Paragu-Rama, der zu d ies er Ze it se ine Mutter getodtet, kamzu d iesemBrahmagenannten gros sen kunda, in demer badete und s ich dadurch v on der Stinde des Muttermordes rei

n igte . Rama machte demBrahmasohne mit se iner Axt (paragu) e inen Weg und ~ fuhrte ihn aus demBrahmakunda zur E rde herab ; er fie l dann in den Lohi ta genannten Teich amFus se des Kailasa .

Rama durchgrub das Ufer de s Sees mi t s e inemBe i le und fij hrte ihn nach der Os tli chen Gegend . An

e iner andern Stel le spaltete V ishnu den Berg Hemagti nga und l e itete ihn nach e inemhe i l igen Orteim Innern Kamarfipa

s ; Brahma gab ihmden Namen Lohitagangaka ; Lauh i tja wird er genannt, we i ler aus demLobita-See geboren i s t.»

3 Burnouf trans lates Manas as : l’

organe de l’

intelligence . Le Lotus de la Bonne Loi , Paris1 852 , II , p . 47 7 .

4 In a s zi tra translated by Burnouf Anav atapta i s said to b e the same lake as Ravana hrada,in Pal i Anav atatta , in Chine se A neou tha . In the said s zi tra Anav atapta i s mentioned in connectionwith the miraculous voyage the Sakya has to make to the lake . For it is a law that a Buddha, beforehe enters complete N irvana, mus t accompli sh 1 0 ind ispensable actions ; h is succe ssor cannot arrive :»tant que réun i a s e s aud ideurs aupres du grand lac Anav atapta, i l n ’

a pas de’ve loppe le tissu de ses

actions antérieures Introduction a l’

hi s toire du BuddhismInd ien par E . Burnouf, Tome I , Paris1 844 , p . 1 7 1 ; and p . 3 3 0 and 3 96 . The same legend is told in the T ibetan D u lv a analysed byCsoma de Koros , A s iatic R esearches , XX , p . 65.

THE KAILAS AND THE SURROUNDING LAKE S . 1 9

shows that this northern highland, on accoun t of its sacrednes s , was in fact the on lycountry outs ide of India which was known to the old Indians . The pilgrims u s ed to

go to the holy lake fromBadari on the sou rces of the Ganga; another road , fromKashmi r

,was also us ed, going ove r the moun ta in Cv eta which was a pas s on the

high mounta in Vatikakhanda and cal led the Gate of Manas a . Anothe r gate was

s i tuated north of V ideha .

I

E . T . Atkinson iden tifie s the Bindu-sarov ara of the Pu ranas w i th Rakas -tal .

The Arunoda lake , which is said to lie eas t of Manas arov ar, he sugge s ts as be ingidentical with Cho Konkyu or Gungyut

-cho (Gunchu-tso), which he places >>near

the source of the Brahmapu tra» . The Ci toda lake , which i s w es t of the Mana lake ,he regards , mos t probably as identica l w i th the Cho Moriri

,»the sou rce of the

wes tern S atlaj » . The Mahabhadra lake he identifies w i th one of the lake s of the

table -land . He finds a mingling of facts fromdi fferen t parts , s o that, for ins tance ,Me ru in some cas es clearly indicates the mountains north and wes t of Kashmi r andin others thos e round the s acred lake . In i ts w ides t s ens e Me ru embraces the wholeof wes tern Tibe t be tween the Kai las and the Kara-korum, the Himalaya and

Kwen—lun .

Regarding the great rive rs Atkinson adheres to Wilson ’

s identification of the

Bhadra as the Obi , the S i ta as the Hwang-ho , the Alaknanda as the Ganges and

the Chakshu as the Oxus . But the loca l tradi tions ident ify the Bhadra w i th the

Indus or »Lion rive r»,that is the S ing-chin-kamba of the Tibetans ; the Chakshu

w ith the Brahmapu tra or »Hors e-rive r» , Tamjyak-kamba ; the Alaknanda with the

Catadru or Satlej or »Bu ll-river» , the Lang-chin-kamba ; the fou rth river is the Kar

nali or Mapchu-kamba , the »Peacock rive r» . »Al l thes e rive rs take the i r ris e in the

Mana and Rakhas lakes or in the mounta ins near themknown as Kai lasa by theHindus .

»2

On a wa l l in the ves tibu le of Tugu-gompa l found a Tibetan inscription ,

which , amongs t other things , conta ins the fol lowing words of the ho ly lake : 3 »Tso

mav ang is the holie s t place in the world . In i ts centre dwe l ls a god in human

»Arjuna kamauf se inemZuge v omWesten her imNorden des Himalaja ti ber den Cv etaBerg zumManasa-See . Der Manasa-See war e in sehr he i l ige s ti rtha und das Baden in ihmge

wahrte dem, der es mi t re inemund erleuchtetemGemuthe that, d ie E rkenntni s s der Wahrhe it.» Ind ische Alterthums kunde , Le ipz ig 1 8 67 , I , p . 4 2 , 57 , 8 5, 1 0 1 2

, etc .

2 In a note Atkinson has brought together the names of the four rivers as given by d iff erentauthors . He mentions that on the great map prepared by order of Emperor Chien Lung the fourcorners or gates of the Mana lake are cal led the l ion , e lephant, hors e and ox gates . An d he quotesthe names given by Moorcroft and the Pund its , by Gerard and J . D . Cunn ingham, and regardsS trachey

s names as the most correct. F inally he gives the trans lation of the T ibetan name s , sayingthat lang i s bu ll , not e lephant ; »e lephant in T ibetan i s great bu ll» . N ow, as the name of the Satleji s real ly Lang-chen , where chen mean s great, the river wil l sti l l b e that of the e lephant. The Himalayan Districts of the North-Western Province s of India . A l lahabad 1 8 8 2 . Vol. II , p . 2 8 5, et s eq .

In Vol. III of h is work Atkinson has an index of names fromthe region of the lakes . But nearlyall geographical detai l in it i s taken fromH enry Strachey .

3 »Trans-H imalaya» , Vol . II , p . 1 54 .

20 SOUTH -WE STE RN THBE T IN IN DIAN ANTIQUITY.form, who inhabits a tent compos ed of tu rquois e and all kinds of precious s tones .

In the mids t of i t grows a tree with a thousand branches , and eve ry branch con

tains a thou sand ce lls in which a thou sand lamas l ive . The lake tree has a doublecrown , one ri s ing like a sunshade and shading Kang-rinpoche , the other overshadowing the whole world . E ach of the 1 02 2 branches bears an image of a god, and

all th es e images turn the i r faces towards Gos sul-gompa , and in former times all the

gods gathe red toge ther he re The lake is the cen tra l point of the whole world.

Sarnb u Tashi grew ou t of the lake tree . SochimPema Dabge is of ve ry holy,

clear, and pure wate r . The Gyagar Shilki chorten s tands in the lake . The palaceof the lake god i s in the lake .»

Baron ANTON VON OW regards this document as very importan t and exclaims»H ier, auf der merkwurd igs ten HOhe des E rdkre is es , haben w ir also den S ee le i b haftig

v or uns , der v or mehreren tau s end Jahren s chon als myth ischer S ee Haomas und Schiwas

gepries en wurde , h ier hab en w ir v or uns den my th is chen Pu shkara , Lotus te ich , aus welchemBrahma s ich erheb t, h ier den my th ischen S ee Ch in der Ch ine s en ,

in des sen Mitte das g '

Ottliche

Knab le in auf Lotus ge bettet ruht ! » 2

He then shows how the legend of the lake-god in the centre of the lake isfami l iar all over As ia and extended even to E gypt and Babylon ia . At anothe rplace v on Ow has the fol low ing interes ti ng pas sage : 3

»Das we s tl iche T i b et und das merkw ii rdige Land der he il igen S een imN orden des H imalaya, wo das Que l lgeb iet v on Indu s , Ganges und Brahmaputra nahe zusammens tOs s t , mu s sden Ahnen der arischen Inder woh l bekannt gewes en s e in . Das Land amob eren Ganges e nt

hal t d ie gehe iligtes ten S tatten ind is cher Gotterv erehrung. D ie Verlegung v ie ler GOttertatennach d ies emTe ile de s Geb irges und das Wallfahrten dah in ge hen in e ine s ehr fruhe Z e it zurii ck .»4

Las s en,

be lieved that the Indians now l iving in the plains , attracted by the

wonde rs of the moun ta ins , placed the abodes of the i r gods in the inacces s ible regionsof north-wes tern Himalaya . N ow v on Ow asks why no sacred tZrc/za s we re placedin the central parts of the Himalayas ; even Nepal does not play any importan tpart in the clas s ica l l iterature . As a ru le peoples of the plains regard the high inhospi table moun tains w ith respect or even fear . And the refore : why shou ld the mos tancien t Indian s , only »attracted by the wonde rs of the moun ta ins » , have placed the i rhol ies t tz'rt/za s , not on ly in the highes t abodes of e te rnal s now and ice in the north

I cannot b e re spon s ib le for the correctnes s of this tran s lation wh ich was carried out by myMunsh i , b ut I hope he has not misunderstood the gen eral mean ing of the text.

2 Anthropos , Bd . V , 1 9 1 0, H e ft. 5, 6 , p . 1 065.

3 Hom, der falsche Prophet aus noachitischer Ze it, p . 1 52 .

4 Atk inson pictures the Aryan immigrants arriving at the Ganges and s end ing some adventurousspirits to explore i ts sources . »A fter travers ing the d i fficult pas s es across the snowy range and the in

c lement table-land of T ibet, they d iscovered the group of mountains called K ai las and the lake s fromwhich flowed forth the great rivers to water and give l i fe to the whole earth . The rugged grandeurof the scene , the awful sol itude and the trials and dangers of the way itse l f natura l ly suggested to animagin ati ve and s imple people that they had at length red iscovered the golden land , the true homesof the ir gods whomthey had worshipped when appearing under mi lder forms as stormand fire and

rain in the pla ins below.»

THE SACRED LAKE OF H INDU MYTHOLOGY 2 1

wes tern Himalaya , b ut even beyond and north of thes e regions ? Von Ow comes tothe conclus ion

»Ich hal te es fii r zwe i fe llos , das s d ie Arier (die S tammvater der Hindu s) uber die Pas s edes nordwes tl ichen Hima laya e ingewandert s ind , und das s a

’a/cer i hre genaue Be kannts chaft mi t

d iesen Ge bieten s tammt, d ie s ie des hal b zu GOtters i tzen machten , we il ihre v ergOtterte n S tammvater das e lbs t gelebt haben .»

And further»Pu shkara (Lotuste ich), e in he i l iger gros s er S ee des Pitamaha Oder Brahma, w ird als das

ers te ti rtha genannt ; es s ind dort ’

dre i s trah lende G ip fe l und dre i Wass erab s ttirze’

; uber d ieLage fe h len nahere Anhaltspunkte ; e s he is s t nur, d ies es ti rtha s e i s ehr s chwer z u erre ichen ,be fand s ich a lso woh l in der zentralas iatis chen He imat der Arier. Das jetz ige Pushkara ,

e in

ktins tli cher See in Agmir, i st n icht damit zu verwechs eln . Das w irk l iche Pushkara nun ,

der Urty p fur d ie h e il igen S een , in deren Mitte der Gott w ie e i n BaumOder e in Lotus wachs t ,haben w ir zwe i fe llos imS ee Manasarov ar zu s uchen .»

The »three shin ing peaks » v on Ow iden tifies wi th the Kai las and i ts ne ighbou rs , and the »three wate r-cas cades » w ith the sou rces of the Indus , Brahmapu traand Satlej .

Like Las s en , v on Ow a lso shows that the ancien t Indians very w e l l knew thatthe Brahmapu tra had its sou rce north of the Himalaya near the holy Ka i las . Lategene rations did not know the origin of the rive r . He re lates the same legend as

Las s en abou t the lake born be tween fou r mountains , which , as i ts originator, wascal led Brahma.

»E s is t d ies nattirlich n ichts andere s als der Gott des Sees Manasarov ar, den uralte Uberl ie ferungen als den Quel ls e e des Brahmaputra (’Sohn des Brahma’) be ze ichn en . Hiemit glau b eich e ine neue S t '

utze gewonnen zu hab en fIi r me ine B ehauptung der Identitat Brahmas und

Haomas : be ide thronen inmitten e ines S ees , und zwar dess e lben S ee s . Las s en b etont, das s inden altes ten V edahymnen das Geb iet des oberen Indu s (s e ine Que lle l iegt noch nOrdlicher als

der Kai las a) als H e ima t gepries en w ird ; der spater so ge fe ierte Gangess tromw ird in dense lbennur als untergeordneter Flus s genannt.

CHA PTE R I I .

TH E GREEKS AND ROMANS .

Tu rn ing our attention to geographica l knowledge Of class ica l antiqu ity we are

s urpris ed to find tha t the names of Ind ia and of the rive r Indus have be en knownand w ritten down so early as abou t 520 or 500 B . C . , name ly , in the works of

HE CAT/EUS , of which some s can ty fragments have be en pre s e rved to our own t ime .

But he had ce rtainly not the s l ighte s t know ledge of the countrie s north of India ,nor of the vas t regions eas t of the rive r Indus .

N or does HE RODOTUS s e emeve r to have heard of the rive r Gauge s , and heknew on ly ve ry little of the we s te rn ha lf of the Ind ian pen insu la . He has des cribedabou t one third of As ia , extending fromthe shore s of the Arabian gu l f, the Med ite rrane an and Pontus E uxinus on the we s t , to the Hype rboreans , and India to the

north and eas t , and he regarded India as the las t inhabited country , bordered eas twards by vas t de s erts , un inhab ited and unexplored . Of China , Mongol ia, S ibe ria ,Tibe t and the e as te rn part of the Indian pen insu la he knew nothing.

He rodotus had heard of the great gold production of India, and te lls us thati t was partly dug out of the earth, partly washed down by the s treams . But the

greate s t pa rt came fromthe s andy de s e rt beyond the Pactyan land , whe re a speciesof ant bu rrowed in the s andy so i l , and threw up large heaps at the mou th of its

burrows , and this s and conta ined gold in abundance . The Ind ians,trave rs ing the

de s e rt upon very flee t came ls , fi lled the i r sacks w ith the precious sand , and thenre treated in all has te , pu rsued by the ants . Herodotus s e ems to have located thiss andy de s ert to the north or north-eas t of the Ind ian tribes to which he re fe rs . But

Herodotus doe s not at all mention the exis tence ofmounta ins in this part of As ia .

ME GASTHE N E S , on the othe r hand, who locate s the gold-digging ants among the

D erdae, knows that the latte r are »a people among the mounta ins towards the eas tof Ind ia» , and de scribes the i r country as a high table -land. The s ame people are

2 The Geographical system of H erodotus , examined and explained By JAME S R E N N E LL.London 1 800, p . 1 64 et seq .

H E RODOTUS AND TH E GOLD-DIGGING ANTS . 23

called Dardae by PLINY, »in regione septemtrionaliumIndorum» . The name Dardsis s ti l l in us e .

I

It is cu rious that not an echo fromthe immens e Hima laya has reached He rodotus , although he has heard the s trange s tory of the gold-d igging ants . This s toryhas be en told ove r and ove r aga in and d is cus sed bymany le arnedmen . Di s regarding the li te ratu re deal ing with the old lege nd, I w i l l on ly quote the las t and mos timportant contribu tion to a solu tion , name ly the article by Dr. LAUFE R , somuch themore , as he the re in a lso touche s upon Tibe tan antiqu ity .

2 Like He rodotus the

Ma/zaa/zamca ment ions the ant—gold, b ut fromwhere does this expre s s ion originate ?G re eks , Ind ians , Mongols and Tibe tans , all us e the expres s ion ant-gold. The

nucleus of the problemwou ld the re fore b e to find out fromwhe re the te rmcomes .

In a tale of Ge se r Khan , Laufe r finds this pas sage : »The re is gold in lumps ,which the king of the ants has collected in his activity .» In the officia l his tory Ofthe Tibe tan roya l dynas ty w ritten in 1 3 2 7 mention is tw ice made abou t s and of

ant-gold. . The Indians have borrowed the s tory of the gold-digging ants fromCentra l As ia . The ants are not Tibe tan gold-d igge rs , normarmots , asmany s cholarshave s uppos ed . On ly ants are meant and He rodotus has given the tale correctly .

Amongs t the Mongols the s to ry may have been ve ry O ld, much Olde r than Herodotu s . When SCHI E RN speaks of human gold-d igge rs of the Tibe tan antiqu ity

,

Laufe r reminds us of the fact that the antiqu ity of Tibe t rema ins unknown to us .

He s ays that one cannot speak of a Tibe t at all fromthe t ime of He rodotus and

the M a/zaa/zara ta . We do not even know at all whe the r the highlands of centra lTibe t we re inhabited in thos e early days , and Laufer fee ls very much incl ined to

be l ieve that no inhabitants then exis ted in the coun try . A s tate of Tibe t, as a

pol itica l and nationa l un ity , cannot b e s poken of be fore the beginn ing of the s eventhcen tu ry A . D .

, and of an old cu l tu re within the boundarie s of Tibe t prope r no

thing is known to u s . Be fore that t ime we have to imagine the exis tence of nume rou s tribe s , who may have con tribu ted to the formation Of the late r Tibe tan nat ion ,b ut who we re not aborigines of Tibe t, as they mu s t have come fromwe s te rnChina at a ve ry early pe riod. Laufer te l ls us that the traditions of the his toricalTibe tans , the Indo-Chines e l ingu is tic and the his tory of the migrations of the IndoChinese tribe s , in this respect agre e w ith each othe r, name ly , that the expans ion of

the Tibe tans has taken place from eas t to wes t, and that the pre s ent Tibe t hasbeen the limi t of the s e wande rings . Tibe tan trad i tion has nothing to te l l of

fights w ith aborigina l inhabitan ts ; the re fore we are a llowed to s uppos e tha t thecountry was un inhabited when the firs t Tibe tan tribe s came in fromthe eas t. On lythe Hima layan tribe s we re pres sed sou thwards by the new immigrants , an incident

2 A History of Ancient Geography byE . H . BUN BURY . Sec . E d . V ol. 1 . London 1 88 3 ,p . 2 29.

2 D ie Sage v on den goldgrab enden Ama isen , v on BE RTHOLD LAUFE R . T’

oung pao. Se' rie 1 1 .

Vol . IX . Le ide 1 908 , p . 4 29 et seq .

24 THE GRE EKS AND ROMANS .

which Laufer be l ieve s mus t have taken place not earl ie r than the 4th or 5th

century A . D . The earl ie s t his torical facts we pos s es s abou t Tibe tan tribe s , whichare known to u s from the Chinese annals , have nothing to do with the countrywhich we now ca l l Tibe t.

The re fore such an Old s tory as that abou t the go ld-digging ants cannot b etrans fe rred to »Old Tibe t» . He rodotus heard it in the fifth century B. C.

,when

the e thnologica l s ituat ion was ve ry d iffe ren t from that of our days , and when the recou ld b e no talk at all of human gold-digge rs . And the re fore , when He rodotusspeaks of the gold-d igging an ts he doe s not a llude to Tibe tan nomads making us e

of ante lope horns as ins truments for s cratching the s oi l. 1

Dr . Laufe r poin ts to the pos s ib il ity of the legend originating fromvery far away ,pe rhaps even from A ltai , one of the oldes t g old-producing regions of As ia . The

re s emblance be twe en the two Mongol words sh iraighol (Ye l low R ive r) and shirgholj i

(ant) may, as Laufe r brave ly s ugge s ts , show the way to the right s olution of a

problem, the popu larity of which has been carried through thousands of years bythe great name of He rodotus . I t shou ld b e remembe red , in this connection , thatthe Rev . A . H . FRANCKE heard the same s tory to ld at Kalats e , on ly a few years ago ,

and that some specimens of the »gold-digging ants » we re even shown to himbynati ve s .

2

CTE S IAS OF CNIDUS who l ived 1 7 years at the coas t of Pe rs iamus t have hadample Opportun ity to acqu ire informat ion abou t the geography of we s te rn and s outhern As ia . H e re turned to G reece in 398 B . C . S till his geographical know ledgeof India s e ems to have be en ve ry l imi ted, i f at all in advance of that of Herodo

tus . Probably he knew nothing more of the countrie s eas t of the Indus than hisgreat predeces sor, and he doe s notment ion the Gange s . He enormous ly exagge rate sthe w idth of the Indus .

3

From the days of He rodotus to thos e Of AR ISTOTLE very l ittle progres shad been made in any defin i te know ledge of the geography of As ia .

Aris totle po ints out that a lmos t all great rive rs find the ir s ource in greatmounta in range s , and adds : >>It is thus that in As ia mos t of the rive rs and thos e of

large s t s ize des cend fromthe mounta in which is cal led Parnasos ; and'

this is gene ra l lycons ide red to b e the large s t of all the moun ta ins that are s ituated towards the w inte rsunris e The Indus a lso take s its ris e in the s ame cha in , which has the large s ts tream of all rive rs .» The Parnas os is probably the s ame as the Paropamisus or

Paropanisus of the Gre eks in late r time s .4

Through the Pe rs ian wars the G reeks rece ived the i r firs t knowledge of As iabeyond what Home r had been able to relate . Then came the great geographe rs

2 HOLD ICH : T ibet,The Mysterious . London 1 904 , p . 243 .

2 A History of Western T ibet. London 1 90 7 , p . 1 4 .

3 BUN BURY, 0p. ci t. p . 3 3 8 .

Op . ci t. p . 3 99.

26 THE GRE EKS AND ROMANS.

ARTEMIDORUS des cribed the Ganges as flowing fromthe Emodian ‘mountains to

wards the south til l it reached a city which was also ca lled Ganges .

I

STRABO w rote about 1 9 A. D . and repre s ents the s tate of geographical scienceas i t exis ted afte r the death of Augus tus . He brought toge the r in a mos t cleve rway all the geographica l know ledge of his t ime , and his work is

, except ing that ofPTOLE MY, the mos t important geography that has come down to us fromantiqu ity.

In re spect to As ia as a whole he fol lowed E ratos thenes , b ut in his book XV healmos t entire ly u s ed as his sou rce Megas thenes , as we l l as N EAR CHUS , ONE S ICRITUS and ARI STOBULUS , who had accompan ied A lexande r down the Indus and

had col lected much hearsay information .

2 S trabo s eems not to have had muchconfidence in the compan ions of Alexande r, for he mentions that only a few Macedon ians eve r s aw India , and that thos e who saw i t, only s aw it partly and hurriedlyand got mos t of the ir news fromhea rs ay . And many contradictory reports hadbe en brought back . If this b e the cas e w ith what they saw , says S trabo, what sha l lwe think of the i r hearsay information !

S trabo u s ed no later s ources at all, as reports fromhis own time . The

merchants who in his days s ailed fromE gypt ove r the Arabian s ea to Ind ia , rare lyproce eded so far as to the Ganges , and we re , as a ru le , not s ufficiently inte l l igen tto make geographical researches .

Like E ratos thenes , S trabo has a great mountain range borde ring India to

the north and extending fromwes t to eas t. He has no de ta iled know ledge of itand only gives us the native appe llat ions , Paropamis us , Emodus , Imaus , and othe rs ,while the Macedonians ca l led i t Caucasus . And he correctly points out tha t thisrange s e rved as a boundary of India to the north . While our Hindu-kush , thecentral portion of the Himalayan sys temwas des ignated w i th the name Emodus or

Emodi monte s of Paropamisus , S trabo appl ied the name of Imaus to the eas te rnmos t part, which ended at the coas t of the eas te rn s ea .

3

Thus S trabo had got much more re l iable information about the mounta insnorth of India than E ratos thenes pos s e s s ed . And s t il l even S trabo had a ve rys canty and confus ed unde rs tanding of the orography. The connection of the diffe re n tranges was only part ly known . A great latitudinal cha in of ranges was , howeve r ,imagined to exis t s tre tching all the way fromPromontorium Sacrumin As ia Mino r ,whe re it was ca l led Taurus , and extending far to the eas t through As ia . Th enfollowed the Emodi Mon tes and Imaus . Some have be l ieved that the name Imau s

2 BUN BURY, op. c it. Vol. II , p . 1 1 et seq .

2 BUN BURY , op. ci t. p . 209 e t s eq.

3 Regarding the names of the mountains bordering India to the north he says : »Montes , quiporro ab Ari is excurrunt, Macedones un iversos Caucas i nomine notav erunt ; apud barbaros autemb o

reales [al ia snorumparte appellab antur ParOpami su s] al ia Emod i monte s , al ia Imaus , et plura ejusmodinomina s ingulis parti bus erant ind ita .» (Book XI, Chapter As to the pos ition of Imaus he says»D icunt etiamultimamTauri partem,

quae Imaiumv ocatur Ind icumpelagus attingentem, neque magisquamIndiamaccedere ad ottum,

neque magis recedere (Book XI , Chapter

STRABO . 27

had be en de rived fromMus (-tag),Iwhich is of cours e an anachronism . PLINY

correctly says : »Imaus , incolarum l ingua n iv os ums ign ificante .» Himalaya -means in

S anscrit »abode of s now» . O ften it is contracted to Himala , and some time s the

forms Hemaghiri , Hemachal, Hemakuta are used , mean ing Mount Hema or »the

mounta in of snow». H imav ata and Himavat are adjective s mean ing snowy , winterly.

In vu lgar pronunciation this be comes Imaot , from which the clas s ical Imaus and

Emodus have been de rived .

2

S trabo pays a good dea l of attention to the two famous rivers Indus and

Ganges . Al l the c las s ical authors agree in placing the s ources of the two rive rs onthe sou the rn s ide of the great moun ta in-wa l l north of Ind ia , proving that absolute lynothing was known of the country beyond, the country which is now ca lled Tibe t .That S trabo place s the sou rce of the Indus not far above the Gange s , whe re the

rive r is sue s fromthe mountains , is s e en by the de s cription of Alexande r’s conque s tof the rock Aornus , which He rcu le s had tw ice in va in attempted to conquer, and

whe re i t i s s aid that Aornus was s ituated near the sou rce of the Indus .3

S trabo te l ls u s that in his time the mos t important of the tributaries of the

Indus we re known , as we l l as the coun try round them, and he give s a ve ry goodde s cription of the hydrography, adding that, for the res t, more was unknown thanknown .

4 And afte r re lating the name s of the principa l tributarie s , he goes on to

s ay that they all join the Indus , and that Hypanis is the farthes t . The re are 1 5tributaries worth mention ing. I t wou ld take centurie s upon centu ries be fore E uropegot at all such a correct de s cription of the gene ral hydrography of India as thatgiven by S trabo .

5

Regarding the gene ral s ituat ion of the Ganges , S trabo quotes Artemidorus ,s aying that the rive r comes down from the Emodian mounta ins . Artemidorusmen tions among its t ributaries the Oedane s and knows that it conta ins crocodile sand de lphine s Dr . A . FORBIGE R is convinced that the Oedane s is the samerive r that CURT IUS cal ls Dyardenes , of which he te l ls exactly the s ame s tory, add

ing that this rive r »minus ce leb er auditu es t, qu ia per u l tima Indiae currit» . Forb iger

cannot agre e with those authors who i dentify the Oedanes with the Iomanes , Jumna .

He cons ide rs i t more l ike ly that it shou ld correspond to the Brahmaputra , which2 FORB IGE R , op. ci t. II , p . 51 . STRAHLE N BE RG gives the same identification , 1 73 0. See be low.

2 V IV IE N DE S A IN T-MARTIN : M emoires présentés a l’

Académie . I Série , Tome VI , 1 . Paris1 8 60

,p . 2 63 .

3 Q uumA lexander petramquandamnomine Aornumprimo adortu cepiss et, cuj us ran/ices [na'zcr

7107: pracu l a f ontc s uo a l luz'

t, glorios e dixerunt H erculemter petramadortum, ter inde repulsumfuis s e .

4 Proinde cogn iti nobis sunt fluv ii memorab iles i i , qu i in Indumi llabuntur; loca etiam, perquae i i deferuntur, nov imus ; de reliquis plus e st ignorationis quamcogni tionis .

5 Tota India flumin ibus irrigatur: quorum quaedam in duo maxima irrumpunt, IndumatqueGangem, quaedampropriis osti i s in mare exeunt ; omn ia e Caucaso primo ad merid iemferuntur, posteaa lia eundems ervant cursum, prae sertimquae in Induminfluunt, al ia flectun tur ad orientem

,ut Ganges .

Hie a montib us descendens, quumin plan itiemperv enit, ad orientemconversus

,e t Palib othra, civ itatem

maximam, praeterfluens in mare quod i s tic es t, efi'

unditur uno ostio, quamquamomniumIndicorummaximumflumen s it. Indus duobus osti i s in mare merid ionale exit

28 TH E GRE EKS AN D ROMANS .

could b e regarded as a tributary to the Gange s .

I I f he b e right , this wou ld probably b e the firs t time that the Brahmapu tra appears in class ica l l ite rature .

It was we l l known to S trabo that the native s worshipped the Ganges . Withadmirable clearne s s and pe rspicaci ty he describe s the monsoon ra ins , and the i r influence upon the volume of wate r in the rive rs , and he knows that the precipitation falls in the mounta ins in the form of s now .

POMPONIUS ME LA , who wrote in 43 A . D .

, has a by far more vague and impe rfect know ledge of India than S trabo and de scribes on ly the country s i tuatedbe tween the Indus and Gange s . To hima lso , both rive rs have the ir sou rce s on

the southe rn s ide of Taurus Mons , which, toge the r with ParOpamis us and Emodus ,cons ti tute s the great partit ion wa ll of the who le continent , fromAs ia Minor to the

eas t coas t of As ia . And how cou ld he have been able to add anything new to

the conception of E ratos thene s and S trabo ! What he has to s ay of the sou rce sof the Gange s may b e regarded as gene ral ly correct.2 H is view on the s ource of

the Indu s is the ord inary one of his t ime .3

In his N acu ra/ [J z'

s cory the e lde r PLINY, 2 3—79 A . D . col lected the know

ledge of his epoch . Through the extens ion of the Roman trade Ind ia had becomebe tter known , in spite of which Pliny de s cribes northe rn India entire ly from write rsof A lexande r’s time or that of the ir s ucce s sors . While S trabo qu ite correctlyplaced the source of the Ganges in the Emodi Monte s , Pliny says that the sourcesof the rive r are unknown , l ike thos e of the N ile , or, as othe r authors wou ld haveit, we re s ituated in the Scythian mountains . As compared wi th S trabo his locatingof the sou rce of the Gange s is a s tep in the w rong direction. And s til l he is not

at all to b e blamed , for cartographers of a much late r t ime have made s imi larmis take s .

4 Pliny also quotes a nothe r account of the source of the Ganges , as

breaking out at once in a violent cascade with a loud no is e , and gradual ly laps ingdown into a gentle and placid s tream, a fte r taking its abode in a ce rtain lake . Alake in connection with the Gange s , one a lmos t hope s a mis take of the s ame kindas that of DE SIDE R I and D ’ANVILLE , and tha t the lake shou ld b e the Manasarov ar!

But no, the re is no s ign of our lake , which , howeve r, long be fore Pliny ’s days hadbeen pra is ed in the R amayana and Ma/zaaaara ta . And Pliny does not give us

any indication of the geographica l s i te of his lake .5

2 STRABo’

s E rdbeschre ibung. Fiinftes Bandchen . Stuttgart 1 8 58 , p . 2 1 1 .

2 l lle multis fontib us in Hemode , India monte , conceptus , s imu l unumalv eumfecit, s it omn iummaximus , et alicub i latins , quando angustiss irne flui t, decemmil ia pas suumpatens in septemora di spergitur.

3 Indus exmonte ParOpamiso exortus , et al ia quidemflumina admitti t, sed clari ss ima, Cophen ,Ace s inen, Hydaspen : conceptamque pluribus alv e is undamlato

'

spatio trahit. H inc paene Gangen

magnitud ine exa quat. Lib . III . Cap. VII .4 We should not forget that some years later, or in 1 64 1 , HON DIUS placed the source of

the Ganges at 4S2

/3°

N . lat. , far to the north of T ian-s han , and in 1 73 0 STRAHLE N BE RG has the riverto ris e fromthe southern s ide of the s ame range , fromthe northern s ide of wh ich the Keriya-daryatakes its origin.

5 The passage concerning the source of the Ganges runs as follows : »Hune al i i incerti s fonti busut N i lumrigantemque vicina eodemmodo, al i i in Scyth icis montibus nasc i dixerunt. Influere in cum

POMPON IUS ME LA , PLINY , CURTIUS AND ARR IAN . 29

Pliny’s hydrography of the Indus sys temis not so good as that of the Gre ekgeographe rs .

Regarding the mountain-sys tems he had nearly the same conception as his

predeces sors , making the Imaus , Emodus,ParOpan isu s and Caucas us to l inks in

one continuous range , fromwhich the country fal ls to an immeasurable pla in s imilarto E gypt. 2 And he even knows China , which RUBRUQUIS , PLANO CARPIN I , and

MARCO POLO had to rediscove r twe lve centuries late r. 3Of DIONYS IUS PE RIE GE TE S who l ived in Domitian ’

s time , and his connectionwith Ind ia , BUNBURY says :4 »The e special importance he attache s to the greatIndian promon tory as the extreme eas te rn l imi t of the world is apparently connectedw ith the poe tica l not ion that Bacchus had e rected the re two columns ‘by the

farthe s t shore of the Ocean s tream , on the remote s t mounta ins of India, whe re the

Gange s pours its white wate rs down to the Nys a an

QUINTUS CURTIUS has not augmented the clas s ica l s tore of know ledge aboutIndia .

5

It is of greate r inte re s t to hear wha t ARR IAN , the mos t brill iant of A lexande r’shis torians , and at the same time philosophe r, s ta te sman , mi l itary commande r , has tote l l u s abou t those parts of As ia which are the Object of this work . He was borntowards the end of the firs t century A . D .

H is geography is chiefly based upon .

E ratos thenes and Megas thenes and the mos t trus tworthy his torians among the con

temporaries of Alexander.6

XIX ammis,

ex 1 15 nav igab ilis praeter jamd ictos Condochaten , E rannob oan , Cosoagum, Sonum. A l i icummagno fragore ips ius s tatirn fontis erumpere de iecturnque per s copulosa et abrupta , ub i primummol l is plan ities contingat, in quodamlacu hospitari , inde leh emfluere» e tc . Lib . VI , Cap. XVIII .

2 In Lib . VI, Cap. XX ‘

he says : »Indus , incolis S ind is appe llatus , in ingo Caucas i monti s ;quod v ocatur Paropanisus adversus solis ortum e fl

'

usus , et ips e XIX rec ipit ammis Again hementions the source of the Indus : »Gens ha c (Bactri) optine t aversa montis ParOpan is i exad v ersusfontis Indi , inc luditur flumine Ocho .» And in L ib . VII . Cap. 1 1 he has a word about the source of

the Ganges , though it does not hel p us to locate it.2 L ib . VI . Cap. XVII .3 »Ultra montis Emodos Seras quoque ab ips is adspici notos etiamcommercio (L ib . VI

Cap . XXII .)4 Op . c it. V ol. I I p . 485.

5 Indi a tota ferme spectat Orientem, minus in latitudinem,quamrecta regione spatiosa. Quae

Austrum accipiunt, in altius terrae fastigiumexcedunt ; plana sunt cetera , multisque inclitis amn ib usCaucaso monte ortis plac idum per campos iter pra b ent . Indus gelidior es t, quamceteri . Aquasv ehit a colore maris haud mu ltumabhorrentes . Ganges amni s ab ortu eximius ad merid ianamregion emdecurri t, e t magnorummontiumiuga recto alveo stringit. Inde eumob iecta tupes incl inant adOrientem. Aces ines cum anget. Ganges decursurum in mare intercipit : magnoque motu amn isuterque coll id itur. Qu ippe Ganges asperumos influenti ob iicit ; nec repercus sa aqua cedunt. Dyardenes minus ce leb er audi tu es t, quia per ulti lna Ind ia curri t ; ceterumnon crocodilos modo uti N i

lus , sed etiam delph ines ignotasque ali is gentibu s b e lluas al it. E rymanthus crebris flexib us sub inde

curv atus , ab accoli s rigantibus carpitur. Qu ippe II I flumina tota Ind ia pra ter Gangemmaximamun imento arc is applicant undas . A septemtrione Indus a llui t ; a merid ie Aces ines Hydaspi confund itur. Lib . VIII IX.

6 J . W. M’

CRINDLE : The Invas ion of India by A LE XAN DE R THE GRE AT . Westminster1 896 p . 9.

30 TH E GRE E KS AND ROMANS .

Arrian te l ls us that he has pre fe rre d to note only such things as describe Ind ia as i t really is . In Book V Chap . IV he a lso s ays : »The fol lowing are s tatements about the rive r Indus which are qu ite unqu es t ionable , and the re fore let merecord them. The Indu s is the large s t Of all the rive rs in As ia and E urope , exceptthe Gange s , which is a lso an Indian rive r. I t take s its ris e on this s ide mountParapamisus , or Caucas us , and d is charges its water in to the G reat S ea which l ie snear India in the d irection of the south wind.»

2

I t i s a lways the s ame s tory abou t the s ource of the Indus : Fonte s ejus ex

Parapamiso v e l Caucas o monte oriunturl Or : »Is (E ratos thenes ) a monte Tauro ,ub i Ind i fonte s s unt , s ecus Indumamnemusque ad magnum oceanume t Indi os tiaIndia la tus s tadia XII I M continere a it (Ind ica , Cap . I II).

In Chapte r I II of his I ndi ca ,Arrian s ays : »In the whole of the re s t of As ia

the re are not s o many rive rs as in Ind ia. The larges t are the Gange s and the In

dus , from the latte r of which the country take s its name . Both of thes e are

large r than the E gyptian N i le and the Scythian Is te r, even i f the i r wa te rs cametoge the r into one .» And in Chapte r IV : »Of the two larges t rive rs thems e lve s , theGauge s and the Indus , Megas thenes has s tated that the forme r exce ls much ins ize ; and so s ay all othe r w rite rs who mention i t. He says that it ris e s great fromi ts sources . »name t grandemjaminde ab ips is fontibus oriri .»

Arrian reports A lexande r to have s a id in one of his spe eche s : »But if anyonedes ires to hear what will b e the end to our warfare its e lf, le t himlearn that thed is tance s ti ll remain ing before w e reach the rive r Gange s and the E as te rn S ea i snot great Book V . Chapt. XXVI . To judge fromthis the Macedon ians mus thave be l ieved that they we re not . far fromthe eas te rn end of the As iatic continent.

But the mos t curious piece of hydrography is the fol lowing , which throws a

b right l ight ove r Alexande r as an explore r in great s tyle : 2 »At firs t he thought.he had discove red the origin of the N i le (N i l i s e capu t repe ris s e arb itrab atur),when he s aw crocodile s in the rive r Indus , which he had s een in no othe r rive r exc ept the N ile , as w e l l as beans 3 growing near the banks of the Aces ines of the

same kind as thos e which the E gyptian land produce s . This conjecture was con

firmed when he heard that the Ace s ine s fa l ls into the Indus . He thought the N i leris es s omewhe re or othe r in India , and afte r flowing through an extens ive tract ofde s e rt coun try los e s the name of Indus the re ; b ut afte rwards when it begins to

flow aga in through the inhabited land , it is ca l led N ile both by the Ae thiopians of

that d is trict and by the E gyptians , and fina l ly empties i ts e lf into the Inne r S ea (theMediterranean). In l ike manne r Home r made the river E gypt give its name to the

country of E gypt. Accord ingly when he wrote to O lympias abou t the country of

2 I am fol lowing the vers ion of E . J. CH IN NOCK in h is ARRIAN’

s A nabas is of A LE XANDE Rand Ind ica . London 1 893 .

2 Book VI . Chapter I .3 The fru it of the s acred Lotus of the Hindus , Ne lumbiumspec iosum.

ALE XANDER AND TH E SOURCE OF THE INDUS . 3 1

India, afte r men tion ing othe r things , he sai d that he thought he had discove red the

source s of the N i le , forming his conclus ions abou t things s o great from such sma l land trivial premis s e s . Howeve r, when he had made a more care fu l inqu i ry into ‘

the facts re lat ing to the rive r Indus , he learned the fo llowing de tai ls fromthe na

t iv es : That the Hydaspes un ite s its wate rs with the Ace s ine s , as the latte r doeswith the Indus , and that they both yie ld up the ir names to the Indus ; that the las tnamed rive r has two mouths , through which i t dis charge s its e l f into the G reat Sea ;bu t that it has no connection with the E gyptian country. He then removed fromthe le tte r to his mothe r the part he had w ri tten abou t the N ile . Planning a voyagedown the rive rs as far as the G reat S ea , he orde red ships for this purpos e to b e

prepared for him.»

Now , as he firs t thought he had discove red the source of the N i le , but late r onwas informed that he had to do on ly with the Indus , he mu s t have s tarted uponhis journey down this rive r w ith the convict ion that he had in rea lity dis cove redits source . For, that he was convinced that he was qu ite near the source ap

pears clearly both from Arrian and S trabo, the latte r saying of Aornus , »cujusrad ices Indus non procul a fonte suo allu it». And the re fore the curious case

occurs , that he claims to have discove red the source of a rive r, which in real itywas di scove red on ly years late r ! For us , when trying to make out whatthe ancien t geographe rs rea lly meant by »the source of the Indus » , it wou ldhave been va luable to b e enabled to locate Aornus . As i t is we mus t s uppos ethat they imagined the whole mighty rive r rushing out fromthe ve ry narrow gorgeitse lf, above and beyond which nothing b ut high inacce s s ible mounta ins we re s e en .

This was the view of the whole antiqu ity, and of E uropean cartographe rs 250 yearsago and les s . And it could not pos s ib ly b e othe rwise , for the country in and beyondHimalaya, was absolute ly unknown .

I

Regarding the s ituation of Aornus in re lation to the source of the Indus , VIVIE N DE SAINT-MARTIN is no doubt right in expres s ing the following view :

On a v u que dans S trab on Aornos est s itué a la source de l’Indus ; il _

e s t c laire que ,dans les anciennes in formations recue ill ie s par les Grecs , ces gorge s étroites ou l

Indus s’

en

gage apres avo ir traversé la contrée montagneus e de Balt i , et d’

ou i l débouche avec un bru itformidab le auprés de Derbe nd , et d e Torb ila , avaient été représ entées comme l’origine memedu fieuv e . C

e s t a ins i qu’

Arrien a pu d ire »I’

Indus est déja un grand fleuv e quand il s ort impétueu sement de ses sources . Tous les auteurs de l’an tiqu i te’ grecque e t lat ine, sans préc i ser la local ité comme le fait Strab on dans cc passage , font, en effe t, naitre l

Indus d irectement au nord , au p ied du Paropan i sus ou Caucase Ind ien .

2

FORBIGE R : Handbuch II , p . 63 .

3 E tude sur la. Géographie Grecque et Latine de l ’Inde , et en particulier sur l’Inde de PTOLEMEE ” dans ses rapports avec la Géographie Sans crite . M émoires pre’ sentes par d ivers savants al’

Académie des Inscriptions et Be lles -Lettres de l’

Ins ti tut Imperial de France . I Série , Tome V, 2 .

Paris 1 858 , p . 43 .

3 2 TH E GRE E KS AND ROMANS .

In expre s s ing hi s view V ivien de Saint-Martin has been influenced by MajorAbbott . Many othe r authors have pa id great atten tion to the Aorne s problema nd s t i ll it has proved impos s ible to de termine the s i tuation of that rock . G eneralCUNN INGH AM ident ifie s it with the isolated rock Ranigat , 1 6 mi le s north of the

tow n of Ohind on the Indus : » I do not ins is t upon the identification , b ut if weadmit that the accounts of the his torians are ve ry much exagge rated , I think thatthe ru ins of Ran igat ta l ly much be tte r w ith the vague de scriptions of Aornos thathave come down to u s , than any othe r pos i tion w i th which I amaccquainted.»

I

BUNBURY cannot accept this identificat ion , which he finds l iable to the insupe rableobjection of be ing so far fromthe Indus as not to agre e at all with the de s cript ions of Cu rtius , D iodorus and S trabo , who all agree tha t Aornus ros e immediate lyabove the Indu s , so that the rive r actual ly washed its bas e . Arrian places it nearto a town cal led Emb olima , in the uppe r va l ley of the Indus , »a pos ition whichexpla ins S trabo ’

s expre s s ion that it was n ea r t/ze s ources of fil e I ndus , a te rmbywhich he undoubtedly meant to de s ignate the place where that rive r firs t is suesfromthe gorge s of the Himalaya . »

2

In h is Book V , Chapte r V , Arrian has the fol lowing important de l ineation of

the mounta in range s of Cen tra l As ia : »Mount Taurus d ivide s As ia , beginn ing fromMycale , the mountain which l ies Oppos i te the is land of Samos ; then cu tting throughthe country of the Pamphylians and C i l icians , it extends into Armen ia . Fromthiscountry it s tre tche s into Media and through the land of the Parthians and Chorasmians . In Bactria i t un ites with mount Parapamisus , which the Macedon ians whos e rved in Alexande r ’s army called Caucasus , in orde r, as i t i s s a id, to enhancethe ir king ’s g lory ; as s e rting that he w ent even beyond the Caucasus with his v ic

torious arms . Pe rhaps i t is a fact that this mounta in range is a continua t ion of the

o the r Caucasu s in Scythia , as the Taurus is of the same . For this reason I haveon a previous occas ion ca l led this range Caucasus , and by the same name I sha l lcontinue to ca l l i t in the fu ture . This Caucasus extends as far as the G reat S eawhich l ie s in the direction of Ind ia and the eas t. Of the rive rs in As ia worth cons ide ration which take the i r ris e from the Taurus and Caucasus , some have the ircou rs e tu rned towards the north , d is charging thems e lve s e ithe r into the lake Maeot is ,

or into the s ea called Hyrcan ian , which in rea l ity is a gu lf of the Great S ea . O the rsflow towards the s ou th , name ly , the E uphrate s , Tigris , Indus , Hydaspe s , Aces ines ,

Hydraotes , Hyphas is , and all those that lie be twe en the s e and the rive r Gange s .

A l l the s e e ithe r d is charge the ir wate r into the s ea , or disappear by pouring thems e lve s out into marshe s , as the rive r E uphrate s doe s .»

In the s e words Arriamaffords u s a ve ry graphic de scription of the greatcontinental mountain sys tem,

running fromwe s t to e as t and be ing a water-parting

The Anc ient Geography of India. London 1 87 1 , p . 76.

BUN BURY , op. c it. V ol. I . p . 496 .

CHA PTE R I I I.

PTOLEMY.

The greate s t geographe r amongs t the Gre eks and Romans of antiqu ity wasPTOLEMY , and it shou ld require age s be fore he , in a mode rn pe riod of his tory ,was surpass ed. The dis cove rie s made during Alexande r ’s campa ign we re a re

ve lation to the w e s t ; at one blow the e as t had been opened up to the G re ek world .

Through the campaigns of S E LE UCUS N ICATOR more re l iable informat ion was won

abou t the countrie s on the Ganges . The journeys and the trade of the Pto lemaeansin E gypt to Ind ia , and the foundat ion of Gre ek kingdoms in Bactria and Cappadocia promoted a be tte r knowledge of the e as t w ith the we s te rn world.

E ratos thenes a lready had a fa irly good idea of the great ou tl ine s of As ia ;he knew ParOpamisus and Imau s , and even the S e res . That know ledge was en

larged by Rome , more e specia l ly by the Syrian and Parthian wars , and by a moredeve loped and w ide spread trade on land and by s ea .

I S trabo s upe rs edes E ratosthenes , b ut is hims e lf in a much highe r degre e supe rs eded by Ptolemy, who isthe firs t to s peak of India extra Gangem, and has col lected a much greate ramount of information about the mounta ins in Centra l As ia than his Greek and Ro

man predece s sors had been able to do .

Ptolemy wrote abou t 1 50 and 1 60 of our era . A ve ry cons ide rable part ofthe know ledge which has made his name so famous , he borrowed fromMAR INUS ofTyre . The backbone in the orographical ske le ton of As ia which had been foundedby E ratos thenes and accepted by S trabo was s trongly confirmed as a fundamentalfact by Ptolemy , in whos e geographical sys temi t formed a partit ion wa l l be twe enthe plains of Scythia on the north and the coun tries of Ariana and Ind ia on the

s outh. Regard ing the name s of the greatmountain sys tem , S trabo had the fol lowingorde r : Paropamisus , Emodi Monte s Imaus . In accordance with him Pliny applied

FORB IGE R even goes s o far as to say : »So batten denn die A lten zur Ze it der hochsten Blii theRoms v on demgrb

ss ten The ile A s iens mit Ausnahme de s nordlichen und nordos tli chen wen igstens e ineob erflachliche , v on den westl ichen und sudwes tlichen The i len aber, d . h . v on den romi schen Provinzen ,

e ine sehr genaue K enntniss , ja the ilwe ise e ine noch umfas sendere und v olls tand igere , als wie in unserenTagen .» Handbuch der alten Geograph ic . Band II . Le ipzig 1 844, p . 4 2 .

PTOLEMY AND MAR INUS OF TYRE . 3 5

the name Imaus to the eas te rn end of the sys tem. In Ptolemy ’s ge ography the

name Imaus be longs to the central part fromwhich the Gange s take s its ris e . H is

Emodus , a part of eas te rn Himalaya , begins fromthe sou rce s of the Ganges and

s tre tche s eastwards . But Ptolemy us ed the name Imaus not on ly for the part ofthe principa l sys tems tre tching fromwe s t to eas t, whe re the Gange s originate s ; hegave i t a lso to the tremendous me ridional branch which goe s out , at a lmos t rightangles , fromthe latitud inal cha in , and s tre tche s far to the north , d ividing the wholenorthe rn half of As ia into Scythia intra ImaumMontemand Scythia extra ImaumMontem. Some 60 or 70 years ago this me rid ional Imaus was s uppos ed to includethe . unfortunate Bolor-tag .

Both we s t and eas t of his me rid ional Imaus , Ptolemy has s eve ra l othe r range s ,the names of which he may have rece ived at s econd or third hand through s ilktrade rs fromthe country of the S e re s , S e rica , or which , perhaps , had be en col lectedalready by Marinus . V ivien de Saint-Martin shows a ce rta in re s emblance be twe enPtolemy and the Puranas ; in both cas es s even principal mounta in range s are ment ioned, and some Puranas , jus t as Pto lemy , enume rate the rive rs which take the i rris e fromeach mounta in-group .

I

The re is qu i te a hos t of s cholars who have wasted the i r brains and the ir ene rgie s in the hope le s s and va in attempts to identify Ptolemy ’s geography with our

pre s ent de ta iled know ledge , and as a ru le eve ry new commentator pu ts forward hisown new sys tem, d iffe rent fromthos e of his predece s sors . E ven s cholars who haveatta ined the highes t s tandard of know ledge , as LASSE N and RICHTHOFE N , haveyie lded to this temptat ion . And, inde ed , i t wou ld b e both inte res ting and of gre atvalue to know how far the information acqu i red by s uch men as Marinus and

Ptolemy reached towards the interior and the E as t of As ia . And s till , S ir HE NRYYULE is right in pointing out the impos s ib il ity of s uch a task, of which a lso Bunbury s ays

Z: »The attempt of Las s en to identi fy the various place s mentioned by

Ptolemy is bas ed throughout upon the fundamen ta l e rror of s uppos ing that thatgeographe r pos ses sed a map of India s imi lar to our own , and that we have onlyto compare the ancient and mode rn name s in orde r to connect the tw

o .» 3

If, in the fol lowing pages , I appear to have committed the same e rror which Iamblaming in othe rs , it is not so much any des i re of mine to try and offe r a new

explanation of a specia l que s t ion , as an attempt to show the enormous importanceof Ptolemy for the cartography of la te r t ime s , and, in one specia l cas e , to provethat the identification brought forward by some mode rn geographe rs cannot pos s iblyb e correct. The problemI amre fe rring to is the Ts angpo

-Brahmaputra . But firs ta few words shou ld b e sa id about the Indus and the Gange s as given by Ptolemy .

Etude sur la Geographic etc. Mémoires p . 4 .

Op . c it. V ol. I I , p . 642 .

3 The last attempt in this d irection has been made by Colone l G . E . GE RIN I in his great workResearches on Ptolemy ’s Geography of E astern A s ia, London 1 909.

36 PTOLEMY .Accord ing to Ptolemy the sou rce s of the Indu s are s ituated in the country

of the Daradra , which is mountainous and highe r than all othe r countrie s . Boththe Indus and the Gange s and the i r tribu tarie s are repre sented by h imas comingfrom the sou thern s ide of ParOpamisus and Imaus . In this re spect Ptolemy hasnot proce eded any farthe r than S trabo and all the re s t of h is predece s sors .

The Satlej is one of the three rive rs which w i ll e spe cial ly occupy our attentionin this work . I t had e scaped the knowledge of Alexander’s his torians as the

G re eks did not reach so far. But Ptolemy has it and locates its s ou rce as the

eas te rnmos t of the tribu tarie s of the Indus . With the sou rce of the Zadadros or

Catadru , w e have , howeve r, only to unde rs tand that part of the rive r which trave rs e sthe mounta inous country immedia te ly above jalandhara .

I I t wou ld requ ire a verylong time be fore E u ropean geographe rs advanced beyond Ptolemy regard ing the

s ituat ion of the s ource of the Satlej . E ven on a map published by Major RE NNE LLso late as in 1 800, the source of the rive r has the same s ituation in re lation to

the mountains as on Ptolemy ’s map .

2

On the accompanying Pl . I is Ptolemy ’s repre s entation of the Indus , and on

PI. I I that of the Gange s . They are facs imi le s fromthe Codex Con s fan tz'

n opo/z'

tan u s

the maps of which , in 1 90 1 , we re photographed by Dr. F . MARTIN for Baron A . E .

NORDE N SKIOLD , and a comple te s e t of photographs is kept in the Roya l Libra ryat S tockholm. The gene ra l ou tl ine s of the Indus and the Gange s on this map are

nearly the same as on othe r Pto lemaean maps,for ins tance the ed ition P tolommzs

R om/oE as t of the Ganges (Pl . II I) Ptolemy has a mounta in range cal led Bepyrrus .

Gene ra lly his ideas of the E as tern Hima laya are confus ed , for this Bepyrrus is no

doubt me ant to b e a part of the E as te rn Himalaya . He a lso show s some of the

rive rs in this region as coming down fromd iffe rent range s , though he has not beenable to place the range s in a correct re lation to e ach othe r . FromBepyrrus two

name le s s Gange tic tributaries flow we s t wards . Saint-Mart in sugges ts that they mayb e the Kaougiki and the Gandaki or the Tis ta. Farthe r eas t we find a rangeca l led Damas i , giving ris e to the rive r Daona , which has a lso a w es tern source inthe Bepyrrus .

4

SA IN T-MARTIN,loc . ci t. Deuxi éme Memoire , p . 3 7 2 . SA IN T-MARTIN adds the fol lowing

reflexion : car la source re’ e l le de ce grand afii uent est tres -loin de la dans l’

in téri eur du T ibet, a pende d is tance des s ources du S indh .

2 The Geographical systemof H E RODOTUS p . 2 29.

3 V ide P l . III, which shows a part of it in reproduction . Themaps of the Codex have not beenpublished before , b ut they have , as Dr. CURT F ISCHE R of Dresden tells me , very l ittle scientific value .

Regard ing the manuscript to which the maps be long Dr. F ischer writes : »D ie 2 Kartenb latter stammenaus e iner Constan tinopoli tanischen Hand schrift (Nr. 2 7 der griechisch en Handschri ften der S erailb ib liothek.)D ie H andschrift gehort demXV. Jahrh . an

, ist auf Papier geschrieben und zahlt 1 1 2 B latter Asi llustration s of the s ituation of the sources of the two rivers in relation to the mountains north of themthe maps w il l a lways b e of a certa in interest to us .

4 Loc . ci t. Trois ieme M emoire : Le bass in du Gange . P . 1 85. V ivien de Sa int-Martin expres ses the fo l lowing opin ion : >Il suffit d’

un coup d ’oai l sur la carte pour voir que , d ’

apres cette

SOUTHE RN TIBET, VOL. 1 .

The Indus of Ptolemy . Code x Cons tantinopoli tanus , 1 sth

centu ry .

GE N . STAB . LIT.AN$T. Srocmouvr

SOUTHE RN TIBE T, VOL . 1 .

-5 59 3 A

N i ' v l l

PLO II .

3 8 PTOLE MY.1 2 5

°E . Long . and 3 7

° N . Lat . , farthe r north than the sou rce s of its tributarie s , exceptthe Koas , and 6

°

above the confluence w ith the latte r. Las s en the re fore be l ieve sthat he does not mean the re a l , uppe r Indus b ut the Shayok , a view that cannotpos s ibly b e correct. For even of Kashmi r, Kaspe ira , he had a rathe r vague idea ,as Dr. STE IN has shown .

We now come to the mos t inte res ting que s tion , name ly , about the Tsangpo .

Las s en regards i t as very l ike ly that the greate s t geographe r of antiqu ity s hou ldhave known the Brahmaputra , and he pos it ive ly affirms that Pto lemy calls thisrive r Baut isu s .

2 The fe ede rs of Baut isus come fromthe Emodus , the Kas ian and

the O ttorokorrhas mounta ins . Las s en identifie s the Kas ian mounta ins w ith the

range s wes t of Kashgar , the Emodus with the ranges be t ween Buthan and Tibe t,

and O ttorokorrhas w i th the Tibe tan Lo-kaha—ptra . I t wou ld take us too far to

fol low Las s en ’

s argumen t ; b e i t s ufficient to s ay that he finds an argument in the

ve ry name as we l l , for the name Bautis us is de rived fromBhota , the Ind ian nameof the Tibe tans ; the re fore Bau tisu s mus t b e identical w ith the uppe r Brahmapu traor Tsangpo .

Amongs t the very great numbe r of au thors who have tried to identify Ptolemy

s geography w ith our pre s ent knowledge of Centra l As ia , I wil l quote R ichthofen. He finds i t e asy to expla in why Pto lemy place s the origin of the Bau tisus

on the Kas ian mounta ins .

3H e agree s with Las s en in identifying the Bautisu s with

the Ye ru-ts angpo or uppe r Brahmaputra , and he finds it natural that the southe rnbranch shou ld come from Emodus .

Both the s ituat ion and the name of the rive r are s triking facts , and at firs ts ight one fee ls tempted to agre e with Las s en and R ichthofen . For my own

part I am, neve rthe le s s , far from pe rsuaded . On Plate I I I we find two sourcebranches which , afte r the ir junction ,

flow eas twards and empty thems e lves in a lakew ithou t outle t . Near the lake is a town , S e ra me tropol is , and sou th of it , qu i teclos e to the lake , is a range of mounta ins , O ttorocoras Mons , which is , it is true ,the immediate cont inuation of the S e rici Mon tes , Emodi i Montes , and Imaus Mons .

This orographica l arrangement inde ed s e ems to indicate the Tsangpo. But it i s morethan doubtfu l whe the r Pto lemy w ith his long range of mountains mean t one s ingles ys tem, which , compared w ith our actua l knowledge , shou ld repre s e nt the Himalayasys tem only. I be l ieve his long range includes the whole mounta inous land northof India all the way up to Kwen-lun . The country s ou th of the range is Ind ia , thecountry north of i t is the de s e rts and s teppe s of Centra l As ia .

M emoir on Maps illustrating the anc ient Geography of Kasmir. Ca lcutta 1 899, p . 9.

2 Ind is che Alterthumskunde I . p . 1 3 1 . »Demersten Geographen des klass ischen Alterthumeswar hoch st wahrsche inl ich , urn n icht zu sagen gewis s , die Kunde v on demBrahmaputra , demgros s tenZ uflusse des Ganges , zugekommen, der d ies en in der Lange se ine s Laufes fib ertrifft, und wie der IndusimNorden des gros sen nordlichen Granzgeb irges entspringt . Ibidem

,III, p . 1 3 1 .

3 China, I p . 48 7 .

TH E R IVE RS BAUTISUS AND OE CHARDE S . 39

Mos t of the late r s ucce s s ors of Ptolemy have not be en able to ident ify the

Baut isus w ith the Tsangpo , for this rive r rema ined unknown ti l l comparat ive ly late ly.

MARTE LLUS GERMANUS , for ins tance , give s the Bau tis is Flu . thre e s ource s , fromChas i i Montes , Se rice Monte s and Octocoras Mons ; they join and flow w ithou t thes l ighte s t hes itation s traight to the G lacial Ocean . On P efrn s Apz

'

a n n s , 1 530 , the

rive r a lso tu rns north. On the P lolemz n s E o’z

'

tz'

o B . Sy l v a n z’

, Ven i ce , 1 51 1 , the

Bautis us amn is goe s eas t and empt ie s i ts e lf into a lake not so very far fromthe

eas te rn coas t of As ia .

1 On the P tol . A rg en t i n e:

1 51 3 , and on G . Merca tor 1 53 8

the rive r aga in goes north to the Polar S ea .

2 The unce rtainty abou t the u ltimatefate of the rive r shou ld not caus e s urpris e , for it exis ted in E u rope and India evenin the time of some geographe rs s til l l iving ; though, of cours e , not to such an ex

tent as 400 years ago .

But how cou ld i t eve r b e expla ined that the S e ra me tropol is shou ld havebeen suppos ed to b e s ituated exactly at the po int whe re the rive r begins to pie rcethe mountains ? If I amright in be l ieving the Emodii Monte s e tc. to repre s ent thewhole Tibe tan highland be twe en Himalaya and Kwen-lun , and i f, as Las sen b eliev e s , the Cas n monte s are thos e of Kashgar, and the north-we s tern branch Of theBau tisus come s fromthem, I shou ld take this branch to b e the Yarkand or Kashgardarya , and the s ou th-we s te rn branch , on which the place Orosana reminds us of

Boras ana near Khotan , the Khotan-darya . Thu s the Baut isus shou ld b e the Tarim,

and the lake in which it come s to an end, the Lop-nor.

But he re we mee t a d ifficu lty . For jus t north of the Baut isus is anothe r rive r,Oecharde s Flu .

, which is exactly l ike the Bautisus , beginn ing w ith two s ou rce branchesfrommounta ins in the north and the south of it, and, afte r the ir junct ion , flowing eas twards to d isappear in a lake . This rive r, Oecharde s , has be en identified with the

Tarim,3 though othe r Opin ions have a lso been expre s s ed . RE NNE LL recogn is e s in the

Oechardm»the O igurs or Yugures of the pre sent time s » .4 FORB IGE R be l ieve s that

the branch of the Oechardes which come s fromthe Auxaci i M .

—A lta i , is identica lwith the S e lenga , whi le the othe r branches shou ld b e some s teppe -rive rs ; MANNE RTidenti fies the wes te rn branch w ith E tz iné ; RE ICH AR D ident ifies the Oe charde s w iththe Tarim fall ing out in the Lop

-nor.5 As to the Bautisus , Forb iger be l ieve s i t is

meant to b e the Hwang-ho . Sa int-Marti n finds it not improbable that the Oechardae represent the te rritory of Skardo in Baltis tan and if this b e right the Oecharde s fluv ius shou ld b e the part of the S indh which trave rs e s the great te rri tory

Facs imi le-Atlas ti ll kartografiens aldsta historia innehallande afb ildningar af de v iktigaste kartor

tryckta fore fi r r6oo . Af A . E . Nordenskiold Stockholm1 8 89, PI. XXXII I .2 Ibidem, P l . XXXV and XLIII.3 R ICHTHOE E N : China, I , Tafe l 8 ; ALBE RT H E RRMAN N : D ie alten Seidenstrass en zwis chen Ch ina

und Syrien , map ; and PE TE RMAN N ’

s M itte ilungen , 1 9 1 1 .

4 R E N N E LL’

s Herodotus , p . 209.

5 FORB IGE R , op. c it. p . 59.

40 PTOLEMY .di rect north of Kashmi r. The l is t of diffe rent opin ions cou ld eas ily b e mu l tipl ied,b ut the task wou ld b e us e les s . On his mos t inte re s ting and ins tructive map

z nr Z ei t o’er a lfen H a nde/soez z

'

e/znng en zwz’

s c/zen C/zz’

na nn a’a’en [ra

n z

'

s c/c a a n z

s c/zen [j i n n/ern » Dr. A . HE RRMAN N iden tifies the S e ra me tropol is w ithLiang—chow and the Cas i i Monte s with the northe rn borde r range s of the Kwen-lun .

Comparing the Cas i i Mon tes w ith the Oecharde s , and accepting this rive r, beyonddoubt , as the Tarim, the Cas n Monte s mus t of cours e b e the Kwen-lun . R ichthofena ls o identifie s the Cas i i Monte s w i th the Kwen-lun , though only the we s te rn part.In fact the Cas i i Monte s are the same as the Chine se Ts ung- l ing or »the On ionmounta ins » . The re fore the space which , on Ptolemy ’s map , is s i tuated be twe en the

Cas ii Mon te s and Emodu s and its eas tern continuation , shou ld be long to the Tibe tanhighlands . And as the re is on ly one rive r, the Bautis us , flow ing fromw e s t to eas t ,on ly the Tsangpo cou ld b e mean t. And s til l I be l ieve that the Bautis us is meantto b e the Tarim, or, in othe r words , that this rive r has been repre s en ted twice0 11 Ptolemy ’s map. Ptolemy has got his informat ion fromme rchants , who , thems e lve s , had been informed by native s and trave l le rs . It wou ld not b e surpris ing ifPtolemy had misunde rs tood his d iffe rent d irect or indi re ct informants and be l ievedthat what he heard abou t the Tarim, fromtwo d iffe rent informants , in fact re ferredto two d iffe ren t rive rs .

3

The re fore , where R ichthofen finds a s triking corre spondence be twe en the

Auxakian , Kas ian and Emodus mounta ins with the Tian—shan, Kwen-lun and Himalaya , this corre spondence is chime rical . The northe rn s lope s of the Cas i i Monte sare the same as the northern s lope s of the Emodus ; the re fore the Tarimrive r appears twice . In real ity the re is no s ign Of Tibe t on Ptolemy ’s map , much le s s of

any T ib e tan’

riv er. The Kwen-lun and the Himalaya , inclus ive , of cours e , of all there s t of Tibe t w ith Arka-tag , Kara-korum and Transhimalaya , or in one word the

Deuxieme Memoire , p . 3 75.

2 D EGUIGN E S is not qu ite clear in h is identification . H e speaks of K hotan-darya wh ich losesitse l f in the de sert,

and of two other rivers , obviously the Yarkand-darya and Aksu-darya. O f thelatter two he says : Le s deux autres vont plus loin, e t apres s’

étre réunies , e lles s e jettent dans nu

grand lac, appe lle Lop , qu i est s itué dans la partie la plus bas s e de tout ce grand terra in . Les an

ciens Chino is penso ient que le s deux fleuv es dont je viens de parler é toient le meme que le Hoamho Cette derniere idée mis e ‘

a part, P toléme

e paroi t avo ir aus s i confondu le H oam-ho avec ces

d eux fleuv es sous le nom d’

Oechardes . Further on he s ays that Bautis us takes i ts origin in the

Montes Cass ii or the Tsung-l ing of the Chinese , s ituated S .W. of Kashgar. Dans les memes montagnesTcung

- l ing vers Yerken i l sort un grand fieuv e qu i v a se rendre dans ce lu i qui part d ’

Aksou,l’

unl ’autre s e jettent dans le lac de Lop . Les Chinois appellent le plus merid ional Cheou-pa-ho, i l do itétre le Bautisus . H istoire générale des Huns , Tome I , s econde partie , p . V and XXXIX . Paris1 756 . In th is s econd case he identifies Bautis us with Yarkand-darya .

3 I wrote to Dr. A . HE RRMAN N to hear h is opin ion in the matter and he answered , amongstother things : »Der Bautisus kann n icht der Tsang-

po se in . Auf der Karte des Marinus i s t, wie ichb e i der R ekonstruktion des selben fand , j ener Name noch n icht entha lten ; erst PTOLE MAUS hat ihn e in

gefuhrt, Ohne clas s er s ich auf bestimmte Nachrichten s ti i tzen konnte ; er hat in demBauti sus nur die

marin ische Darstellung de s Oechardes Tarimund s e ine Forts etzung der Hwang-ho) schematischnachgeahmt. E s ware daher fals ch

, in demBauti s us e inen tibetan ischen Fluss suchen zu wo llen ; er

existierte nur auf der so fehlerhaften Karte des PTOLE MAUS .»

TH E TSANGPO WAS UNKNOWN TO PTOLEMY . 4 1

whole Tibe tan highland be twe en Kwen-lun and Hima laya has been pres s e d toge the r,and forms on ly one s ingle mounta in wall runn ing eas t and we s t. This wal l s eparate s the pla ins of India fromthe plains of Centra l As ia .

1

The re fore it is not surpris ing that Pto lemy , at so e arly a date , made the

s ame mis take fromthe unre l iable and unce rta in information he cou ld gathe r. Ind iawas comparative ly we l l known . I t was bounded on the north by a high range of

mountains . The Tarimbas in was , on account of its trade roads , a lso known to a

ce rta in extent . I t was bounded on the sou th by a high range ofmounta ins . The rewe re no trade roads through Tibe t, and Tibe t i ts e lf rema ined comple te ly unknown .

The refore the northe rn l pes s e en by s ome informan ts , and the sou the rn s lope ss ee n by othe rs , we re be l ieved to be long to one and the same range . To Ptolemy,Tibe t did not exis t at all and Bautisus cou ld not b e the Tsangpo .

I F ifteen hundred years later we fi nd exactly the same representation on many E uropeanmaps of A s ia . On GASTALDI

S map the Gange s takes i ts origin fromthe same range which formsthe southern boundary of Diserto de Camul. The French trave l ler BE RN IE R has on hi s map of 1 699only one s ingle range north of India, fromthe southern S lopes of which the Indus and the Ganges

take the ir ris e , just as in PTOLEMY’

s days , and immed iate ly north of the same range i s Zagathay, Tar

tarie and Turquestan , that is to s ay no S ign of Tibet. In 1 73 0 STRAHLE N BE RG , re lying upon otherauthors , begins to open up some space for T ibet, b ut on ly in the east. Three years later D ’AN V ILLEpubl ished his map , which was indeed a reve lation to the geographical world . The s ecrets which , evers ince E RATOSTHE N E S ’ days had been kept within the narrow wal ls of the Imaus and Emodus were at

once d iscovered , the mounta in ranges opened up and separated and T ibet occupied its due s ituationnorth of the H imalaya. And amongst the new labyri nth of mountains , lakes and rivers in T ibet theTs angpo was

, perhaps , the most prominent feature .

6—1 3 1 3 8 7 I

CHA PTE R I V.

TH E ARAB IAN GEO GRAPH ERS .

Geography ente red upon a flou rishing era at the time of the highes t deve lopment of powe r in the Empire of the Ca l iphs , and from that time a whole s e rie s ofgreat Arabian s cholars wrote the ir de scriptions of the d iffe rent countrie s of the earth .

Ptolemy’s as tronomy was trans lated unde r the ti tle of A lmages t and became a

kind of fundamenta l codex to the Arabian geographe rs . Great parts of the i r worksare devoted to As ia , and even the kingdom of Tibe t did not e scape the i r attention ,though they had on ly a ve ry vague concept ion of that country. The Arabians whohave s o much to te l l u s abou t Ind ia , Iran and Tu ran , and even some thing abou tChina , know Tibe t only fromhears ay , and have often rathe r fantas tic ideas abou tthe mounta inous country

,its s i tuation , and its characte r.

The Arabians reckoned the Tibe tans amongs t the Turks , as they did with allmore or les s unknown nations and tribe s in the eas t and north of As ia .

R ichthofen has shown that much of the informat ion which the Arabian write rsattributed to Tibe t , in real ity re fe rred to Khotan ; for even E dris i ’s source s be longto a time , when the name of this place was not known , and, as the neare s t greatcity of trade , it a lso repres ented the kingdom of Tibe t to the conception of the

Arabs .

2

In the fol lowing page s I have tried to col lect some extracts from the moreimportant Arabian write rs , in which wil l b e found how far they knew ourmountainsnorth of India , the kingdomof Tibe t, and the sou rce s of the great rive rs .

2 Or as RE INAUD puts it : »Chez les e'

criv ains de l'

antiqu ité , les peuples du nord de l ’As ie e t

d e l’

E urope éta ient des Scythes ; chez les anc iens écriva ins arabes , cc sont des Turks ; chez le s écriva inspostérieurs , cc sont des Tatars . De leur cOté , le s Chinois se sont fa it umpla is ir de donner a ces po

pulations des noms que lquefois bizarres , souv ant arbitraires .» R E IN AUD could eas ily have added thateven so late a s in his own days many E uropean trave l lers who approached the frontier of T ibet fromthe Ind ian s ide

,called the T ibetan s Tartars . Re lation des Voyage s faits par le s Arabes et les Persans

dans l’

lnde et a la Chine dans le IXes iecle “

de l’

ére chretienne par M . R E IN AUD . Tome I .Paris 1 845, p . CXLI .

2 China , I , p . 566.

SULE IMAN TH E ME RCHANT . 43

Abbé RE NAUDOT is the firs t s cholar who made E urope acqua inted w ith the

works of the Arabs upon India and the countrie s adjacen t. And curious ly enough ,his trans lation deals w i th the earl ie s t of the Arab geographe rs abou t Ind ia .

1 In his

remarks Renaudot te l ls u s that me rchants fromMe sopotamia and Pe rs ia u s ed to

trave l overland v ia Couz is tan to Tibe t or China , and that in his own days the greattrade with Tibe tan musk took place through the kingdomof Bu thah , »which mus tb e a part of ancient Tibe t , or the country unde r the Khacan of Tibe t». Regard ingthe s ituation of Tibe t in re lat ion to China he says of his two au thors : »Ils remarquent aus s i que le Royaume de Tibe t , Tobit, ou Tobat , comme prononcent le s Arabes ,n’

en e s t pas fort e s lo igné : 81

que le pai‘s des Tagazgaz le borne du cos te del’

Orient .» 2

In 1 845 RE INAUD published his edition of the same work which is e spe cial lyremarkable for its exce llent introductory commentary . He showed that the work didnot conta in the narratives of two trave l lers as Renaudot had be l ieved, but only of

one , the me rchant SULE IMAN , whos e account was compiled in 851 A . D .,jus t at the

time when the comme rcial commun ications be tween the Empire of the Cal iphs of

Bagdad and India and China we re at the i r highe s t point of act ivity. The s econd

author, ABU S A ID , neve r vis ited Ind ia and China . All that he had to re late hadbe en reported to himby othe rs .

That part of India w i th which the A rabs had the leas t commun icat ion was

Hindus tan prope r, or the country trave rsed by the j umna and Gange s , that is to

s ay, from Panjab to the gulf of Bengal . 3 There fore the. Arabs had only a vague

idea of Assamand the Brahmaputra . The principal object of Su le iman ’

s narrativewas to make known the road which was trave l led by the me rchants of Bagdad ,Bas ra and S iraf on the ir way to China , a road which a lso had been taken byS u le iman .

The me rchant Su le iman ment ions Tibe t in the follow ing pas sagesE n deca de la Ch ine sont le pays des Tagazgaz , 4 peup le de race turke , e t le Khakan

du T i bet. Vo i la ce qu i termine la Ch ine da cOté du pay s cle s Turks . La Ch ine , du c6té du

s ole i l couchant, a pour l imite la v ille appe lée Madou ,sur les frontieres du T i b et. La Ch ine e t le T i be t

sont dans nu état d ’ hos til ités continue l les . Q uelqu’

un de ceu x qu i ont fa it le vo yage de Ch inenous a d it y avo ir v u nu homme qui porta it sur s on dos dumus c dans une outre ; cet homme étaitpart i de Samarkand , et ava it franch i a p ied la d is tance qui s épare s on pay s de la Ch ine . I l

éta it venu de v i lle e n v ille jusqu ’

a Khanfou ,5place on se d irigent les marchands de Syraf.

Le pay s v it la che vre qu i fournit le mu s c de Ch ine , et le Ti b et, ne forment qu ’

une s eu le e t

2 Anciennes Re lations des Indes et de la Chine , de deux Voyageurs Mahometans , qu i y allerent,dans le neuvieme s iecle ; traduites d ’

arabe : avec des Remarques sur les principaux endroits de ces Re

lations . Paris 1 7 1 8 .

2 Op . cit. p. 1 79, 2 2 2 and 252 .

3 R E IN AUD , Op . c it. p . XLVIII.4 The country of the Turkish Hwei-hu ; R ICHTHOFE N op. ci t. , p . 565.

5 R E IN AUD identifies Khanfou with Hang-tcheo-fou, and Madou with Amdou or Amdo, a namehe knows fromde lla Penna .

44 TH E ARABIAN GE OGRAPH E RS .

meme contrée . Les Ch inois att irent a eux le s chevres qu i v i vent pres de leur territo ire ; i l enes t de meme des hab itants du T i bet. La s up ériorité du mus c du T i bet sur ce lu i de la Ch inet ient a deux cau s es 1 1 y a an T i be t des hommes qu i font me’ tier d ’

al ler a la recherche dumus e , et qu i pos sedent, a ce t égard , de s conna is s ances particulieres .

x

The re is much more of musk than of Tibe t in Su le iman ’

s narrative , b u t themusk was more important for the me rchants .

IBN KHORDADHBE H , who d ied in 9 1 2 A . D . only ment ions Tibe t in the

fol low ing pas sage s 2

Le T i b et, les contrées li ab i tées par le s Turcs , la Chine e t al Mangoura c’

e s t-a-d ire tousles pays s itués a hu it degrés au de la du centre de l’orient

,ont leur Kibla (l orientat ion dans

la priere) tres proche de la p ierre no ire .

Speaking of the titles of the kings of the world , he s aysLe s rois des Tu rcs , de s T i béta ins et de s Khazare s portent tous le t itre de Khakan , a

l'

except ion du roi des Kharlokh (tri bu turque) qu ’

on appe l le Djab ghouya .

As to China he knows that it is bounded by the s ea , Tibe t, the country of

the Turks , and, to the we s t , by Ind ia .

Unde r the title : S ource a nd w on t/cs of r i v ers , he write s»Le Djaihoun (O xus ), fleuv e de Balkh , s ort des montagnes du T i be t Le Mihran

,

fleuv e du S ind , sort des montagnes de Sch ikinan , et c’

e s t proprement une branche du Dja ihoun . Une part ie du royaume de l’Inde porte le nomde cc fleuv e (S ind). Apres avo ir formép lus ieurs des ri v iere s de l’Inde , i l pas se par al-Mangoura e t s e jette dans l’Océan oriental .Une pers onne d igne de confiance qu i a fréquenté les pay s lo inta ins m’

a appris que le Dja ihoun es t une des deu x branches d ’

un fle uv e qu i sort des montagnes de la Ch ine et au -de ladela Ch ine et qu i coule s ur des p ierres énormes et des rochers de sorte qu ’ il es t impos s i b le d ’

y

nav iguer e t meme de la travers er, s i l’

on n'

y es t pas accoutumé . Une des deu x branches s e

d irige vers le ‘S ind , l’autre e s t le Djaihoun . A tro is journ e' es de marche e t demie a partir dela separat ion des deu x branches , i l y a s ur le Djaihoun umgu é condu isant au pay s des Turcsnommés Sch ikinan . Les marchands partent avec leurs marchand is e s de la v ille de Khottalanpour un ri bat (re la is ) dont j ’a i oub l ié le nom

,a 1 parasange de d is tance , 011 s

élev e,s ur le s

bords de ce grand fleuv e , une montagne que nu l ne peut franch ir sans l’a ide des ind igene s ,qu i y s ont accoutumés . Le s marchands ayant fa it pri x avec ceu x -ci pour le trans port desmarchand is es , i ls grav is sent la montagne , chaque homme charge’ d ’

un fardeau de trente l ivres(mann) . Le s entier es t s i étro it, qu ’

on a jus te as s e z d ’

e space pour mettre le p ied . Arri vés ausomme t i ls élév ent les s ignau x con venus e ntre eux et les Schikinan pour le ur annoncer queles marchands s ont arrive’ s . Pu is i ls des cendent, s u i v is des marchands , vers la riv iere . A lav ue des s ignau x , les Sch ik inan travers ent le fleuv e avec de s chameau x accoutumés au trajet , e tmun is d ’

une es corte . Arrivés a l’autre rive , i ls font un contrat forme l avec le s marchands pourle trans port de leurs marchand is es e t bagages , e t ayant chargé les chameaux , i ls repass ent lofle uv e e t condu is ent les marchands s ur la route qu ’

i ls dés irent prendre , les uns vers la Ch ine ,le s autres vers Moultan . A dro ite du pays de s Toghozghor ve rs le s ud d emeurent lesT i béta ins .

2 R E IN AUD , Op. c i t. p . 60 e t 1 1 4 .

2 Kitab almasali k wa’

lmamalik , auctore IBN KHORDADHBE H etc . M . J . DE GOBJE . Lugdun i

Batav orum,1 8 89, p . 3 , 1 2

, 49 , 1 3 5, 1 3 9 , and 2 04 . IBN KHORDADHBE H’

S »Book of Roads and K ingdoms» has also been trans lated by BARB I E R DE M E YN ARD and publ ished in Journal A s iatique 1 8 65.

46 THE ARABIAN GE OGRAPH E RS .

tion beyond Ptolemy. He is not s atis fied to place the source on the sou the rns lope s of the mounta ins north of Panjab ; he looks beyond the cre s ts and locate s i tin the v e ry heart of the mounta in knot , at the same place whe re the Amu-daryatakes its ris e .

Ib n Khordadhb eh’

s Tibe t , in s pite of his own ignorance abou t its rea l s ituat ion ,mus t no doubt co incide fa irly we l l w i th the country be twe en Panj , Khotan , and

Panjab . E ven nowadays Ladak is ca lled Tebbe t by the Turks of E as te rn Tu rkes tan .

E DR ISI who w rote more than 200 years afte r Ib n Khordadhb eh confounds Tibe tw i th Khotan .

The informat ion our au thor rece ived froma »re l iable pe rson» , about the difficulties in ford ing the rive r which s ent off the Oxu s and the Indus , as we l l as whathe was told of the mounta in on the bank of the Oxus which cou ld not b e cros s edw i thou t gu ides , and which may have been any of the high pas s e s at the Panj , thede s cription of came ls us ed for transport ove r the rive r, and fina lly what he re late sabout the caravan roads branching off to China and to India , all this prove s that heis speaking of the country be twe en Pami r and we s te rn Tibe t , and that a very graphicand trus tworthy account of a me rchant trave lle r has been his s ource .

Ibn Khordadhb eh does not s eemto have heard of the Ganges . The Satlejand the Brahmaputra he ignored comple te ly.

AL-JA‘

KUBI who wrote abou t 8 80 has not much to say of Tibe t. So , for

ins tance , that ware s come to Bagdad fromHind , S ind , China , Tibe t , and the province s of the Turks . He compare s the air of Bagdad w i th that of Tibe t, which iss o b ad as to change the skin of the face of the Tibe tans , and keep the i r bod ie ssmal l and make the i r ha ir curly . The city of Ds irmis , fromthe eas t, the las t citywhich borde rs upon the province of Tibe t . He has a lso some thing to re late aboutan embas sy from Tibe t asking for a teache r of Is lam. The »Bis amof Tibe t» isment ione d , and the nord is a gras s that grows in India and a lso in Tibe t.

IBN ROSTE H w rote be fore 9 1 3 , and borrowed some information d irectly fromIbn Khordadhb eh. He mentions Tibe t and Kabu l amongs t those place s whichduring the s umme r ge t no ra in b ut in winte r are snowed ove r on account of thegre at co ld of the a ir. To hima lso the Oxus has its source in Tibe t. In the uppe rpart Of the prov inces E l-Khottal, along the cours e of the rive r Vakhkhab , whichcomes from Tibe t and forms the source of the Dse ihun . Thes e are places whe rego ld i s dug out , and from which gold is exported in piece s not bigge r than pinheads . The fourth cl imate begins in the eas t, and s tre tches a long the provinceof Tibe t .

ABU DOLE F began in 94 1 his journey eas twards from Bokhara . He s aysthat he reached the regions inhab ited by Tibe tan tribes , where he trave l led for 40days withou t d ifficult ies . The Tibe tans l ive on wheat , barley , beans , diffe rent sorts

MASUD I’

S M EADOWS OF GOLD . 47

of meat, milk , grape s and apple s . They have a gre at town bu i lt of reeds , w ith a

temple . Mohammedans , Jew s , Chris tians , Magicians and Indians l ive in Tibe t . Theypay tribu te to the prince s of Bagdad.

MASUD I is one of the greate s t Arabian trave l le rs in As ia and one of the mos tadmi red au thors in the Arabian language . H is work is a compendiumof the Arabianknow ledge of the world in his time , and of As ia he has a much fu l le r and cleare ridea than Ibn Khordadhb eh , though his imaginat ion in some case s takes him a

l ittle too far. In 9 1 2 he vi s ited Mu ltan and Mansurah ; in 9 1 5, afte r mee t ing ABUSAID in Bas ra , he trave l led through Fars and Ke rman , en te red India and probablyvis i ted Ceylon in 9 1 6 , fromwhe re he sai led to Madagascar and Oman . He s eemsto have navigated along the As iatic sou th coas t to China . With the Caspian and

Red S ea he was famil iar. H e d ied at Fos tat , old Ca iro, in 956 . E ven for our t imehe wou ld have been ca lled a trave lle r of a ce rta in fame . But he l ived and worked a

thou sand years ago , and his achievements are the re fore worthy of our admirationand as ton ishment . The follow ing are some extracts fromhis great work : M nrnj n

-Z'

Z a/zao or JlI eadozos of Gold .

2

Chapter IX has the t itle : »R ens e ignements généraux s ur les migrations des .

mers , e t sur les principaux fleuv e s » , and the re an in te re s ting pas sage runs thusL

origine des fleuv e s e t de s sources a s ou levé des d is cus s ions . S e lon les uns , i ls prov iennent tous de la grande mer, c’

es t-a-d ire de la mer d ’

eau douce , qu ’ i l ne faut pas confondreavec l

Océan . D’

autres prétendent que l ’eau s e trouve dans la terre,comme les ve ine s dans

le corps On a cherch é depu is longtemp s la s ource , l’emb ouchure e t l’

étendue du parcoursdes grands fleuv es , te l que le N il

,l’

E uphrate , le T igre , le fleuv e de Balkh ou Dje ihoun , loMe hran , qui arros e le S ind ; le Gauge , fleuv e important de l’Inde , E l-Djahez pretend

que le Mehran (Indus), fleuv e du S ind , provient du N il, e t donne comme preuve l’exis tence de s .

crocod iles dans le Mehran . J’

ignore i l a été ch e rcher un pare il argument. I] a avancécette thes e dans son l ivre des Grandes m'

lles et des merv ez'l l es de [a terre . C’

es t un e xce llenttrava il ; mais l

auteur, n’

ayant pas navigué , u i as s e z voyagé pour connai tre le s royaumes e t

les c ités , ignora it que le M e hran da S ind sort de sources b ien connues , s ituées dans la hau teregion du S ind , le territo ire de Kanoudj , le royaume de Baourah , les pay s de Kachmir, deKandahar e t de Tafen , e t qu ’ i l e ntre ensu ite dans le Moultan , i l regoit le nomde M ehran

d’

or, Le M ehran , apres avo ir trav erse le pay s d ’

e l-Mansourah , s e jette dans la mer del’

Inde 3

These quotations , fromJA’

KUB I to And DOLE F , are taken fromKUUN GEza : I smerete inkT ib etrol, in E loadasok Koros i Csoma Sandor Emlékezetére . I Szam. Budapest, 1 900, p . 57—68 .

The whole article of KUUN was trans lated for me by one of the pupils of my friend Profes sor E UGE NVON CHOLN OKY at Kolozsvar, a kindness for whi ch I b eg to expres s my deep gratitude .

2 MAgOUD I . Les Prairies d ’or. Texte et traduction par C. BARB IE R DE M E YNARD et PAVE T

D E COURTE ILLE . Paris 1 86 1 . Another ed ition bears the title : E l-Mas’

ud i’

s historical E ncyclopoedia, eutitled »Meadows of Gold and M ines of Gems » . Trans lated fromthe Arabic by ALov s SPRE NGE R , .

M . D . London 1 84 1 . I amquoting the French ed ition .

3 Op . ci t. Tome I p . 2 03 et seq .

48 TH E ARABIAN GE OGRAPHE RS .

He re Masudi g ives us the inte res t ing news that in his days , as during class ical an tiqu i ty and Ale xander’s time , geographe rs we re s earching for the sou rceso f the mos t famou s rive rs , amongs t themthe Indus and the Gange s . In a somewhat new formw e aga in me e t the e x traordina ry theory of the connection be twe enthe N ile and the Indu s . Bu t while Alexande r, according to Arriam, s uppos edthe Indu s we re the s ou rce of the N i le , E L-DJAH E Z sugges ts that the N ile is thesource of the Indus .

I In both case s the crocod iles are re spons ible for this mos tcurious mis take . Masud i knows the absurdity of the theory , b ut is hims e lf ve ryvague in placing the sou rce of the Indu s , in spite of his as s e rting the s ituation of

the sources to b e w e l l known . H e locate s themOve r a very cons ide rable are a ,

fromKandahar in the wes t, which re fe rs to the Kabu l rive r , indeed a tribu ta ry of

the Indus , to the te rri tory of Kanauj in the eas t , which is absurd , as Kanaujfal ls w i thin the dra inage a rea of the Gange s . If he had satisfied hims e lf w i th the

a s s e rtion that the s ources we re s i tuated in the high region of S ind , he wou ld havebeen neare r the tru th . Masudi doe s not point ou t one principa l source b ut thinkso f the s ou rce of eve ry s eparate tributary in the Panjab ; in Kashmi r for ins tance hehas one source , re fe rring to the Jeblum, and in this case he is right.

But what doe s he mean by the »great fre shwate r s ea that shou ld not b e con

founded w ith the ocean» ? As he specia lly po ints out this d is t inct ion he can onlymean a lake in the inte rior of the continent. The Caspian and Lake Ara l are a

prz'

orz

excluded, for Masudi knew themand they conta in sa lt wate r. During hisjourneys in India he may e as i ly have heard that the Indus and the Gange s and twoo the r rive rs came froma great fresh-wate r lake in the north, and with the usua le xagge ration of his time he has changed the information to embrace a ll rive rs . I tis d ifficu l t to s ee what e ls e his fre sh-wate r s ea cou ld b e than our Old Manasarov ar.

Unfortunate ly he doe s not mention the name , and the re fore , as in so many othe rcase s , we are confined to conjecture .

2

Furthe r on he aga in touche s the que s tion of the source of the Oxus and its

connection with the IndusLa v ille de Balkh pos sede un pos te (ri bat) nommé e l-Akhcheban , et

,

s itu é a v ingt joursd e marche env iron . E n face v i vent deu x tri bus de Turcs infidele s , les Oukhan et les T i bétainse t a leur dro ite d

autre s Turcs nommés Igan . C’

e s t dans le territoire de ceu x -Ci qu ’

e s t las ource d ’

un grand fle uv e nommé aus s i fleuv e d ’

Igan . P lus ieurs personnes ins tru ites pre nnent ccfleuv e pour le commencement du Dje ihoun, ou fleuv e de Balkh . Le Dje ihoun a un parcoursde cent c inquante parasanges , s e lon les uns , e t de quatre ce nt parasanges s e lon ceu x qu i leconfondent avec le fleuv e des Turcs ou Igan . Quant aux auteurs qu i avancent que le Dje ihouns e jette dans le Mehran (Indu s), i ls sont dans l’erreur.

2 In SPRE N GE R’

S trans lation th is pas sage run s thus : »E l-Jahit suppos es that the river M ihran ine s -S ind is the N ile , al leging as a proof that crocodiles l ive in it.» Op . cit. Vol . I , p . 2 3 3 .

2 In the following century ALBE RUN I , quoting the Matsya and Vaya-Purana , said of anotherlake : »In the mountain Ka i lasa there i s the pond Manda, a s large a s a sea , whence comes the riverM andakin i .»

MASUDI’

S GE OGRAPH Y . 49

Mas ud i thus places the Tibe tans much furthe r wes t than Ib n Khordadhb eh .

The Oxus is repre sented as confounded w ith the rive r Igan , whe the r this b e the

Yarkand-darya or any othe r rive r in that region . He po ints out indirectly that Ib nKhordadhb eh was w rong in combin ing the Oxus and the Indus .

In the fol low ing pas sage he again take s up the que s tion of the Indus and the

Panjab , and giv e s a rathe r good des cription of the mountains round Kashmir :Le roi de Kandahar, l ’un des ro is du S ind et de s es montagne s , porte toujours e t

généralement le nomde Hahadj ; c’

es t de son te rrito ire que coule le Ra‘

i'

d,l’

un des c inq fleuv e sdont la reunion forme le Me hran Umtrois ieni e fleuv e sur les c inq sort de la montagneappe l ée B ehati l

,dan s le S ind , et travers e le territo ire des Rahpout ou le Kandahar. Le qua

trieme fleuv e pre nd s on origine dans les montagnes de Kabou l Le cinqu ieme prend na i ss ance dans lo Kachmir Kachmir fait au s s i part ie du S ind ; c’e s t un pays montagneu x , formant un grand royaume , qu i me renferme pas mo ins de s oixante r ou s o i xante e t dixmi l le v i llese t v il lages . I l es t inacce s s i b le , e xce pté d ’

un cOté , e t l’

on n’

y p e ut pén étrer que par une s eu leporte . E n e ffe t, i l e s t renferme entre des montagnes e scarp ées e t inabordab les , que personnene s aura it grav ir, pu is que le s betes fau ves meme 11

en atte ignent po int le sommet, e t que le s

O is e au x s eu ls peuvent y parven ir . La 01) les montagne s ce s s ent, i l y a des vallée s imprat icable s , d ’

épais s e s forets , des jungle s et de s fleu v es dont le cours impétueu x e st in franch is s ab le .

Ce que nous d isons ic i de l’ impos s ib i lité de grav ir ce s remparts naturels du Kachmir e s t connude tout le monde dans le Khoracan e t a i l leurs , ce . qu i fait d e cc royaume une de s merve ille sde la terre .

He re is , at any rate , an attempt to locate the origin of the d iffe ren t rivers ofthe Panjab . The on ly gate to Kashmi r is Obvious ly the Jehlumand Baramu la .

He doe s not s e emto have heard of the roads fromKashmir to Ladak and E as te rnTurke s tan , as we l l as in othe r d irections ove r the mounta ins , as he make s Kashmirs urrounded by impract icable mounta ins on all s ide s . He kn ow s the Gange s and its

re l igi ous importance b ut is not qu i te sure of its s ituationLe Gange e s t umfleuv e de l’Inde qu i s ort des montagnes s ituées dans la partie la p lus

recu l ée de l’

Inde , du cOte’

de la Ch ine , e t pres du pay s hab ité par la p eup lade turque desTagazgaz . Apres un parcours de quatre cents parasanges , i l s e jette dans lamerAby s s in iennes ur la cote d e l ’Inde .

2

Fu rthe r on he te l ls u s that some people locate the origin of the »He rment»in the mounta ins of S ind and India , while o the rs be l ieve that it come s fromthe

s ame place as the Ganges , a rive r which runs at the s ide of the greates t part ofthe S ind mounta ins .

C’

es t un cours d ’

eau rap ide et impétueu x ; s es bords sont rougis du sang de beaucoupd e dévots ind iens qu i se mu t i lent avec le fer e t s

engloutis s ent dans s es flots , pous sés par leuré loignement pour ce monde et le dés ir de le qu itter. Ce s fanatiqu es remontent le Gange e t

arr ivent a un endro it 0111 s e trouvent de hautes montagnes e t de s arbres s éculaires s ou s lesque lsd e s hommes sont as s is .

3

2 Op . c i t. p . 3 7 2 .

2 Op . cit. Tome I p . 2 1 4 .

3 Op . c it. Tome II p . 80 e t seq .

7—1 3 1 3 8 7 I

50 TH E ARABIAN GE OGRAPHE RS .

In the s e words Masud i te lls u s of the pilgrimage s to the s acred temple of

Gangotri near the s ource of the Bhagirathi Ganga .

Speaking Of the des cendants of Amu r, of whomone fraction emigrated to

the frontie rs of India , he continues 2

Une autre portion encore alla s e fi xe r dans le Th ib et et s e donna un roi qu i éta it soumis a l

au torité du Khakan ; mais de pu is que la suprématie de ce souvera in a ces s é, le sh ab itants du T h ibe t donnent a leu r che f le titre de Khakan , en mémo ire de s anciens ro is turcs ,

qu i portaient le t itre de Khakan des Khakans .

He knows that in China the re are rive rs as large as the Tigris and the E u

phrate s , and that they have the i r source s in the countries of the Tu rks , in Tibe tand in S ogd be tw een Bokhara and Samarkand. The s tatement about Tibe t i s cor

rect. In the country of the Sogds we re the moun ta ins producing s al-ammon iac,which SPRE NGE R identifie s with the volcano introduced to the notice of E urope byKLAPROTH , ABE L-RE MUSAT and HUMBOLDT .

3 Mas udi saw , as he s ays , fire ris ingabove the s e mounta ins vis ible at n ight at a d is tance of 1 00 parasanges .

Fol low ing the rou te of the s e moun ta ins the dis tance fromKhorasan to Chinamay b e cove red in 40 days , partly through cu ltivated land, partly through sand

de s e rts . The re is anothe r route for caravans , which requ ires fou r monthsJ

a i rencontré a Bal k h un beau v ie i llard , qu i ava it fa it p lus ieurs fois le voyage dela Ch ine , sans jama is prendre la vo ie de mer; j ’ai connu également, dans le Khoracan , p lus ieurspers onnes qu i s

éta ient rendue s du pay s de Sogd au Th ibet e t en Ch ine , en pas sant par lesmine s d ’

ammon iaque .4

Mas udi’

s description of Tibe t, quoted be low , is highly inte re s ting, as muchof what he s ays agre e s w ith the rea l ity we know nowadays , while othe r parts are

somewhat exagge rated . The geographica l s ituation he fixe s is also ve ry good , re

memb ering that the name Khorasan in his time had probably a much w ide r s ensethan nowadays .

Le Th i bet es t un royaume d is tinct de la Ch ine ; la popu lat ion s e compos e , en grandepartie , de Himiarites meles a qu e lques des cendants des Tobba Parmi le s Thib étains , le suns sont s édentaires e t hab itent dans les v il les , les autres v ivent s ous la tente . Ces d ern iers ,Turcs d

origine , s ont le s p lus nombreu x , le s p lus pu is sants e t le s p lus i llustres de toutes lestri bus nomade s de la meme race , parce qu e le s ceptre leur appartena it autre fo is , et que lesautres peup lade s turques cro ient qu ’ i l leur rev iendra un jour. Le Th ibet es t un pay s priv i légiépour s on cl imat, s es eau x

, son s ol, s es p laines e t s e s montagne s . Le s hab itants y s ont toujours

s ouriants , gais e t contents , et on ne les vo it jama is triste s , chagrins ou souc ieux . On ne s au

The same story was told 700 years later by Father AN TON IO D E AN DRADA who trave lled thisway to Tsaparang. The d escription MASUD I gives of the place surrounded by high mounta in s coinc ides exactly wi th the narrative of Capta in HODGSON , the first v is itor in modern times . In this particu lar point MASUD I i s more correct than Major RE N N E LL , who derived the Ganges fromthe Manasarov ar.

2 Op . c it. Tome I p . 2 8 9 et s eq .,and 3 4 7 e t seq .

3 E l-Mas’

udi p . 3 59. R ichthofen makes the fol lowing comment : »Daraus geht deutl ichhervor, dass j ene Berge mit e inem v on den Chin es en mehrfach beschriebenen Ort, we lcher 200 ll

nOrd lich v on Kutsha l iegt, identisch s ind .» China, I . p . 560 .

4 Th is statement does not at a ll agre e w ith R ichthofen ’

s identification as given above .

ISTAKRI AND IBN HAUKAL . SI

rait e'

numérer la vari été merve illeu s e des fru its e t des fleurs de cc royaume , non p lus quetoute s le s riches s es de s es paturages e t de s es fleuv es . Le climat donne nu temperaments angu in a tout ce qu i a v ie , so it parmi les hommes

,s o it parmi le s an imau x , La douceur

du nature l , la gaieté , la v ivac ité qu i sont l’apanage de tous le s Th ib étains les portent a cu ltiverla mus ique avec pas s ion , e t a s

adonner a toute e spece de dans es Ce pay s a été nomméT h ib e t a cau se de l

ins tallat ion des H imiarites qu i s’

y s ont étab l is , la racine moa t s ign ifiant s efi xer, s

étab lir. Cette étymologie e s t encore la p lus probab le de toutes ce lles qu i ont été propos ées 2 Le T h ibe t touche a la Ch ine d ’

un cOté , et des autre s cOté s a l’Inde,au Khoragan ,

et aux dés erts de s Turcs . On y trouve beaucoup de v i lles po pu leus es , floris s antes et b ienfort ifiée s Le canton v it la chevre amu sc du Th ibet et ce lu i v it la chev re amu s cd e la Ch ine sont contigus l ’un a l’autre e t ne forment qu ’

une s eule e t meme contrée ; toute fo isla superiorité du mu s c du Th ibe t e st inconte s tab leand then follow s word for word the same account of the mu sk as given in the

work of the me rchant Su le iman .

Fina l ly Masud i te lls u s of a corre sponde nce be twe en ANUSH IRVAN , king of

Pe rs ia , and the king of Tibe t. The le tte r of the latte r was accompan ied by manycu rios itie s fromTibe t , amongs t othe r things ma n ofmusk .

2

AL ISTAKRI wrote abou t 95 1 . H e apportions the Turkish province s and a

part of Tibe t to China . S ind , Kashmi r and a part of Tibe t be long to the kingdom of India . The re fore , he s ays a part of Tibe t s tre tche s towards Ind ia ,

whileanothe r s tre tche s into China . Othe rwis e , the way in which he located Tibe t is not

clear. I t i s s ituated be tween the country of the Tagazgaz3 and the Pe rs ian S e a ,

and if one is sa i l ing a long the coas t of India and that of Tibe t one arrives in China .

Fina lly,he knows that the Tibe tan and Chine s e language s are d iffe rent fromthos e

of the Turks .

IBN HAUKAL is the famous au thor of anothe r K z

'

la’

on -l M as a’

l zlé zoo-l M o nm’lz'é

or B ook of R oads a nd K i ngdoms . He le ft Bagdad in 943 and had not aecomplished his work before 976 . He brie fly re fe rs to Tibe t and the source of the

Indu s withou t enla rging the horizon of information of his time .4

2 In Tome III , p . 253 we find the fol lowing explanation by Masudi : »De toutes les tribusturques la plus noble e s t ce l le qui habite le Thibet, puisqu’e lle descend de H imyar, comme nousl ’avons dit plus haut en parlant des Tobba, qui l’etab li rent dans ce pays . » Re inaud says that Masud i

s

statement regard ing the es tablishment of peoples of Turkis h race into T ibet and the valley O f the Indusi s confi rmed by a Pers ian work , Mojme l-al-te v arik, and by ALBE RUN I .

2 Op. cit. Tome II , p . 2 02 .

3 >D ie arabis chen Schriftste l ler rechnen T ibet te i ls zumChines is chen, te ils zumInd ischen Re iche ;s ie wiss en , das s ihre nOrdli chen N achbaren d ie Tagazgaz en , d . h . Toguz-u iguren s ind , wie s ie v on

Abu l-Ghaz i und Rasch id ed-din genannt werden , und deren Hauptstadt je ne Kuschan war, in we lcherehemals d ie H andels ' Magas ine der tibeti schen Iss edonen gewesen waren Sie sagen das s d ie Karluken

auch N achbaren der Tibetaner gewesen , we lche Ostl ich v on Jaxartes b i s zur we stlichen Grenze de schines ischen R e iches s ich herumtrieben.» Count KUUN , 0p . cit. p . 4 1 . Compare note p . 4 7 .

4 The Oriental Geography of E BN H AUKAL , an Arabian trave l ler of th e tenth century . Translated by Sir WILL IAM OUSE LE v . London 1 800 ,p . 1 0 , 1 2 , 1 55, 2 3 3 and 2 39. I do not change the

curious spe ll ing of some of the names .

52 TH E ARABIAN GE OGRAPHE RS .

»The country of T ibet i s s ituated b etween Khurlhiz and the emp ire of Cheen . Cheenl i e s be tween the s ea and the land o f Ghu z and T i bet ; and Cheen its e lf cons t itutes th is cl imate(or d iv is ion) ; b ut the other parts of T i bet were anne xed to it S ind i s the s ame as Man

s urek ; and the region of Lattian , as far as Cheen , e x tends along the coas t of H indoos tan , toT i bet

, and Cheen Macheca ,beyond w h ich no one pas s e s . O f the Mihran it is s a id that the

s ource i s the ri ver Jihoun ; it comes out at Moultan , and pas s e s on to the borders of Be sme id ,and b y Mansourah , and falls into the s ea on the eas t o f Dambu l . The waters of the ri ve rMihran are p leasant and w holesome , and they s ay it is l iab le to t ides , or flux and reflu x

,l i ke

the N ile , and that it is in fe s ted b y crocod iles . The S ind Rud,at threemerh ileh fromMoultan ,

i s of p leasant water, and jo ins the ri ve r Mihran . Mus k is brought fromT ib b et , and s ent toall parts The mounta inous country , bordering upon T i bet, i s very popu lou s , we ll cu lt iv ated

, abound ing in fru its , and e xce llent cattle ; and the cl imat is very pure and health y .»The pas s age abou t the Indu s is so d iffe rent in E LLIOT ’

s »H z

'

s tory» that I amnot

s ure whe the r it i s the s ame as the above in OUSE LE Y ’

s trans lation . It runs : »The Mihranis the chie f rive r of thos e part s . Its sou rce is in a mounta in , fromwhich a lso s ome of

the feede rs of the Jihun flow . Many great rive rs incre as e its volume , and it appearsl ike the s ea in the ne ighbou rhood of Mu l tan . It then flows by Basmad, Alruz , andMansura , and fa lls into the s ea , to the eas t of Deba l . Its wate r is very swe e t , andthe re are s a id to b e crocodi le s in i t l ike thos e of E gypt . I t equa ls the N ile in v ol

ume and s trength of current. I t inundate s the land du ring the s ummer ra ins , and

on its s ubs idence the s e ed is sown , as in

The latte r ve rs ion regard ing the source of the Indus is rather good for a t imewhen so l ittle was known of the inte rior of this world of mounta ins . The usua lcomparison w ith the N i le re turns , e specia l ly on account of the crocod i les . At a fewplace s , excep t those quoted above , Tibe t is only ju s t ment ioned . E ll iot has alsothe pas s age :' »Fromthe s ea to Tibe t is fou r months ’ journey The musk is a lsoan object which has attracted the attention of mos t of the Arabian wri ters . Ge o

graphical ly Tibe t rema ins inacce s s ible and almos t unknown .

N or does ALBE RUN I he lp u s to fix the pos i tion of Tibe t , and i f he ment ionsour lake , Manasa , it is on ly becaus e he has got it fromthe Puranas . He quote sa long l is t of S ans crit books in h is work on Ind ia, and his geography is to a greatextent pure ly Sanscrit. And s ti l l he cou ld not , be ing a barbarian , approach the

centre s of Indian learn ing, Benares and Kashmi r .When Alb eruni s ays : »The people inhabiting the mounta ins s tre tching from

the region of Panchir into the ne ighbourhood of Kashmir l ive unde r the ru le tha ts evera l brothers have one wi fe in common», one fe e ls incl ined to think of the Tib e tans , b ut he means the tribe s of H indu-ku sh , Hazara, Svat , Chitral and Kafiri s tan .

2

2 The History Of Ind ia , as told by i ts own Historians . E d ited fromthe Posthumous Papersof the late S ir H . M . E LLIOT , by Prof. JOHN DAW SON . V ol. I , London 1 8 67 , p . 40.

2 Alb erun i’

s Ind ia, about A . D . 1 03 0 . By Dr. E Dw. C . SACHAU . London 1 9 1 0. Vol. I, p . 1 08 .

54 THE ARAB IAN GE OGRAPHE RS .

are pres s ed l i ke ol ives or grapes , in cons equence of w h ich the ra in pours down, and the cloudsnever pas s be yond the mounta ins Th is rule s e ldomhas an e xce pt ion ; however, a certa inamount of e x traord inary meteorolog ical occurrance s i s pecu l iar to every prov ince of Ind ia .»

He has not much to give u s as far as pos itive new orographica l know ledgeis concerned . H e be l ieves in the de ta iled orography of the Matsya-Pu rana , and

enume rate s the great moun ta ins round Me ru : the H imav ant , a lways cove red withs now , Hemakuta , the go lden , N ishada , N ila , the peacock-l ike , of many colours ,Sve ta and Sringav ant ; the region be tween the H imav ant and the Sringav ant is

cal led Ka ilas a , the play-ground of the Rakshasa and Aps aras .

2

Aga in having quoted the Matsya-Purana and the Vaya—Pu rana in connectionw i th the rive rs ris ing in the moun ta ins of H imav ant , he re turns for a while to the

sol id ground of sound real ity , s aying»The reader mu s t imagine that the mounta ins formthe boundaries of Ind ia. The north

ern mountains are the s nowy Himav ant . In the ir centre l ies Kas hmi r, and th e y are connectedw ith the country of the Turks . Th is mounta in region b ecome s colder and colder t ill the end

of the inhab itab le world and Mount Meru . Becaus e th is mounta in has i ts ch ie f e xtens ion inlongitude , the rive rs ris ing on i ts north s ide flow through the countries of the Turks

,T ibetans ,Khazars , and S lavon ians , and fall into the S ea of Jurjan (the Casp ian Sea), or the s ea of

Khwarizm(th e Aral S ea), or the S ea Pontus (the Black S ea), or the northern S ea of the S la

v on ians (the Balt ic) ; w h ils t the r i vers ris ing on the s outhern S lope s flow through Ind ia and fallinto the great ocean , s ome reach ing it s ingle , others comb ined . The rivers Of Ind ia comee ither fromthe cold mountains in the north or fromthe eas tern mounta ins , both of w h ich inreal ity formone and the same chain , e x tend ing towards the east, and then turn ing towards thesouth unt i l th ey reach the great ocean ,

whe re parts of it penetrate into the s ea at the p lacecalled the D i ke Of Rama . O f cours e , thes e mountains d iffer very much in co ld and heat .»

This exce llent re sume’ does not need any commen tarie s . Unde r the name of

H imav ant he means the whole Himalaya . Speaking of rive rs en te ring Ind ia fromthe eas t , he may re fe r to the Brahmapu tra without having any neare r informationabou t that rive r. As the fifth rive r of the Panjab he correctly men tions Shatladaror Satlej .

2 And of the Ganges he s ays tha t the Hindus be l ieve it flowed in ancienttimes in Paradis e . He re aga in he quotes the Matsya-Purana . In his chapte r abou tp ilgrimage 3 he gets his kn owledge e n tire ly fromthe Puranas . The holy ponds are

s ituated in the mounta ins round Me ru . Thus , for ins tance , the pond V ishnupada isnear the mountain N ishadha and give s ris e to the rive rs S arasvati and Gandharv i .

»In the mountain Ka ilas a there i s the pond Manda , as large as a s ea , w hence come s theri ve r Mandak in i South -e as t of Ka ilas a there is the mountain Loh ita, and at i ts foot apond called Loh ita . Thence come s the river Loh itanadi . South of Kai lasa there i s themounta in Sarayusati and at its foot the pond Manas a. Th ence comes the ri ver Sarayu. Wes tof Ka i las a there i s the mounta in Aruna , always covered w ith s now , wh ich cannot b e ascended .

At i ts foot is the p ond Sa i léda, whence comes the r iver Sai lOda.»

2 Ibidemp . 2 47 .

2 Op . c it. p . 2 60 .

3 Op . cit. V ol . II , p . 1 4 2 .

ALBE RUN I’

S H IMALAYAN GE OGRAPHY . 55

SO long as this kind of symme trica l and e xa lted geography is confined to thePuranas it is inte re s ting ; b u t eve n a gl imps e of fre sh obs erva tion by an Arabiangeographe r wou ld b e of greate r va lue . Such a gl imps e Alb eruni give s u s of the

town and fortre s s of Nagarkot , at the Mohammedan conque s t of wh ich he was

pre s ent. The town was s ituated at the foot of the Himalaya and famou s on accoun tof i ts Old temple . At a ce rta in epoch the name Nagarkot became famou s even in

E u rope , when the whole central H imalaya was cal led the mounta ins of Nagarkot ,as , for ins tance , on ORTE LIUS ’ map of 1 579 .

Anothe r ins tance of firs t clas s geographical information is the fo l low ing .

‘ One

marche s 77 fars akh e as twards fromKanauj ; furthe r on the coun try of Tilwat is to

the right»Thence you come to the mountains of Kamrfi, wh ich s tretch away as far as the s ea .»

And then he goes on to s ay : »Oppos ite Tilwat the country to the le ft is the realmof N a ipal.A man who had trave lled in those countries gave me the follow ing re port ; ’W h en in Tanwat he le ft the eas terly d irect ion and turned to the le ft. He marched to N aipal, a d is tanceof 20 farsakh

,mos t of w h ich was asce nd ing country . FromN aipal he came to BhOteshar i n

th irty day s , a d is tance Of nearly 80 farsakh , in w h ich th ere is more as cend ing than d e scend ingcountry . And there is a water w h ich i s s e veral times cros s ed on bridges cons is t ing of p lank st ied w ith cords to two canes , wh ich s tretch fromrock to rock , and are fas tened tomi le stonescons tructed on e ither s ide . Peop le carry the burdens 0 11 the ir s hou lders ove r s uch a bri dge ,wh ils t be low , at a depth of 1 00 yards , the wate r foame s as w h ite as s now

,threaten ing to

s hatter the rock s . On the other s ide of the bridge s , the b urdens are trans ported on the backof goats

We sha l l have to re turn to this road late r on . I t has been trodden by the

Pundits in recen t years . I t was u s ed by the firs t E uropeans who eve r vis ited Nepa l ,name ly GRUE BE R and DORVILLE in 1 66 2 . Be fore the ir journey eve n the name of

Nepal was unknown in E u rope . In the firs t ha l f of the 1 8 th cen tu ry it was trave l ledmany t ime s by the Capuchin Mis s ionarie s s tat ioned in Lhasa . I t is e ithe r fromPadre Horatio de l la Penna or Fra Cass iano Beligatt i that Fathe r GE ORGI has gothis picture sque de scription of this s ame road fromNes ti to Ku ti , of which a few

sentences shou ld b e atten t ive ly compared w ith the above quotation fromAlb erun i

»Altero ab hoc rure mi lliario , pe r angus tis s imas s calas ex s ecti s mob ilib usque lapidib uss tructas ascend is des cendisque ex e d iti s s imis rupib us ad orams empe r immani s prae cipit i iD irectus e s t ite r propius ad Boream. V imin prmcipiti pos ita angus ti s s imee sunt, circumquealtis s imorummontiumlatera perpe tuo serpunt . R upes pers ze pe d isjunctmponticulis pens i lib us

s ine lateral i fu lcro , junguntur. Duodecies per tremulos hosce angus tosque ponticulos ex pertic is arb orumque ramis conte x tos transeundum e s t . Te rroremaugent tumimmensa barathratraj i cientiumoculis ad perpendicu lumu trinque subjecta , tums trep itu s fragor ingens aquarumin imo per saxa ruent ium Scalpro lap idemmed iumexcav arunt per interstit ia ad gres s umhomin is accommodata

, ut v iatores hab erent, quo n i totumvest igium, calcaneumsaltemfigere

caute

Op . cit. Vol . I , p . 2 0 1 .

2 Alphab etumT ib etanum, Romae MDCCLXII , p . 43 7 , 43 8 .

56 THE ARABIAN GE OGRAPHE RS .

It runs as if the Latin ve rs ion we re partly trans lated fromthe Arabian , whichis 700 years Olde r .

G rueber and Dorv ille came fromthe north , and to themKu ti , now cal ledN ilam-dsong and s i tuated on the Tibe tan s ide Of the front ie r , was »la premiere V il ledu Royaume de N e cbal» (Kircher).

To the Chine s e , Nepal , unde r the name of N e -po-lo , had be en known hun

dreds of years be fore Alb eruni ’s time . H iue n-tsang, the famou s and admi rable pilgrim,

whos e jou rneys fal l w ithin 629 and 644 , s eems to b e the firs t Chines e trave lle rwho gathe red information abou t Nepal , a country which he did not vis it

, and whichhad not be en mentioned by the great Fa—hian who trave l led to the we s tern countrie s200 years e arlier.

I

In a book which ought to b e cal led I-ts ing H is Pilgrims , I-TS ING ,hims e lf

a p ilgrim,te l ls us in ve ry short and pregnan t words the expe rie nce s and peregrina

tions of the Chine s e Buddhis ts , who ,in the s econd ha lf Of the 7th century , trave l led

to India , and of whommany , e ithe r coming fromthe north or fromthe sou th ,cros sed Nepal on that horrible road , s o graphical ly de scribed by Alb eruni 400years late r. S eve ra l of thos e Chine s e pilgrims who wou ld re turn v ia Nepa l d iedthe re , as they cou ld not s upport the hard cl imate of the moun tains . Unfortunate lythe re is ve ry li ttle geography in the i r de scriptions , or rathe r in the annotations of

I-ts ing . He re is an ins tance : 2 The Mas te r of Law , H iuen-t

a i , in 650 to 655 »pritle chemin de s T

ou-fan (Tibéta ins), trave rsa le N i-po-louo (N e’pau l) e t a rriva dans

l’

Inde du centre » , and then re turned to his home the s ame way3

Alb erun i’

s account Of this Old road i s one of the mos t pre cious pearls inthe geographica l l ite ratu re of the Arabs .

E DRI S I was born at Ceu ta about 1 1 00 and had fin ished his gre at geographyin 1 1 54 . He neve r vis ited the e as t hims e lf, b ut he was a ve ry learned man and

made us e of all the geographica l knowledge of his time , though , as Re inaud s up

pos e s , he has not known the account of Su le iman and the remarks of Ab u S aid ;fromothe r works he has borrow ed who le page s , and he consu lted narrative s which

2 Le Nepal, E tude historique d ’

un royaume hi ndou par SYLVAIN LEV I . Vol. I . p . 1 52 .

2 M émoire compose a l’

époque de la grande dynas tic T ’

ang sur les R e ligie ux Emi nents quiallerent chercher la loi dans le s pays d ’

occident, par I

-ts ing, traduit en frangais par E DOUARD CHAVAN N E S . Paris 1 8 94 , p . 3 5.

3 When I was in Shigatse in the early spring of 1 907 , and the Chinese authorities would haveme return to Ind ia , I made up my mind that the on ly way I would take in such a case , wou ld b e

the famous road of the Jesu its and Capuch ins and of the Chinese invas ion into Nepal in 1 792 , famousalready in the days of H lUE N -TSAN G and ALB E RUN I . I therefore questioned the N epalese consu l inShigatse and he gave me a description wh ich in every detai l coinc ides with that of the great Arabianwriter. Nothing has changed s ince then . Coming fromthe T ibetan s ide one can use yaks or sheepfor transport, or goats as ALBE RUN I has it. But where the gorge begins with i ts dangerous ga l leriesabove the precip ices

,one has to walk on foot.

58 THE ARABIAN GE OGRAPH E RS .

In his 9th s ection E dris i deals with Tibe t , Bagharghar and the country of

Kh iz ildj is , b ut he re it is as d ifficu l t to know one’

s whe reabou ts , though in some cas e s

we mus t think that he real ly means regions w e l l-known in our own days :Dans ces d ive rs e s contrées on trouve de s lacs d '

eau douce , des riviere s , des paturageset des l ieu x de campement d '

ete pour le s Turks La Ch ine e x térieure a pour l imites lepays de Bagharghar, lequel e s t vo i s in de la mer orie ntale ; du cOté du Ferghanah , le pay s deT i bet, lequel touche a la Ch ine (proprement d ite) e t a d i vers es part ies de l’Inde , e t du cOté

du nord ,le p ay s des Kh iz i ldj is .

The capital Of Bagharghar is Tanb ia ’

, s itua ted on the bank s of a rive r run

n ing to the e as t ; JAUBE RT thinks he means Kashgar .La v ille de T i bet e s t grande , e t le pay s don t e lle e s t la cap itale porte s on 1 i om. Ce

pay s e s t ce lu i des Turk s T i béta ins . S es hab itants e ntre tiennent de s re lations avec ceu x duFerghanah , du Botme t avec les s ujets du Khakan; ils voyagent dans la majeure partie de ce s

contrée s e t i ls y porte nt du fer, de l’argent, des p ierres de cou leur, de s peau x de léopard et

du musc du T i be t. Cette v ille e s t batie s ur une e’minence au p ied de laque l le cou le une ri v iere

qu i v a s e j eter dans le lac de Berwan , s itu é vcrs l ’orient ; e l le e s t ce inte de forte s murailles e t

s ert de res idence a un prince qu i a beaucoup de troupes et beaucoup de cava lerie rev étue de

cottes de mai l les e t armée de p ied e n cap .

Coars e s ilk s tuffs are manufactured , Turkish s lave s and musk are sold to

Ferghana and Ind ia . The country of Bagharghar is s i tuated be twe en Tibe t andChina , and is borde red to the north by the country of the Khirkhirs , which Jaube rtsugges ts as s tand ing for Kirghis .

Au nombre de s dépendances du T i bet es t Buthinkh , v i lle de moyenne grandeur, baties ur une éminence , ce inte d ’

une forte mura ille en p ierre et mun ie d ’

une s eu le porte ; i l y a de sfabriques e t i l s’

y fa it un comme rce tres-acti f avec le s pays env ironnants , c’

es t-a-d ire avec leKabou l, le eWakhan ,

le Dj i l, lo Wakhch, et le pay s de Ras et ; on en tire du fer renommé et

du mu s c. On rapporte que lo nard ind ien cro it en grand abondance dans les montagne svo is ine s de Buth inkh , e t qu ’

au s e in des forets qu i les couvrent, on trouve des chevrettes amus c en quantité ; on ajoute que ces an imau x broutent la c ime de la plante , bo i vent de l ’eaude la riv iere qu i coule a B uth inkh , On vo it aus s i

,dan s ces montagnes , une grotte ex

trémement pro fonde au fond de laq ue lle on ente nd le bru it d ’

un torrent ; i l e s t absolument impos s i b le d

atte indre le fond de ce t ab ime , e t quant au bru it que font le s eau x , on l’entendtres -d is tinctement

C’

es t également la que cro it la rhubarbe de Ch ine Chermakh est le nomde la

ri v 1e re qu i coule 51 Buth inkh , (v i l le ), éloignée de 5 journées de d is tance du lac de Berwan .

Cet inte rvalle es t couvert de paturages , de forets e t de Chateau x -forts appartenant aux TurksT i béta ins . Le lac s

'

étend , en longueur, s ur un es pace de 40 paras anges ; s a largeur e s t de 72mille s ; s es eau x s ont douce s ; les habitants de Berwan e t d

Oudj y péchent beaucoup de pois son .

Ces deu x dern ieres v ille s , compris es dans le T ibet, s '

ont s itu ées s ur le s bords du lacbaties s ur des coll ines ri ve ra ine s du lac, dont le s b ab itants de ces deu x v i l les bo i vent les eaux

Le lac de B erwan rego it de tous cOtés un grand nombre de ri v iere s cons id érab le s .

N on lo in de s v illes de Berwan e t de Oudj , da cOté du mid i , e s t unemontagne recourb 'éeen forme de e t te llement haute

, que cc n’

e st qu ’

av ec beaucoup de pe ine qu ’

on peut atte indre s on s ommet dont le re vers touche aux montagne s de l’Inde . Sur cc s ommet es t un

p lateau ferti le ou l’

on vo it umédifice carré dépourvu de porte

MANASAROVAR 1 N E DRISI’

S GE OGRAPHY. 59

One is los t in this s treamof pos it ive geography which has s uch a fascinatingre s emblance to the truth . E dris i provide s an ind ication abou t the source s originallyu s ed for the de ta ils he give s regard ing Tibe t , Bagharghar and the country of the

Khiz ildj is : »Nous en parle rons d ’

apres ce qu’

offrent de plus ce rta in e t de plus au

thentique les l ivres écrits e t compos és sou s la dictée de Turks qu i , ayant trave rs évces pays ou ayant habité dans leur vo is inage , ont pu rapporte r ce qu ’

ils en s ava ient.» 2But cou ld these Turks rea lly de scribe what they had s een ? And we re thos e who

took down the notes not l iab le to misunde rs tandings ? Such an able man as Grueb er who saw a good dea l of the s e countries w ith his own eye s hundreds of yearslate r , cou ld not de scribe thems uffic iently we l l to make it pos s ible for othe rs to fol

low his route s i n de ta i l on the map. E ven Abbé HUC , 700 years afte r E dris i , isve ry vague in the mos t inte re s ting portion of his journey , and s til l he had s tudiedthe country from autopsy , while E dris i pre sents us w ith third or fourth handknow ledge .

I t may s eem audacious to try and bring his geography in accordance w ithour pre s ent map of Tibe t , b ut I cannot he lp making a few s ugge s tions .

2

A5 1 have pointed ou t be fore , he very like ly refe rs , as a ru le , only to Ladakand we s te rn Tibe t , or to the region which in about 1 1 50 obeyed the king of Ladak .

Ladak is even now cal led Tibe t by the Mohammedans ; and Leh is s imply ca l ledT ebbe t by the native s or E as te rn Turke s tan . E dris i says that the great city of

Tibe t is the capital in the country of Tibe t which is inhabited by Tibe tan Turks ,a ll of which co incides exactly with the actual s tate of things

,remembe ring that the

2 Op . cit. Tome I p . 490 .

2 I must a lso us e this opportun ity to make good a grave mistake I made years ago be fore Ihad ever s een T ibet. Only on account of a certain re semblance of words , I wrote to Baron v on R ichthofen , quoting E dris i : »Sollte nun unter Berwan der Karaburan Prshe v alsky

s zu verstehen se in , sokonnte man das halbkre is rmige Gebirge als den A ltyn-tagh deuten . A lle in , wenn die Stadt Buthinkhder Araber Khotan s ein soll, so genugen die funf Tagere isen n icht, 1 lmden See an erre ichenDr. Sven Hed in’

s Forschungsre ise nach demLop-nor. Januar b i s Mai 1 896 . Briefliche Mitte ilungen

an H errn v . R ichthofen . Ze itschri ft der Ges e l l schaft fiir E rdkunde zu Berl in , Bd . XXXI . 1 896 , p . 3 52 .

T his is of cours e abs urd and should never have been printed . With a feeling of consolation and

s urprise I read the fol lowing words by Sir THOMAS H OLD ICH : »The cours e of the river on whi ch thetown (T ibet) i s bu ilt, no less than the name of the lake (Berwan) into which that river falls and the

descri ption of the Turk s lave girls , i s quite inapplicable to anything to b e found in modern T ibet. Ihave little doubt that the T ibet of Idris i was a town on the h igh-road to China , wh ich fol lowed the

TarimR iver eastward to i ts bourne in Lake Buthan . Lake Burhan i s now a swamp d istinct fromLob ,b u t 1 000 years ago it may have been a part of the Lob system, and Bagnarghar a part of Mongo

lia It is impos s ible to place the ancient town of Tibet accurate ly : There are ruined s ites innumbers on the Tarimbanks , and amongst thema place cal led T ippak, b ut it would b e dangerousto as sume a connection between T ibet and T ippak . Th is i s interesting, becau se it indicates thatmodern Chines e Turkistan was included in T ibet a thous and years ago The Gates of Ind ia be ingan h istorical narrative by Colone l Sir Thomas Hold ich . London 1 9 1 0 , p . 2 82 . Regard ing the identification of Berwan as Kara-buran, it is curious that two writers cou ld fal l upon such an extraord inaryidea .

The matter should not b e compl icated more than neces sary . When E dris i happens to saytquite correctly that T ibet borders upon China proper and on d ifferent portions of Ind ia it is hardto see how it could b e a town on the banks of the Tarim.

60 THE ARABIAN GE OGRAPH E RS .

Arabs reckoned the Tibe tans amongs t the Tu rks . E dris i’

s Tibe t , i . e . Ladak, is

in commun ication W ith Ferghana and the Turks furthe r north , which a ls o very like lywas the case in a much highe r degree than now . Amongs t the products exportedfrom Tibe t, are mentioned musk and rhuba rb , both ve ry characte ris tic things forTibe t. The capital of Tibe t, i . e . Leh , is we l l fortified and bu i lt on a hi ll

, and at

its foot a rive r runs , the Indus .

The King of Tibe t has a great army. Of course E dri s i cou ld not know his

name . But the Rev . A . H . FR ANCKE te lls us that about 1 1 25 to 1 1 50 the King of

Ladak was Lha chen Utpala , son of Lha chen rgyalpo . »He united the forcesof Uppe r and Lowe r Ladakh , s ubjected all the vas sa l chie fs , and even conque reda new province , Lowo , eas t of Purang ; s o that his empire was pe rhaps eve ngreate r than that of Nyima gon He a lso invaded Kulu , and the King of Ku lubound hims e lf by oath , ’

s o long as the glacie rs of the Ka i lasa do not me lt away,

or the Manasarowar Lake dry up, to pay his tribute to the King of Ladakh ’

Nyima gon ’

s empire as shown on a map by the Re v . A . H . Francke, embraced

eas twards , the whole uppe r bas in of the Indus , S atlej and Brahmaputra . And

now we he ar that E dris i ’s king of Tibe t was pe rhaps mightie r s till . Lha chenUtpala

s Empire is very like ly the same as E dris i’

s Tibe t.Anothe r que s tion is whe the r E dris i ’s fortified city Tibe t, can b e ident ica l with

Leh . Francke has found that the dynas ty of the ancient k i ng s of L e/z had takenthe ir name afte r Ke sar, an old s aga or epic song which enjoyed great popu larityin abou t 1 000 A . D .

3 The firs t royal pa lace at Leh was bu il t by Trashi namgyal , abou t 1 500

—1 530 , and the pre s en t cas tle of Leh, the favourite res idence of

the late r kings , was comple ted in three years during the re ign of S engge namgyal ,abou t 1 590

'—l ozo .

4 But there may have be en bu i ld ings on this picture sque hil l already in the time of E dris i . 5A d ifficulty is pre sen ted in the s tatemen t that the rive r Indus flows eas t

wards , and emptie s itse l f into the lake of Berwan . But this may eas i ly have been a

misunde rs tand ing of E dris i , and the original in formation mus t have been that therive r came fromlake Berwan and flowed we s twards .

6

2 Sir H E N RY YULE says : »Ladakh i s probably ’

the city of T ibet, bu i lt on an eminence over ariver’ of E dris i » , Cathay and the way thither, V ol. I , London 1 8 64 , p . LXX.

2 A H istory of Western Tibet. One of the Unknown Empires . London 1 907 , p . 64 .

3 Op. ci t . p . 56 .

4 Op . cit. p . 84 and 99.

5 In a letter, dated March i st 1 9 1 3 , the Rev . Francke sends me the following commun ication»Leh war erst e ine dard ische S iedelung, deren Name noch unbekannt is t. T ibetische Nomaden battendort ihre Hurden , welche s ie ge legentl ich besuchten . Diese Hurden nannten s ie G las Oder H las lHas ),aus we lchen spater G les , lHes , sLe s wurde ; d ies s ind tibetische Beze ichnungen v on Leh . Hauptstadtwurde Leh erst im 1 4ten Jahrhundert.»

6 E ven E uropeans may sometimes ge t alarmed about the directions of rivers . Lord DUNMQR E

w as astonished that the Chahlung river of the Pamirs did flow to the east instead of the west and callsthis very s imple fact »a geographical problemimposs ible for us to solve» . And he had seen the riverChahlung, while E dris i had never seen the river of the city of T ibet. The Pamirs , Vol. II , p . 3 .

62 THE ARABIAN GE OGRAPH E RS .

J e su i t Mis s ionaries (probably Andrade and Grueber), TAVE RN IE R and THEVE NOT ,

in whose accoun ts I have not been able to find any ment ion of lake Berwan or

Be ruan . But I sha l l have to re turn to this map late r on.

2

The fol lowing are some extracts from Tome I I in Jaub ert ’s trans lat ion.

2

He de scribe s the country of Adhkach. To the eas t it has the mounta ins whichs urround Gog and Magog . I t is rich in bu tte r and honey and has innume rableflocks of sheep and oxen .

Au mid i de cette contrée , i l e x is te u h lac dit de Tehama dont la c irconférence e s t de

250 mi lle s . Les eau x de ce lac sont d ’

un vert foncé ; e lles e x halent un parfumagréab le e t

s ont d ’

une saveur e xce llente . On y trouve une s orte de po is son p lat de cou leurs variées .

Au mi l ieu de ce lac i l e x is te une e spece d ’ i le dont le sol es t extrémement fertile et toujoursc overt d ’

une abondante v egetation . Le s Turcs y font paitre leurs troupeau x et y campent duranttoute la b e lle s a ison . Au centre de l ’ i le e st un pu its s ans eau dont on n

a pu trouver le fond .

L’ ile produ it, a ce qu

on d it, une p lante dont les feu i lles ress emb lent a ce lles du s o’

ad , qu i

s’

étendent b eaucoup , e t qui s ont de cou leur verteQuatre fleuv es ont leur embouchure dans ce lac. Le premier e s t le Tehama

,cons ider~

ab le , mais peu rap ide , et tres -pro fond . S es sources sont a 6 journées de d is tance du lac , e t

2 E dri s i’

s Berwan has a striki ng resemblance to Ibn Batuta ’

s Barwan, of which the latters ays : »A fter this (Kundus and Bagh lan) I proceeded to the c ity of Barwan , in the road of which i s ahigh mountain ,

covered with snow and exce ed ingly cold ; they cal l it the Hindu Kush , i . e . H indoos layer, because most of the s laves brought thither fromInd ia di e on account of the intenseness of theco ld .» (The trave ls of Ibn Batuta tran s lated fromthe abridged Arabic manuscript copies by theRev . SAMUE L LE E , London 1 8 29, p . Ib n Batuta proceeded fromBarwan to E l Jarkh and C hizna ,which was only 1 0 stages fromKandahar. Hi s Barwan is the same of which Alb erun i speaks in

conn ection w ith the river Ghorwand (the present Gorb end) which he says ris es in themountain s bordering on the kingdom of Kabul , and wh ich rece ives several affluents , amongst them»the river of thegorge of Panch ir

,be low the town of Parwan» . (Sachau

s Alb erun i , Vol. I , p . W . TOMASCH EKidentifies Parwan with the station Parthona in the Tabu la Peutingerana . (Zur historischen Topographiev on Pers ien ,

Wien 1 8 8 3 , I . p . This place is stil l in existence , between Hindu-kus h and K abul,s traight north of Kabul . But E dri s i

s description does not agree with the s ituation of Parwan . Thisvillage i s s ituated on a brook , which , li ke the Gorb end , is an affluent to Panj ir or Panjjar and flowsfromnorth to south . E dris i cal ls both the town and the country T ibet, whi le Parwan in the days of

E dris i be longed to the kingdomof the Ghaznawi , who could hardly b e called T ibetan Turks . The

river on wh ich E dris i ’ s city of T ibet i s s ituated , is in connection with the lake Berwan , and on ly ”zeroi s the city of Berwan . Alb erun i

'

s and the present Parwan , i s on a rivulet, which is in no connectionwhatever w ith a lake . The c itie s ofBerwan and Oudj of E dris i are »be longing to TMel» , and » s ituatedon the shores of the lake» , which i s supposed to have the enormous d imens ions of 40 farsakh x 7 2league s .

On Cante l l i ’s map, Pl . XXXIII , be low, both the lake Beruan and the c ity of Paruan are entered ;s uch is also the case w ith D E Witt’s map , Pl. XXXV . On many other maps , as , for instance , P l’XXIX , XXXII , XXXIV, and XXXVII I , Parvan is to b e found . In some cas es , as on PI. XXXI , bothParvan and Kabu l have been placed in the country north of the mounta ins , fromwhich the Indus andthe Ganges take the ir ris e , a country which by no means is identical wi th what we ca l l T ibet.

According to TOMASCH E K Parwan was we l l known by the Arab s : »D ie arabischen Geographennennen oft den zu Bamiyan gerechneten Rus taq Parwan mit demgle ichnamigen Vororte , we lcher inder alteren arabischen E poche e in befestigtes Heerlager, ’

askar’ , b ildete , woraus die Wichtigke it derLage erhe l lt» , (I. E dris i cannot have ignored Alb erun i ’s detailed description of the s ituation of

the place . Still he does not mention Parwan , only Berwan . Fromthe Turkish in formants he refersto, he has very l ike ly got a description of the Manasarov ar, the Indus and »T ibet»

,and confounded

.themwith Parwan , which sounded like Purang.

2 N euvieme Section,p . 3 44 e t s eq .

LAKE TEHAMA IDE NTICAL WITH MANASAROVAR . 63.

e lle s s e compos ent de tro is fontaine s s urgis sant e n abondance , a la d is tance de 2 journée s les .

une s des autre s . Uh de ces cours d ’

eau coule a l’

occide nt , e t un au tre a l ’orient de la v i l leau -des sou s de laque l le i ls s e réun is s ent

,forment une mas se d ’

eau cons iderab le e t s e jettent dansle lac. Les hab itants du pay s d ’

Adhkach y condu is ent leurs en fants pour les purifier dans le seau x de ce fleuv e , a l

époque de la pube rté Ils d is ent qu ’

en bu vant pendant s ept jours del ’eau de ce fleuv e , on a la certitude d ’

étre guéri de qu elque malad ie que ce pu is s e etre ; qu ’

en

s e lavan t la téte avec cette eau ,on e s t e x empt de migraines pendant un an . E nfin i ls ajoutent

a ce sujet tant d ’

as s ert ions et tant de particu larités merve i lleus es , qu ’ i l (nou s) para it convenab lede les pas s er s ous s ilence .

Le s econd de s fleuv es qui s e j ettent dans ce lac prov i e nt de lamontagne de Djenf. Son

cours est rap ide e t s on li t remp l i dc ca i llou x rou lés Ses eau x s ont douce s , ma is exece s s iv ement fro ides .

Le tro is i eme prend s a source dans la montagne d ’

A scaroun , coule vers l ’orient, jus qu a

Re chaca,pas se au mid i

,baigne les murs de cette v i l le , pu is s e d irige vers l ’orient, cou le au

nord de Baknoun ,v il le aupres de laque lle i l d étourne son cours ve rs le mid i , longe la bas e de s

montagnes de Ras , s e j ette dans la ri v iere de Djenf, pu is dans le lac.

Le quatrieme v ient du mid i .Au nord de ce lac i l e x iste une colline de terre rouge parsemée de trous de toutes .

Sur le s ommet de la coll ine i l e x is te une v i lle du nomd e Chanderan .

A 4 journée s de cc l ieu e s t la montagne de Kharda , tres -haute e t de toutes parts inacces s i ble , car s e s flancs s ont a pic e t s emb lab les a de s murail les un ies ; ma is au -des s ou s de

cette montagne 011 a taillé une grande porte , et , ap'res avo ir fa it d e s e x cavations,on y a pra

t iqué un chemin ,ou plutOt u h e s cal ie r avec des marches , par leque l on parv ient au somme t de ~

la montagne et a une v i lle extrémement forte et capab le de rés is ter,quand meme i l n ’

y re s tera it qu ’

nn s eu l homme . Au centre de la v i lle es t une source d ’

eau douce tres -abondanteL

excédant de cette eau s’

écoule auprés des murs , sans qu ’ i l e n res te aucune trace , dans nu

trou dont personne ne connait la pro fondeur.I t is s even days ’ jou rney fromFort Kharda to a mounta in cal led Coca i a .

Cette montagne e s t celle qu i ce int le pay s de Gog et de Magog. E l le es t te l lementabrupte qu ’ i l est impos s i b le de la grav ir, e t quand méme on y parv iendrait, i l s era it impos s i bled’

atta indre son sommet, a caus e de la quant ité de ne iges éterne lle s qu i y s ont amonce lées e t

de s b roi llards e’

pais qu i toujours l’env ironnent . .Au de la sont de nombreu s es v i lles dependente s .

du Gog et du Magog. 1 1 y a dans cette montagne nomb re de s erpents e t de rept i le s énormes .

qui v i vent au fond des valléesFinally he te l ls u s that north of the ir coun try is a range of mounta ins cal led

Farghan , 1 8 marche s long fromwe s t to eas t. In the s e mounta ins there is a roundplateau with a lake of unknown depth. At the foot of the mounta in ,

towards the

sou th and oppos ite the lake , the re is a cave rn fromwhich a te rrible no is e oftenris e s . The Adhkach Turks have large face s , big heads , much ha i r, ve ry l ive ly eye s ,speak a particu lar language and worship the fire and eve ry brill ian t thing .

An inte re s ting pas sage in E dris i ’s rathe r confus ed account is that abou t lakeTehama and its fou r rive rs . The s tory of the fou r rive rs has ancient root in As ia ,all fromGene s is , 2 : 1 0 , whe re -a s treamis s aid to have is sued fromE den , spre adingits e l f into fou r great rive rs , Pison , G ihon , Hiddeke l or Tigris , and Frat or E uphrat .On his map FRA MAURO identifies Ganges w ith Pis on , at leas t he has the name

6 4 TH E ARABIAN GE OGRAPH ERS .

Phison entered at the s ide of Gange s .

2 The V ishnu Punana te lls us how the

G ange s , afte r encircl ing the capita l of Brahma ,d ivides into fou r mighty rive rs , flow

ing in oppos i te d irections , S i ta , Alaknanda , Chakshu and Bhadra .

The names and the d is tances of E dris i do not , of cours e , play any part whate ve r in an accoun t whe re so much is in confus ion . The s ame is the cas e with the

d irection of the s e rive rs . E dris i makes themente r the lake ins tead of is su ing fromi t. In the follow ing pages we shall often have to deal w ith the four rive rs is su ingfrom the Manasarov ar. DE L ISLE and TIE FFE NTHALE R be l ieve , at leas t partly, inthe i r exis tence and hav e themon the i r maps ; De l is le , howeve r , has only two . E venin our own days one hea rs , fromtime to time Tibe tans as se rting that four rive rsfl ow unde rground to the mou ths of the Lion , the E lephant , the Peacock and the

Hors e , the s e rive rs be ing the Indus , S atlej , Map-chu , and Ts angpo

-Brahmapu tra .

The firs t of E dri s i ’s fou r rive rs may b e the s acred Gange s , »heard of, bathedin , sanctifying all be ings » , for he says the inhabitants of Adhkach take the ir childrento bathe them and pu rify themin its wate r. The native s regard the ablu tions inthe Tehama as incon tes table . Drinking the wate r heals eve ry s icknes s , and washing the head w ith it ensu res aga ins t headache for a year. The e te rna l wonde rs of

the Gange s -wate r be l ieved in by the Hindus have been d ige s ted in a more practicalway by the Mohammedan w rite r.North of the lake are s ome famous mounta ins , one of which , Coca

'

i'

a , is de

s cribed as ve ry abrupt and cove red w ith e te rna l s now . This may b e the Ka'

i' las a .

2

To this d is cus s ion i t might b e objected that E dris i should have de scribed the

Manas arov ar as if i t we re two di ffe ren t lake s , the Berwan and the Tehama , the one

in Tibe t, and the othe r in the country of the Adhkach . But this is not at all s ur

pris ing, for’

E dri s i is , as a ru le , ve ry confus ed and having taken his informationfromtwo diffe ren t sou rce s about one and the s ame lake , he ve ry l ike ly be l ieved thathe had to do w ith two qu i te d iffe rent lakes in d ifferent countries . And we shou ldnot forge t that both the s e countrie s we re practical ly unknown in his t ime . The s ame

2»Mappamondo di Fra Mauro. Da una copia fotografi ca de ll’originala vene ziano, es istentepresso l

Istitu to Geografico M i l itare .» Stud i Ital ian i di F ilologia Indo-Iran ica, D ire tt : da FRAN CE SCOL . PULLE . V ol . IV . Atl. de l la Parte I , F irenze 1 90 1 . See be low P l . XIII.

2 In this mountain numbers of enormou s s erpents are l iving. Such tales were told 1 50 yearslater by M arco Polo, though in one case h is s erpents were crocod i le s . Yule : The Book of SerM arco Polo . Vol . II, p . 76. O f a qu ite d ifferent part, name ly, a place be longing to the Pres idencyof Madras , Marco Polo s ays : » in thos e mountains gre at serpents are ri fe to a marve l lous degree ,bes ides other vermin

,and this owing to the great heat. The s erpents are also the most venomous in

existence , insomuch that any one going to that regi on runs fearfu l peril ; formany have been destroyedby these evil repti le s » . Op . ci t. V ol. II , p . 360. There is a great resemblance between this and

E DRI S I’

s : »Il y a dans cette montagne nombre de serpents et de reptiles énormes d’

o1‘

1 résu ltentdes dangers et des obstacles pour qu iconque voudra it la gravir.» The following legend on GASTALDI

s

map of 1 56 1 , placed in a region far to the east of the source of the Gange s , probably refers to

Marco Polo’

s s erpents as described in his V ol. I I, p . 76 : »qui u i sono s erpe grandi s ime ) ; and the

same mus t b e the case with M ercator’s legend at the very source of the Gang es , on his map of 1 569 :

» In hi s mon ti bus sunt serpentes maximi .» I do not know wh ether E dri s i ’s serpents are taken fromt he same source as those of Marco P010 and the old

’maps .

66 THE ARABIAN GE OGRAPHE RS .

Lahawar (Lahor). Then there are the mounta ins of B illur, in the d irection of Turk is tan,which

are denominated Shami lan .

2

The j eb lum he cal ls the rive r of Tibe t, a new indicat ion that by the nameTibe t was gene ra l ly unde rs tood only the we s te rn part of the moun ta inous world .

The rive rs of Panjab are all sa id to combine with the Satlader or Satlej . Fromthe same mounta ins which s end down the »panj -ab » or fiv e rive rs to the south , the»haft-ab », or s even rive rs , go to the north and formthe rive r of Ba lkh , Amu-darya.

The city of Kanauj s tands on the we s te rn bank of the Ganges ; to whichBINAKITI, who quotes portions of this chapter , adds : »which come s fromthe ci ty ofTurmuz , through the mountains of the eas t» (E ll iot) ; a curious s tatement, regard ingRashideddin

s view that the s even rive rs of the Amu—darya jo in at Turmuz .

Ras hideddin hims e lf is be tte r informed than B inak it i . He s ays : »The Gange sis sues fromits source , ca lled Gangdwar, and wate rs many of the citie s of India

Rashideddin quote s Alb erun i word for word in gi ving the itine rary fromInd iathrough Nepa l to Tibe t.

Of the s ource of the Indus we learn : »The S ind rises in the mounta ins of

Amak , on the borde rs of the Turkis h country .» Alb erun i has Unang.

2

In the geography of the great prince ABULFE DA , who was born at Damascusin 1 2 73 and died 1 3 3 1 , and who d is t ingu ished hims e lf as much in scho larship as in

the wars agains t the crusade rs and Mongol ians , we do not , however, find much of

value as to the regions tha t are of specia l inte re s t to us . H is work was aecomplished in 1 3 2 1 . Of the Indus , Nahr Mihran , he s ays it has a lso the name S ind ,and gi ve s the same descript ion of its cours e in India as we know fromhis predece ssors .

C’

e s t '

nu grand fleuv e , dont l ’eau e s t tres -douce ; il res s emb le au N il d’Egypte, en ce

qu’ i l nourrit dans son se in des crocod i les On l it dans le R esm-A lmamour que le fleuv eMihran pre nd s a source s ous le 1 z6e degré de longitude et le 36

c degré de lat itude ; qu ’ i l cou leau sud-oues t 3

He has nothing abou t the source of the Gange s .

Le fleuv e du Gange , N ahr Gang, e s t appe lé par les Ind iens Gangou . Il s’

av ance a

l ’orient de Canoge (Kanauj ) Le Gange e st nu fleuv e tres -res pecté de s Ind iens ; les Ind ienss’

y rendent e n pelerinage ; ils s e préci p itent dans s es eaux on s e tuent sur s es ri ve s .

The gre at latitudina l sys temof ranges in the inte rior of As ia he s imply cal lsDjebe l , or »the Mounta in» par preference :Dje bel e s t le nomd

une cha ine , qu i partant des front i eres de la Ch ine s e tend a l’

occi

dent jusqu’

au pay s de Fergana et de Osrouschna (dans le -Transoxiane) ; e lle pas se entre les

2 The B illur-tagh or »crystal mountains » runn ing north through Badakhshan . Shami lar isprobably fromthe Arabic »Shamn» , north . (DOW SON ). Sachau writes Bolor.2 O f Amak CUN N INGHAM says : »Thi s i s apparentl y the T ibetan Gya-nag, pronounced Ganak ,

which means the B lack Plains , and i s the name for Chinese Tartary where the Indus actually rises .»

E LL IOT, op. cit. p . 64 .

3 Geographic d’

Ab oulféda par M . Re inaud. Tome I I , Paris 1 848 , p. 79.

ABDULFE DA . I BN BATUTA . 67

v i lles de Kes eh e t de Samarcand , et s e réun it a la montagne de Bokhara, connue s ous le nomde Ouarka .

Re inaud thinks that e ithe r Abu lfeda did not know the name of the s e mounta ins or he be l ieved that the nat ive name was synonymou s with »mounta im. At oneplace I Re inaud w rites »Djebe l (Thibe t)» as if Abu lfeda had especial ly meant thatpart of the sys temwhich borde rs upon Tibe t.Ind ia i s borde red on the we s t by S ind, on the eas t by the de s e rts which s ep

arate it fromChina , and on the north by the country of the Turkish nomads .

Regarding Tibe t and its s i tuation ‘

he only quotes ISTAKRI , and MOHALLAB I

who s ays : »Le Tibe t (Tob b at) tombe au nord du royaume de Canoge : une granded is tance s épare ces deux pays .»

N or has the famous trave l le r IBN BATUTA much to te l l abou t our parts Of As ia .

He l ived from 1 304 to 1 3 7 7 . Like nearly all othe r trave lle rs and geographe rs heavo ided Tibe t, the inacces s ible country beyond the mounta ins . Of the Indus he s aysthat » it is the greates t rive r in the world , and ove rflows during the hot weathe rjus t as the N ile doe s And of the othe r great rive r he s ays : »Some of the

Hindoos drown themse lves in the rive r Gange s , to which they pe rformpilgrimage s ;and in to which they pour the ashe s of thos e who have be en bu rnt.»

In 1 340 A. D . we find Ib n Batuta at Sadkawan, a place in Bengal : »FromSadkawan I trave lled for the mounta ins of Kamru , which are at the d is tance of one

month from this place . The s e are e xtens ive mounta ins , and they join the mounta insof Thibe t, whe re the re are musk gaze lle s . The inhabitants of thes e mountains are ,

l ike the Turks , famou s for the i r a ttention to magic.» This is obvious ly the eas te rnHimalaya , for in Alb eruni we have found that from the point whe re one turnsnorthwa rds to Nepa l the mounta ins of Kamru »s tre tch away as far as the s ea .»

3

Then it »happened» that he ente red the country of China , and came to a citycal led Jab nak ; » it i s d ivided by the rive r which descends fromthe mounta ins of

Kamru , cal led the Blue R ive r. By this one may trave l to Benga l and the countrie sof Laknouti .» FromYunnan and the Blue R ive r one may indeed trave l to Benga l,though Ibn Batuta place s the sou rce s of the rive r too far sou th , ins tead of in the

range s north of Tang—la .

The knowledge of the world regarding the geography of Tibe t cannot b es aid to have been more augmented by the journey of Ibn Batuta than by that of

his contemporary Friar Odorico de Pordenone , though the latte r trave lled s tra ightacros s Tibe t.

2 Op . c it. p . 8 9.

2 The Trave ls of Ibn Batuta ; trans lated fromthe abridged Arabic manuscript copies ,the Rev . SAMUE L LE E . London 1 8 29 , p . 1 00 .

3 KUUN says that Mount Kamru, or Kamaru as he has it, is s ituated in the province of As sam.

Op . cit. p . 78 .

CHA PTE R V.

LATER MOHAMM EDAN WR ITERS .

In the Ma ls zdt-i 77 772a , or the au tobiographical memoi r of the great conqueror TI MUR , which origina lly was w ri tten in the Jagata i Turki language , translated into Pe rs ian and ded icated to Empe ror Shah Jahan who began his re ign in1 62 8 , we find on ly a few re fe rence s which may b e of inte res t in this connection .

2

During his campaign the Empe ror rece ived information that Nagarkot »was a largeand important town of Hindu s tan , and s ituated in the se moun ta ins As

soon as he heard this he decided to carry on war aga ins t the infide l Hindus of

Nagarkot, and the city was conque red . >A party of the Hindus fled towards the

mounta in , and I taking a body of soldiers purs ued themup that lofty mounta in , andput themto the sword . Afte r mount ing to the summi t I halted .»

In the city of Jabhan »I made inqu iries abou t the coun try and city of

Kashmi r frommen who w e re acqua inted with it, and fromthem I learned thatKashmi r is an incomparable country». Its capita l was called N aghaz and a largerive r flowed through the city . »The source of this rive r is within the l imits of

Kashmir in a large lake , some parasanges in length and breadth , which i s cal ledV ir-nak The rive r pas ses on and jo ins the Chinab above Multan . The un itedwate rs pas s be low Mu ltan and then join the Rawi . The rive r Biyah come s downthrough anothe r part and jo ins them, and the three un ited rive rs fa l l into the S indor Indus in the ne ighbourhood of Uch . All thes e (un i ted) rive rs are ca lled the S indor the Panjab , and this rive r fal ls in to the Pe rs ian Gu lf near Thatta.» H e doe s notmen tion the Satlej in which the Bias fal ls be fore jo in ing the Indus . The Ravi jo insChenab above Mu l tan .

For a conque ror as Timur the inacce s s ible , and to a great extent un inhabitedcountrie s to the north had no va lue and no inte re s t whateve r, and he doe s not evenme ntion them, though he who had drenched half As ia in blood ce rta inly knew the

geography be tte r than anyone of his time . His hi s torian , SHE RE FE DDIN from Yesd,2 E LLIOT

s Hi story of Ind ia , Vol. I II p . 465 et seq.

SHE RE FE DDIN FROM YE SD . 69

who d ied in 1 446 , at leas t mentions Tibe t in his Zafar-N a'ma or B ook of Vi ctory .

And he knows the source of the Ganges , at the Cow ’

s HeadLe De’ tro it de Coupe lé (Montagne s ur le Gange) es t s itué au p ied d ’

une montagne parpas se le Gange , a qu in ze mi lles de chemin p lus haut que ce Detroi t, i l y a une p ierre

ta i llée en forme de Vache , de laque lle p ierre sort la s ource de ce grand F leuve ; c’

e s t la cause

pour laque lle le s Ind iens adorent cette p ierre , dans tous les pay s circonv o is ins jus ques aune

année de chemin ,i ls se tournent pour prier du cOté de ce Detro it de cette Vache de p ierre .

As Sherefeddin hims e lf says , he has de rived what he te lls abou t Kashmir, partlyfromnative informat ion , partly fromwhat he had be en able to ve rify pe rsona lly :

La forme de ce Pay s e s t ovale : i l e s t entouré de haute s montagnes de tous les cOtés

ce lle ‘

du Mid i regarde De l i la terre de s Indes : ce lle du N ord regarde Bedakchan la Co

rassane : ce lle d ’

Occident regarde les H orde s de s Ougan i s , ce lle d ’

Orient regarde Tob bot , ouTheb et

He finds it surpris ing that a rive r w ith such a swift current as that of N agaz ,can come down from one s ingle origin , the source of which is in the country its e lf.Of the rive r of Kashmi r he knows

i l s e jo int aude s sous de Moultan au Fleuve Genav i (Ab -i-Chenab , Chenab i), tous deu xens emb le ayant pas s é Mou ltan , tombent dans le Fleuve de Rav i , qu i pas se par l’autre Mou ltan ;

ensu ite le Fleuve B iah les jo int, tous aupres de la Ville d ’

OUtche s e j e ttent dans le grandFle uve Indus nomme’ Ab s end (Ab -i -S ind)He knows three roads to Kashmir, one rathe r difficu lt fromIndia , and

ce lle de Tob b at ou Theb et e s t p lus fac ile ; ma is durant p lus ieurs journées on trouvequantité d ’

herb es ven imeus es qu i empo isonnent les che vau x des pas s ants .

2

HAFIZ ABRU, who died in 1 430, gives in his Tci r i k/c a few particulars abou tthe s ource s of the great rive rs : 3

2 H istoire de T imur-Bee . E crite en Persan par SH E RE FE DD IN ALI , nati f d ’

Yezd , auteur contemporain . Trad . par PE T I s D E LA CROIx. V ol. III , Paris 1 7 2 2 , p . 1 3 1 and 1 60 et s eq.

2 Sherefeddin’

s description of the rivers of the Panjab i s far superior to anything written byE uropeans even some 3 00 years later, as can b e seen if h is re lation b e compared wi th the l ittle mapaccompan ying PE TI s D E LA CROIx

s edition of 1 7 2 2 (PI. IV) . For there i s noth ing wrong in Sheref

edd in ’

s hydrography. The s ingle curious th ing i s that the name B ias i s given to the whole courseof the Satlej , fromi ts junction with the B ias to i ts junction with the Chenab. The upper course of

the Satlej , from the Manasarov ar to the junction of the B ias , he does not mention . On Petis de laCroix ’s map the Jehlamjoins the Chenab ; bu t the n ext river eastwards , i s not the Ravi as it ought tob e , but the B ias , which i s shown as join ing the Chenab be low Multan instead of the Satlej . The

easternmost river of the map i s called Caul or Dena and corresponds to the Satlej as can b e seenfromMt. Couke’

, correspondi ng to the present Guge . Mansar may stand for Mansaroar or Manasarov ar. Lahor is there , but placed on the B ias instead of the R avi . Peti t T ibet and Grand T ibet arenot badl y placed . Sherefeddin says correctly that the Ganges takes its origin fromthe Cow

s Head .

But on the E uropean map the source of the Ganges i s placed some 1 3 days ’ journey above the Cow

which i s even entered on the map an d provided wi th the legend : »Vache de piere adore’

e par les

Guebres» . The Indians , otherwise the Hindus of Sherefedd in have , 3 00 years later, been improved to

fireworshippers . The li ttle map has no original value whatever ; on Del is le’s map of 1 7 2 3 the Panj abrivers are arranged in qu ite a d iff erent way.

3 E ll iot, op. cit. Vol. IV ,p . 4 et seq .

70 LATE R MOHAMME DAN WRITE RS.

»Th is ri ver (S ind) has i ts s ource in the s k irts of the mountains of Kas hmir, and .runs

from the wes tern s ide o f thos e mountains into the country of Mansura . Its cours e is fromnorth to south , the e nd turn ing to the eas t.» And of the Ganges : »Its s ource i s on the east ofthe country of Kanauj There are other large rivers to the eas t of th is wh ich are ment ioned b y men who have travelled in Ind ia ; b ut the ir name s , s ource s , and embouchure s , havenot been accurate ly s tated . So also there are many large rivers in Ch ina , b ut it i s not known

whether the y run to the e as t or to the wes t, nor where they ri s e , nor w here they d ischarge .»

A hundred years late r KHONDAMIR says :»The ri ver Gange s is a river s ituated to the eas t of Kanauj , and the Hindus are of

opin ion that the water of th is ri ver s prings fromthe fountain of Parad is e ; hav ing burned the irdead , the y throw the as hes into the s tream, and th is practice the y hold as puri fy ing themfromthe ir s ins .»

2

In his autobiography the great BABAR , one of the mos t fascinating leade rs ofmen who eve r l ived. in As ia (1 48 2 makes a short re fe rence to the mountains in the north , from the wes tern continuation of which he had such wonde rfu lp e rsona l expe riences

»The north e rn range of h i lls has be en ment ioned . Immed iate ly on cros s ing the ri verS ind , we come upon s e veral countries in th is range of mounta ins , connected w ith Kas hmir,s uch as Pakhali and Shamang. Mos t of th em,

though now independe nt of Kas hmir, wereformerly included in its territories . After leav ing Kashmir, thes e h ills conta in innumerabletribes and s tates , parganas and countries , and e x tend all the way to Bengal and the s hores of

the Great Ocean . About thes e h i lls are o ther tri be s ofmen .»2

Shah Babar has trave rs ed and described the Hindu-kush ; we now come to a

Mohammedan who has trave l led s tra ight through thos e regions of s ou th-we s te rnTibe t which i nte res t us he re .

MIRZA HAIDAR was de spatched by the Khan of Yarkand to invade Ursangand des troy its idol-temple , which was the point of adoration of the whole of Khita i . 3H is jou rney l ie s be tween j uly 4th and Octobe r 2 2nd, 1 53 3 and s eems to b e the firs tin this part of Tibe t, from which any re l iable information exis ts .

’ It is , howeve r ,very meagre in geographica l de ta i l . He s tarted fromM51q or Leh and for 20

days found no othe r s igns of infide ls than a few fortre s se s . Leaving the heavy partof his army behind he wen t on »in all has te» w ith the s tronge s t hors e s and the be s tpart of his cava l ry . Afte r 50 days he came to a place called Barmang whe re hehad a s ucce s s fu l fight w ith the Champa people of Tibe t. The part of the armythat followed finally approached a fortre s s on the way, ca lled Kardun . The de feu

2 Op . ci t. p . 1 8 1 .

2 Op . c it. V ol. IV , p . 2 2 1 .

3 N . E LIAS and E . DE N ISON Ros s : The Tarikh-i-rashidi of M irza Muh . Haidar. London 1 895,p . 454 et seq .

72 LATE R MOHAMME DAN WRITE RS .

sounding in the leas t l ike Luk-u-Labuk . Only the latte r hal f of this name has a

vague l ikenes s w ith S e-Lib uk (S e lipuk), which , howeve r, shou ld change the lake intoNganglaring

-tso . From S e lipuk the re is the Serpun-lamto Lhasa , b u t it is not

l ikely that an army wou ld have chos en s uch a way. The heavy loss es in hors e scanno t have be en cau s ed by »dam-giri» as he says , s pecial ly not i f the hors es werefrom Ladak. They mus t have died fromfatigue , which po ints to the hardships inve ry high regions . But then it is d ifficu lt to account for Hari . The re is nothingto b e su re of, even not the s tatement that he arrived wi thin 8 days fromUrsangor Lhasa. In S aka-dsong the re is , at leas t nowadays , a l itt le temple which may havebeen called Saka-labrang, which has a fa in t l ikenes s with Askab rak . But this placeis much more than 8 days fromLhasa , though it can b e sa id to b e 8 days fromthe wes tern frontie r of the province of Utsang.

The A z

'

n-z

'—A ,ééarz'

compiled unde r the supe rintendence of ABUL FAZL ,prime

min is te r to Empe ror AKB AR , has rightly been regarded as a gold mine of in

formation regarding Hindus tan , as it was in the days of Akbar. Of geographica lmatte r it doe s not, howeve r, conta in so much as one cou ld have expe cted , althoughi t is known how Akbar took s pe cia l inte res t in a t leas t one of the great rivers ,name ly, the Gange s , even to s uch an exten t that he sent a s pecia l exped ition ins earch of the sou rce of the s acred rive r, 1 and that he , as we are told by Abu l

MAN OUCH I , a Ven etian phys ic ian ,who served at the court of Emperor AURAN GZE B , tells us

how the »native explorers » of Akbar d i scovered the source of the Gange s : »Another of Akeb ar’s Curios itie s was , that of knowing the Source of the Ganges . That R iver the Most E astern of all Indoustan

runs i ts Course fromNorth to South I t'

s Source has always been the Ground of a Dispute b etween the Brachmane s of these , and the Gymnos ophists of former times . In the days of A kebar itwas still unknown , as the Source of the N ile was unknown not above an Age ago. The Emperortherefore spar’d no Cost to d is cover the Head of a R iver which was the best Jewel in hi s Crown . He

Commiss ionated certain Persons , with orders to steer Northwards by the Banks of the Ganges , til l theyarriv

'

d at i t’

s Fountain . H e furnish’

d’

emwith Provis ion s , Hors es , Mony, and Letters of Recommendation for pass ing un dis turb

d thro’

all the Countries bordering upon the Gange s , and which werenot of h i s own Dependance . They kept on the ir Course towards the North , and the nearer theyapproached the Source , the narrower the R iver grew. They pass

d through untrodden Fores ts , wherethey were forc

d to cut out Paths for themse lves . At last they arriv ’d at a high Mountain , whichs eem’

d to b e shaped by Art into the Formof a Cows H ead. Out of it issue s a vast Quantity ofWater, which the Commiss ioners supposed to b e the Source of the Ganges . They penetrated nofarther, they re turn

'

d after having run through vari ous Dangers , to give the Emperor an Account ofthe ir Trave ls . The ir Re lation was inserted in the Chron icle fromwhence I have taken it. And aftera ll , th ey have told us noth ing new. Long be fore Akeb ars T ime , the People in the Ind ies were perswaded that the Ganges took its Source in a h igh Mountain ,

whose Figure resemb l’d that of a CowsH ead The Ind ians s ince the Days of Akebar have made some further Discoveries , and found thatthe Ganges on ly forms a Cascade on the Mountain fromwhence they be l ieved it took its Source ; b utrises much h igher in the Country towards the middle of great Tartary .» The General H istory of theMogu l Empire , Fromi t

s Foundation by Tamerlane , to the Late Emperor Orangzeb . E xtracted fromthe M emoirs of M . M anouchi , a Venetian , and Chie f Phys it ian to Orangzeb for above forty years .

By F . F . CATROU. London 1 709, p . 1 52 et s eq. On E uropean maps fromAurangzeb’

s time the

source of the Ganges is indeed placed far away to the north , in Tartaria Magna .

TH E AIN—I-AKBARI . 73

Faz l , 1 a lthough a Mohammedan , only us ed to drink Gange s wate r bo th at homeand whi le trave l l ing.

To judge from the A z'

n-i -Aé éarz

'

, Tibe t was known and valued chiefly on

account of its products . »Although gold is imported into Hindu s tan , it is to b e

found in abundance in the northe rn mounta ins of the country, as a lso in Tibe t.Go ld may a lso b e ob ta ined by the Salon i-proce s s fromthe sands of the Gangesand Indus , and s eve ral othe r rive rs , as mos t of the wate rs of this country are mixedw i th gold : howeve r , the labou r and expense gre atly exceed the profit .»

2 Or whenit i s s a id that me rchants bring to court good horse s froms eve ra l diffe rent countries ,Kirgis , Tibe t, and Kashmir amongs t othe rs . Or when the »katas » or Tibe tan yakis mentioned , »an an imal of extraordinary appearance , occu rring in the ne ighbou rhood of Tibe t and Kashmir» .

3

Akbar was told that »in Tibe t the re was even now a clas s of Lamahs , or

Mongolian devotee s , and reclus e s , and he rmits , that l ive two hundred years,and

more>,4 informat ion that mus t have arous ed the curios ity of the Empe ror, who was

particu larly inte re s ted in re l igion ,not le as t in that of the Je su it Fa the rs at Goa .

The following piece of hydrography is important ; it occurs unde r the title»S zi baé of La

é or fi

»It has s ix princi pal ri vers wh ich all flow fromthe northernmounta ins . I . The Sutlej theancient name of w h ich i s Shattudar and whos e s ource i s in the Kah lor h i lls 6 i t rece i vesthe B iah at the Bauh ferry . 7 I I . The B iah (B ias) was anc iently cal led B ipasha .

8 Its source i snamed B iahkund in the Ku llu mounta ins in the vic in ity of wh ich the town of S u ltanpur s tandsabove the r i ver. III. The Rav i

,the ancient I ’rawat i , ris e s in the Bhadral h ills . Lahor the

cap ital, i s s ituated on i ts banks . IV . The Chenab,anciently Chandarb haga V. The

B i hat,9 anciently ca lled Bidas ta

,has its ris e in a lake in the parganah of V er in Kas hmir, flow s

through Srinagar and enters H indus tan V I . The s ource of the S indh (Indus ) i s placed b ys ome b etween Kas hmir and Kashghar, wh i le others locate it in Ch ina . It flows along the

borders of the Sawad territory . into Baluch is tan .»1 °

1 FRAN QOIS BE RN IE R writes that Aurangzeb also used Ganges-water, as we ll as all members of

the court ; when trave lli ng several came ls laden with Ganges -water accompan ied the camp . Trave lsin the Mogu l Empire A . D . 1 656

—1 668 Con stable ’s ed ition,London 1 8 9 1 , p . 2 2 1 , 3 56 , and 3 64 .

9 The A in i Akbari by Abu l Faz l ’Al lami , trans lated fromth e orig ina l Pers ian , by H . BLOCH

MAN N M . A . V ol. I . Calcutta 1 8 73 , p . 1 2 .

3 Ibidemp . 1 3 2 and 1 49 .

4 Ibidemp . 2 0 1 .

5 Op . cit. V ol. II , trans lated by Colone l H . S . JARRE TT, Calcutta , 1 89 1 , p . 3 1 0 .

6 I do not know whether the Kahlor hil ls stand for Kulu or Kailas or anything else . The

s ource of the river in Manasarov ar seems to have been unknown , at least to the compilators of the

Ain-i-Akbari . In a note the translator says : »It ris es l ike the Indus on the s10pes of the Kailasmountain s The twin lakes of Manasarowar and Rakas -tal, un ited w ith each other, are i ts d irect source» ,a statement which is on ly partly correct.

7 P . 3 2 6 i s found a more detai led description of the junction of the B ias and Satlej .3 Sanscrit Vipas a , Greek Hyphasis .

9 Sanscrit V itasta , PTOLE MY’S BidaSpes .

The uncerta inty betrayed by this passage becomes so much the stronger when it is sa id late ron

, p . 3 64 , that the river called S ind rises in Tibet.1 0—1 3 1 3 8 7 I .

74 LATE R MOHAMMEDAN WR ITE RS .

Unde r the t itle »S s a/z of B eng a l»‘we read

»Its ri vers are countle s s and the firs t of themin this prov ince is the Ganges : its sourcecannot b e traced . The H indu sages s ay that it flows down fromthe ha ir ofMahade va’s head .

R is ing in the mounta ins towards the north , it pas s e s through the province of In

prais e of th is s treamthe H indu sages have w ritten volumes . Fromi ts s ource to its mouth iti s cons idered s acred b u t some s pots have pecul iar sanctity .»I t is added that this sacred wate r is swe e t, l ight , whole some , andmay b e kept

in a ve s s e l for years w ithou t unde rgoing any change . The compilator of A imiAé barz

s e ems to have be en very unce rta in abou t the source of the Gange s , for, atanothe r place

,

2 dea l ing w ith the S u éci /z of a dés he speaks of a holy place calledChikar Ti rth , w ith an image of Mahadeo, and adds

»N ear it a s pring ris e s wh ich i s he ld to b e the Ganges . An as cet ic b y the power ofthe Almighty was in the hab it of go ing to the Ganges da i ly fromth is s pot. One n igh t theri ve r appeared to h imin a dreamand s a id : ’Undertake thes e fat igues no longer ; Imys e lf w i llr is e up in thy ce ll ’ . Accord ingly in the morn ing it b egan to w e ll forth and is flow ing at the

pre sent t ime .»

At las t we me e t with the Brahmapu tra : »Anothe r rive r (in Benga l) is theBrahmaputra . I t flow s fromKhata to Kuch and thence through the Sarkar of Bazoha and fe rt i l iz ing the country, fa lls into the s ea .

» 3 Jus t as the eas te rnmos t rive rof the Panjab , the e as te rnmos t rive r of Benga l has e s caped the attent ion of the Mo

hammedan geographe rs for hundreds of years . S uch was the cas e e specia l ly w iththe Brahmapu tra which was s ituated far away fromthe centre of Mohammedanpowe r in India .

4

Kashmi r was of cours e very we l l known in Empe ror Akb ar’s days , he

h ims e l f, is sa id to have trave lled the Fir Panja l-road thrice . Great Tibe t is placedto the N .E .

,and Litt le Tibe t to the N .W . of Kas hmir . Lar is the d is trict which is

sa id to borde r on the mounta ins of Great Tibe t . To the north of Lar the re is a

lofty mounta in which domina te s all the su rrounding country, and the ascent of whichis arduous .

5

The old Sans crit orography of the mounta ins north of India is re lated by,

Ain- i-Akbari , and aga in we mee t Sumeru , Himachala , H emakuta , N ishada and the

res t, without becoming the wis e r as to the ir s i tuat ion in re lation to each othe rE ven Ka ilas a is mentioned, though the re are manymountains w ith this name . Amongs tthe sacred place s of pilgrimage of the Hindus , the Ganges is enume rated as No . 1 ,

and the Satlej as N o

1 Op . c it. p . 1 2 0.

2 Op . cit. p . 2 24 .

3 To which the trans lator has a note : >Its rise is supposed to b e fromthe SE . base of the

sacred Kai las h ill , on the oppos ite s ide of the waterparting in which the Sutlej and the Indus alsotake the ir rise a view that is not in accordance with real ity.

4 In h is Akbar Nama Abul Faz l re lates war operations on the banks of the Brahmaputra,»which i s a large river, that flows fromKhata», E lliot’s History . V ol. VI . 1 8 75, p. 73 .

5 Op . c it. p . 3 63 .

6 Op . ci t. Vol . III . Calcutta 1 894 , p . 3 0, 290 and 3 03 .

CH IN E S E ND T I B ETAN

G E PH E RS

80 CH INE SE \VORKS ON TH E HYDROGRAPHY OF SOUTH -WE STE RN TIBE T .

w ri tten in 1 792 by MA SHAO-YUN and ME I HSI-SHE NG , for this work was foundto conta in nearly all earl ier Chines e information . Rockhill supplemented the in

formation of We i-tsang-t ’u-chih by extracts fromall Chines e works published downto 1 89 1 , thus add ing s everal itine raries and o the r facts which had not been mentioned in older books . Rockhill gives the titles of thirteen different works whichhave been his principal sources .

I

BRE TSCHN E IDE R regards the Ta -on’

i ng- i -t

nng or Grea t Geograpny oft/ze C/zi n es e empi re n n a

’er tne Ta -o/t

i ng (Manchu) a’

yn a s ty as the mos t importan tof the Chines e works prov id ing us w ith geographical know ledge of the Chines e Empire and its dependencies . Its firs t edition appeared in 1 743 , while the s econdmuchenlarged edition was published in 1 764 and con ta ins 500 chapters . The bes t mapof the Chines e domin ions is the Ta -on’

i ng—i—t’n ng—y i ’t-t’n , which was published in Wu

ch ’ang-fu inHowever, before I proce ed to quote s ome interes ting pas sages fromthes e and

othe r more recen t works , I may b e allow ed to enter af'

few quotations from what theearly Chines e chron icles have to s ay abou t the re lations be tween China and Tibe t .

Du ring the T ’

ang dynasty Tibe t is mentioned as one of the four great powerswhich made war aga ins t China , and fromthe early part of the same dynas ty the

firs t embas sy fromTibe t to the cou rt of the Chines e Emperor is men tioned : »The

firs t day of the fifth moon , 634 , an ecl ips e of the sun took place . The emperorrece ived an embas sy fromthe king of Tibe t. The name of this coun try was T’

u

fan (old pronunciat ion T ’

u-

po). This king was cal led Tsan-p’

u . He had formidablearmies and was feared by all his ne ighbours .» 3

E DOUARD CHAVANNE S reminds u s of the great exten t of the Tibetan dominions in the latter ha lf of the s even th century . In 663 the Tibe tans had ga ined a

great victory over the T’

u-

yij-hun

,4 a Tungus people who had es tablished them

s e lves ih the region of Koko-nor. In 670 the Tibe tans conqu ered the »fou r garrisons »,Kucha, Khotan , Kashgar, and Tokmak . The Chines e came to the as s is tance of the

Tungus tri be , b ut the imperial tr00ps were complete ly defeated in the val ley of

Bukha in-

gol. Thenceforward the Tibe tans , be ing sove re igns of Koko-nor and E as te rnTu rkes tan , became ve ry pow e rfu l and mixed thems e lves cons tan tly -in the affa i rs of

the wes te rn Turks , the ir ne ighbours to the north , while India was the ir ne ighbour

Loc . cit. p . 3 .

2 Mediaeva l R esearches FromE astern A s iatic Sources Fragments towards the knowledge of the

geography and history of Centra l and We stern A s ia fromthe 1 3 th to the 1 7th century. By E . Bretschne ider. V ol . 1 . London 1 9 1 0 , p . 2 06 .

3 Abrégé de l’

histoire Chinoise de la grande dynastie Tang. Mémoires concernant l ’histoireetc. des Chinoi s , par le s M is s ionnaire s de Pékin . Tome qu in z ieme , Paris 1 79 1 , p . 444 .

4 I amextreme ly indebted to Doctor A LBE RT HE RRMAN N for h is kindnes s in read ing two proofsof this cha

pter. For the transcription of the Chin es e names Doctor Herrmann has advi sed me tofollow Wade

s system; it should , however, b e remarked that on ly the pronunc iation of the present dayis re ferred to . Only in some case s , where it has been poss ible and nece ssary, the old pronunciationhas been added within brackets .

TA-T’ANG-HSI-vu-cnr. 8 1

to the s ou th . To the eas t they he ld the country to Liang-chou in Kansu and to

S zechuan . In 692 the Chines e , having defeated the Tibetans , conque red the fourgarrisons .

In the C/zi n T ang-s ine and the T ang

-ska there is to b e found a de ta iled

chron icle of the pol itical and military re lations be tween the Tibetans and Chines eand s urrounding peoples , though there is nothing of geographical interes t to us . In

Chavannes ’ trans lation we are told that during the i—feng pe riod, 676—678 , Tu-chitook the ti tle of Kagan of the ten tribes , concluded an all iance with the T

u-fan

(T’

u-po) or Tibe tans and ravaged An-hs i or Kucha . In the s ame period the Tibetansare . s aid to have ravaged the Kingdomof Kas hgar.

Great P ’

o-li i or Baltis tan is said to b e direct w es t of the T ’

u-fan (T

u-

po) orTibe tans , and borde ring upon Litt le P ’

o-l'

n, G ilgi t, and wes twards upon the te rritoryof Wu

-ch ’ang or Oudyana. It be longs to the T

u-fan (T

u-

po), and from696 to

74 1 it is reported to have s ent ambas sadors to the court thrice ; l i to the

eas t, a l itt le to the s ou th , fromLittle P’

o-lii one arrives at the encampment of T ’

u

fan (T’

u-

po) tsan-p’u , the b ts anpo of Tibe t ; 500 l i sou th is Ku-shih-mi or Kashmir.In 747 Kao Hs ien-chih made an expedition to Li ttle P ’

o-l'

u.

S eve ral features of the ancien t Sanscri t geography of S .W . Tibet are to b e

found in Ta-t’

ang-li s i -y i

t-oni , R ecords of t/ze Wes tern World (compiled during) theGreat T’

ang dynas ty (A . D . 6 1 8 trans lated by Impe rial command by HSUANCHUANG (H iuen Ts ang), a Doctor of the three Pitakas , and ed ited by PIE N CHI ,a Shaman of the Ta-tsung-chi Temple . In the In troduction to Book I of this work ,written by Chang Yueh , who flourished as min is te r of s tate unde r T ’

ang Huan Tsung(A . D . 7 1 3 we read the following pas sages , which for the sake of completenes s shou ld b e ins erted in this connection 2

The mounta in called S uméru s tands up in the midst of the great s ea firmly fi xed on

a circle of gold , around wh ich mountain the sun and moon re volve ; th is mounta in i s pe rfectedby (composed of) four prec ious sub s tances , and i s the abode o f the Devas . Around th is are

s even mounta in-ranges and se ven s eas ; betw e en each range a flow ing s ea of the e igh t pecu l iarqual ities . Outs ide the s even golden mountain-ranges i s the salt sea . There are four lands(countries or is lands

,dv ipas) in the s alt sea, w h ich are inhab ited . On the east, (Purva) v idéha ;

on the south , Jamb udv ipa ; on the west, Godhanya ; on the north , Kurudv ipa .

In the middle of Jamb udv ipa there is a lake cal led Anav atapta, to the south of the

Fragrant Mounta ins and to the north of the great Snowy Mountains ; it is 800 l i and more inc ircu it ; its s ide s are compos ed of gold , s ilver, lap is-lazul i , and cry stal ; go lden sands lie at the

bottom, and its waters are clear as a mirror. The great earth Bodh isattva, b y the power ofhis v ow , trans forms h ims e lf into a N aga

-raja and dwe lls there in ; fromh is dwell ing the coolwaters proceed forth and enrich Jamb udv ipa (S han-pu-chao).

Shorni k trudofi‘

orkhonskor ekspedi ts iy, VI Documents sur les Tou-kiue (Turcs) occidentaux .Recue i l l is et commentés par E douard Chavanne s . S . Petersburg, 1 903 , p . 74 et seq.

2 Si-yu-ki . Buddh ist Records of The Western World . Trans lated fromthe Chinese of H iuen

Ts iang (A. D . by Samue l Beal . Vol. I , London 1 8 84 , p . 1 0 et seq.

1 1—1 3 1 3 8 7 I .

82 CH INESE WORKS ON THE HYDROGRAPHY OF SOUTH -WE STE RN TIBE T.

Fromthe eas tern s ide of the lake , th rough the mouth of a s ilver ox,flows the Ganges

(King-ch ’ ie h , old pronunc. Kang-ga) r i ve r ; encircl ing the lake once , it enters the south-eastern s ea .

Fromthe s outh of the lake , through a golden e lep hant’s mouth , proceeds the S indhu(Hs in-tu) river ; enc ircling the lake once , it flows into the s outh-wes tern s ea.

Fromthe wes tern s ide of the lake , fromthe mouth of a hors e of lap is -lazu li , proceedsthe ri ver V akshu (Fu-ch

u,old pronunc. Fok and encircl ing the lake once , it falls into

the north -wes tern s ea . Fromthe north s ide of the lake , through the mouth of a cry s tal l ion ,proceeds the ri ver S ita (H s i -to), and enc ircl ing the lake once , it falls into the north-eastern s ea .

The y also s ay that the s treams of th is river S ita, entering the earth , flow out beneath the Ts ihrock mounta in , and give ris e to the ri ver of the midd le country (Ch ina).

In Book IX of the same work it is s aid 1

To the we s t of the north gate of themounta in c ity is the mountain called P ’

i -pu-lo

(V ipu la-giri). Accord ing to the common report of the country it i s said, ’

Ou the northern s ideof the s outh-we s tern crags of th is mountain there were formerly fiv e hundred warms prings ;now there are on ly s ome ten or s o ; b ut s ome of thes e are warmand others cold , b ut none

of themhot .’

Thes e s prings have the ir origin to the south of the Snowy Mounta ins fromtheAnav atapta lake (Wu-jo-nao and flow ing underground , b urs t forth here . The water is veryswee t and pure , and the tas te is li ke that of the water of the lake . The s treams (fromthe lake)are fiv e hundred in number (branches ), and as the y pas s b y the les s er underground fire-abodes(he lls), the power of the flames ascend ing caus es the wate r to b e hot . At the mouths of the

various hot s prings there are p laced carved s tones , somet imes s haped l i ke l ions , and at othertimes as the heads of wh ite e lephants

In a note to the Chines e work Anav atapta is defined as »whithout the annoyanceof heat» , i . e .

, cool . BE AL s eems to b e unce rta in abou t the identificat ion of Anavatapta ; in the Index he has : »Anav atapta (

O-neu

,

O-neu

-ta, and O-na-p

o-ta-to), in

Chines e Wu-

jeh-no

-ch

i , a lake ,» withou t saying whe ther it is Manasarov ar or ariyother lake . At anothe r place (Vol. I , p . 1 3) he says that the Vakshu or Oxus flowsfrom the Sarik-kul lake in the Pamir plateau ; »it is we l l called , therefore , ’

the coo llake ’ (Anav atapta). The Oxus is su es from the wes te rn end of the lake .) Here ,therefore , Anav atapta shou ld b e identical with the Lake Victoria ofWOOD . Regard ingthe great dragon lake (Nagahrada) he identifies it w ith the S arik-kul lake (Vol. II ,p . 297) or the lake of the Great Pamir, and continu es : »The great Naga lake iss ometimes cal led the Rav anahrada .» Vol. II , p . 4 1 Beal aga in expla ins Anav atapta

thu s : »In Tibetan Ma-dros—pa , the king of the Nagas (snakes ) of the lake of the

same name , and (p . 1 55) Anav atapta is expla ined thu s : »Rav anahrad ; in Pal i,Anav atatta , in Tibe tan , Ma

—dros , in Chines e , Wu-

jo-nao.»

In the ir explanatory notes to Fo-kuo-chi ABE L REMUSAT and KLAPROTH enterthe problemof Anav atapta or Anav adata and its rivers . Speaking of Fa Hian ’

s rive rHs in-t

ou Rému sat s ays : 2

1 Op . cit. Vol . I I , p . 1 55.

2 Foé Koue'

K i ou re lation des royaumes bouddh iques : voyage dans la Tartaric, dans l’Afghan istan et dans l

Inde , execute, a la fin du IV e S iecle , par Ch‘

y F5. H ian . Tradu it du Chinois et commenté par M . Abe l R ému sat. Ouvrage posthume rev u , complete, et augmente d ’éclaircissementsnouveaux par MM . Klaproth et Landresse . Paris 1 8 3 6, p . 3 6 et seq .

84 CH INE SE WORKS ON TH E H YDROGRAPH Y OF SOUTH -WE STE RN TI BE T.

Its title is : >>M af of Me Wes tern R eg i ons and I ndi a » (Pl . V). It was publishedin 1 7 1 4 in a Japanes e E ncyclopedia : »0 nan-

s a n-ts

a i -t’

n-nn i .» Klaproth who cal ls

it an important work, be l ieves that it was compos ed in the 1 5th century , a lthoughpublished some three hundred years late r. But the mate ria ls mus t b e much olde rs ti l l , at leas t thos e regard ing India mus t b e even fromthe 7th cen tu ry . Klaprothpo ints out that the Chines e , who originally col lected the material , at that epoch hada much cleare r knowledge of the interior of As ia than the Greek geographe rs

,

especially Ptolemy.

The source of Hwangho is repres ented in accordance with the ancient Chines ebe l ief. It is formed by two rivers join ing N .W . of Khotan . The sou thern one comesfromthe mounta ins wes t of Yarkand, called Ts ’ung-l ing. This river is the Yarkanddarya. The northe rn one

'

comes from the lake Lung-ch ’ ih or Dragon lake , theKara-kul, which, as Klaproth be l ieved, gave ris e to Kas hgar-darya .

‘ Thes e two rive rsformed the Tarim, which runs eas twards to the Lop

-nor, repres ented on the map,

and then con tinued as the Hwangho.

2

The kingdomof Khotan and the great s and des e rt are there . North of Tarimis Hami . The Hindu-kush of the map is cal led Hs '

ueh-shan or Snow Mountains anda range further sou th He i- l ing or Black Mountains . It wou ld s e emmore plau s ib le ,however, to identify the Snow Mountains w ith Kwen-lun , the Black Mountains wi ththe Kara-korumand the range sou th of this with the Himalaya. There is ve ry littlespace le ft on the map for Tibet, ju s t as on so many E uropean maps , where , as

we shal l find, this coun try some times a lmos t d isappears .

As to the Manas arov ar and the rivers Klaproth s ays in h is text»L

Indus , le Setledj et le Gange ont tous leur s ource au s ud du lac O neou tchy,3qu i

es t le Manas sarov ar. Au nord-oues t du lac O neou teb y 3 e s t le pays de Po lou qu i e st

le Bolor de nos jours . Le s montagnes qu i avo is inent ce lac sont appe lées en tub e tain Gangd isdans le Chou i k ing, 5 qu i es t une tres -anc ienne h ydrograph ic ch ino is e , e lles portent le nomd ’

O

neou ta.

6 Les commentateurs de ce l i vre d is ent que cette denominat ion s e retrouve dans lesl i vres bouddh iques . E n e ffet

,on l it dans le s ouvrages des C ingala is , que le grand lac Anotatte

w ille , s e trouve dans un vas te dés ert, e t que les quatre princi pa les fleuv es dumonde en s ortentpar quatre porte s , dont une forme un gueu le de l ion , l ’autre ce lle d ’

un éléphant,la tro is ieme

ce lle d’

un che val , et la quatrieme celle d’

un boeu f s auvage . E ncore aujourd ’hu i les quatre1 Doctor Albert H errmann in forms me that the Lung ch ih or Dragon lake i s not, as Klaproth

supposed , the Kara kul, b u t beyond doubt the Dragon lake of the P ilgrimo

H sti an chuang (Hiuen tsang) ,i . e . the present V ictoria lake of Great Pamir (compare M . A . Ste in, Ancient Khotan I , p .1 Doctor Al bert H errmann has proved beyond doubt, that the Chine se opin ion , that the Tarim

shou ld b e the upper cours e of the Hwangho, s imply depends upon a misunderstanding of a few words

in Chang ch ’ ien ’

s text. He trave lled 1 3 9—1 2 7 B . C . Soon the Chinese began to doubt the veracity

of this statement, for already in the Hon Han -shu the passage in question has been omitted . D ie

al ten Se idenstrass en , e tc. p . 1 6 et seq .

3 In Wade’

s transcription A-nu-ch ’ ih (ch ’ih lake) .4 Po-ln-lo .

5 Shui -ch ing.

6 A-nu-ta.

86 CH INE SE WORKS ON TH E H YDROGRAPH Y OF SOUTH -WE STE RN TIBET .

Au mi l ieu du continent merid ional du monde sont le s plus haute s c imes des grand smonts ne igeu x , appe lés Thsoung l ing (montagnes d e a l

orient de ce s c imes s e

trouve l ’emp ire de Tch in tan“, ou la Ch ine ; au s ud es t le Th ian tchu 3 , ou , l’H indous tan , a l’

oue st

le royaume de Pho s zu 4,ou la Pers e ; et au nord sont les pay s de s barbares nomades .

5

In BUSHE LL’

s trans lation from the officia l his tory of the T ’

ang dynas ty w efind a few glimps es of the geography of E as tern Tibe t. 6 Here we are told of an

ambas s ador, LIU YUA N -TING ,who, in 8 2 2 A . D .

, was s ent by the empe ror on a

mis s ion to T’

u-fan (T’

u-

po). Some 300 l i fromthe point where he cros s ed the Yellow R iver (Hwangho), there w e re three mounta ins l ike circu lar, flat-bottomed coppersin shape ; they w e re cal led the Tzu (Purple) Mountains and bound the Greater Yangt ’ung country .

»Thes e are the ancient K ’

un lun Mountain s . They are called by theTibe tans the Ménmoli Mountains The sou rce of the Ye l low R iver is in the

mids t of thes e mountains . The water is very pu re and clear, b ut as it flows pas tother s treams i t changes to red ; and afterwards , when it has be en joined by moretribu taries , i t gradually becomes ye llow and turbid.» Bushe l l locates the GreaterYang-t ’ung on the plateau to the sou th of Khotan , be twe en the Les s er Yang—t’ungto the ir wes t and the T

u-fan (T

u-

po) on the i r eas t.Fu rther the Chines e text runs : »The country to the sou th-wes t of the Ye llow

R ive r is flat , l ike a Whets tone an uncu l tivated plain , richly covered with gras s ,w ith many tamarisk and w i l low-trees on the banks of the river , whi le themounta insare cove red w ith cypres s es .» The Ts angpo is called »the Tsang River» , b ut onlyoccas ional ly men tioned .

In the book on pilgrimage , written by I-TS ING , who was born in 634 , on lya ve ry few geographica l h ints are to b e found . It is rathe r a col lection of shortbiographies , of chiefly Chines e Buddhis ts who , in the s econd half of the s eventhcentury A . D . , wen t in s earch of Sanscri t books contain ing the dogmas of the i r law .

One of thes e pilgrims was I-Ts ing hims e l f. Chavanne s who has trans lated the

work is indeed right in s aying that i t is s urpris ing that in on ly one generation no

fewe r than s ixty persons cou ld find i t worth while to vis i t the wes t , and he thinks1 T s

ung-ling.

1 Ch ' in-tan .

3 T’ien-ch ’

u .

4 P’

o-szfi.

5 Humboldt points to the same fact when d iscuss ing h is great central A s iatic range which eastof the meridi an of K hotan gets the name of A-néu-ta . >Man konnte s ich wundern , dass d ieser Namegle ichfal ls sowohl fur den b eruhmten A lpensee Tub e ts , aus we lchemder Sutledj entspringt , als we iterhinnoch fur e inen Berg im Suden de s H ima laya gebrauchlich is t ; aber man muss beachten , dass d iebeschre ibenden Namen b es tandig vera llgeme inert werden ,

was fur die Geographie sehr gefahrlich i s t,und das s die H indus , ehe s ie noch die Topographie Ttib ets kannten , s ich mogen eingebi ldet haben ,das goldfuhrende Land des Norden s zwischen demKuen-lun und den H e i l igen Seen (R avana-hradaund Manas a) b ab e man s ich als e ine wen ig bre ite R egion ganz d icht amH imalaya vorzustel len .» A .

v on Humboldt, Central—A s ien , Berl in 1 844 . I . p . 600 .

6 The E arly H istory of T ibet. FromCh inese Sources . By S . W. Bushe l l , Journal of the Roya lA s iatic Society . N . S ,

Vol. XII . London 1 8 80 , p . 4 3 5 et seq .

I -TSING AND H IS PILGRIMS . 87

that before and after I-Ts ing hundreds of trave llers went out with the same object.He shows that there mus t have been a continual s treamof pilgrims to India bravingthe dange rs of land and s ea . Thos e who trave l led by land had to cros s the

Gobi des e rt.On his firs t jou rney (abou t 650—664 A . D .) one of the Pilgrims , HSUAN-CH AO

(Hinen-chao), has on two d i ffe rent occas ions vis ited Southe rn Tibe t. The firs t time

he en te red T’

u-fan (T

u-po) from the wes t ; after which que en Wen-ch ’eng, a Chi

nes e princes s , ordered that he shou ld b e brought to Northe rn India . H is re turnjourney brought him back to Tibe t, b ut this time v ia Nepal . Aga in the Tibe tanquee n gave himan honourable reception ; afte r which he returned to China s traightacros s Tibe t . On both occas ions he mu s t there fore have be en at the Ye re-tsangpo .

On his s econd jou rney , which he s tarted in 665, he was not fortunate . Once he wasattacked by Tibe tan robbe rs , and when it was his intention to re turn home v iaNepaland Tibe t, the Tibe tans prevented him fromentering the ir country . The refore hewas obliged to re turn to India whe re he final ly died .

I

1 M émoire Compose a l’

époque de la grande dyn as ti c T’

ang sur les Re l igieux Eminents qu iallerent chercher la loi dans les pays d ’

occident par I-Ts ing. Tradui t en francais par E douard Chavannes .

Paris 1 894 . For the above vers ion of H stian-chao’

s Trave ls I amindebted to Doctor A lbert H errmann . His vers ion difi

'

ers in many respects fromthat of Chavannes . Doctor Herrmann confirms hisview in the following wordsChavanne s , hat s ich , wi e ich glaube , b ei der E rklarung der v on H suan-chao be suchten Ortl ieb

keiten in e in igen wesentl ichen Punkten ge irrt. R ichtig ist, dass d ie »Sables mouv antsz , die der

Re isend e zu allererst durchzog, die Sandw i i s te wes tl ich v on Tun-huang s ind . Dann is t aber die folgende Gegend »le Desert pierreux» s chwerl ich das Que llgebiet des Hwangho ; denn v on Tun-huang aus

wird H suan-chao s e inen Marsch doch woh l durch das Tarimb ecken fortgesetzt haben . Ich suche die>Steinwi i s te» in Sarikol oder sonst irgendwo in Pamir. H ierzu passt auch d ie we itere Beschre ibung :

dans les montagnes e scarpées i l marcha sur le bord de s pas sere lles de bois il fut balance’sous les ponts de corde S ie kann namli ch n ichts anderes darstel len als d ie in den chine s ischenAnna len v ielgenann ten »Hangenden Pas se (H ien die auch v on Fa-hian und anderen P ilgern benutzts e in mus sen . Das i st der Weg uber den K i l i k-Pas s und durch Kandjut (auf me iner Karte zu d en

»alten Seidens trassen» falschlich auf e inen westli cheren Gebirgsweg fib ertragen , Begrii ndung folgt inme inem2 . Band) .

Das Land T’

u -f on (alte Aussprache T ’

u-po) sche int H . damals hochs tens an den Grenzen b etreten zu haben . Chavannes ’ Annahme , das zwe imal bes uchte Luo-c/z'a s e i Ladak , halte ich fur v erfehlt. Denn wenn es he isst, dass Luo-ch ’a imwestl io/zen Ind ien auf demWege nach Fo-k

o- luo (B a te/z)und zugle ich in der N '

ahe v on S ind l iegen so ll,dann darf es nun imaussers ten NW. Ind iens gesucht

werden ; oi ellez'

e/zt war es das v on den P i lgern wegen e ines gros sartigen Buddhakultus v ielb esuchte

Udydna (imTale des Swat) .Fur T ibet s e lbst kommt eigen tlich nur d ie erste R e ise des Hsuan-chao in Betracht (bald nach 650

n . Auf demWege v on Tokharestan (NO . Iran) kamer, wie es he i sst, in das Gebiet der T ’

u

f a n (T’

u-po) und wurde dort v on der Kon igin Wen-eff i ng nach N . Ind ien geleitet. Diese , eine chi

n es i sche Prin zess in,war 64 1 n . Chr. mi t demtibet. Kon ig v ermahlt worden (Naheres s . Chavannes ,p . 1 4 , n . Dass H . schon b ei d iesemersten Male b is in das Quellgebiet des Yeretsangpo gekommen zu se in sche int, durfte besonders aus den Angaben hervorgehen ,

wo se in 2 . Besuch , namlichauf se iner Ruckre i se behande lt wird . Als er Ni—pu-lo (Nepal) durchzog, »le roy de ce pays envoyaune escorte qui l

accompagna jusque che z les T ’

ou-fan ; z'

l v i t do nouv eau la prz'

nees se Wen

e l le lui ofi'

ri t beaucoup de presents et le tra ita avec honneur. E l le lui donna des moyens de reven irdan s le pays des T ’

ang (China) . A lors i t paroorrut [e terri toire des Fan (Tu-fan) , de Z’

onest et ar

ri v a en C/zine (664 n .

88 CHIN E SE WORKS ON TH E HYDROGRAPHY OF SOUTH -WE STE RN TIBE T.

Anothe r of thes e re l igious travel lers was the mas te r of the law ,Hsuan-t’a i ,

who in 650 to 655 went through Tibet and Nepal to India and returned the s ameway. Many others are mentioned as having vi s ited Nepal and some of themdiedthere , the climate be ing too cold after the heat of India .

In the short biographical notes of I-Ts ing the Ch’ iang-chia or Ganges is sometimes mentioned b ut w i thou t any geographical detai ls .

90 WE I-TSANG-T’

U-CH IH .

au sud du Lo so tchhouan s’

appelle aujourd ’hu i Zzang ’

; c’

es t de ce tte riv1 e re quele Tubet a regu le nomde S i tsang.

3 »

»Sous la dynas tie des Ming, ce tte contrée porta le nomde 011 5 tsang (Ou1Zzang) 4 . Ou

1'

Zzang4 est la meme chos e que Thou pho 5, ou Tube t ; mais sous le

premie r nomles provi nces de Ou'

1‘ 6

et de Zzang2e’ ta ient s eu lemen t compris es .»

The Chine s e au thor’s des cri ption of Wes tern Tibe t is ve ry short : »E n a llantdu H

las s e i -ts io-k ’hang, a l’

oues t , on arrive , par Djachi i-lounbo , a S an-sang e t a la

frontiere du Nga-ri . D e S an-sang on v a, par le mon t Gangtes , au Nga

-ri, jusqu

a

la frontiere de Gardou . Un au tre chemin condu i t de San-sang, par les mon ts Mer-la

e t Thoung-la , et par Hiegar a Gnialam, ville s ituée sur la frontiere des Gork ’ha .

Le Tube t s e partage en quatre provinces , dont la 4° es t le Nga-ri (les grands monts ),contrée as s ez étendue , et s i tuée au nord-oues t ; e l le es t con tigue aux deux tribus deLatak e t de Gougou-dze

The map which accompan ies this article , 7 and which is reproduced here as

Pl . VI, is trans lated fromthe Chines e original by Klaproth . It gi ves a very cu rious

repres entati on , or rather de formation of Tibe t. One s ees how the e rrors grow towardsthe north and wes t, and how the northern and wes te rn parts are qu ite out of pro

portion as compared wi th the sou th-eas tern part or the part through which the authorhas trave lled and the t1 00ps were brought. The d is tance between Lhas a and Tas hilunpo is greater than the d is tance between Tashi-lunpo and the Manasarov ar. Tashilunpo is as far fromTengri-nor as Keriya-la . In fact Tengri-nor occupies the centreof Tibet. The whole Kwen-lun is there , and in its eas tern part »the s ources of the

Hwangho» are placed . South of Tengri—nor we recogn is e a part of the eas te rnTranshimalaya . L. Daryouk may b e meant as Tarok-tso , as a rive r flows into i tfromthe sou th . The Chaktak-ts angpo , i f i t i s meant to b e tha t river comesfrom far in the north . The inte rior plateau- land i s fu ll of lakes . S an-sang (Sangsang) is N E of Manasarov ar; Gartok, Ngari and Ladak are all on the map, althoughplaced in a perfectly fantas tical way. The re is a name les s lake N .W . of Lac Map

ham,

which mus t b e the Rakas -tal, a lthough a river falls i nto it fromthe wes t. There isno channe l be twe en the two lakes . NE . of the lakes the re is a mountain range ,M . Gandis , and on i ts north—eas tern s ide the F . Zzang ts iou (Tsang-chiu) takes its ris e .

Thus this map places the source of the Brahmapu tra on the eas tern s lopes of the Kai las ,a view that not ve ry long ago was not unfami l iar to E uropean geographe rs . Mou rouou s sou is regarded as the sou rce branch of S in-sha-chiang which is correct . Northof Tengri-nor i s »Pays de Dam» , a name that re tu rns in Nain S ing ’s Dam-largen-la.

1 In Wade’

s transcription Lo-so-ch ’uan .

1 Tsang.

3 H s i-tsang.

4 We i -tsang.

5 T’

u-fan (old pronunciation T’

u-po) .6 We i .7 Nouveau Journal A s iatique , Tome VI , p . 3 50 .

CHA PTE R V I I I.

TH E CH IN ESE GEOGRAPHERS ON TH E SOURCE OF TH EBRAHMAPUTRA.

In another article Klaproth reminds u s of the fact that in the beginn ing of

the 1 7 th cen tury Tibe t was ve ry li ttle known and that E u rope ’s knowledge of the

country was due to some vague information by mis s ionaries who tried to find a

'

new

road to China .

I The great river which du Halde calls Yarou Tsanpou on his

maps travers es Tibet from w es t to eas t, b ut the maps leave u s in

doubt as to its fu rthe r cou rs e . As to its Tibetan cou rs e , however, Klaproth givesthe following des cription , which is enti re ly bas ed on extracts fromChines e geographers :

2

Son v eri tab le nomtub etain es t Dzangb o tchou . Le premie r de ces deu x mots, Dzangb o ,

s ign ifie l impide et pa r, et le s econd tchou , es t le nomord ina ire de l ’ean et des ri v ieres . Le

Dzangb o tchou v ient du Tub et occ idental , ou de la prov ince de Ths ang 3; i l y porte le nomde

Yaerou dzangb o tchou , c’

e s t-a-d ire , ri v i ere cla ire e t pare de la f ron tiere g a uc/ze on de l’

oues t.

Ce fleuv e prend s a s ource par d e lat . nord , et de long. E ., dans le Damtchoukkabab , montagne ne igeus e s ituée s ur le s confins de la prov ince d ’

Ari . Apres un cours d ’

env iron

dix l ieues a l ’es t,i l rego it a gauche une ri v iere qu i sort du Dj imagoungroung, p etit lac s itu é

a l ’e s t du Langt s ian kabab , bou che de l’éléphant, autre montagne ne ige us e . Ce nomlu i v ientde ce que les e au x produ ite s par la fonte des ne iges sortant de s e s flancs crevas s e’ s avec lameme impe

tuo s ite’

que de la trompe d ’

un e' lephant . Le premier affluent cons idérable du Ya rou

dzangb o tchou es t a gau che , le N aouk dzangb o tchou , qu i v ient du nord-est . P lus b as i l recoi tdu coté oppos é le Gouyang, qu i d écou le de l’H imalaya . I l travers e apres

,de l

ouest a l ’e s t,la prov ince de D z ang, p as s e de vant Jikadze et le con v ent de Djach i loumbo , e t reco it encorea gauche le Djochotgh i choung ,

le D zaka dzangb o tchou , l’

Outchou ,le D ok tchou e t le Chang ;

et a dro ite le Ghiazoungrou ,le Chord i , le Manggar, le Chab tchou e t le Djouangdz e , que lu i

1»Memoire sur le cours de la grande riv1 ere du Tubet; appe lée Iraouaddy dans le royaume

des B irmans» . Mémoires re latifs a l’As ie , par M . J . K laproth , Tome III , p . 3 70 e t seq . Paris 1 8 2 8 .

The same article i s publi shed under the heading : »Mémoire sur le cours du Yarou dzangbo tchou , ou

du Grand F leuve du T ibet ; su iv i de notices sur la source du Burrampouter» , in Magas in A s iatique ou

R evue Géographique et historique de l’

As ie Centrale publié par Mr J . Klaproth , Janvier 1 8 2 6.

Republished Paris 1 83 5 in one volume together w ith other articles of the s ame author.

1 Mémoires relatifs a l’As ie , loc. c it. p . 3 75.

3 T sang in Wade’

s Transcription .

THE SOURCE OF DAMTCHOUK KABAB . 93

envo ie l’

H imalaya . A Djachi loumbo , s on lit es t tres -large , et partagé par des i les en un

nombre infin i de canau x , dont le principal , pres de ce couvent, e s t e’ tro it et profond

,e t ne

de v ient jamais guéab le . La cc fleuv e n’

e s t p lus d és igné que par le s eu l nomde Dzangb o tchou .

No trave l ler before 1 907 has given a more correct des crip tion of the s i tuationof the sou rce of the Brahmapu tra than the Chines e geographers fromwhos e worksthe above pas sage is taken . The Chines e des cription is concis e , clear

, short,and

correct.Yere—tsangpo was the firs t name I eve r heard for the river, when I came down

to Ye .

‘ The as tronomical s i tuation as given in the Chines e text is w rong, as usual ,which , however, does not matte r a bit in this connection . For here the qu es tion iswhich rive r in the reaches of the uppe r Tsangpo is the sou rce of the Brahmapu tra ,and not : which are its co

-ordinates ? And i t is s aid that the rive r Yaerou dzangb o or

Brahmapu tra takes its source in the Damtchouk kabab snow-mounta in s i tuated on the

fronti er of the provi nce of Ari . 2 E ven oppos i te to Shigats e the rive r is cal ledTamchok-kamba , 3 and at the confluence with Raga-tsangpo I found it cal led Damchu , corresponding to the French spe ll ing Damtchouk . Kabab means >>sou rce of» .

Afte r a cours e of s ome 1 0 l ieu es to the eas t i t (the Ye re-tsangpo) rece ivesfrom the left a river which comes from a l ittle lake , cal led Dj imagoungroung .

Klaproth can on ly mean the ordinary lieu e of which 25 made one degree of the

equator and which was mete rs . Thu s the d is tance to the confluence shou ldb e ci rca kilome te rs . In fact i t i s 63 kilometers fromthe very source . But i fwe take the text l ite ra lly and on ly regard the part which flows to the eas t thedis tance i s 44 kilome ters . But the Chines e text obv i ou s ly means the d is tance fromthe sou rce to the confluence . Ve ry like ly Klaproth has made the e las tic 11 too short.This does not matte r, for the principa l thing is to note , that the rive r originatingfromDamtchouk kabab is regarded as the main branch of the Brahmapu tra, whereasthe rive r coming from the l i ttle lake Djimagoungroung i s on ly a tribu tary . The

las t-mentioned is the one which was s tyled tome bymy three Tibe tan gu ides Chemayundung

-chu . The main rive r

,the one coming fromDamtchouk kabab or »the

Hors e rive r’s sou rce» , i s my Kubi-tsangpo .4 The Chines e text, trans lated by Klap

roth thus agrees exactly w ith my view,that the Kubi-tsangpo and not the Chema

yundung is the sou rce river of the Brahmaputra . Very like ly the re is a sma l l lakein front of the glacie rs fromwhich Chema-yundung comes , for the re is s uch a lakein front of the glaciers of the Kubi-tsangpo. This l i ttle lake of Chema-yundung(al ias Dj imagoungroung) is said to b e s i tuated eas t of the Langts ian kabab or

E lephan t’s mou th or the s ou rce of the Satlej . A glance at my pre l iminary map i s

1»Trans-H ima laya», Vol. I , p .

1 Ari is the Chinese name for Ngari-khorsum.

3 >Tran s -Himalaya» , V ol. I , p . 403 .

4 See Chapter XLII : »In Search of the source of the Brahmaputra», »Trans-Himalaya», Vol. II,p . 89 et seq .

94 THE CHINE SE GE OGRAPH E RS ON TH E SOURCE OF THE BRAHMAPUTRA .

s ufficient to show that i f such a lake exis ts , i t mu s t b e s i tuated eas t of the Ganglung glacie r or sou rce of the S atlej , for so far as I cou ld s ee the cours e of the riverChema-

yundung no lake was vis ible . Here i s a lso anothe r proof thatmy view abou tthe identi fication of the sou rce of the S atlej was correct, or, at any rate , in harmonyw i th the sound and unprejud iced opin ion of Chines e geographers .

This des cription is even be tte r than d ’Anv i lle’

s map of this regi on . H e s eemsso far to have mis unders tood the Chines e hydrography, that he has placed the nameYarou Ds ancpou ou Ts anpou R . along the rive r which corres ponds to Chema-

yundoung.

That th is is s imply a s l ip of the pen i s proved by the fact that the othe r, sou the rnbranch , corresponding to Kubi—ts angpo or Tamchok-kamba , really comes d irectlyfrom MM Tamtchouc . In the same region (Pl . VII) d ’Anv ille has a M . Cocoun

Kentchian or, as on his gene ra l map of Tibe t, s imply Cocoun M ., which , as far as

the s ituation is conce rned , exactly corres ponds with the Kubi-gangri .Further the text te lls us that the firs t cons iderable tribu tary to the Ye re—ts angpo

is the Naouk-dzangb o. The Maryum-chu is not even mentioned , and indeed, as

compared w i th the Kubi-tsangpo, i t is a mere brook. The N aouc Tsanpou R . is

entered on d’

Anv i lle’

s map and may b e identical with Nain S ing’s Chu Nago R .

The Gouyang ,which jo ins the Tsangpo fromthe Himalaya lowe r down , is probably

my Gyang-chu Iand KAWAGUCHI

s Kyang-chu .

2 Of the fol lowing rive rs men tionedthe Djochotghi choung is one of the rive rs fromToshot , the Dzaka dzangb o chu is

Chaktak-tsangpo , which can eas ily b e called the rive r of Dzaka or Saka-dsong.

The Dok tchou is my Dok-chu 3 or Raga-ts angpo, the Chang is Shang-chu and the

Djouangdze is Nyang-chu coming fromGyangts e .

After havi ng examined the exis ting mate rial Klaproth arrives at the resu lt thatthe Ya rou dzangb o tchou which had been regarded as the uppe r cours e of the

Brahmapu tra, was in real ity the uppe r cours e of the Irrawaddi . He bu i lds up his

theory on a mis take made by the Chines e authors z4

11 1 1 y a encore une autre ri v iere qu i pas se par l ’extréme front iere de Yun nan, c’

es t leP in lang k iang (fleuv e de l’Arec). S a s ource es t dans le Ngari ou Ari , prov ince du Tubet , al’

e s t du mont Gangd is , s ur le mont D amdjouk S kabab , c’

es t-a-dire bouche de cheva l . Ce

fleuv e reco it p lus b as le nomde Yaerou dzangb o ; i l coule généralement vers l’e s t, en dév iantun peu au s ud ; pas s e par le pay s de Dzangghe et la v ille de J'

I'

kargounggar6

The map accompanying his article is ve ry we l l made for the time .7 On the

fan tas tic hybrid of a river that is drawn on i t we find the follow ing names : Yaroudzangb o tchou R . ou Grand Kin-cha-kiang

,Fleuve a S able d ’

Or, Pin lang kiang on

1»Trans -H imalaya» , V ol . II , p . 8 8 .

1 Compare below,V ol. II .

3 »Trans -H imalaya» , V ol. I , p . 4 1 7 .

4 Op . cit. , p . 3 94 . With this shou ld b e compared the new trans lati on by E . H aenisch (p .5 Tamchok .6 This i s from an order is sued by Emperor Kang H i in 1 7 2 1 and bas ed upon information

given himby high T ibetan lamas .

7 Carte du cours inférieur du Yarou dzangbo tchou , par J . Klaproth , 1 8 2 8 .

CHA PTE R IX.

TH E NORTH ERN TRIBUTAR IES TO THE TSANGPO.

Al l the northe rn tribu taries of the Tsangpo , known to the E uropeans s inceNain S ing ’s journey , we re already marked on the Ta-ch ’ing map with the i r correctnames and to the i r ve ry sources . So far as I have had an opportun ity to con trolthe Chines e s tatements I .have found the ir draw ing of these rive rs ve ry incorrect.Some of thes e rivers may b e recogn is ed on earlie r editions of S t ieler

s H and-A tlas ,

for ins tance for 1 8 26 and 1 834 and on H . BE RGH AUS ’ map, where two of themrise north of the Indo-Tibe tan watershed. As a ru le they are much be tter on the

Ta-ch ’ ing map than on d’

Anv ille’

s .

Firs t Of the more cons ide rable tribu taries is the Naouk tchou repres entedas coming from the lake Dzang ri . This rive r has not been changed on d

An

vi lle ’s map, where it is cal led Naoue Tsanpou R , and the lake Sanc-li , whichmay b e suppos ed to b e iden tica l wi th the S en i t-tso , which , as I was told, shou ld b es i tuated somewhere he re . If the Chines e map we re correct , the N aouk tchoushou ld b e longer than the Kubi-ts angpo and the Chema-yundung. I cannot u tte rany Opin ion regarding the probabil ity of this rive r-cours e , as my rou te goes northof i t. But beyond doubt i t exis ts and is identical wi th Nain S ing’s Chu Nago.

Ryder has ce rta inly drawn it too short as all the othe r northe rn tribu taries , and

the Chines e map has probably made i t too long. In this part of the TranshimalayaI have , already , on my pre l iminary map, wri tten some notes of interrogation ,markingout that I do not know the geography of this particu lar region . I therefore amnot

able to draw any conclus ions fromthe orographical s tructure e ither.The text of the Ta-en

i ng-i -t

n ng-enin as re lated by DUTREUIL DE RH INS ,

s ays of this rive rLe N aouk tchou ou Lhab ouk dzang b o tchou a sa source au nord-es t (des Djochot)

dans le lac Dzang r i , coule d ’

ab ord 250 l i (55mi lles) a l ’oues t e t regoit sur la dro ite deu xr iv iere s qui v ie nne nt des monts D z iang ri ou Sang ri et Mou k rong. Au s ud , il regoit tro isautres ri v iere s qu i décou lent de s monts Tchang la tchouk , Dzoreng ri e t Yang b ou ri . Le

N aouk tourne au s ud , fait 80 11 ( 1 7 mil les ), reco it a dro ite les eau x des montagnes Yala et Ta

TA-CH ’ING-I-T’UNG-CH IH . 97

krong . Il coule encore 60 l i (1 3 mi lles) e t s e jette dans le Tsan po 21 3o h (6 milles ) dans les ud-oues t de s Djochot. 1

As a ru le the Ta-ch ’ing map becomes more and more unce rta in the moreone proceeds to the north fromthe Tsangpo. The refore the upper cours e of the

Naouk-tchou s e ems to b e exaggerated in lengt h. On the other hand, the TchaTchou (Tsa-chu), which is s ituated comparative ly near the Ts angpo , is remarkablygood , and very like the repres entation given to this rive r on Ryde r’s map . In bothcas es i ts ma in branch keeps almos t paral le l w ith the Tsangpo and rece ives s eve ra ltributari es from the north, N E . and eas t. The s ame may b e said of d

Anv i lle’

s

drawing of th is rive r.The Chines e text says»Le Ts a ou Tcha tchou e s t a 1 00 11 (or rather 200 11) a l’oues t de la horde des Saga .

S ix riv ieres de’

coulant de la chaine des montagne s de ne ige s appe lées Yoro gang ts ian se

d irigent 1 00 l i (22 mi lles) vers le sud , s e réun is sent e t forment le G iab lar tchou . Ce lu i-ci ,apres umcours de 50 11 ( 1 1 milles), recoit les eau x de tro is ri v ieres venant du nord-oues t e tdes monts Chang la tchouk ri

, Tzoreng r i et Yang b ou ri . Alors , le Giab lar tourne au s ud-e s t

et recoit deux ri v ieres : l ’une venant du nord-e st de la montagne Djedze goung gar tang, I’autredu sud -oues t du mont N iamri ; pu is i l prend le nomde Tcha tchou e t cou le 70 11 (I4 mi lles )au sud-oues t pour s e j eter dans le Ts an po.»

z

The Chines e river Man-ch

u is identica l w i th Nain S ing’s Minchu Sangpo,which is , of cou rs e , a lso on Ryder’s map, and which I fol lowed to i ts source at

Kilung-la , probably identi cal with the Mt. K irgang ts ian on the Ta-ch ’ ing map . I ti s , a lthough not improved, to b e found on d

Anv ille’

s map as we l l .Of this rive r the Chines e text says»Le Man tchou es t in l

oue s t de s Saga. Deu x riv iere s venant du nord e t des montsS ier tchoung e t Pi loung r i coulent pendant 200 ll (44 mi l les) au sud ; tro is autres v iennent del ’es t de la montagne Gang tchoung djadak ri e t une décou le de l’ouest , de la montagne Lakdzang djora ri . Ces s ix riv ieres s e réun is s ent et forment le Man tchou qu i cou le enco

’re 40 l i(9 milles) au sud et s e jette dans le Ts an po . Le mont Kiemla tchour mou es t

a 1 90 II

(40 milles ), le mont S ier tchoung a 1 10 l i (24 mi lles) et le mont K io our tchoung a 38 l i (9mi l les ) au nord-oues t de s Saga .»

Then fol lows one of the mos t important rive rs , Saga dzang b o , the length of

which is ve ry much exaggerated , whe reas its lowe r part is remarkably correct. The

great bend towards the wes t, before the junction with the Tsangpo, is to b e s een .

The Sa tchou comes in fromthe eas t . This is the tribu tary on the bank of whichS aka—dsong is s ituated . Sa-chu is its rea l name , the only name I heard for thisriver. Nain S ing cal ls it Chaka Chu . Above the junction w ith the Sa tchou the

main rive r is cal led Djadak tchou , in which we bette r recogn ise Chaktak-chu thanNain S ing’s Charta . S ti l l highe r up the rive r is cal led Lab tchou , coming from a

'

1 L’

As ie Centrale . Paris 1 889, p . 53 8 .

1 Ibidem, p . 54 1 .

1 3—1 3 1 3 3 7 1 .

98 TH E NORTH E RN TRIBUTARIE S To TH E TSAN GPO .

lake cal led Rab mtso . This lake is of cours e my Lapchung-tso and the rive r Labtchou is Lap

-chu , a lthough my gu ides a lso called i t he re Chaktak-ts angpo. From

the ri ght, or north , the rive r rece ives a tr ibu tary coming froma lake called Rargong,which I d id not hear of. But as this lake is entered on the Chines e map we mayb e sure that i t exis ts somewhere in this region . Wes t of the lower part of thisrive r is a Mt. Samoié , obvious ly my Samye-la . E as t of lake Rab mts o is a Mt.Loumbou , which sounds l ike my Lunpo-gangri , which may, perhaps , b e wri tten morecorrectly : Lumbo . The high range Lunpo

-gangri i s in real ity wes t and S .W . of

Samye-la and he re indeed we find a name which a lso reminds us of Lumpo-gangri ,name ly, Mt. Rong pou gangri . O ther names in the ne ighbourhood of the Chaktak

tsangpo are d ifficu l t to identi fy.

This rive r has been s eri ous ly spoi l t on d ’Anv ille ’s map. The great bend to

the wes t has disappeared a ltogether. But the two lakes are the re , Lap-chung-tsocalled Lio L .

, the othe r Raron hakon L . The rive r its e lf is cal led Sanki Tsanpou .

Then fol lows the Raka dzang b o, which correctly flows fromwes t to eas t,al though i t was very much in need of the improvement given i t by Nain S ing. In

i ts uppe r cou rs e i t i s cal led Raka Dzang b O tchou ; then it correctly rece ives the

tri bu tary fromAmtchok mts o, which is too big and too far north . On my pre l iminary map to Trans -H ima laya I have cal led the lowes t part of Raga-tsangpoDok-chu , and i t is inte res ti ng to find exactly the same name on the Ta-ch ’ingmap, name ly Dogh tchou . Two tri bu taries come down here fromthe north . The

eas te rn of them corresponds in reality with my M '

u-chu , b ut is called Ou tchou .

Be low i ts confluence the Dok—chu or Raga-ts angpo has got the name 01 tchou .

Then fol low two re lative ly short northern tribu taries : K ie tchou or Joung tchouand Rong ; tchou . The junction of the firs t men tioned is a l ittle be low Pinsol ing .

Bu t in real i ty this tribu ta ry is the Rung-chu and the next is Tanak-chu or Tanakpu

-chu . Final ly we have the Chang tchou coming from the N .N -W. ins tead of

N E . Naml ing is cal led Dz iang amring. He re two rive rs mee t. The one whichis drawn as a tributary to Chang tchou comes fromN .N -E . froma mountain calledMt D zo ri (Sok b ou ri), which shou ld correspond to the Khalamb a-la. Anothe rtri bu tary to the Shang-chu , Daghri phou tchou , is the longes t in the whole sys temof Shang, b ut whe the r i t exis ts or not cannot as yet b e s tated .

Of the sou the rn tri bu tari es Kouoyang may b e iden tical with my Gyang-chu ,es pecial ly as the re i s a place cal led G iangb en on i ts left bank . Chordi tchou isobv ious ly Nain S ing ’s Shorta Sangpo and Sab tchou is Shap-chu .

On d’

Anv i lle’

s map mos t of the las t-mentioned rivers have been s erious lys poi lt. The Raga-ts angpo, e special ly , is not l ike the real ity at all and in the lake at itswes tern end we hardly recogn is e the Amchok-ts o. Shang-chu is more l ike the re

pres entation on the Ta—ch ’ ing map. The names are a lways be tter on the Ta

ch ’ing than on d’

Anv i lle . For ins tance in the firs t cas e PhoumSoukling, in the

latter Ke ltan pousoucliamfor Pinsol ing.

CHA PTE R X.

TH E SOURCE O F TH E BRAHMAPUTRA ON TH ETA-CH ’ ING MAP .

At the time when DUTRE UIL DE RHI NS published his work on Central As ia , Isome revi ewers were rathe r unkind to himand did not think much of a work whichwas to such a great exten t founded on geome tri cal cons truction and calcu lation . Iagree with themthat the me thod is insu fficient and that on ly new exploration mayclear up the geography Of an unknown country. D e Rhins hims e lf is aware of thisfact, for he s ays : 2 »E n As ie Cen trale , l’inconnu cons e rve de te l les proportions q ’unedouzaine d

explorateurs a l’

Oeuv re pendan t umdemi-s iécle me s uffiraient pas aux

travaux prél iminai res d ’

une topographie s érieuse S ti l l,I th ink, L ’

A s i e Cen tra le

i s an admi rable work. It is fu l l of knowledge and erudi tion frombegi nn ing to end .

It i s s cien tific and cri tica l and d isplays a great amount of pe rspicacity . It shows no

end of patie nce and love for the work . It als o shows a great bibl iographic knowledge , although i t is su rpri s ing that he gives on ly the ti tles of s uch German worksas R i tter’s and R ichthofen ’

s , in which he shou ld have found so much information important for h is res earches . One reads the book w ith sympathy for i t wasde Rhins ’ preparati on for the great and beau ti fu l jou rney on which he los t his l ife ,and which afterwards has be en so we l l des cribed by F. GRE NARD .

The principa l work fromwhich de Rhins has found hi s materials is the greatChines e atlas of the dynas ty Ch ’ ing or Manchu , a work which conta ins also Tibe tand is founded on Chines e and Jes u i t inves tigation and afte rwards comple ted and

con trol led . It is called Ta-o/t’

i ng—i—t ’n ng -

y i i-t

n,and published 1 744

—1 756—1 76 1

1 862 . H e does not s e em to value the Ta—ch ’ ing map so much as i t is worth»La carte des Ta-ch ’ ing repre’ s ente la parti e occidental e du Ngari de la facon la plusincomplete et la plus inexacte On the othe r hand he has too great a vene rati on for the We i -tsa ng

-t’

n-o/zi/z of 1 792 , fromwhich some extracts are given in a

1 L’

As ie Centrale , Paris 1 889 .

1 Op . c it. , p . 589.

3 Op . cit. , p . 52 2 .

1 02 TH E SOURCE OF THE BRAHMAPUTRA ON TH E TA-CH ’IN G MAP.

noms de hameaux , de campements , ou de riv 1 eres que Na in S ingh indique ne s e

re trouvent pas sur la carte chinois e ; mais i l s ’accorde avec ce lu i-ci s ur la d is tancedes poin ts extremes .» The fact that Nain S ing’s names are not to b e found on the

Chines e map only proves that the Pundi t’s map is ri che r than the Chines e in thisres pect. But the fact that he , and after him Ryde r and Rawling, fol lowed the

Maryum-chu , whereas the two Chines e i tine raries men tioned su rveyed the s ou rce

branches from the S .W . explains the misunders tandings abou t the source of the

Brahmaputra , which re igned in E u ropean geography unti l I was able to clear up thehydrography. It a lso proves that the problemcou ld not b e s olved and the branchesnot even roughly su rveyed froma d is tance . It is almos t a consolati on for E uropeangeographers , that the confus ion preva i l ing in di fferent Chines e works is as great asin the E uropean . The d i fference between the We i-ts ang—t ’u-chih and the Ta

ch ’ ing- i-t ’uirg-

yu-t

u rs even greate r than be tween S ir R ichard S trachey and Na inS ing . The Chines e have , howeve r, neve r fal len into the mis take of call ing the

Maryum-chu the main branch of the Tsangpo. Some au thors give i t the same rank

as the southern branch, others do not even men tion it. Some cal l the Chemayu ndung a tribu ta ry , others make i t the ma in branch . The fol low ing is the Ta

ch ’ ing text as gi ven by de RhinsLe Tsan po es t formé de la réun ion de deu x riv i eres princi pales : le Tamtchouk ou

Tamdjamkamba (source du cheval) dont les s ources s e trouvent dans les monts Koub en gangts ian , 1 Goumang et Tamtchouk khabab

,et le G ianggh ia s umla ou D z ian dz ia tchou .

Le D z ian dz ia es t forme' de quatre ri v ieres qu i na is s ent au nord-oues t,dans les monts

Maryoung, Changou yarak ri , Chadz iar r i . E lle s cou le nt environ 1 00 11 (22 milles ) vers l’e st .Le D z ian dz ia, qu ’

e lles forment en s e réun issant, parcourt 50 l i ( 1 1 mi lles) au s ud. Il regoit a

dro ite le Tchar tchou qu i v ient de l ’oues t ; e t, a 60 l i (I 3 mi lle s) en aval, il s e jette dans lagauche du TSan po.

Le mont Koub en gang ts ian (Cocoun K en tch ian de d ’Anv i lle) e st a 258 l i (55mi l le s )a l

oues t des Djochot (Daksam). I l e s t tres élevé , couvert d ’

un énorme glac ier et fait part iede la meme cha ine que le mont Goumang et le Tamtchouk khabab qu i e s t s itué a 340 l i(75mil les) des Djochot.

Les monts Maryong, a 280 11 (62 milles) des Djochot , re l ie nt les Tamtchouk k hababaux monts Changou yarak ri s itués a 280 ll (62 mille s) au N .O . des Djochot .

This pas sage is extreme ly important and inte res ting, b ut i t has to b e we l lcompared with the Ta-ch ’ ing map and s ti ll more with my map of the country as itreal ly is . D iscuss ed alone i t leads to a misunde rs tanding which othe rwis e may eas i lyb e avoided .

Trans lated in an in te ll igible way i t runs as follows : The Tsangpo is formedby two principa l rivers : The Tamchok-kamba, or sou rce of the Hors e , the sourcesof which are s i tuated in the mountains of Kubi-gangri , Chema—yundung—gangri and

Tamchok-kabab , and the Maryum-chu .

1 Op . cit. , p . 51 7 e t seq .

1 Kubi-gangri !

TAMCHOK-KAMBA AND MARYUM-LA . 1 03

The Maryum-chu is formed by four ri ve rs which ris e in the N .W . , in the

mounta ins of Maryumand Transhimalaya . They flow some 2 2 French mi les to the

eas t. The Maryum-chu which is formed by the i r confluence flows 1 1 Frenchmi le s to

the south . I t rece ives fromthe right the Chema-yundung which comes fromthe wes t ;and 1 3 French mi les fu rthe r down i t throws i ts e lf into the left s ide of the Kubi-tsangpo .

We begin with the Maryum-chu . In the Chinese text only the mountains are

called Maryum, and in thes e mounta ins our we l l-known Maryum-la is s ituated . But

the ri ve r is cal led Dz ian dz ia tchou , or G iangghia s umla , which on d’

Anv ille’

s maphas been improved to Kiankia somla. Dz ian dz ia , G iang ghia and Kian kia are

d iffe rent deformations of the same Tibe tan word , whateve r its mean ing may b e .

[(yang ,wild as s , may ente r in the name . The sumla of the Chines e text means ,

in Tibe tan, the three pas s es . The name is thus something like »The kyang-rive r ofthe three pas s es » . One of themmay b e Maryum-la . But the two principal pas s es ,fromwhich the river gets the greates t part of its wate r, are , undoub tedly, s i tuated inthe Transhimalaya . Mode rn explorers comple te ly ignore thes e two pas s es , and s implydescribe the rive r as ris ing fromthe Maryum-la

, a vi ew which , as shal l b e shown ,was un iversally accepted even in E u ropean handbooks of geography . And s ti ll thed i fference be tween thes e three pas s es mus t b e very great. The Maryum-la

,a l though

extreme ly important as a watershed , is on ly a comparative ly low transvers e thresholdin a longi tud inal val ley , whereas the two othe r pas s es mus t b e much highe r andS i tuated on one of the principal cres ts of the Transhimalaya , and perhaps su rroundedby fie lds of e ternal snow . On Ryder’s map there is an indication of the brancheswhich form the Maryum-

chu , although the sources of the branches coming from the

north , may b e s i tuated—

fu rthe r north than he has marked . In the Chines e text i t iss aid that the fou r rive rs flow to the eas t . This is fai rly tru e for the one we know,

name ly, the brook fromMaryum-la . As to the three others theymore probably flows ou thwards or to the S .E .

, as repres ented on d’

Anv i lle’

s map. The Ta—ch ’ ing maphas on ly three source branches , b ut d’

Anv i lle has four. The mos t interes ting is ,howeve r, that the Chines e text, a lthough knowing the name Maryoung for the

mountains , does not cal l the rive r Maryoung-tchou , b ut derives its name fromall

three pas ses where the sources are s i tuated. It is a lso worth while to remembertha t the Chines e geographe rs have correctly noticed that this northe rn branch , whichfor wan t of a be tte r name

, so farmus t b e cal led Maryum-chu , comes frompa s s es

and not fromg angri s .

After the confluence of all four branches , the river, we are told, flows sou thwards as far as to a poin t where it rece ives a tributary coming fromthe wes t andcal led Tchar tchou . Which is this ri ve r ? As I have not been up the Maryum-

chu

I cou ld not te l l frommy own experience . Ryde r has two tributaries , both comingfromthe wes t and join ing the Maryum—chu . And s ti l l I be l ieve that the Tchar tchou(Char-chu) is nothing e ls e than the lower cours e of Chema-yundung , a confus ionwhich we recognis e ve ry we l l fromothe r maps , discus s ed above . Both on the Ta

1 04 TH E SOURCE OF TH E BRAHMAPUTRA ON THE TA-CH ’IN G MAP .

ch’ ing and on d’

Anv i lles map the two Chines e it ineraries are la id in. The northerni tinerary does not at any point, wes t of Lhats e -dsong, cros s the Tsangpo . But i tcros s es the Maryum-

chu and its tribu ta ry Tchar tchou before it reaches Gunchu -ts o.

Thus i t follows a road which runs s ou th of the ordinary road ove r Maryum-la . I tmus t b e the same road which was fol lowed by Kawaguchi . For he came fromGunchu-ts o and kept sou th of Maryum-la and »went on s eve ral low undu lating hills ,and final ly reached the lowe r cours e of the rive r Chema Yungdung» , afte r which hecros s ed the »Kobe i-chu» . Thu s he left the Chines e i tinerary on his le ft and had not

to cross the Maryum-chu . The Chines e it inerary cros s ed the Maryum-

chu , Kawaguchithe Kubi-ts angpo . But on the road to Gunchu-ts o both mus t have cros s ed the sameriver. And as this is cal led Chema Yungdung by the Japanes e , i t mu s t b e the samein the Chin es e cas e , which a lso corres ponds pe rfectly wi th real facts , so much the

more as there is no other S ign of a junction be twe en the Chema-yundung'

and

Maryum-chu on the map .

The firs t-mentioned of the two principal rivers , the Tamchok-kamba , has itssources in the mounta ins of Koub en gang ts ian , Goumang and Tamtchouk khabab .

The firs t of thes e is Kubi-gangri , which is ve ry we l l placed on the Ta-ch ’ing map .

The s econd , Goumang , corres ponds exactly to Chema-yundung-pu , i . e . the mounta inmas s ive fromwhich the Chema—yundung comes down

,and which i s probably some

times cal led Chema-yundung-gangri , a lthough I on ly heard i t cal led »pu» or »uppe rvalley» . Fromthe eas te rn s lope of the moun tain a rive r goes down to a l i ttle lake ,on the Ta—ch ’ing map called L . Dj ima Young rong, which also proves that Mt .Goumang mus t b e identical wi th the Chema-yundung-pu . The thi rdmountain , Tamtchouk khabab , has , cu rious ly enough, been placed north of Mt. Goumang and a

l i ttle tribu ta ry, pos s ibly the Angs i-chu , may b e sa id to ris e fromits eas te rn s lopes .

Thus Langchen-kabab and Tamchok-kabab become , on this map, ve ry clos e ne ighbou rs , and the source of the Hors e is regarded as s i tuated at the head of the rive rwhich i s farthes t wes t, and , probably , somewhat longe r than the other branches . In

this particu lar de tai l the di fferent Chines e ve rs ions do not agree . For in the vers iontrans lated by Klaproth it is sa id that the Tsangpo comes fromTamchok-kabab and

after a cours e of I O l ieu es rece ives a tribu tary fromthe lake Chema-yundung. Herei t is s pecially pointed out that on ly the head river comes fromTamchok-kabab .

With this d’

Anv i lle’

s map agrees for i t places Tamchok-kabab as be longing to or

be ing a part of Kubi-gangri , and Dutreu i l de Rhins says of the Tamchok-kambathat i t is xla branche principale du Tsan po» .

I

The mis take of the Ta-ch ’ing map has of cours e been accepted by d’Anv i lle ,for in the gene ral hydrographical arrangement both maps are practically like eachother. It is ve ry easy to find out how this mis take cou ld ris e , i f we only cons ide rthe itineraries on the Ta-ch ’ing map . The southe rn i ti ne rary has crossed the Kubi

1 Op . cit.,p . 51 7 .

1 06 TH E SOURCE OF TH E BRAHMAPUTRA ON TH E TA-CH ’IN G MAP .

Mandata wi th the Kubi-gangri , th e northe rn pierced by a problematic river whichis marked as a sou the rn tri bu tary to the Tage-ts angpo. I cou ld not te l l whe thers uch a river exis ts or not , as I never have been beyond the Ganglung-gangri . Asmal l tributary comes from the s outh no doubt, b ut I be l ieve i t comes fromthe

range which joins the Gurla wi th the Ganglung-gangri . If Webber’s map cou ld b etrus ted in th e leas t, which is not the cas e , his Dak E o pas s of fee t shou ldb e s i tuated exactly where Ryder has his transvers e va lley cut through the samerange . At the firs t s ight the Ta-ch ’ ing map s eems to corroborate the vi ew of the

exis tence of s uch a sou thern river, for i t repres en ts the Tage-ts angpo double . But

this argumen t los es all of i ts force as both branches are drawn as s i tuated north of

the Langchen-kabab or Ganglung-gangri . As Ryder fol lowed the northern rou te he

cou ld not poss ibly make out the orography a long the southe rn rou te , and much lessthe rive r-cours es and the i r val leys . Fromhigh s tations of triangu lation all the dominating peaks were vis ible b ut some times they may have been formed into rangesin a wrong way. But Ryde r has qu i te correctly left the qu es tion open by markingboth the Tage—ts angpo and its sou thern tributary w i th do tted l ines . Only a new

s urvey in the country S .E . of Gurla can so lve the problem.

The Chines e documen ts de Rhins has us ed for south-wes tern Tibe t cons is t ofthe Ta—ch’ ing map, some Chines e i tineraries (Wei -ts a ng

—t’n -eli i n

,trans lated by Klap

roth) and the general geography of Tibe t (part of the Ta -en

i ng-i -t

nng-e/zin trans

lated by the mis s ionaries in Peking and Klaproth). In his atlas de Rhins gives a

reproduction of the Tai-ch ’ ing map ; 1 on his map of cons truction , N zo 2 2 , he putsdown in red the i t inerari es of the Pundi ts . He tries to ge t rid of the e rrors of the

Ta—ch ’ ing map by means of trans formation , as indicated on his map N zo 2 1 (Pl . VII .)H is geome trical deductions and comparisons do not inspire much confidence in spiteof the e rudi te and hones t work he has s acrificed . He compares his own cons tru etions with the Chines e map to find out the di ffe ren t values of this . FromNainS ing’s map from Shigats e to Maryum-la he ge ts a control of the Chines e map ,which , howeve r, does not at all agree wi th the Ta-ch ’ ing text . The tex t is correct,the map wrong in dis tances .

In the following quotations from de Rhins , bas ed on the Chines e text 2 wefind that the Indus i s regarded as taking its ris e fromthe moun ta ins N E . of the

Gang dis ri (la montagne de cou leur de ne ige), which was transcribed Kaintais sé bythe mis s ionaries : »Dans le nord-es t du Gang dis ri s ’élév ent les pics Imarou , Patchoung et S enghé khabab .» I cannot identify Imarou , b u t Pa tchoung and S enghé

khabab are he re gi ven exactly as they are pronounced by the Tibe tans . »Le S enghé

khabab gang ri mon tagne ne igeus e de la source du Lion es t a 3 60 l i (80mi lles )au nord-es t de la ville de Goughé djachi loumbo ; e t au nord-es t du Gang dis ri .

1 M ap N zo 2 0.

1 Op . cit., p . 51 8 e t seq .

THE SOURCE S OF THE GRE AT IND IAN R IVE RS. 107

A ins i que son noml’

indique , les ru is s eaux qu i en décou lent a l’oues t donnent na iss ance au S enghe tchou ou a l

Indus .» A description of the source of the Induscou ld not b e given in a shorte r, more precis e and more correct way, showing thatthe Chine se had a much clearer know ledge of this part of Tibe t than some E u ropeanseven so late as the beginn ing of 1 900. The Chines e text clearly s ays that thesource of the Indus is s ituated in the mountains N E of Ka i las and not on the

Ka ilas its e lf. As to the source of the Satlej the text says : »Le Lang ts ian khababgang ri , a 250 l i ou 55 mi lles nord-es t de Tak la kar; le Lang tchou y prend sa

source .» This orien tation w il l coincide w ith the M . Lantchia Kepou of d’

Anv ille’

s

map , which is also perfectly correct . On his cons truct ion map ‘ de Rhins has repre

sen ted eve rything from Chines e sources in black , and eve rything fromthe Punditsin red and has tried to bring thes e two pieces of information into harmony w itheach othe r. H e arrives at the resu lt that the Lang ts ian kabab is s i tuated e xactlywhe re I s tated i t to b e . The Some tchou (Samo-tsangpo) comes fromL . Gonghioud ,

which may have been the cas e in thos e days , although it is not at all probable , thelake be ing salt . But as I have not been at the place I cannot add to the solution of the

problem. But even i f the Samo-ts angpo in olden times came fromthe Gunchu-tso i tcou ld not b e cal led the source of the Satlej , so long as the Chines e and Tibe tansplaced the source of the rive r at Langchen-kabab , or the source of the Tage—tsangpo .

The on ly mis take in this connect ion committed on the Chines e map i s that Tagetsangpo is repres ented as a left tributary to the Samo-tsangpo, which , of course , hasnever been and can never have been the cas e .

On the cons truct ion map S eng ghe kabab is very correctly placed NE . of

the Ka i las . For this mountain de Rhins has the names : Ka in ta is e , Gang dis ri ,Aneouta or Ka i lasa , all except the mos t common one which is Kang—rinpoche .

Thus Du treu i l de Rhins has no Opin ion of his own regarding the hydrographicalproblem, he on ly quotes native reports .

1 N zo 2 2 : Thibet sud-occidental , Carte de construction .

CHA PTE R X I.

SOME DESCR I PT IONS OF TH E MANASAROVAR AND

ITS SURROUND INGS.

E . HAE NI SCH has trans lated and publ ished a short Chines e des cription of the

sacred lake and the sources of the great rivers under the fol low ing heading : AmTage hs in-ssu e rgi ng e in Ka is e rl iches E dikt an die Gros s s ekretare , H ii lfs s ekretareund die neun Min is te rien , we lches s ie beauftragte , genaue E rhebungen tibe r dieNamen der s ii dwes tlich der aus s ers ten Grenzen ge legenen Berge , FltISs e und Orts chaften anzu s tellen .

I The pas sage runs as fol lows :Dann g i bt es in demGeb iete v on JOn-nan (Yunnan) noch e inen Flus s namens B in-lang

giyang (Pin-lang-ch iang). S e ine Quelle e ntspringt auf demGeb irge Damju k ’abab (Tamchokkabab) Os tl ich v on Gangdis e in (der Prov in z) Ari . Das Wort bedeute t P ferdemau l . D ie Quelle ,welche dort herv ors trOmt

,fuhrt den N amen Jaru-tsangb u-Flus s . D ieser F lus s kommt v on Suden

und fliess t dann,indeme r e inen Bogen macht, in Os tlicher R ichtung durch d ie Lander Tsang

und We i an der Stadt Zi g’

a gungg’

ar vorbe i , worau f er s ich mit demGaljoo muren (Fluss e)vere inigt. Dann richtet er s e inen Lauf w ieder nach Suden , durchs trOmt das Land des GunghoS tammes und betri tt in C u-yung jeo (Ku -yung-chou) das Geb iet v on JOn-nan, wo er den N amenB inlang giyang (Pin-lang-ch iang) fuhrt. Dann tritt er aus demPas s e Tiye i

-b i (T’ ieh-p i) h inaus

in das Re ich Miyan (Mie n) e in . ImS tiden v on Gangdrse l iegt e in Berg mi t N amen Langci yan k ’abab (Langche n kabab). Das Wort bedeutet E lefantenmaul. Dort kommt e ine Quellehervor, we lche e inen Te ich Mampimdala i durchs trOmt und dann auf i hremwe iteren Laufe inden See Langk ’a noor e intritt. D as Was ser d ieser be iden S een ergi es s t s ich in we s tl icherR ichtung in das Land Sangnan . ImN orden v on Gangd ise l iegt e in Geb irge , welches denN amen Sengge k ’abab tr 'agt. Das he is s t LOwenmaul. Dort ents pringt auch e ine Que l le , welchein we stl icher R ichtung in das Land Sangnan flie s s t . D ie s e be iden (letzterw '

aihnten) Gewas servere in igen s ich dort in e inemLaufe , welcher nach Suden ge ht, dann w ieder e inen Bogen nach

Os ten macht und imLande N akra S umdo s ich mit demFlus s e vere inigt, welcher v on demGeb irge Mabgiya k 'abab wes tl ich v omG

angdi s e kommt. D er N ame Mabgiya k ’abab (Mab

gya-kabab) bedeutet Pfauenmaul. D ies es Gewas ser, w e lches nach Sude n s trOmt , vere inigt s ich

1 Bruchs tticke aus der Ges chichte Ch inas unter der gegenwartigen Dyn astic v on E . Haenisch .

I . D ie E roberung v on T ibet, aus dem >Fe ldzug gegen die Dsungaren» aus zugswe ise fib ersetzt.

T’

oung pao V ol . XII . Le ide 1 9 1 1 , p . 4 1 5 e t seq . As the text that has been used i s M anchu the

names al so appear in the Manchu transcription . The origi na l Chinese or rather T ibetan denominations have in the quotation be low been added w ithin brackets by Doctor A lbert Herrmann , so far

as it has been poss ible to identi fy them.

1 1 0 SOME DE SCRIPTIONS OF THE MANASAROVAR AND ITS SURROUNDINGS .

It is curious that the fres h—wate r lake Manas arov ar cou ld b e s uspected as be ingrespons ible for the Tibetan borax . The borax was known to come fromTibe tanlakes , and as the Ma

-

pin-mou-ta-lai or Manasarov ar was more famous than all othe r

lakes in Tibet it was particu larly men tioned as the place fromwhere the boraxwas brought .

In Klaproth ’s trans lation of the Grea t Imper i a l Geograpny the Maphamdala i and its we s te rn ne ighbour are des cribed thu s

Le mot maphams ign ifie , en tub etain , ce qu i surpas s e tout , dala i , en mongol, es t mer.

Les H indoux appe llent ce lac Manas sarov ar. Il es t a 200 l i nord-es t de la v i lle de Dakla de

la prov ince de Ari , e t a 65 l i sud-e s t de la montagne Gangd is 1 1. Il e st nourr i par les eau xqu i décou lent de la montagne de ne ige , nommée Langts en kabab gangri , s ituée au s ud-es t ,

e l le s coulent nu nord-oues t . et s e jettent dans le lac, qu i a 1 80 l i de c ircu it. La cou leur d es es eau x es t verte

,e t leur gout pur e t dou x . Apres mid i , e lles prennent d iflére ntes cou leurs

,

et réfléchi s s ent une lumiere s emb lab le a celle des éclairs . Ce lac es t en fermé , des quatre cOtés ;

par des montagnes qu i en forment comme de s portes . Les gens du pay s cro ient que , pourpu iser de s es eau x , i l faut néces s a irement pas s er par une de ces portes . La premiere porte

,

nommée en tub eta in Arab ko , ou la porte occ identale , es t a l’

oues t du lac, e t au sud du

courant d ’

eau qu i le joint au Lang mths o. La s econde e s t ce l le du nord ; e lle s ’appe lle Dadzanloung ; e lle es t au nord de ce courant. La tro is ieme , ce lle de l ’es t, porte le nomd e Tou igo

char ; e lle est au sud de la ri v iere D s iema’

r’ tchou , qu i v ient du nord-es t et s e jette dans le lac.

La quatriéme , enfin,e s t ce lle nommée Gh iour go, ou la porte du sud ; e lle est sur le bord

me’ r id ional du lac, e t au nord-oues t de la montagne de ne ige Manak n il gangri .Le Lang mthso , ou Langga mthso, c’es t-a-dire le lac du boeuf, e st a 1 70 l i au nord de

Dakla,e t 30 l i s ud-oue s t de la montagne Gang d is ri . II a p lus de 300 11 de c ircon férence ,

e t regoit les eau x du Maphamdala'

r‘

, par le courant duque l on v ient de parle r. L’

A ltan go] ,

on la riv 1ere d '

or, nommée aus s i Menghe djas s on altan gol, es t formée par la fonte des ne igesdu Gang d is ri , cou le env iron 20 l i au sud-oues t, prend apres la d irection du sud , et s e jette ,apres 40 l i , dans l ’angle s eptentrional du lac. Du Lang mthso , sort la riv iere Lang tchou , oudu boeu f

, qu i forme , avec le La tchou , la riv 1e re S etledj , laque lle v a rejo indre l’Indus . Le La

tchou,ou l

eau d iv ine , a s a s ource dans la montagne Sengghe Kabab . L’

eau du lac e st de

cou leur tirant sur le no ir ; e lle a un gout agréab le . Il s’

appe lle che z les Hindous Rav anhrad .

Klaproth give s the descrip tion of Gangdis ri or Kai las and its s urroundings ,comple ting partly w ith his own words , in the following way

E lle e s t entourée par d’

autre s p ics tres—éle vés , ma is e lle les surpass e tou s de p lus decent tois es . E lle e st couverte de ne ige s et de glaces perp étue lle s qu i y forment des bancse scarpés , e t lu i donnent un éclat b lanc e t re s p lend is s ant. D e son flanc cou lent un grand nombrede sources et de torrents , qu i s e réun is s ent a s on p ied , dont l’accés es t tres-d ifli cile , ‘

a ra isondes hau ts rochers e t des crevass es imme ns es qu i I’entourent . E n s e d irigeant de la Ch ine au

sud-oues t, on v a toujours en montant jusqu ’

a ce tte montagne , qu i e s t le po int le plus élevéde tou te la cha ine dont on s u it la d ire ction . Ce tte chaine

'

s e s ubd iv is e e n plus ieurs branchesqu i travers ent le pays dans des d irections tortueu s e s . Au nord-oues t du Gangd is ri , e s t la

c ime ne ige us e appelée Sengghe’ kabab gangri . D

autres montagnes couvertes de ne ige s perpétue lle s qu i , pour cette ra ison , portent e n tub etain le nomde Gangri e t en mongol ce lu i deMous soun , entourent la prov ince d '

Ari , s ur une étendue de 11, e t entrent, vers le nord-es t,dan s le pay s de Kats i . Du cOté nord -es t du Gangd is r i , sont les nrontagne s appe lées Dabr ics i erké , Ghioouke

n iants ian tangla, Samtan gangdza, N omkhofrn ombach i e t Ba'

r‘

n khara, qui

THE GRE AT IMPE RIAL GE OGRAPHY . I I I

entourent la prov ince de Ou i , du cOté du nord, et pas sent par le s sources de Houang ho et

le lac Khoukho noorAnothe r ma in branch of mountains s tart s fromGangdis-ri to the S .E .

, beginn ingwith Tamchok kabab gangri and othe rs and con tinu ing the whole way to Y '

unnan

and S zt’

rchuan . Klaproth goes on to s ay that according to the commentary of Shu iching, which is a phys ical geography, thre e rive rs begin S .W . of Mount Oneuta

(A-nu

-ta)-shan , and form the Heng-shu i or Ganges . In his his tory of Fu-nan K’ANG

T’AI s ays that the Ganges comes fromthe farthes t north-wes t ofmount Kwen-lun ,and that i t has fiv e great sources . Fromthe mounta ins which formthe boundaryof Hindu s tan begin the rive rs Lang-chu , La-chu and Mar

-chu , which firs t flow to

the wes t and then return to the S .E . to formthe Ganges which takes its d i rectionto the sou the rn s ea . Klaproth points out to this : Ici les auteurs chinois confonden tle Gange avec l’Indus , and continues

Il parait donc que le mont Oneuta e st la meme montagne que le Gangdi s 1 1 de nos

jours . Oneuta e st le nomque le s l ivre s bouddh iques lu i donnent , tand is que les anc iens auteursch ino is l

appellent Kuen lun . Ma is toute s ces trad itions sont tres -peu sfires , et l’

élo ignementdes l ieu x rend une recherche approfond ie sur cet obje t d ifli cile . Cependent la nouve lle e

'

d i

t ion du Tha i ths ing y thoung tch i 1 , rapporte qu ’

e n 1 783 , Kh ian loung d is ait aux Grands de sa

Cour, que , d ’

apres les l ivres fan ou sans krits , quatre grands fleuv e s prenaient leur origine s ur

le mont Oneu ta, au p ied duque l se trou v ait le lac du meme nom. Quand on exanrine cett econtrée , ajouta l ’empereur, on vo it que l’Oneuta es t la meme montagne qu i , a prés ent, portele nom tub etain de Gangdis ri , c

es t-a-d ire Orig ine de toutes le s montagnes e t r i v ieresManak ni l gangri , montagne de n e ige de la d iv inite’ Manak n il, qu ’

on cro it y hab iterE lle forme le bord s ud-oue s t du lac Mapamdala i , ou Manas sarov ar.

The re are a lso descrip tions of the ord inary fou r »kababs » and of Koubungangts ian ri , which borders upon Tamchok kabab gangri , very high , and crownedwith an enormous glacie r ; the latte r is obvi ous ly Kubi-gangri . Mar young la or

Maryum—la is ment ioned as a branch fromthe Ka i las . 2The same geography of Tibe t had a lready be en trans lated by P . AM IOT and

published in the Vol. XIV of the M emoi rs of the mis s ionaries at Peking, Paris 1 789 .

I ins ert what he says of Lake Ma-pin-mu-ta-lai-chih , only by way of an example ,

p . 203 :

Il est a deu x cens lys nu nord -e st de la v i lle de Ta-ko-la, du département des Nga-ly,a so i xante-c inq lys au sud-est de la montagne Kan-ty

-see 3. Il e s t formé par les eaux qu i

cou lent du p ied de la montagne Lang-ts ien-ka-pa-pou -chan 4. Ce lac passe pour etre la source

princi pale du Gange , c’

e s t pour cette ra is on qu ’

on l’

a honoré du titre de Ta-la i-tché 5. Il peu tavo ir cent quatre-v ingts lys de tour. S e s eaux sont propres , douces s alutaires

,quo ique d ’

une

1 Ta-ts ’ ing-i -t’ung-chih .1 Description du S i dzang ou Tubet, D ’apres la grande géographie impériale de la Chine et leDictionnaire géographique de l

As ie centrale , publ ié a Pékin , en 1 7 75. Magas in A s iati que , Tome II,N zo IV . Publié par M . J . Klaproth . Paris 1 8 2 8 , p . 209 et seq.

3 Kang-ti-s su s han .

4 Lang-Chien ko-pa

-po shan .

5 Ta-lai-ch ’ ih .

1 1 2 SOME DE SCRIPTIONS OF THE MAN ASAROVAR AND ITS SURROUNDIN GS .

cou leur t irant sur le vert. E lles prennent d ifi'érentes cou leurs , su ivant la d ifférente élévat iondu sole il s ur l’horison ; mais vers le mi l ieu du jour, e lles bri llent de toutes les cou leurs ens emb le

,

réfléch is s ent une lumiere s emblab le a ce lle des ecla irs . Ce lac e st ferme' des quatre cOtés

par des montagnes qu i en s ont comme les porte s ; cc qu i a donné l ieu au d icton de s gens dupays : Pour en avo ir de l ’eau ,i l faut entrer par le mil ieu des portes , c

es t-a-d ire que , pourpu is er dans la s ou rce meme du Gange , i l faut franch ir quelqu ’

une de ces montagnes .

And of the Lang-ka-ch ’ ihIl es t a cent so i xante -d ix lys au nord de la v ille de Ta-ko-la

,a trentequatre lys au

sud-oue s t de Kan-tyo s ee

1. Sa largeur la p lus grande e s t d

env iron tro is cens lys . C’

e s t un de spremiers bass ins du lac Ma-p in-mou ~ta-lai-tche 1

, dont le s eau x , apres avo ir cou lé que lque tempsvers l’ouest , s

amas sent forment le Lang-ka-tché 3 . D e ce lu i-ci s e forme la ri v iere qu i coulevers l ’oues t sous le nomde Lang-tchou-ho 4

. L’

eau de cc lac es t tres -s aine ; e lle es t de cou le urt irant sur le no ir.

In Amiot ’s ve rs ion the Ku-pén—kang—ch ’ ien s chan or Kubi-gangri is thus

des cribedE lle est au sud-ou est de Tchouo-chou-té-pou-lo, dont e lle es t eloignée de deux cens

cinquante lys . E lle es t cont igue a la montagne du Cheval. E lle es t tres -haute ; comme e llea p lus ieurs po intes in égale s qu i s ont presque toujours couvertes de ne ige , e lle pres ente umobj e tdes p lus agréab le s a vo ir.

Final ly the fo llowing is an extract fromRockhill’

s important article on the

geography of Tibe t derived fromChines e sources : 5Kang

-t i—s sd s/zan . NE . of the Ngari K ’asumd is trict (O-l i) of Central T i bet. Its

circumference i s over 1 44 11. On all s ides of it ris e ridges and peaks, the h ighes t in T i bet,

and great mass es of snow hang over the ir edge s . On the summit of the mounta in are manys prings , wh ich all flow into a depres s ion

,and there the water rema ins . Th is i s unque s tionably

the greates t of all mountains . In Sanscrit books it is called the A-o-ta (Anav atapta)mountain .

Ta -ma-e/zn -ko-pa-

po s/zan : The mounta in res embles in s hape a hors e (ta , hence i ts name).Ta-mu-chu-ko k

o-

pa-

pu ,in the I -t’ung—ch ih , S .W . of Cho-shu-t

é , 340 11, near Men-na-ko -n ir

s han. and fac ing Ka ilas on the S .E . It i s one of a grou p of four h igh mounta ins . The Yarats ’ang-po flow s from the E . of th is mounta in . Lang

-ol den ko-pa -po s/zan : The mounta in i ss haped l i ke an e le phant (lang ch

é, hence its name ). Sneng-t o k

o-pa-

po s/zan : The mounta ini s s haped l i ke a l ion (sengge, hence the name). M a -po-cnia ko-pa-po s lzan : The mounta in i ss haped l i ke a peacock (ma—ja , hence its name). All the s e (four)mountains are conterminou s tothe Kang-t i-s sr‘i s han . The total length (of th is cha in) is over 800 11, and is called A-li ta s han .

Ts’

ang c/zi ang ,also cal led the P o olm. 6 It has three sources . One flows out in three

channe ls , and falls into the Po chu ; the s econd comes out of a cle ft in the Kang-la s han , and

also flows into the PO -chu . The th ird comes out of Ln -ma l ing, enters the Wu-s su’

ch iang,and flows into the PO—chu . The waters Of th es e three r i vers hav ing met, flow on in a mightymass , and thos e who want to cross it to go to Lh

asa have to pas s it in wooden or h ide boats .

1 Kang-ti-ss ii shan .

1 Ma-pin-mu-ta-lai-ch ’ih .

3 Lang-ka-ch ’ ih .

4 Lang-chu-ho .

5 Journa l of the Royal A s iatic Society, 1 89 1 , p . 255 et seq .

6 Rockhi l l identifies Po chu with Bod ch’

u or R iver of T ibet, generally cal led Ya-lu ts ’ang-pu

ch iang in Chine se works .

Map of Tib et , fromthe work edite

CH INE SE MAPS OF TIBE T . 1 1 3.

The Manasarov ar is men tioned unde r the name Of A-o-ta chih or Lake Anavatapta , s ituated to the south of Kang-ti-s s fr or Kai las . Rockhi ll te l ls us that theI-t’ung—chih mentions , among the lake s of Tibe t, »Ma

—pin

-mu ta—la , 200 l i N E . of

Ta—ko-la in Ngari , and 65 l i S .E . of mount Kai las . It is the source of the Ganges .

and is ove r . 1 80 l i in extent . Lake Chi-ka, 1 70 11 N . of Ta-ko-la and 34 l i S .W .

of the Kai las . It is ove r 300 l i in extent , and is connected w ith Ma-pin-mu-ta-la .

»

The Chi-ka is the refore our Langak-ts o .

Rockhill i l lus trates his important article wi th a Route map of Ulterior Tibe t,

.

taken from Hs i-chao-t ’u-l ij eh , of which Pl . VIII is a reproduction . Unfortunate ly itdoes not reach sufficiently far to the w es t . But Kang-ti-s s ir come s in , and is indeeddrawn

.

as the highes t mountain of Tibe t . We s t of it is La-ta—ko (Ladak), south is .

A-l i (Ngari ), and N E . i s S a-ko (Saka). To the north is a S alt Lake , b ut the re is .

no Tso-mav ang . The Kai las is correctly shown as s i tuated sou th of the Transhimalaya . In two cas es the tribu taries are in connection with each othe r, forming rings .

Final ly Pl . IX is the reproduction of a map for which I amindebted to

Profes sor E douard Chavannes who accompan ied it with the following words : »Je vous .

commun ique le fascicu le contenan t de s cartes , de S i tsang t’

ou k’

ao’

Ca rte s avecnotes cri tiques sur le Tibe t ’ . Ce t Ouvrage a été rédigé en 1 8 86 par Houang T

ei

k ’ iao.» The hydrography of this map has Obvious ly been de rived fromthe samesources as d

Anv ille’

s map of 1 73 3 , of which a reproduction w il l b e found in v o

lume II I of this work (compare a lso Pl . LI be low). In the we s t of the map we

find how the Ganges i s formed by two branches , afte r which the join t rive r in a

sharp bend turns eas t and sou th-eas t. The northern branch is in rea l ity the uppe rIndus , the s outhe rn is the S atlej . Be tween both , near the i r junction , is a lake whichon d

Anv ille’

s map is cal led D sak ion Somtou . The S atlej branch comes fromtwo

lakes , on d’

Anv ille’

s map Mapama Tala i and Lanken , or Tso-mav ang and Langaktso. Other great re s emblances wi th d ’

Anv ille’

s map may b e found , for ins tance inthe repre s entation of Tengri-nor and Jamdok-tso .

1 5—1 3 1 3 8 7 1

CHA PTE R X I I.

TH E ‘ SH U I-TAO-T I-KANG ON TH E SOURCE OF TH EBRAHMAPUTRA AN D TH E SATLE J .

In the S/i n i—tao-ti—t a ng or Ou tl in es of Hydrograp/zy , Book 2 2nd,compiled

by CHI CH AO NAN in the 26th year of Emperor CH IE N LUNG ( 1 762 A . we find

the follow ing des cription of the s ou rce of the Brahmapu tra, which was trans lated forme by Profes sor OGAWA during my s tay at Kyoto .

I Under the heading : Wa ters

of H s i—ts a ng » or Tibet , the author says»The Ya -ln -ts

a ng-

pn-o/zi a ng i s the Ta -

o/zi n -sna -

oni a ng Great Gold S andRiver). It is suppos ed by some geographers to b e the P a -

ou-cli na n of olden times .

S ome cons ide r i t to b e the H ei -s/zu i (2 Black Wate r) in V iz-ka ng , b ut it is too

far s ituated . Its sou rce s come out fromTa -ma -onn -éo—t ’

a-

pa-

pn-snan 2

s tand ing at

a d is tance more than 340 11 northwes t of Cao-s/i n -te‘3 tribes in the w es te rn frontie r

of Ts ang . [The north-wes tern part of the mounta in is near the sou rce of the LakeM a

-

fii n-ma -ta -la ‘l in the L ang

-on

i en -n’

a-

pa-s/zan .

5 This mounta in (Ta -mac/znré -t

a-

pa-

pn- s/zan ) is s i tuated 300 l i S .E . of Kang

—t i—s sa—saan (Ka i las Mountain). The mountain is ve ry high and great

,and its formres embles a hors e , whence

it is named . The source of the Ya -ln -ts a ng-

pn-o/zi a ng

6 is 3 5° W . (of Peking) and

(the altitude of the north) Po le 7 There are three sources , all of which flownorth-eas twards and un ite into one rive r. The river flows at firs t turn ing eas twards ,then sou th-eas t wards for more than 200 l i . A s treamof fore mounta ins of K i t

-mak a ng comes from S .W . to mee t the rive r. The rive r now turn ing to the northeas t for 1 00 l i , rece ives the Cai ang—oni a -s n-ma -la -no which flows sou th-eas twards from

1 To avoid misunderstand ings I give Ogawa’s trans lation l iterally , as I got it fromhim.

1 Tamchok-kabab .3 Choshut.4 Mapam-tala

,Manasarov ar.

5 Langchen-kabab-mounta ins .

6 Yere -tsangpo-river.

7 Professor Ogawa te l ls me that the passages within brackets J in the original Chinese text aregiven with smal ler characters as annotations or expl ications to the geographical names mentioned inthe text.

11 1 6 THE SHUI-TAO-TI-KANG ON THE SOURCE OF THE BRAHMAPUTRA AND THE SATLE J .

Furthe r eas t for 30 l i the M a nm/zn-ts a ng-

yfin-no comes from north tomeet the rive r.

[The ri ver M an-c/zn -tsang

-

pn , otherwis e cal led the M a-o/zn-tsang u—p i—la , l ies inthe sou th-wes t of the Sa-ko tribes . Two s treams flow for more than 200 l i sou thwards fromH s i en-ér/z-onnng

-saan and P i e/z-lnng—snon on the north . Afte r rece ivi ng

three s treams coming fromK a ng-oknng—cna -ta -t o-saan on the eas t, and further one

from La -éo-ts a ng—ono-l i -s/i a n on the w es t , the rive r becomes the M an

-onn -tsang

pn-li o

,and flows south-eastwards for 5o li before ente ring the Ya-ln-ts a ng

-

pn R iver.]The rive r flows further eas twards for 60 l i , and rece ives a s tream coming from

sou th . Fu rthe r eas t for 70 l i a s treamcomes fromA -l i -ts nng—onéng on the sou th

to meet it . Furthe r eas t for 80 l i , i t turns to the north-eas t, and me e ts the S ant o

ts a ng-

pn-no which comes fromLake L a

-

pn on the north-eas t , and which at firs tflows south-w es twards b ut afte r gathering s everal s treams turns sou th-eas twards to enterthe rive r. [The river S a -t o-ts ang

—pn , othe rwis e cal led S a -e‘rn-éo-p i -l a , l ies on the

sou th-eas t of S a -t o Province . Its sou rce is Lake La -pn on the north-eas t , and at

firs t flows unde rground, and appearing on the su rface , it has a s inuous cours e to

wards sou th-wes t for more than 400 l i . A s treamcomes fromP i -fiu-ta-éo-la—éo-s/zanon the north , and taking a south-eas terly cours e and rece iving a s treamfromLa

lnng-a

-e‘

r/z-s/za n on the north-eas t, flows sou thwards to meet the river. Flowingfu rther sou th-we s t for 1 50

—1 60 l i , anothe r s tream comes fromnorth-wes t, and further1 70 l i south-wes t a s treamfromLa -

pn-éang

-e/znng

-saan on the north-eas t. Further

1 00 l i south-wes t , a s treamfromL nng-t an -

s/zan comes fromthe wes t to meet theriver. The river now flows sou thwards for 60 l i , then turns sou th-wes twards , and

rece ives a smal l s treamfromthe north ; further to south-wes t it rece ives anothe r fromthe wes t. The rive r then takes a sou therly cou rs e , and afte r rece iving two s treamsfromKa ng

—lung-s/zan and two fromC/zao-y n

-la -ennng

-saa n on the north-eas t, turns

to the s outh-wes t , and rece ives fu rthe r a s treamcoming fromCli i -pn-lnng

-saa n on

the south-eas t w ith another fromthe eas t . The rive r flows fu rther for 80—60 l isouth-wes twards , then south-eas twards . The river is called the S a -éo-tsang n

-no fromhe re down to the conflu ence w ith the Ya -ln -ts a ng

-

pn R iver for 1 50—1 60 l i . The

length of the river is l i long .»1

Here aga in , in the description of the rive rs of Hs i-tsang or Tibe t, the sourceo f the »Yaru-tsangpo

-chiang» or Brahmaputra is placed in the »Tamchuk-khapap

shan » . We are told that the north-wes te rn part of the Tamchok-kabab mounta in isnear the »Langchien

-khapap» mounta in , which , in this description is cal led the sou rce-of the lake Mapam or Manasarov ar. This is also interes ting becau s e it does not

leave any doubt as to'

zo/zi en affluent to the lake the Chines e text regards as the

s ource or main feede r of the lake , name ly Tage-tsangpo. The s tatement is im1 According to Rockhil l Chi Chao N an was also one of the principal editors of Ta-ch ’ ing-i

t’ung-chih ; Shu i-tao-ti-kang was , accord ing to Rockhill , written in 1 776. The resemblance betweenthe two works depends , there fore , on the fact that the author i s the same , or that he , in both cases ,h as used the same material .

DISCUSSION OF TH E CH INE SE H YDROGRAPHY . 1 1 7

portant, as the origi nal Chines e author has be en at the place . Tamchok-kabab iss a id to b e 300 l i S .E . of the Ka i las and is ve ry high and great . The latitude isgiven more than one degree too far south .

When it i s s aid that »there are thre e sources » , one shou ld at firs t think of

Kubi-tsangpo , Chema-yundung and Maryum-chu , b ut this is not right, for with »there»i s meant the as tronomical ly gi ven source of the Ya-lu-ts ang-pu-chiang , and then itis sa id that all three sources flow north-eastwards and un ite in to one rive r

,which is

the Kubi-ts angpo or uppe rmos t Brahmapu tra . At firs t s ight this s tatement s e ems tob e correct . ‘ For, indeed , the three principal sou rce branches of the Kub i-ts angpocome out of three mighty glaciers , of which the one farthes t wes t is double , a lthoughi t gives ris e on ly to one s tream, name ly the sou rce of the Brahmapu tra . But in

real ity the Chines e text does not at all mean the three sources of the Kubi . Itmeans the Angs i

-chu , Chema and Kubi , which w e have found joined much too earlyon the Ta-ch ’ ing map . But the Chema s eems to b e regarded as the principal source .

The next pas sage of the Chines e text te l ls us that the rive r »at firs t» turnseas twards and then south-eas twards , of which the firs t is true for the Kub i-tsangpo ,and Chema-yundung the s econd for the Ts angpo . Fromthe S .W . a tri bu tarymee tsthe Ye re-tsangpo and comes froma mounta in cal led Kumu-gangri or some thing likei t. As this tribu tary mus t b e identi ca l w ith the Kubi-tsangpo , it may b e that thisri ver i s not included in the three source branches men tioned above . Then Tamchokis obvious ly placed at the head of Chema-yundung , as appears clearly fromthe Ta

ch’ ing text , where Koub en gang ts ian is placed at 258 l i and Tamtchouk khababat 340 l i wes t of Djochot . The Shu i-tao-ti -kang , which has the same d is tance ,makes the bearing, in Ogawa ’

s trans lation , N .W . of Cho-shu-té (Djochot).The next tri bu tary mentioned comes fromthe left s ide and is cal led Kiang

chia—su-mu-la—ho . It corresponds to Kiankia somla of d’

Anv ille’

s gene ral map and

K iankia Somla R . of his de tai led map , and G iangghia sumla of de Rh ins . Its

s ource comes froma mounta in cal led Sha-ku-

ya-la—ma—la-shan

,which may b e a tran

s cription of Shakya-lama-la or »the pas s of the Buddha pries t ». The Sha-ku-ya

—lama-shan and Cha-ko-

chia—la—shan are identi cal with de Rhins ’ Changou Yarak ri andthe Chadz iar ri . Comparing the Chines e text and d

Anv ille’

s map with Ryde r ’smap we find again that this river comes fromthe Transhimalaya and that the Ma

ryum—chu , or rathe r the min imal brook coming fromthe pas s Maryum-la,is only one

of the smal les t tribu taries to the Chiang-chia—su-mu—la—ho ,which its e lf i s on ly a trib

u tary to the Yere-ts angpo . The fact that it, in its lowe r parts , is cal led Maryumchu , as I was told by the Tibe tans , probably depends upon the tasamwhich follows up the brook to Maryum-la . The Chines e text does not evenmen tion Maryumchu . So all attempts to procla im the Maryum-

chu as one of the sources of the

Brahmapu tra has no foundation whateve r.1 Compare »Trans-Himalaya» , V ol. II , p . 96 and 1 0 1 , and the map.

1 1 8 TH E SHUI-TAO-TI -KANG ON THE SOURCE OF TH E BRAHMAPUTRA AND THE SATLE J .

As to the Chema-yundung it is not mentioned in Ogawa ’s vers ion e ither.The follow ing tributaries fromthe north s e emve ry doub tfu l , at any rate the i r

de rivat ions . A-la-chu-ho is d’

Anv i lle’

s Artchou K , b ut the Sha-la-mu-kang cannotb e identified . The N a-yu

-ko-tsang-pu is d ’

Anv ille’

S N aouc Tsanpou R . and is probablyident ical with Ryder’s Nakchak and Na in S ing’s Chu Nago . It is sa id to comefromthe lake Sang-Ii , d ’

Anv ille’

s Sanc-l i , which , perhaps , may b e S en it-tso, a lakes i tuated on the southern s ide of the Transhimalayan water-parting as the Tibetanstold me . But the rive r N a—yu-ku-tsang-pu cannot b e S .W . of Cho-

shu-té , if this isthe dis tri ct of Toshut or Hor-toshu t as I be l ieve , for then it is N .W . of that d is trict .

The mountains fromwhich the tributaries come are as yet impos s ible to ident ify ,un les s Ya—la—l ing is Yor-la , one of the chief pas s es of the Transhimalaya .

In the des cription of the many tributaries of Sa-chu-tsang-pu or Tsachu we

eas i ly recogn is e the real ity. Mos t of the water of this rive r s e ems to come fromLunpo

-gangri and flow to the S .W . The Man-chu-tsang-pu is my Men-chu .

The genera l des cription of the cou rs e of S a—ko-tsang-pu is admirable . It isthe same river which Nain S ing more than a hundred years late r called ChartaS angpo and which I call Chaktak—tsangpo . The Chines e author s imply gives it thesame name as the province in which it is s ituated , S a-ko , the S arka-jong of NainS ing, S aka—dzong of Ryder and mys e lf. It is s a id to come fromthe lake La-pu on

the N .E .

, and indeed , as I found in 1 908, it comes froma lake Lap-chung-tso s itu

ated N .N -E . of its junction with the Tsangpo . It is a lso pe rfectly right to say thatthe river firs t flows south-wes twards and , afte r rece iving s everal tribu taries

,turns

south-eas twards to join the Tsangpo . But when he comes to the explanatory detai lswithin brackets , he i s d ifficu lt to fol low . The La-pu

-kang-chung-shan is all right,for the whole country round the lake i s called Lap-Chung , and the re may eas i ly b ea Lap

-chung-gangri . Kang- lung-shan may b e my Kan-

chung-gangri . The lengthhe gives to the river, l i , is ve ry much exagge rated , i f the dis tance be tweenthe Kai las and the Tamchok-kabab shal l b e u s ed for comparison and which is givenas 300 11. For the dis tance fromthe Lap

-chung-tso to the mouth of the Chaktak

tsangpo i s not even so much as 300 11. But the genera l descript ion of the river isincomparably be tter and more correct than the fantas tical repres entation given on

d’

Anv ille’

s map , whe re the rive r in a s tra ight l ine goe s to the S .W . the whole way.

The Lio L . of d’

Anv ille is meant to b e the La—pu lake of Chi Chao N an . Abou thal f way be tween the lake and the mouth of the rive r d ’

Anv ille has a range of

mountains he cal ls Lop M .

,an echo fromthe Lap

-Chung mountains . He ca l ls the

rive r the S anki Tsanpou .

We have s een that some of the Chines e writers make the Chema the principa lb ranch , coming from Tamchok-kabab , others s ay that Chema is on ly a tribu taryjo in ing the Kubi . In all ins tances

,both we s tern and eas tern , the Kubi-tsangpo has ,

howeve r, been almos t ignored . The Chines e au thori ties do not mention its name ,a lthough at leas t in one cas e

,it is cal led the V ere-tsangpo . Only Kawaguchi s eems

1 20 TH E SHUI-TAO-TI-KANG ON THE SOURCE OF TH E BRAHMAPUTRA AND THE SATLEJ .

mounta ins 80 l i north-eas t of Lang-chuan-ka—pa—pu . Two s treams flow wes twardsfrom the mountain and tu rn north-wes twards after the i r junction . It now takes a

s inuous cours e for 60 l i , turns south-wes twards , and jo ins the ma in river. This is as ource ]

The rive r flows further to the w es t-by-north for 40 l i , then to the north-eas t,

to b e met by the wate r of Lake K n ng-s/zeng

‘ which s inks underneath the groundof the lake bas in , b ut which , afte r reappearing , and after rece iving three northernaffluents , runs south-wes twards to the rive r .

[The Lake of K'

n ng—s/i eng

-o-mo has two sources , one coming from the north

eas t, fromTa -leo-la -énng

-ma -sna n ,

2and flowing 1 50

—1 60 l i , the othe r fromthe eas t,

fromthe wes tern foot of M a-e

r/z-yo-ma -l i ng

3 in the w es te rn frontie rs of Cli o-snn -te‘

.

This las t-men tioned mountain forms the eas tern boundary of A-l i and is the chiefrange going south-eastwards fromKang-ti-s su. The water (of the Lake Kung-sheng)flows wes twards for more than 50 l i and forms anothe r lake , 80 l i w ide and withou tan outlet . Howeve r, more than 1 0 l i farthe r to the wes t , there is a third lake w itha s ubte rranean source and with a length of 30 l i . A s treamcomes fromnorth tothe lake . The river now flows south-w es twards for 60 l i , and rece ives a s treamcoming fromthe north-eas t . 40 l i farthe r south-wes twards it rece ives a s treamcomingfromthe northern mountains , fu rthe r south-we s twards , the rivermee ts the water fromLang-chuan-ka-pa—pu-shan .]

The wate r forms Lake III a -

p i n-ma -ta-la i .4 [Fromsou th to north it is 1 50 l i

long, from eas t to wes t 80 or 1 00 l i wide , and has a ci rcumference more than200 l i . On the northern s ide of the lake there are two s treams coming fromthe

north . The lake is s ituated 1 20 l i to the south of Kang-ti The water flowsout from the wes t of the lake into Lake Lang-ka s in a d is tance of 60 l i . The

latter lake rece ives a s treamcoming fromthe N E [Lake L ang-t a has a narrow

rectangu lar shape , pointed and e longated , the length from south to north be ing1 70 l i and the width fromeas t to wes t 1 00 l i . Its northern pointed corne r has thes treamcoming fromnorth-eas t . The re are three sources on the sou thern foot at a

dis tance of 70 l i froma southe rn branch of Kang-ti-s s fr; they flow sou thwards, un ite

into a s tream,which takes a sou th-wes te rly cours e for 1 50

—1 60 l i b efore enteringthe lake . The lake is same 6 in ci rcumfe rence and area , b ut d ifferent in outl ine .]

The water (of Lake Lang-ka) flows out fromthe wes t, and afte r runn ing wes twards for more than 1 00 l i , i t tu rns to the S .W . It i s now called the L a ng

-enn

/zo,7 and takes a s inuous cours e for more than 200 li . Then it rece ives the

1 Gunchu-tso.

1 Tacra Concla on d’

Anv i lle’

s map.

3 M aryum-la .

4 Mapama-tala i , or Manasarov ar.

5 Langak-tso or Rakas -tal.6 i . e . of the s ame s ize as M apama-talai .7 i . e . the Satlej .

LAKE MA-PIU-MU-TA-LAI AND LAKE LANG-KA . 1 2 1

Can-t a -la -no coming from N B [The Chu-ka—la-ho comes out fromthe s outhe rnfoot of Ts ’ang-wen-l ing, and firs t flows southwards , then sou th-wes twards , and pas s ingto the S .E . of K n

-éo-o/za -s/zi—ln -ma -

po-ts e-eni ng , turns sou th-westwards and joins the

Lang-chu-ho . The length of this s treamis more than 200

This description is , according to Profes sor Ogawa , published in the 26th yearof CHI E N LUNG ’

s re ign . When reading it again and aga in I cannot he lp ge tting theimpres s ion that it dates fromthe s ame documents and ske tches which we re de l iveredto the Jesu it Fathe rs and by thems ent to d ’

Anv i lle . For the des crip tion is in per

fee t harmony with d ’

Anv i lle’

s maps in du Halde and even the same names are to

b e found in both cas es and wri tten ve ry much like each othe r, d is regarding the

French transcription of the Chines e syllables . And as the Shu i-tao-ti-kang is onlya compilation its au thor had to us e any re liable mate ria l he cou ld get hold of. Fromwherever i t comes , this des cription is admi rable and dis tingu ished by the same careful conformity wi th the truth and conscientiou snes s as all othe r Chines e geographica lde scriptions .

1

The cas e is the same so far as the Shu i-tao-ti-kang i s conce rned . A des cript ion in a few words of the Kang-ti-s su (d’Anv i lle has Kentais s e) cou ld not pos s ibly b emore graphic and correct . The same mis take about the Ganges as on the Lamamap i s made here , when the author thinks the Kang—ka-

chiang or Ganges i s a con

ti nuati on of the rive r which »comes out from» the Kai las , and which , on d’

Anv i lle’

s

map is cal led Latchou R ., or one of the fe ede rs of Gar-chu , the S .W . branch of

the Indus . In a preceding chapter I have tri ed to e xplain why the Lamas confoundedthe Indus and the S atlej with the Ganges , which gives an example of a geographica lObject they had not s een wi th the i r own eyes , and in which they there fore weremis taken .

The mos t interes ting pas sage is , howeve r, the one about the »high moun ta in»Langchen

-kabab which , qu i te correctly, is said to b e S .E . of the Ka i las and eas t ofTaklakhar. It res embles an e lephan t , a qua lity which is now trans fe rred to a

l i ttle hill on the bank of the Satlej at D'

Olchu-gompa , wes t of Langak-tso . Thenfollows a des cri pt ion of the uppe rmos t Satlej on its way fromthe northe rn foot ofLangchen

—kabab towards the N .W . When the au thor says that this rive r, which i s1 Nobody has ever known Chinese cartography better than the late Baron v on R ichthofen who ,

speaking of the maps of the Je su its , says : »Mehr und mehr waren s ie mit demti b eraus re ichen Materia lder e inheimischen Kartographie bekannt geworden ; s ie mus sten sehen , das s das selbe hins ichtl ich der E intragungen v on Flti ssen und Ortschaften den Ansprtichen an Gewis senhaftigke it und Treue in so we itentsprach , als nur wirkl ich Vorhandenes aufge ze ichnet war , dass aber den Chine s en das Gesch ick inder richtigen Zusammenste l lung des M aterials abging, da s ie n icht fahig waren , astronomische Ortsbestimmungen mi t Genauigke it auszufiihren .» (China , I , p . In hi s c lass ica l article on Lop

-nor

the same great authority has po inted out the fact, that the Chinese topographers never enter any geograph ical feature upon the ir maps un less they have themse lves actual ly s een it. Verhandlungen de r

Ges . f. E rdkunde , Berl in , V ol. V , 1 8 78 , pp. 1 2 1 e t s eq . Compare also my book Through A s ia, London 1 898 , I , p . 1 8 and II , 8 67 , as we l l as my Sc ientific Re su lts of a Journey in Centra l As ia 1 8991 902 , Vol . II , p . 2 63 e t seq . And everybody who has had an opportun ity to trave l in parts of A s iawhich have been surveyed by Chinese explorers wi l l have been struck by the ir re l iabil ity .

1 6—1 3 1 3 8 7 I

1 22 TH E SHUI-TAO-TI-KANG ON THE SOURCE OF TH E BRAH MAPUTRA AND TH E SATLE J .

the Tage -tsangpo , is met by a rive r fromGunchu-ts o , he is wrong, for the Tagets angpo goes its own way to the Manasarov ar and the s tatement that the wate rfrom the Gunebo-ts o s hou ld » s ink unde r the ground» for a certain dis tance is of

cou rs e impos s ible as the Gumcho-ts o is s alt .

I The Guncho-tso is said to have t wo

source s treams , one fromthe N .E ., fromthe mountain Ta-ko-la-kung-ma , which is

also to b e found on d’

Anv ille’

s map ,under the name of Tacra Concla ; the othe r

comes fromthe w es t s ide of the pas s Maryum- la , which agrees w ith Ryde r’s map inall particu lars . Maryum-la i s s a id to b e on the w es te rn fronti e r of Cho—shu-té , a d is tri cton the southe rn s lope s of the Transhimalaya , and ce rtain ly iden tical w ith the Hor

toshu t or Toshu t-horpa of which I heard s eve ra l t imes in 1 908 . Of the two lakess aid to b e s i tuated wes t of Gumcho-ts o , on ly one i s marked on d

Anv ille’

s map , b utboth may b e qu i te small and temporary .

It is worth whi le to notice that the compi lor of the Chines e hydrography re

gards the Mapama-talai s imply as a formation of the water fromLangchen-kabab or

as a part of the hydrographica l sys tem, which froms everal po ints of V i ew i s correct .At the wes tern s ide »the wate r », i . e . the water fromLangchen

-kabab flows in to the

lake Lanka , or Lanken as d’

Anv i lle w rites . The dis tance be t ween the lakes is s aidto b e 60 l i , which corresponds to my mi les . As a ru le the dis tances are ve ryunre l iable . For if i t i s 60 l i bet ween the lakes along the channe l , i t shou ld b e 1 80 11

and not 1 20 be t we en lake Mapama and the Ka i las . And i f the lake i s 1 50 and

80 or 1 00 l i acros s , its ci rcumference mus t of cours e b e much more than 200 l i .At any rate we have here a pos it ive s tatement regarding the channe l , and a per

fectly t rue vi ew of its characte r of be ing the cont inuation of the Satlej fromLangchen-kabab .

Final ly. »the water» , 2 i . e . the wate r fromthe uppermos t Langchen-kabab or

S atlej flows out from the wes te rn s ide of the lake and i s now cal led Lang-chu-ho ,

the Lanctchou R . of d’

Anv i lle . Fromthe Chines e text i t s eems that this name wasnot u s ed for the uppe rmos t part of the rive r-cours e . This is doubtful . For the nameLangchen

-kamba i s s ti l l u s ed for a Spring on the Tage-tsangpo. But Tage -ts angpoi s at leas t nowadays the name of the uppe rmos t Satlej .

Final ly i t i s s a id that the Chu-kar (Chu-ka-la) comes fromthe N E . and joinsthe head river. Unles s the re are two rive rs Of this name , the s tatement is w rong,as the Chu -kar of S trachey comes from the south . But even he re it i s pointedout that the S atlej is the head- river and the Chu -ka-la a tributary .

1 The Gunebo-tso cannot have had fresh water and an underground outlet 1 50 years ago ; theChines e explorers have made a mistake in this point.

1 In h is trans lation Professor Ogawa has put between brackets : (of lake Lan-ka), wh ich of cours ei s a lso correct.

1 24 MANASAROVAR AND SURROUNDING RE GIONS IN TIBE TAN WORKS.

In chapter 2 3 is a s tory told abou t fou r Bonpo , who went i nto a boat andhad it driven into the middle of the Brahmapu tra. And furthe r

Da s tie s sen s ie e inen Fluch au s : »rJe i lha , lenk e den Lauf des Stromes ab ! Tre i be dasWas s er des Brahmaputra nach aufwarts zur

'

uck ! Las s imLande T i bet de ine mann igfaltigenZaub erkuns te s p ie len ! ) Mit d ies en Worten flogen d ie v ier Bonpo gen H imme l, s t iegen imLande gYa gon gyimb u he rn ieder und we ilte n dort imBes itz e cle r S e l igke it. A ls s ich nun

der Oberlau f des Brahmaputra nach s e iner Que lle zurfickwandte , wurden vie le B egle iter desPadmas amb hav a v on der S trOmung fortge ris s en . ImWe s ten war der Berg Yar lha Sampo ,

imSuden der lDon lha , imN orden der T!

ah lha, imOs ten der sBomra s ichtbar. D er er

hob ene Turkisglanz nahmab , der erhabene S ee trocknete aus .

Many othe r disas ters happened, which are told in this s tory, b ut after prayersto ce rtain gods the Brahmaputra aga in re turned, and flowed downwards , an event,which fi l led the people with happines s .

Dr . Berthold Laufe r has gi ven anothe r contri bu tion to our knowledge of the

popu lar re l igi on of Tibe t, the E on-re l igi on which preva i led in that country be forethe Buddhismu s was introduced , and of which so ve ry l i ttle is known , except that i twas fu l l of an imismus . The work in ques tion , which be longs to the E on-l iterature ,has the ti tle : K l n oonm os dns ~

fia i snin-po, and i s an abbreviated ve rs ion of the

work on the hundred thousand N '

ci ga’

s .

I In a ve ry cleve r and sharp way Laufe rproves that this work i s an inten tional ly made Lamais tic fals ification . He s ays : D er

das Bonwerk umarbe i tende Lama benu tzte fur s e ine Zwecke e in S anskritwerk jenesNamens (Krahamantanama Dharan i), da dies es auch v on Naga ’

s hande lte , s chmugge lte s e inen Tite l e in und fiigte wahrs che in l ich auch al lerle i v omInhalte in das O rigi nal hine in .

The s econd s ection of this wonderfu l work is cal led : The black s ection of the

work abou t '

the hundred thousand Nagas , and there is a confes s ion of s ins of the

Nagas , in ve rs e , which has also be en trans lated into Ge rman vers e by Laufer . Icannot he lp quoting a few pas s ages in which both the Manasarov ar and Mount Tis eare men tioned

,su rrounded by mys te ry and legend

T ie f ruh t der Palas t der Naga’sIn demgros s en S ee Ma-p

am-

pa ;

Gros s is t er,der Tri rkis s ee ,

Gros s der S ee mit neun der Inse ln .

Auf den grunen A lpenauenRotkuh , Rots t ier murmeln le ise ,Schwarmen Gans e , s chwarmen E nten ,Flattern feders tolze Pfauen ,

Tont die Cymbe l i hre We is eAch , v on so lchemPrachtpalas teHatten w ir ja ke ine Kunde ,Trub ten drumden S ee und warfen ,

1 M emoires de la Socie’te finno-ougrienne , XI , Hels ingfors 1 898 , p . 4 1 et seq .

TH E HUNDRE D TH OUSAND NAGA’

s . 1 25

Was der Brand vers engt, h ine in ;Doch w ir bergen n ichts imHerzen ,W ir be kennen und w ir s ii hnen .

Nagaschloss iml ichten Hain ragt,Wo amS ee s ich dehnt das D ickicht ; etc.

Wenn des Nagaschlos s-see

s Dammwerk,D ie Kanale , die den S ee , ach , truben ,

S ind erlegen der Zers tOrung,Murmeln Rotkuh , etc .

Warfen, was der Brand versengt ,Auf den t ie fen Grund des S ee ’

s ; etc.

Ragt da e in Palas t der N aga’sAuf des T ise we is sen Gletschern ;Rotkuh , etc.

Ragt da e in Palas t der N éga’sAuf der B erge grij nen Almen ; etc.

Gri in S ind dort d ie Schieferb erge ,Baume prangen in der Fe ls s ch lucht, etc.

Ragt da e in Palast der Naga’sAn demblauen T ‘

Urk is s ee ,

An demgros s en See Ma-dros -pa ; etc.

In demNagaschloss e s chmuckre ichSch lecht is t unsrer Hande WerkWuhlten in der s and

gen SteppeUnd den gros s en Alpenw ies en ; e tc.

Ragt da e in Pal as t der Naga’sAuf des Hoch bergs bre iten Hangen ; etc .

S ie h , w ie s ich der V ielkopfnaga w indet,Und d ie V ielkopfschlange h in und herwogt ,

Und der blaue S ee v on T ij rkis s ch immert !The White s ection of the work of the hundred thousand Nagas beginsThe Nagaraja

s s amt eurer Ge folgschaft, haus end auf den imKosmos , E rdall und in derSchopfungswelt ge lagerten Meeren und S trOmen ,

Flu s s u fern , S eeen, Quellge b ieten, Bachen ,

Te ichen und andern (Gewass ern), auf den s ieben Bergen , Fe ls enb ergen und erd igen S te inen , inW ind , Feuer, Was ser, Ather, in allen jenen E lementen

,auf, ers che inet und empfangt hier d ie

Opferspende !

1 26 MANASAROVAR AND SURROUNDING RE GION S'

I N TIBETAN WORKS .

According to Laufe r the s even mountains are obviou s ly iden tical with the s evenranges which are s uppos ed to su rround the Sumeru in concentric ci rcles . Laufe rfinds it not un l ike ly that thes e r 'i -b du n correspond to real geographical facts as

conce ived by the Tibe tans , though th is cou ld not b e proved in 1 898 . He refe rsto the s ingle then e x is ting s tatement about the orography of the inte rior Of Tibe t ,name ly, the one gi ven by CSOMA fromnative sources , that s ixmountain ranges werecounted fromHima laya to the pla ins of Tartary.

I

One shou ld expect that a learned Tibetan Lama ’

s description of Tibet shou ldcon tain some important and s tri king facts unknown to the E u ropean geographers ,b ut this can hardly b e said of MINCHUL KHUTUKTU , whos e rathe r Short des cript ionhas been trans lated by Profes sor V . VASILIE V of S t . Pete rsburg.

2 Al l that isknown of the au thor , says Vas ili ev

,is , that in the twenties and thirties of the las t

centu ry he was an As s is tant to the Cons istorial Admin is tration of the Lamas and

temples in Peking , for he d ied in 1 8 39 . He wrote a geography of the whole world,for which he got mate ria l fromE uropean sources , b ut to us the only interes ting thingis what he has to say fromhis own expe rience about Tibe t

,a part which Vas iliev

trans lated in Peking . The origi nal Tibetan text is now in the l ibrary of the Un i

vers it y Of St . Pe tersburg .

To begin with the Khrituktu gives a gene ra l view of the geographical s i tuation of Tibe t as compared w ith India . One

t>

has to cros s 9 black mountains beforeone arrives at a snowy mountain (Himalaya). »In the s ame way it is Said , that onthe other s ide of thes e mountains there is a s nowymountain ,

although s eve ra l peopleregard this las t-mentioned one as one and the s ame with Gans -ti-s e (Gandes ), b ut i ts eems more su itable to attribute this name , as a nomen appe l lativum,

not only tothe Gans -ti-s e , b ut a lso to many thou sands of othe r mountains , which are s i tuatedbetwe en Kapes tan (Kabu l) to the north , and Tson-ke

a in the eas t, in K’

ams , thenative country of the King of fai th

,Tson-k’

ap-a .»

He gives a genera l description of the country, which is rather chilly , andmuchhigher than the s urrounding countries . »The magn ificent snow-mountains ris e l ikemonumen ts of the pures t crys tal .» H e gives the names of some of them, as Ti-s eand gNan

-c

i

en-t

an-lha . »And the re are innume rable blackmountains a lso .3 A great

part of the s urface is occupied by pu re and transparent lakes : Ma—p

amgYu-mts ’o ,

1 In h is »G los sar»,p . 7 9 and 93 , Laufer has the fol lowing explanations : T i -se Oder Te-se = S .

Kailasa ; gans dkar, womit man den Be inamen des Berges gans ri vergle iche . T ib . auch Ke-la-Sa,Kai-la-s i . M ilaraspa nennt ihn rii rgyal

-po Kon ig der Berge .

—'

Ma-dros -pa Anav atapta , der Ma

nasarOv arasee . Auch Ma-p’

am,M a-

p’

am-pa , M i

-p’

am-

pa S ein Be iname i s t yyn-mts o

TUrkis see oder Tyn-mts ’o snon -

po blauer TUrkis see . »Der b eriihmte See v on Turkis » wird er i n e inemHymnus des M ilaraspa genannt E ine Angabe , clas s der See neun Inse ln hat

,habe ich n irgendwo

gefunden . Nowadays it i s we ll known that the lake has not a s ingle is land .

1 Geografia Tib eta , perevod i s T ib etskago sochinen iya M inchul Khutukty. V . Vas il ieva . St.

Peterburg 1 895.

3 I . e . wi thout eternal snow.

1 23 MANASAROVAR AND SURROUNDING RE GIONS IN TIBETAN WORKS .

mountain , cal led mGon-po-ri . Not far in front of it to the south is s ituated the lakeMa-p

amand to the wes t fromthis lake the l ittle lake Lan-dar-mtso .»

Then he gives some re l igi ous legends about Ka i las and continues : »To the

eas t fromGan-ri are s ituated ’Brog-2ad (D shoshot), Luri -k’

a and other dis tricts .» In

Ladak he knows Leh and K’

ri g-s e . He a lso knows mT ’

o-ldin-gs er-k

an (Totl ing S erkang) in Gu

-

ge . Near the front s ide of Gangri he places the dis trict of Ngari Pu rang,whe re the temple Shin-Pel-gLing

I is s i tuated . He correctly obs erves that eas tof Purang is a country called kLo-b o-sman-t ’an, whe re Indian tribes predominate ,although the re are Tibe tan tribes as we ll . In old times the country be longed to

Tibet, in his own days to Nepal . Din-ri (Tingri) and S a-skya-gompa are a lso knownto him. »North of the temple of Sa-skya , on the bank of the rive r Tsan-b o and at

othe r places are s ituated Lha-rts e , N am-rin,

P’

un-ts

og rdson, and othe r towns .»

He talks of a magn ificent s nowy range to the north,and on its othe r s ide is

the sTod—hor ai’mak . Pas s ing this moun tain and s ome nomad grounds and s teppes

in which the re is no gras s , no wate r , no trees , one comes to the Mohammedancountry of Ants ian ,

2and from the re , that i s to say fromthe north , nobody can

make an invas ion in Tibet . »But in olden times , when the re was war be tween the

S a—skya-pa and’Bri-gu ir~ pa , the

’Bri-gun-pa’

s brought the Kashgarian army to S akiaacros s thes e mounta ins and, it s eems very like ly , even now the re are to b e foundindependent Mohammedans on the othe r s ide of these mountains .»

Tashi-lun-b o, (e a-S is -lhun-po) whe re the Banchen—E rden i (Mongol ian trans la

tion of Pan-c’

en rin-

po-c’

e , common in Rus s ian books) res ides , has monks .

He gives a des crip tion of Lhasa and its su rroundings , dGa-ldan , S e—ra ,

’Bras—spuns ,Gyan- rts e , Yar-’brog-mts °o , dPa-sde rdson.

»North of dBus—gTsafr (U—tsan) is s i tuated the great country of the nomads , inwhich the fou r great northe rn tribes wande r about : name ly Nags -ts i an, gNam-

ru ,

Nag-c’

u,

and Yans -pa-can ; who , however , are a lso cal led Hor, b ut they are

Tibetans In that country is s ituated the lake Namtsochugmo ; one of the fou rs tormy , snowy mountains , gNan—c’en-t’ar'r-lhai-gans -ri and some othe r inte res t ingplaces .» Aga in he cros s es the Tsan-po and describes rTs e Jar

-lun etc. Onlyen pas s a n t he mentions S a-dga, which Vas iliev iden tifies with Sa—skya , b ut whichmore l ike ly is meant to b e Saka-dsong . Finally he des cribes Konbo and goes on

to K‘

am, Li-t ’an, sKu’

b um, S i-lin K '

oko-nor etc. The re is a good deal of othe r

geographical matte r in his account, which is not easy to make out , and I havequoted fromV as iliev

s trans lation on ly such places which are of special interes t to us .

It is not surpris ing that the Manasarov ar, as be ing the object of so manypilgrims ’ journeys shou ld attract the attent ion of Minchu l Khu tuktu as we l l . But he

has not much to s ay of the lake . The spring in its ne ighbourhood , which he cal led

1 S ibl ing or Shib eli ng-

gompa .

1 This V as i liev trans lates with Kokan , although Andi shan seems to b e more probable .

SARAT CHANDRA DAS ON A TIBE TAN GE OGRAPHY . . 1 29

’Te

un-grol , may, pe rhaps , b e the same as GE RARD’

s Chomik Tingdol, for »chumik»means spring .

I He iden tifies the Manas arov ar with the Anudata and re ite rates the

old saying of the fou r rivers beginn ing fromthe Ka i las . He te l ls us , as it werefromhis own expe rience , that the Indus begins in Ba lti , wes t of the Ka i las . Con

ce rn ing the rive r Pakshu , which ris es fromthe north-wes tern s ide of the Ka i las , heknows that i t goes to Bokhara and Khiva and fal ls into the s ea called Manas arv ara ,

which mus t b e identical with the Mav arannar or Mavara—un-Nahr or Transoxiana .

The rive r he made origi nate fromthe Ka i las is thus the Oxus and the s ea is the

lake of Aral .The S i ta which he makes origi nate fromthe north-wes tern ramifications of the

Kailas goes to the salt lake of Tu rfan , that is to say Lop-nor. Thence it went to

China . This is the old s tory of the Ye llow rive r, the uppe r cou rs e of which wass uppos ed to b e the Tarim. WILFORD at one place says the S ita might b e the

S itoda , suppos ed to commun icate with the Satlej , and at anothe r he tries to identi fyit w ith the Ye llow rive r .

M inchul’

s des cription of the Ka i las is rathe r good and ve ry picturesqu e . H is’Brog-i ad (Dshoshot) and Lun-k

a may b e identified with Toshot and Lunkar.

2

Amongs t the fou r great northe rn tribes we mis s the nomads of Bongba , forhe on ly mentions Nags -ts ’an, gNam-

ru , Nag-c’

u and Yans -pa-can .

Final ly a few extracts w i l l Show how S AR AT CH ANDRA DAS has dealt w iththe Tibetan geography in his art icle : A bri ef account of Tibe t from»D s amL i ngGy es/ze» , the we l l-known geographical work of Lama Ts anpo Nomankhan of Amdo.

3

Fromthe sacred Buddhis t scri ptu res cal led Cn/zosmngon -

parmdsod (Aonid/zarmaȎos/za) he quotes the fol low ing vers e

Hence northward there lie b lack mountains n ine ,W h ich pas s ed , the lofty snow-clad peaks appear

,

Be yond w h ich e xtend Himavat, the realmof snow .

According to Sarat Chand ra Das thes e n ine blackmountains al lude to the

long ranges of both low and lofty mountain s which intervene be tween Uddayanaand Yun-

nan . The coun try of Himavat i s known by the genera l name of P0. On

the N .W . it is touched by the Hor country of Kapis tan , which shou ld b e Yarkandand Kashgar. At anothe r place Hor is trans lated by the more general s ignification Tartary . Then Sarat Chandra Das makes us acqua inted with the Tibe tan textin the follow ing ve ry free way :

To the eas t of Upper T i bet are the snowy mountains of Tes i (Kallas h), lake Mapham(Mansarov ara), the fountain c fing

-

grol , w h ich has the reputation of e xtend ing sal vation to

1 Passt ganz gut, da grol wi e do! au sgesprochen wird , (Grunwedel) .1 Chandra Das has Dragsho and Lungkha .

3 Journa l of the A s iatic Society of Bengal , V ol . LV I . Part I . N zo 1 . Calcutta 1 8 8 7 ,p . 1 e t seq . The T ibetan geography of Nomankhan as it has been presented to us by Chandra Dasi s s imply a b ad trans lation of the s ame text which had been used by V as iliev . (Gri inwedel .)

1 7—1 3 1 3 8 7 I

1 30 MANASAROVAR AND SURROUNDmG RE GIONS IN TIBE TAN WORKS .

those that drink i ts water N owaday s the pi lgrims and devotees of Gangs-ri des ignate the

s nowy mountain ment ioned in Mngon-mdsod and other s acred books b y the name of Kang

Tes i , and the lake Mts ho ma-dros -pa b y the name of M tsho Maphan. The commentators of

Mngon-mdsod des cri be the four great ri vers of Upper T i bet as is su ing fromrocks , w h ich

res p e cti ve l y have the appearances of an e lephant, an eagle , a hors e and a l ion . According to

other writers the rocks have the appearances of the head of a bull, a hors e , a peacock and a

l ion ,fromw h ich s pring Ganga , Loh ita, Pakshu and S indhu . E ach of the se great rivers i s sa id

to flow into one of the four oceans , after rece i v ing more than fiv e hundred tri butaries . The

great lake Mtshoma-dros -pa is mentioned as e x te nd ing over an area of 80 leagues . Thes eaccounts vary very much when compared w ith w hat i s now s een. T h is may b e attri buted to

the d ifl’

erence of mora l merit in the d i fferent generat ions of mank ind . It is probably ow ing to

th e smallnes s of moral merit in us , that we do not s ee the se s acred places in the ir original stateas our ancestors s aw them. There i s no other e x planation e xcept th is why great th ings s hou ldlook small. The mighty river S indhu , is su ing froma glac ier on the western s lopes of KangTes i , flows wes tward through Balt i and afterwards in a southwestern d irection through Kapis tan,

Jalandhra, and Panchanad , til l jo ined by the rivers Satadru , e tc. On account of thereexi s ting numerous s nowy mountains in the countrie s of Panchanad

,Kas hmir, Ushmaparanta ,Kab e la and Jalandhra, w hi ch s end the ir water into it, the ri ver S indhu i s very powerful , and in

fact it i s the greates t of the four ri vers ment ioned above . The ri ver Pakshu s pringing fromthe glaciers on th e north-we s tern s lopes of Kang Tes i , and flow ing through the country ofTho-kar in a wes terly d irection , and Balag Bhokar and Hiva, and also through the s teppespeopled by the Turushka hordes , d ischarges i ts contents into the great lake Mansarov ar (l)The ri ver S ita, is su ing fromthe glaciers of the Te s imounta ins on the ir eas tern s ide , flows throughthe country of Yar-khan and Thokar to empty itse lf into the lake Tsha~mtsho. Trad ition s ay sthat formerly th is river, flow ing through the centre of the Hor country , d is charged its elf intothe eas tern ocean

b ut on account of its be ing drained b y means of aqueducts , cut fromit toirrigate both Ch inese and Thokar countries , its progress to the s ea was arres ted Confrontingthe sacred mounta in, at a s hort d is tance i s s ituated the famous lake M tsho-mapham, to the wes tof w h ich there is another b ut smaller lake , cal led Lagran-mtsho FromPal Sakya (thefamous monas tery ) i f you go northward for a full day ’s march , you arrive at Khahu TagJong To the north of Khahu Jong there is a very lofty s nowy range on the back of wh ichi s the d is trict inhabited b y a tri be of Hors cal led Toi Hor, s aid to b e de scended fromSrinpo

(cann i bal hobgobl ins) Th is mountain i s e xtremely h igh . B eyond thes e snow y mountainse xi s t many Dok tri bes . Thes e Lalo (Mohammedan people) are subject to Kas hgar To the

north of the famous monas tery of Pal Sakya flows the ri ver Ts angpo, on the bank of w h ichs tand Lha-rts e , Ngam-ring and Phun tshogs

-l ing Jong, wh ich all now belong to the Governmentof Ts ang.

CHA PTE R X I V

EUROPE ’S KNOWLEDGE OF T IBET IN TH E M IDDLE AGES .

The enormous deve lopment of CHI NGIS KH AN ’

s Empire which took place inthe firs t ha lf of the 1 3 th century was a ve ry importan t factor in draw ing the

E as t neare r to the We s t. In a short time the Mongo ls became mas te rs of the

greate s t part of As ia and eas te rn E urope , and at the cu lmen of its powe r the irEmpire s tre tched fromJapan to Katzbach , fromFa rthe r India to lake I lmen . Nearlyall the nations of the old world became more or le s s dependent upon , or ente redinto commun ication w ith the conquerors , and at the i r court the G reat Khans rece ivedambassadors fromPope s and Cal iphs , fromByzantine Empe rors and French Kings

,

fromSu l tans of Rumand Grand Duke s of Ru s s ia .

I

The Mongols afte r having conque red all the northe rn ha lf of As ia , brokethrough the Caucasus in 1 2 2 2 , and s pread te rror ove r Kipchak , as southe rn Rus s ia .

was then ca l led. In 1 2 3 8 a new Mongol devas tation of E as te rn E u rope began to

awake the comprehens ion of the prince s in the w e s te rn hal f of our continent. In

1 24 1 S i le s ia , Moravia and Hungary we re ove rwhe lmed. On ly pol itica l reasonsand the death of the Mongo l Empe ror Ogodai put an end to furthe r progre s s we s twards . The dange r cou ld re turn , and Pope Innocent IV procla imed a crus adeaga ins t the Tartars , as the Mongols we re ca lled the envoys of Satan and min is te rsof the Tartarus . But he a lso us ed more civil ized means , and found amongs t the»Ta rtars » an exce l lent fie ld of labour for the membe rs of the Mend icant Orde rs of

S t . Domin ic and S t . Francis ? The Friar JOHN OF PIAN DE CARP INE was des patchedto the great unknown E as t , and his name has become famous in the his tory of

1 KOPPE N : D ie Lamaische H ierarchie und Kirche , Berl in 1 859, p . 8 6. A BE L-REMUSAT says , inhi s M e

’moires sur les re lations pol itiques des princes chrétiens , et particuli érement de s rois de France ,avec les empereurs mongols , Paris 1 8 24 , p . 5: »Les év énemens qui rapprocherent, au XI II 1 s iecle , des .peuples jusque-la séparés par l’étendue enti ere de notre continent

,n’

ont point d ’

exemple dans les an

nales du genre humain . La grandeur Mongo le , qu i fail l it embrasser le monde entier, fut cre’

ée en

moins de temps qu ’ i l n ’

en faut d ’ord inaire pour fonder et peupler une s eule cite.»

1 W. W. ROCKH ILL : The Journey of Wil liamof R ubruck to the E astern Parts of the World , .

1 2 53—55, London , 1 900, Introductory Notice .

1 34 E UROPE ’

S KNOWLE DGE OF TIBE T IN TH E M IDDLE AGE S .

A s iatic exploration . He s e t out fromLyons in 1 245 and de l ivered the le tte r fromthe Pope to Kuyuk Khan , not far fromthe city of Karakorum. Two years late rhe was back with the answe r fromthe Mongol Khan.

Friar John is special ly inte res ting to us as be ing the firs t E uropean to

mention Tibe t. Afte r te l l ing how Chingis Khan ’

s army had been at war, the textruns , in Hakluyt

s ve rs ion of V incen t of Beauva is ’ abridgment, as follows»And in traue i ling homewardes , the sayd armie of the Mongals came v nto the lande of

B urithab eth (the inhab itants w hereof are Pagans) and conquered the peop le in b attell. The sepeop le haue a s trange or rather a mis erab le k inde of cus tome . For when an ie mans fatherdeceas eth ,

he ass emb leth all h is kindred , and the y cate him. T he s e men haue no beards at

all, for we saw th emcarie a certa ine iron ins trument in th e ir hands , w herew ith , i f any ha iresg rove upon the ir ch inne , the y presently plucke themout . The y are als o ve ry deformed .»

1

The re is no doubt that this pas s age , which is ve ry much in accordance withthe narrative s of Rub ruck and Marco Po lo , real ly re fers to Tibe t. The name Burithabe th is us ed some 60 years late r by an orienta l write r , name ly , Rashideddin , in

h is yami n—t Tawar i én, re lating of S ingun , the son of Ong Khan that he »tookhis re fuge in Bouri-Tibe t» .

1 ROCKHILL thinks it is a hybrid word , compos e d of“

the native appe l lation Bod and the word Tibe t , while d ’

Av e zac s ugge s ts i t to b e a

corruption of the Baron-Tala , by which name the Mongols de s ignate Tibe t . 3 Boththese e xplanations s eem,

howeve r, to b e wrong , for the name Buri-Tibe t also occursin the Ts

i n oneng la ,whe re it is s a id that Ong Khan

s son fled to the people of

Bo- l i t ’u-fan ; and t’

u-fan i s one of the Chinese de s ignations for Tibe t. 4

Rockhil l doe s not be l ieve in Friar John ’

s charge of cann ibal ismaga ins t theTibe tans . If 770 ye ars ago they had the same habit as nowadays of cu tt ing up

the ir dead,

and throw ing the piece s to the dogs , the des cription of this proceduremay have been e xagge ra ted be fore it reached Mongol ia , or, pe rhaps mis unde rs toodb y early E u ropean trave l le rs .

5 The habit of plucking out the ha irs of the i r beardsi s s ti l l en vogue in Tibe t, and nearly eve ry Tibe tan , even amongs t the nomads , hasa n i ron pince r in his be lt exclus ive ly for this purpos e .

S eve ral o the r mis s ionarie s we re s ent out to As ia in subs equen t years , and

amongs t themthe French Franciscan WILLI AM OF RUBRUCK is the greates t of all.Yu le ca l ls his narrative one of the bes t in exis tence . As an explore r, this admi rableFriar cou ld not e as i ly b e su rpass ed, for he indicated the sou rce s and the cours e of

t he Don and the Vo lga , he showed that the Caspian was a lake , and that Cathaywas the s ame as the clas s ical coun try of the S e re s , he made Balkash and the city

1 The Texts and Versons of John de P lano Carpini and Will iamde Rub ruqu is , ed ited by'C . R . BEAZLE Y London 1 903 , p . 1 1 8 .

1 C . D’

OH SSON : H istoire des Mongols , Tome I . La H aye e t Amsterdam,1 8 3 4 , p . 8 2 .

3 ROCKH ILL , Op . c it. p . 1 51 .

4 E . BRE TSCHN E ID E R : Med iaeval Researches from’

E astern A s iatic Sources , Vol. 1 1 , London-1 9 1 0, p . 2 5.

5 See al so Yule : Cathay and the way th ither, p . 1 51 , and Marco P010 1 , p . 3 1 2 .

1 36 E UROPE ’

S KNOWLE DGE OF TIBE T IN TH E MIDDLE AGE S .

»The s orcerers who do th is are called Te be t and K e s imur, wh ich are the names of two

nations of Ido laters T hey alway s go in a s tate of d irt and uncleannes s , devo id of re spectfor thems elves , or for thos e who s ee them, unwas hed , unkempt

, and sord id ly attired . The sep e op le also have a cus tomwh ich I mus t te l l you . I f a man i s condemned to death and exe

c u ted b y the law fu l authority , the y take h is body and cook and e at it. But i f any one d ie a

natural death then the y w i ll not eat the body .» 1

Furthe r Marco Polo te lls us that afte r a fiv e days ’ march we reach a provincecal led Tebe t

»After those fiv e day s ’ march that I S poke of, you enter a prov ince which has beens ore ly ravage d ; and th is was done in the wars of Mongu Kaan . There are indeed towns and

v i llages and hamlets , b u t al l harried and des troyed You r ide for 20 day s w ithout find inga ny inhab ite d s pot

,s o that trav e llers are obl iged to carry all the ir prov is ions w ith them, and

are cons tantly fa l l ing in w ith thos e w i ld beas ts w h ich are s o numerous and so dangerou s .

After that you come at length to a tract wh ere there are towns and v illages in con s iderab len umbers The peop le are Idolaters and an evil ge nerat ion , hold ing it no s in to rob and

maltreat : in fact, they are the greates t brigands on earth . The y l i ve b y the chas e, as we ll as

o n the ir cattle and the fru its of the e arth . I s hou ld te ll you also that in th is country thereare many of the an imals that produce mus k , wh ich are cal led in the Tartar language Gudderi .

T hos e rascals have great numbers of large and fine dogs , wh ich are of great serv ice in catch ingthe mus k-b eas ts , and s o the y procure great abundance of mus k . The y have none of the GreatK aan

s paper money, b ut us e s alt instead of mone y . They are very poorly c lad , for the ir

c lothes are on ly of the s kins of b eas ts,and of canvas , and of buckram. The y have a language

of the ir own, and the y are cal led Teb et. And th is country of Tebe t forms a very great prov

i nce,Of w h ich I w ill g i ve you a brie f account.»The »D i s cours e concern i ng Tol et» runs as follow s»T h is prov ince cal led Te bet, i s of very great e x tent. The p eop le, as I have to ld you ,have a language of the ir own , and th ey are Idolaters , and the y border on Man z i and sundry

oth e r reg ions . Moreover,th e y are very great th ie ves . The country i s , in fact, so great thatit embraces e ight Kingdoms , and a vas t numb er of c ities and v i llages . It conta in s in s e veral

quarters ri vers and lake s , in wh ich gold-dus t is found in great abundance . C innamon alsogrows there in great p lenty . Coral i s in great demand in th is country and fetches a h ighprice , for the y d e l ight to hang it round the neck s of the ir women and of the ir idols . Theyhave also i n th is country p lenty of fine woo llens and other s tufi’s , and many k inds of s p ice sa re produced the re w h ich are never s een in our country . Among th is p eop le , too , you findthe bes t enchanters and as tro logers that e x is t in all that quarte r of the world ; they performs uch e x traord inary marve ls and s orceries b y d iabol ic art, that it as tounds one to s ee or evenh ear of them T hes e peop le of Te be t are an i l l-cond itioned race . The y havemas t iff dogsas b ig as donke y s , wh ich are cap ital at s e i z ing w ild b eas ts and in particu lar the w i ld o xenw h ich are called Beyamini , ve ry great and fierce an imals . The y have als o s undry other k indso f s porting dogs , and e xce llent lanner falcons and s akers

,sw i ft in fl ight and well-trained , wh ich

are got in the mounta ins of the country A s regards Te bet, you s hou ld unders tandthat it i s s ubject to the Great Kaan .»

1

1 The Book of Ser M arco P010, 1 , p . 3 0 1 . London 1 903 , where the neces sary remarks con

c ern ing this pas sage , by S ir Henry Yu le and Henri Cord ier wil l b e found .

1 Op . c it. Book I I , p . 42 et seq .

MARCO POLO . I 37

This is the firs t re l iable account eve r written on Tibe t by a E uropean . As

Marco Polo approached the inacce s s ib le country much neare r than Friar John and

Friar Will iam, and probably got information fromnat ive s on the trade roads be twe enTibe t and wes te rn China , he has more to te l l , both of the inhabitants , the ir cus tomsand the i r country . Though he neve r vis ited Tibet prope r , on ly its eas te rn borde rland, his de s cription of Tibe t is in many re spects ve ry characte ris tic, and ce rtainportions of it cou ld as we l l have be en written in our own days . While the Arabgeographe rs moved in a lmos t complete darknes s , Marco Polo has firme r groundunde r his fee t and de a ls with rea l itie s , and speaks of a nation w ith whomhe has

bee n in contact . And s ti l l , i f w e compare his s tore of Tibe tan information with thatof the Chine s e of h is time and earl ie r, we mus t confe s s that he knew ve ry l ittleinde ed.

H e knew ,howeve r, that Tibe t was a country of ve ry great extent , embrac ing

e ight kingdoms , subject to the G re at Kaan , a fact that was comple te ly unknownto cartographe rs even some 400 years late r. R ichthofen is of the Opin ion thatMarco Po lo ’

s de script ion of the rude nat ive s applie s on ly to the inhabitan ts to

wards the Chine s e frontie r , and Yu le agre es with him that Marco Polo ’

s Tibe tcommence s with the moun ta inous region near Ya—chau , s ituated north of the countryof the Lolos , and that his 20 days ’ jou rney took himto N ing—yuan fu , fromwhe rehe trave l led to Yunnan fu .

I Frompe rs ona l expe rience he ce rta inly knew only thos eeas te rn regions , which are now parts of Szechuan and Yu nnan . Bu t the re are in

dications in his account po inting to a wi de r know ledge , pe rhaps embracing the greatw e s tward portions of Tibe t. He knows the s ituation of Kashmir in re lation to

Badakshan , India and the s ea

'

of India , 2 and , as we have s e en , he men tions Kashmi rin connection w ith Tibe t . In fact he approached what w e cal l Tibe t both fromthe we s t, Pamir and Bo lor, the north

,Khotan , and the eas t, Yunnan . From

the s outh he did not touch Tibe t at all, as his knowledge of Ind ia , as a who le ,was ve ry inexact. 3 He doe s not men tion the Hima laya , nor the Indus nor

the Gange s , and is in this re spect far behind the G re eks 1 600 years e arlie r,and Ptolemy a thous and years be fore his time , though , of cours e he neve rhad the intention to w rite a geography as the clas s ic s cholars did . But he

mus t have heard of Tibe t fromso many s ource s that he fe lt convinced of its greatexten t .

When he says of Tibe t that it conta ins lakes i n s ev era l qu a rters , this can

h ardly apply to anything b ut the vas t inte rior plate au land and sou the rn Tibe t . On

the othe r hand i t is su rpris ing that he has not heard of the capita l of Tibe t, thoughLhasa mus t have be en in trade conne ction with the limitrOph -dis tricts vis i ted by

1 Yule ’s Marco Polo , Book II, p . 48 and 70 ; map of the route I I , p . 1 3 1 .

1 Op . c it . Book I , p . 1 66 .

3 Op. ci t . Book II , p . 42 6.

1 8—1 3 1 3 8 7 I

1 38 E UROPE ’

S KNO\VLE DGE OF TIBE T IN THE MIDDLE AGES .

Marco Polo. Wh at he te l ls u s of corals ,Iwool lens , 2 enchan te rs and as trologe rs , 3

mas t iff dogs , 4 and mu sk , 5 agree s perfectly w ith wha t is we l l known of Tibe t prope rin our time , all the way to the frontie r of Ladak and eas twards to the boundaryof Szechuan and Yunnan. Only his opin ion of the Tibe tans is unjus t , for so far as

my expe rience goes a more good-tempered and kind-hearted people can hardly b efound in the inte rior of As ia .

6

Marco Polo has obse rved that the Tibe tans u se sa l t ins tead of money. All

the people of the province of Carajan are s a id to make a l iving by sa lt made frombrine -we l ls .

7 In the southe rn parts of Tibe t Prope r , s alt is inde ed us ed as money.

In the village of Pas aguk I saw a trad ing-hou s e with a large s tore of s alt in bags .

He re a marke t is he ld fromtime to time , s alt be ing the mediumof exchange .

8

The few glimps e s“

Marco Polo give s us of what lze calls Tibe t are , the refore , so characte ris tic and true for what we cal l Tibe t nowadays , that he very like lyhas kn own the far we s tward exten t of this country and may even have be l ieved thati t borde red upon Kashmir. 9 Accord ing to himTibe t was subject to the Great Kaan ,

1 See Trans-H imalaya, London 1 909 , I , 55, 3 80 , 3 8 1 , 3 85, 3 86 , and III, London 1 9 1 3 , p . 1 0 1,

and several other places .

1»The val ley of Jh ansu (Gyangtse) I understood to b e particu larly famous for themanufacture

of wool len cloth , for which there is a very great demand . T hes e c loths . which are confined to two

colours , garnet and white , s e ldomexceed hal f a yard in breadth : They are woven very thick and c lose ,l i ke our frieze ; they are very soft to the touch , for the fleece of the ir s heep appears to b e remarkablyfine , and supplies an exce l lent material . SAMUE L TURN E R : An Account of an Embassy to the Courtof the Teshoo Lama in T ibet, London 1 800 , p . 2 25. Speaking of the market of Gyangtse , thethird largest in T ibet, next after Lhas a and Shigats e , WADDE LL says that it i s »espec ially celebratedfor i ts woollen cloth and carpet manufactures » . Lhasa and i ts Mysteries , London 1 905, p . 1 96 .

3 »Der Ge isterglaube und die Geisterb annerei , d ie Magie fib erhaupt, spie lt in ihm(demLamaismus) e ine grOs sere Rolle und hat e ine ausgedehntere Praxi s , als in jeder andern bekannten Ge

staltung der Buddhareligion KOPPE N , op. ci t. II , p . 8 2 .

4 FromChumbi WADDE LL writes : »The watch -dogs cha ined up at the doors of the houses gaveus a fierce reception . They are huge Tibetan mastiti s ’

the mastiff dogs’

of which M arco Polowrite s , ’

as b ig as donkeys , which are capital at se iz ing wi ld Op . ci t. p . 89 . In »Trans

H imalaya» I have often mentioned the T ibetan dogs , for instance , fromShigats e : »In the court . alarge black watch-dog, with red eye s and a red swol len ri ng round his neck , is cha ined up , and i s so

savage that he has to b e he ld while we pass », I , p . 3 85.

5 The T ibetan musk i s famous , and has been mentioned by all early travel lers , fromSULE IMANthe merchant

,to the Swedis h prisoners whom Tsar PE TE R he ld in S iberian captivity after 1 709.

And even now the chie f th ing that the Chines e get good fromLhasa is mus k . Waddel l , op.

crt. , p . 3 59 .

6 Wadde ll says of the T ibe tans at Lhasa : »Their friendly demeanor d id not bear out MARCOPoLo

s wholes ale denunciation , that ’

The people of Tebet are an in conditioned Op cit. p . 3 45.7 Yule’s Marco Polo . Book II

,p . 66 .

6 Trans -Himalaya, I I , p . 65. In the German trans lati on more correctly fromthe Swedis horiginal : Sche idemt

'

rnz e , wh ich i s exactly the same as in Marco Polo. See a lso op. ci t. E ngl ishedition I I , p . 75, and II I

,p . 5, 2 3 and 1 8 3 : »F ive sheep ’s loads of salt were equ ival ent to four

sheep’s loads of barley, and the value of every s ixth load of sal t was the duty demanded by theGovernment» .

9 In connection with the pas sage where he talks of 8 kingdoms Yule has a note : »H ere M arco

at least shows that he knew T ibet to b e much more extens ive than the smal l part he had seen . Butbeyond this his information amounts to l ittle .» Op . ci t. Book II, p . 52 .

CHA PTE R XV.

FR IAR ODOR IC . S IR J OHN MANDEV I LLE.

FRI AR ODOR IC is the firs t E u ropean who has eve r vis ited Tibe t Proper andeven Lhasa , and s til l his de s cription of the coun try is ve ry meagre if comparedw ith the info rmation brought back by Marco Polo . From Cathay he trave l leds eve ral day ’s jou rney we s t wards through many countrie s and citie s . Thus he cameto the country of Pre s te r John cal led Penthexoire w ith the capital Cosan . Fromthisprovince he came to anothe r called Cos samw ith the capital Cas sam, which was one

of the twe lve province s of the Great Kaan .

The s cholars who have tried to make out this route , Yule , Klaproth , Pauthie r, and Cordier, have not, as cou ld b e expected froms uch scan ty mate rial , beensucce s s fu l . According to Cordie r the re l igiou s trave lle r went fromPeking towards thecoun try of Tangut, pas s ing through Kan-

su .

Whe reve r his Cos s ammay have be en s ituated , he trave lled fromthe re to

anothe r ve ry great province which has the name Rib oth and borde rs upon India .

Of this kingdom, Tibe t, he saysCe l royaume e s t sub giet au grant Caan , et y treuv e -on pa in e t v in en tres p lus grant

hab ondance qu e en nu l le part du monde . Les gens de ce pais demeurent en tentes de fuerrenoir . Leur mais tre c ité es t mou lt be lle , toute de b lanche p ierre , e t le s rues b ien pavées . E l lee s t appe llee Gota . E n ce s te c ité nu l n ’

os e e spandre s ang humain ,ne aus s i de que lconque be ste

pour la révérence d ’

un ydole qu ’

on y adoure . E n ce ste c ité demeure l ’ob as sy, c’es t a d ire leurpape en leur langaige . Il e s t ch ie f de tous les ydolatre s et donne les b enefices du pay s a

s a gu is e .

1

1 Les voyages en A s ie au XIV° s iécle du bienheureux frere Odoric de Pordenone R eligieux deSaint-Francois publi és avec une introduction e t de s note s par Henri Cord ier. Pari s 1 89 1 , p . 449 e t seq .

1 In HAxLUYT’

s vers ion this passage runs thus : »Ultra ven i ad unumregnumv ocatumT ib ek

quod e st sub jectum Can i , in quo e s t major copia pan is vini , quams i t, in toto nrundo ut credo .

Gens i ll ins terrae moratur commun i ter in ten tori i s factis ex feltri s n igris : Pri ncipal is c ivitas s ua murataes t pu lcherrime ex lapidib us alb iss imi s n igerrimis interescalariter dispos iti s curiose compositis ,omnes v ise ejus optime pav atae . In ista contrata nullus audet efi’undere sangu inemhomin is nec ali

cujus an irnali s , ob rev erentiam un ius Idoli . In ista c ivi tate moratur Abass i I , Papa corum, qu i e s t

caput 81 princeps omn iumIdolatrarum; qu ibus dat distri b uit b eneficia secundummoremeorum; s ieutnoster Papa Romanus est caput omn iumChristianorum.» (Vol. IV, p .

TH E JOURNE Y OF ODORICO DE PORDE NONE . 1 4 1

Then fol lows the Old s tory of the treatmen t of the dead, the head be ing cu t

off the dead fathe r and given to the son who eats it, while the body is cu t to piece sand given to eagle s and vu ltu res .

CORD IER s ugges ts that Gota may b e de rived fromBod , Pot-

pa , Buddha-la ,Po-ta—la , and finds it, at any rate , to b e identical w i th Lhas a , Odoric thus be ingthe firs t E uropean who has eve r vis i ted this city . In his introduction Cordie r s aysthat Odoric re turned from Cathay to E urope v ia Shan-s i , Shen-

s i , S zechuan and

Tibe t, whe reas the res t of h is journey rema ins in darkne s s ; Badakshan , Khoras an ,

Tab ris and Armen ia may have be en on his route .

I H is jou rney was comple tedin 1 3 30 .

The firs t E uropean who eve r ente rs Tibe t and reache s its , late r on , s o de s i redand mys te rious capita l , cal ls the country Riboth and the capital Go ta , and tha t i sall the geography we find in his narrative ! Howeve r, Odoric knew that Tibe tborde red upon India , and he te l ls us that the native s l ived in tents of black fe lt ?The s tone hous es of Lhasa we re white as nowadays , 3 and the s tree ts w e re we l lpaved , a comfort that was abandoned long ago .

4 He was famil iar w ith the firs tof the E ight Pre cepts of Buddhism: ’

One shou ld not de s troy life ’

. And he had in

Lhasa s e en or heard of some high prie s t, »the i r pope » , though not ye t a DALAILAM A .

S ir JOHN MANDEVILLE was a contemporary to Odoric , and pre tended to haves tarted on a jou rney of 34 years in 1 3 2 2 , the gre a tes t part of which was accomplishedon ly in his own imagination . H is narrative was published be tween 1 357 and 1 3 7 1 ,

and he was the he ro of the great uncritica l publ ic for centurie s . More critica l spiritswe re s truck with the s imilari ty , often word for word , be tween Mande v ille

s accountand othe r books of trave l , amongs t themthe narrative of Odoric.

5 Some write rs

1 Op . cit. p . XXIX .

1 H e has there fore trave l led through d i stricts inhabited by nomads , before he reached d istrictswith citie s and temples .

3 Ab b e Huc s ays of the hous es of Lhasa : »elle s s ont entrerement blanch ies a _ l’

eau de chaux ,a l

exception de que lques bordures e t des encadrements des portes e t des fenetres qui sont en rouge

on en jaune Les habitants de Lha-Ssa e'

tant dans l ’us age de pe indre tous les ans leurs maisons ,el les sont habitue l lement d ’

un e admirable propreté , e t para is sent toujours baties de fraiche dateSouven irs d

un voyage dans la Tartarie , le Thibet e t la Chine , 1 844—1 8 46. Paris 1 853 , TomeII

,p . 24 7 .

4 »If the original description of the place by Father Andrade had any real foundation ,the

capita l of T ibet has changed sadly for the wors e , for not even the kind l ie st advocate could find in

the s losh and fi lth of every street, or in the ramshackle structures which cumber every avai lable inchof ground bes ide the heavier houses , the we l l—paved thorough-fare s and dign ified architecture which hedescribe s .» PE R CE VAL LAN DON : Lhasa, London 1 905, V ol. I I , p . 2 06. He obvious ly means Odoric,for Andrade was never in Lhasa .

5 In BE RGE RON’

S work we read : Ces deux voyages d ’

Oderic de M andeville sont si semblablesl’

un a l ’autre , so it aux choses vraies . soit aux fabu leuse s , dont ils sont rempl is , qu ’ i l sembla qu ’

i ls

a ient été pris l’

un de l ’autre ; ma is i ] y a plus d ’

apparence que M andevi l le l ’a it pri s d ’Oderic , qu imourut des l

an 1 3 3 1 Voyages faits princ ipalement en A s ie dans les XII, XIII , XIV e t XVS iecles Tome I , a la Haye 1 73 5, p . 53 .

1 42 FRIAR ODORIC . S IR JOHN MANDEVILLE .

there fore be l ieved that the Francis can Friar had been a compan ion to Mandeville ,and plagiated his diaries . Purchas would not reckon the Friar amongs t his pilgrims ,b ut en terta ined a ve ry high opin ion abou t his countryman .

‘ ASTLE Y says that afte rMarco Polo’

s t ime »we mee t w i th no E uropean trave lle r into Tartary for abovethree hundred years , excepting our Mandeville , who wen t thithe r fifty years afte r».

2

He doe s not be lieve at all in Odoric : »This is a mos t supe rficia l Re lation and fu l lof Lies In short, though he re late s s ome Things of the Tartars and Manci ,which agre e w ith Po lo ’

s Account ; ye t it s eems pla in , fromthe N ame s of Placesand othe r C ircums tance s , that he neve r was in thos e Countries , b ut impos ed on the

Publ ic the few Informations he had fromothe rs , mixed w ith the many Fictions of

his own . » 3

As a matte r of fact Odoric was a s imple b ut honourable man , and Mandeville a we l l-re ad and inte l ligen t impos tor. H e has s to len eve ry word in his narrat ive , except the des cription of the Sul tan of E gypt and the l ife at his cou rt. He

has , in a mos t uns crupu lous way, s tolen whole pas sage s , ve rbally , fromOdoric,Haiton ,

Plano Carpin i , and even from Pliny and othe r clas s ics , as we l l as frommany books of his own time . He neve r made the journey he de scribes .

4

One place , howeve r , he gene rous ly confe s se s not to have vis ited ; the talk isof the firs t Pe rs ian kingdomwhich begins toward the eas t, toward the kingdomof

Turke s tan , and s tre tche s toward the we s t , >unt0 the Ryv ere of Phison , that is on

of the 4 Ryv ere s , that comen out of Paradys » , b u t : »of Paradys ne can not I spekenpropurly : for I was not the re So much is su re that in the middle of paradis e ,and at its highes t place the re is a we l l , >>that cas te the out the 4 E lode s , that rennenb e dyv ers e Londes : of the whiche the firs t is clept Phison or Ganges , that is al leon ; and i t

'

renne the thorghe out Ynde or Emlak : in the whiche Ryv ere b en manyepreciou se S tone s , and moche l of Lignu A loe s , and moche grave l le of Gold» .

The Gange s wate r is in some places clear, in othe rs troubled , in some placeshot , in othe rs cool . And no mortal man can approach Paradis e ; »and b e the Ry

vere s may no man go ; for the wate r renne the s o rude ly and so sharply , becaus ethat i t come the down s o outrage ous ly fromthe h ighe place s ab ov en

Unde r s uch cond itions it is clear that the wonde rfu l , though brie f obs ervat ionsOdoric brought back fromTibe t wou ld not b e le ft a lone by the noble kn ight, in

1 H enri Cordi er in Yule ’s M arco Polo , I I , p . 602 .

2 A new Genera l Collection of Voyages and Trave ls , London 1 747 , V ol. IV , p . 542 .

3 Op . c i t. IV,p . 6 20 .

4 Untersuchungen fi ber Johann v on Mandeville i 1nd die‘Quellen se iner Re isebeschre ibung. Von

ALBE RT BOV E N SCH E N . Ze its chr. d . G esellsch . f. E rdkunde 7 u Berl in . Band 2 3 . Berlin 1 8 8 8 , p 3 05.

»Comme on le sait, Mandeville parai t devoir etre l’ouv rage d nu habile géographe en chambre qui ne

s era it autre que le médecin Jean de Bourgogne ou Jean 9. la Barbe , d apres umpas sage d ’

nu chron iqueurde Liege , Jean d ’

Outremeuse , découvert par le Dr. S . Bormans . Cord ier, T oung pao, V ol. I , Le ide1 8 o , 3 45.9 P

5 The V oiage and Trav aile of S ir John M aundev i le , K t. R eprinted fromthe E dition of A . D .

1 72 5. J . O . HALL IWE LL . London 1 8 8 3 , p . 2 58 and 3 03 e t seq .

CHA PTE R XV I.

EAST IND IAN TRAVELLERS IN TH E S IXTEENTH AND

SEVENTEENTH CENTUR IES .

A long t ime shou ld e laps e before the E uropeans , ANDR ADE , GRUE BE R and

DORVILLE ,DE SIDE R I and FRE YRE ,

and the Capuchin Mis s ionarie s vis ited Tibe t. And

w i th some reason Andrade cou ld b e sa id to have d iscove red the mys te rious country ,for at this time nobody s e ems to have remembe red the vague information broughtback by Odoric. N or had the wonde rfu l tale s of Plano Carpin i , Rub ruck and MarcoPolo made any de ep impre s s ion in E u rope , and Tibe t had inde ed to b e red iscove red.

The name of the country , howeve r, was some times , though ve ry rare ly, me nt ioned by trave lle rs who vis i ted India even before Andrade ’s journey , though theyd id not know whe re i t was s i tuated any more than did the Arabian geographe rs .

So much the more atten tion was attracted by the sou rce of one of the threerivers , which are the object of this his torical account, name ly, the Indus , and gene ra l ly in connection with specu lations regard ing the sou rce of the Gange s . The following repre s entative s of the great clas s of E as t India trave l lers wil l give an idea of the

know ledge abou t the countries jus t north of Ind ia as pos se s s ed by E urope in the

s econd half of the 1 6th and the firs t hal f of the i 7th century . Al l the informationthey we re ab le to gathe r was founded on hearsay , and there fore froma geographica lpoint of view of very little va lue .

When ANTHONY JE NKI NSON , in 1 558 , vis ited the wes te rn parts of Centra lAs ia , he only heard that the Oxus had its s ou rce s on the moun tains of »Paraponisu s

» in

India and in Bokhara he me t me rchants who came from»the furthe s t parts of India ,even from the country of Bengala , the river Ganges » ;

Iand amongs t the note s

R ICH ARD JOHNSON , a year late r, gathe red of d iffe rent roads fromRus s ia to Cathay ,we find one , given by a Tartarian me rchant in Bokhara that twenty days ’ journeyfromCathay was a country , »whe re l ive th the beas t tha t beare th the be s t Muske » , 2a s tatement in which we fee l , as it we re , a sme l l of dis tant Tibe t .

1 R ICHARD H AKLUYT : The Princ ipal Navigations Voyages Traffiques Discoveries etc. G lasgow 1 903 , V ol. II, p . 46 1 and 4 7 2 .

2 Op . cit. p . 48 2 .

JE NKINSON , LINSCOUTE N AND FINCH . 1 45

Mos t of the trave l le rs to India vis ited on ly the s ea ports and had nothing to

do in the northe rn parts of Hindos tan ; the re fore the Indus often escaped the ir attention ,

while the more famous Gange s is remembe red in the ir narrat ives . Thus forins tance , CE SAR FRE DE RI CKE ( 1 563 has a good dea l to say of the Gange s ,

I

and RALPH FITCH ,who in 1 58 3

—1 59 1 trave l led to Goa , Benga l and Pegu , evenknows that the Ganges come s from the N .W .

, and runs eas twards to the Gu l fof Bengal afte r rece iving many tributarie s as large as the ma in rive r its e lf. 2 JA NHUYGHE N VAN LLN SCOUTE N ,

1 58 3 , devotes a s pecia l chapte r to Benga l and the rive rGange s , in which he s ays that the source of the rive r is unknown , and whe re we

be l ieve we recogn is e a re fe rence to Empe ror Akb ar’s -exploring expedition to the

source , mixed up with the old lege nd of the Pison of Paradise , and some clas s ica lnotion of the Ganges .

3

On his journey round the world from 1 586 to 1 605 PE DRO TE IXE IR A vis i tedGoa, the capita l of Portugue s e India , whe re Linscouten , who vis i ted many place son both Indian pen insu las had been re s iding some years . In his chapte r on the

»K z

ngs of P ers i a » , Te ixe ira has a digre s s ion on India , in which he re fe rs to the

Indus and its affluents , and e numerates variou s kingdoms in the N .W . of India .

He mentions the crocod ile s of the Gange s , and has heard tha t »the mos t and be s tof the rhubarb come s fromGax Ghar or Ka

'

x Ghar (Kashgar), a city of Usbek , a

province near Ke thao KothanWILLIAM FINCH , the me rchant, knew , in 1 6 1 0 , that Ravi is »a goodly R ive r

which fal le th into Indus » , and that the Gange s is three quarte rs of a mi le broad ,rece ive s 30 tribu tarie s , swe l ls ove r its banks a fte r heavy ra ins and falls into the Gu lfof Benga l . 5 Of Ka shmir he has learn t that i t is a s trong ci ty on the rive r Bahat ,and that the country is a goodly pla in , lying on the mounta ins . »This Country iscold , s ubject to fros ts , and great snowe s , neare to Cas car, b u t s eparated w i th suchmountaine s , that the re is no pas sage for Caravans : ye t the re commeth oft- time sMusk , w i th S ilke and othe r me rchandize this way by men and goods Upon

HAKLUYT, op. cit. Vol. V , p . 3 65.

2 Ibidem,p . 4 76.

3 N av igatio ac itinerariumJohann i s H ugonis Linscotani in Orientalems ive Lu s i tanorumInd iam.

Hagae-Comitis 1 599. The passage about the Ganges runs as fol lows (p . Ad finemregion is Orixae

tractusque Choromande l , flumen Ganges inc ipit in regno Bengalae . Hoc fluentuminter orbis aquasclari s s imum es t. Origo ignoratur, quamqu idamex Paradiso repetun t fabulae intu itu , quamBengalaehomines referunt. Rex qu idam (i ta ajunt) cupidus cognoscendi ortus , quosdameducari juss it piscecrudo , ut i is commodius u teretur. Hos nav i b us impos itos advers o flumine exploratores mi s it, qui demumpost quorundammens ium nav igationé dev enére in loca , ub i odore suav is s imo aermi tis s imus luceret,aqu is leni ter stagnantibus , impleta fide Paradi s i . Cumautemu lterius tendere conaren tur, ips i qu iden ihi l profecerut, repugnante val ido fiurn ini s cursu : Ideoque revers i ad Regem,

quae visa fueran t expos uerunt, s imu lque credi tumes t, Gangemex Paradiso aquas in nostrumorhemferre .

4 The Trave ls of Pedro Te ixe ira ; w ith hi s »Kings of H armuz» , and extracts fromhi s »Kings ofPers ia» . Trans l . and annotated by W. F . S IN CLAIR . London

,H akluyt Society 1 902 , p . 2 09 e t s eq .

Some of the passages which wou ld have been most intere sting to us are mis s ing in thi s edition .

5 Hakluytus Posthumus or Purchas H is P ilgrimes , by SAMUE L PURCHA S . V ol . IV . G lasgow1 905, p . 52 .

1 9—1 3 1 3 8 7 I

[46 E AST INDIAN TRAVE LLE RS IN TH E S IXTE E NTH AND SEVE NTE E NTH CE NTURIE S .

the s e mountaine s ke eps a smal l King called T ibb ot , who of late s e nt one of his

daughte rs to Sha S e l imto make affin it ie .» By Tib b ot he obviou s ly means Ladak ,

as it is mentioned in connect ion w ith the d ifficu lt road ove r the Kara-korumto Kashgar.The next trave l le r of fame is E DWARD TE RRY, Mas te r of Arts and s tudent

of Chris t Church at Oxford, who has given us a re lation of a Voyage to the E as te rn India , undertaken in I 6 1 6 .

I In this he te lls u s that the large Empire of the

great Mogol is bounded on the north by the mountains of Caucasus , and Tartaria ,which shows that he e l iminate s Tibe t , and make s Caucasus , or Himalaya , a d ividing wal l be tween Ind ia and Tarta ria . He mentions Kashmir (Chishmeere) and its

capita l S iranakar. Amongs t the countrie s of N .W . India he mentions Jengapor withi ts capita l of the same name and s i tuated on the »R ive r Kaul , one of the fiv e R ive rsthat wate r Penjab » , and probably corresponding to the Satlej as , on maps fromthe

time , it is marked as the las t great tribu tary to the Indu s fromthe eas t. 2The ne x t province is »Syb a , the chiefe Citie is ca l led Hardwa ir, where the

famous R ive r Gange s s eemed to begin , is su ing out of a Rocke , which the s upe rs titiou s Gentile s imagine to b e l ike a Cowe s he ad , which of all s ens ible Creature sthey love be s t . Kakares is sa id to b e >>ve ry large and exceeding moun ta inou s ,d ivided from Ta rtaria by the Mountaynes of Caucas e s ». And Gor is a ls o fu l l ofmounta ins .

As in the clas s ica l write rs the Indus and Ganges are mentioned toge the r : »ThisRegion is wate red w ith many goodly R ive rs , the two principal] are Indu s and Gange s ,whe re this thing remarkable mus t not pas s e , that one pinte of the Wate r of Gange sw e ighe th le s s e by an ounce , then any in the whole K ingdome , and the re fore the

Mogol whe resoeve r he e is , ha th it brought to himthat he may drinke it .» 3 »Negracut» is mentioned amongs t the mos t famous places of pilgrimage .

Much the same geographical horizon i s commanded by S irTHOMAS ROE in 1 6 1 7 .

Kabu l borde rs upon Tartaria . Through >>Kyshmier» w ith the chie f city »S iriv aker» runs therive r of Bhat »and finde th the S ea by Gange s , or some say of i t s e lfe in the north part ofthe Bay of Bengala Gor l ie s toward the head of the Gange s . The Empireof Mogor i s very great and »s tre tche th to the Mountaines of Taurus North» . In a

le tter of 1 6 1 5 S ir Thomas Roe cri ti cis e s the exis ting maps : »I have one Obs e rvationmore to make of the fals ene s s e of our Maps , both of Me rcator, and all othe rs , and

the i r ignorance in this Countrey. Firs t, the famous R ive r Indus doth not empt iehims e lf into the S ea at Cambaya , as his chiefe mou th , b ut at S inde He

mis s e s Agra on the maps , and knows that it is a rive r that fal ls into the Gange s .

The Ganges i s e special ly attractive to the imagi nation of the world , as can

b e s e en in a letter of 1 6 1 5 by THOMAS CORYAT ,who, in an oration to the Great

I Purchas H i s P ilgr imes , V ol. IX , p . 1 3 et seq .

2 Compare for instance , N ICOLAUS V I SSCH ER ’

S map of about 1 680 (See below P l . XXXII) .

3 Compare AIN -I -AKBAR I , supra,p . 73 .

4 PUR CHAS , V ol. IV , p . 43 1 et seq .

TERRY , ROE , HE RBE RT AND MANDE LSLO . 1 47

Mogol s aid that one of the four caus e s which had brought himby land fromJe rusalemto India was : »to see you r famous R ive r Gange s , which is the Captayne of all

the R ive rs of the World» .

I

Regarding the mounta in wal l to the north, PIE TRO DE LLA VALLE had the

s ame opin ion as his prede ces sors . He says of Shah Se l imthat he was king of the

greate r part of India »be tween Indu s and Ganges , and whose Coun tries are extendedNorthwards as far as the cl i ffs of Mount Taurus or Imans , whe re i t d ivide s IndiafromTartaria» .

2

Nor did THOMAS HE RBE RT increas e the s tore of geographical knowledge of

his_time , so far as India is conce rned . He began his great journey in 1 626 , afte r

Andrade had re turned fromhis firs t vis it to Tibe t . What he says of the Gangeswate r i s taken di rectly fromE dward Te rry. Conce rn ing Cas s imer and its Me tropol isShyrenaker he has nothing new to te l l . Panch—ob i s a Pe rs ian word mean ing fiv ewate rs , from the rive rs Rav ee , Behat, Ob -Chan , Wihy, and Synde . Some w ri te rs ,he thinks , have given too great l imi ts to the garden of Paradis e in making both theN i le and the Gange s ris e the re . The Gange s ris e s fromImaus in S cythia .

3

It is a lways the same s tory that is told, and no new trave l le r dare s to s e e

anything that has not been noted by his predecessors . The re are long de scriptions of the marve ls witne s s ed in India and long dis cuss ions upon his torical events ,of cus toms and manne rs of the natives , and even of an imals and plants . But the

ge ography is a lways the same . There are always the two great rive rs w i th or w ithout tribu taries , there is the wal l of mountains to the north , s eparating Hindos tanfromTartaria or Scythia , nothing more than was known , and some time s be tte r knownby clas s ica l anti qu ity. In eve ry new narrative we recogn is e Ptolemy ’s geography .

The grip of the great Alexandrine ge ographer is s ti l l so s trong, that he ove rru le sthe common s ens e and the facu lty of obs e rvation of trave l le rs in so recent time s .

Ptolemy , and even earlie r writers are often quoted . The northe rn mountains are

cal led Imans , Tau rus or Caucas us . The Satlej was much be tte r repre s ented by Ptolemy than by anyone of the trave l lers menti oned above , who as a ru le did not evenknow the exis tence of this rive r. The trave lle rs do not s e emto have had any con

fidence in thems e lves . And if they went beyond Ptolemy , they s ti ll had to re fe rto somebody e ls e , for ins tance Te rry or Roe , who we re often s imply copied.

So is also the cas e with JOH AN ALBRE CHT VON MANDE LSLO , who trave l ledin India in 1 63 8

—39. He firmly be l ieves in the au thority of S ir Thomas Roe , though

in one important po int he is independent, name ly , regarding the sou rce of the Indus .

Speaking of the rive rs of Penjab he s ays that the firs t, Bagal or Begal , has its sourceIbidemp . 442 .

2 The Trave ls of Pietro de l la Va lle in Ind ia. E d ited by E DWARD GRE Y. London ,Hakluyt

Society,1 892 , V ol. I , p . 48 .

3 Some Years Trave ls into A frica A s ia the Great. E spec ial ly Descri b ing the Famous Empi res of Pers ia and Indus tant . As also Divers other Kingdoms in the Orientall Ind ies , and I ’ les Adjacent.By THo. H E RBE RT E sq . London 1 63 8 , p . 65 et seq .

, and 2 2 1 . Pl . X i s themap i l lustrating H erbert’s journey .

1 48 E AST INDIAN TRAVE LLE RS IN THE SIXTE E NTH AND SEVE NTE E NTH CE NTURIE S .

near Kabu l ; the s econd , Chanab , takes i ts origin in the provi nce of Q ue smir or Cas samie r ; the third , Rayy , has its s ource in the ne ighbou rhood of Lahor, while the fou rthand fifth , V ia and S ind, »come frommuch farther Two pages fu rthe r onhe s ays that S rinagar is on the rive r Be zat or Badt , which , afte r a great détour,falls in to the Ganges . Kashmir borders on Kabu l , and is fai rly cold on account ofthe mountains ; b u t cons idering that »the Kingdomof Thibe t s e rves as its frontie ron the northe rn s ide >, the cl imate of Kashmir is rathe r tempe rate .

R egarding the northe rn provinces he give s the s ame information that had bee ngi ven by Te rry . The province of S iba with its capital of Hardware , is s ituatedbe twe e n N augraku t and Pitan : »The re the rive r Ganges takes its source .» And thenfol low s the ordinary s tory abou t the divine rock w i th the Cow

s Head , fromwhichthe rive r is sue s . E ve n in the new ed ition of 1 7 27 Mount Caucasus s eparates the

Empire of Mogol fromTartaria .

In Be rge ron ’

s edition of VINCE NT LE BLANC ’

S voyage s we meet again the

ordinary s tore of know ledge regarding the two famous rive rs . India has de rived itsname fromthe ce lebrated rive r Indu s or Schind , which take s its origin fromthe highmoun tains of Caucasus and ParOpamisus , »which are sa id nowadays to b e the Nau

gracot l’

V s sonte .

2

S ome people be l ieve , though w i thou t reas on , that the rive r Ganges is one of

the fou r rive rs of Paradis e cal led Philon or G ihon . There is a great di ffe rence of

Opin ion amongs t mode rn s cholars , he says , whe ther this river is the real Ganges of

the ancients , or whe the r the ancient Gange s shou ld not rather b e the same as the

rive r of Can ton in China , or any s ti l l farther eas tern rive r. He , howeve r, leaves thisques tion to b e solved by others , and s atisfies hims e l f by saying, that all the Portugue s e and 'many othe rs regard the two rive rs as identical , as the name is s ti l l thes ame as in an tiqu i ty . The s ame opin ion is confirmed by the Je su i t fathe rs in the irnew reports on the great Kingdomof Tebe t or Tibe t and Cathay . The river has itss ource on the mountains of High India not far fromthat of the Indus , and the na

t iv es think it is unknown , as coming fromParadis e .

A »certain dutchman» in Pegu gave himsome information about Great Tartary , amongs t the nat ions of which are s uch as the Mongal , Tangor and Theb et or

Tibe t . North of Pegu are desolated regions and s andy des e rts , where we l ls are to

b e found on ly by the he lp of gu ide s ; in High As ia one cou ld march 2 2 days s ee ing

2 Voyages Ce lebres e t remarquable s , faits de Perse aux Indes Orien tales par le Sr. JEAN -ALBE RTD E MAN DE LSLO , Gentilhomme des Ambas sadeurs du Duc D E HOLSTE IN en Moscov ie Perse , etc.

Mis en ordre publ ie z , apres la mort de l ’I llus tre Voyageur, par le Sr. ADAM OLE ARIUS , Tradu itsde I

Ori ginal par le Sr. A . de WICQUE FORT Tome I , Amsterdam 1 7 2 7 , p . 46 . The origina l narrative of Mande lslo

s journey to India which he undertook after the embas sy to Pers ia , has the titleDas H och E delgeb ornen Johan A lbrechts v on M ande ls lo Morgenlandische Reyse

-Beschre ibung .

H erausgegeben durch ADAM OLE AR IUM , Schleszwi g 1 658 .

2 Les Voyages fameux du S ieur V incent le B lanc Redigez fidellement sur ses MemoiresRegistre s , tire z de la B ibl iotheque de Mon s ieur de Pe iresc par P ierre Bergeron . Paris 1 649,p . 67 , 1 25 e t s eq. and 260 e t s eq .

1 50 E AST INDIAN TRAVE LLE RS IN THE SIXTE E NTH AND SEVE NTE ENTH CE NTURIE S.

short , though very interes ting account of what he has heard of the mys te rious and

inacces s ible country .

»The Kingdomof Bou tan is of ve ry great extent , b ut w e have not ye t be enable to acqu ire an exact know ledge of i t. » I t was only on his vari ous journeys inInd ia that he met people from»Bou tan» , especial ly at the marke t of Patna whe re theme rchants u s ed to s e l l the musk . Gorrochepour i s the las t city of Great Mogol inthe d irection of Boutan . He has heard of a road which is ve ry di fficult »on accountof the mountains which nearly always are covered with snow , and of the great dezs e rts one has to trave rs e in the pla ins .» The caravans ne ed three mon ths fromPatnato Bou tan »Five or s ix l ieues beyond Gorrochepour one enters the te rritory of the

Raja of Nupa l , s tre tching to the ve ry fronti e r of the Kingdomof Bou tan» The

capital of Nupal is also called Nupal . The caravans arrive at the »foot of the highmountains known nowadays unde r the name of N augrocot , and which cannot b ecros sed in les s than n ine or ten days , as they are ve ry high and ve ry narrow ,

withprecipices Afte r pas s ing thes e mountains one us es oxen , came ls , and hors es fortransport to Bou tan . »

To the south fromBoutan are high mountains and narrow pas sage s ; »to the

north the re is nothing b ut fore s ts , nearly a lways s now ; 81 both to the E as tand the We s t the re are vas t des e rts , where one hardly finds anything b ut saltwate r

This des crip tion of Tibe t, which he picked up at Patna , has a good deal ofrea l ity . It is on ly curious that he has not heard the name of Lhasa, for his Bou tani s obviou s ly Tibe t Prope r. The tale abou t the fore s ts is much exagge rated , b utthe re is e te rna l s now on the mounta ins

,and furthe r north, in the des erts of the high

plateau-land, nume rou s s alt-lakes shou ld indeed b e dis cove red in late r years . The

road Tave rn ie r des cribes is the s ame which was already known by Alb eruni ,going through Nepal v ia Katmandu to the Tibe tan frontier and further throughsou thern Tibe t .Abou t the s ame time WALTE R SCHOUTE N trave lled to the E as t Indie s , and

heard that Kashmi r s tre tches towards the eas t betwe en Great and Little Tibe t . 2The city of Cas s imir is des cribed as s i tuated in the middle of a great plain , s ur

rounded on all s ides by high mounta ins , »which s tre tch at leas t 9 or 1 0 cos to the

north» . During Novembe r , Decembe r, and January there are con tinual rains and snow ,

and the mountains are cove red with s now which give s ris e to some great rivers .

S chou ten the refore be l ieved that the be lt of mountains north of Kashmi rwas ve ry narrow , more l ike a wal l s eparating this

country fromTartary . Somefifty years e arlie r Finch had a more correct view , for he speaks of nearly insurmountable moun ta ins ou the caravan road be twe en Kas hmi r and Kas hgar. On the

Supra p . 55.

2 Voyage de Gautier Schouten aux In des Orientale s , Commence l’an 1 658 81 fini l’

an 1 665.

Tradu it du Hol landais . Tome premier. Amsterdam1 708 , p . LXIX.

SCHOUTE N , CH ARDIN , THEVE NOT . 1 51

o the r hand S chouten s e ems to b e the firs t to point out the fact , ve ry natural ini tse lf, that the snows in the s e mountains give ris e to great rive rs .

The Ptolemman partition of India in intra and extra Gangemis accepted byS chou ten in a more modern form. R egarding the identification of the Gange swith the Pison of Parad is e , and the country of Havi la w ith India , he on ly quotesthe opin ion of the interpre ters of the Holy S cript . »The Ganges takes its rise inthe mountains which are s i tuated in the northe rnmos t parts of India . Some havewritten that it begins in Mount Caucasu s , othe rs as s ert that i t come s fromthe

high mounta ins of Theb e t , which are enti re ly cove red with snow . » He a lso ta lks of

the »northern mounta ins which are along the Ganges » , 2 which doe s not we l l agreew ith the orie ntation of the rive r fromnorth to sou th , nor w ith his s aying that therive r flows a long the mounta inou s province s of Mev at and N ahracut , the northe rnmos t provinces of G reat India .

3

But when he te l ls u s the old s tory of the Ganges taking its origin fromthe

rock w ith the Cow’

s Head, he quotes VAN TWIST , who has borrowed his s tatemen tfromE do ii ard Terrijns , as E dward Te rry is cal led in the book .

4

CH ARD IN , 1 664 to 1 68 1 , s pecial ly famou s for his jou rneys i n Pers ia,has a

rathe r antiquated view of As iatic geography , and quotes Pliny, Cu rtius and S trabo ,5

while THE VE NOT , who arrived at Su rat in '

1 666 , has more independent notions of

our regions . He has , however, not much roomle ft for Tibe t, which as so often1 Op . cit. Tome II, publ . 1 7 25, p . 1 60.

2 Op . ci t. I , p . 2 2 4 .

3 Op . c it. II , p . 1 60 .

4 Tweede Dee l v an het Begin ende V oortgangh der V ereen ighde N ederlantsche Geoctroyeerde

C ost-Indis che Compagn ie 1 646 such i s the title of a great book contain ing the descriptions ofs ome ten d ifferent journeys to the east, each with s eparate pagination . There is noth ing b ut ships ,coasts and trade , and what the different capta ins have seen in seaports and on is lands , and there isv ery l ittle of the interior of India . The Ganges i s on ly mentioned once or twice , the Indus on ly once .

A part of the work , occupying 1 1 2 page s , has the promis ing title : Generale Beschrijv inghe v an In

d ien Door Johan v an Twist, Ghewesen Over-hooft vande N ederlantsch Comptooren 1 63 8 .

But nearly all he knows he has borrowed fromothers , special ly Te ixe ira , Terry and Roe . Thus weagain get the description of the fiv e rivers of Penjab , of the Gange s and the rock with th e Cow’

s H ead ,a nd Kashmir wi th i ts capital Syranakar. Mandels lo who trave l led in 1 63 8 , had heard that Bezat or

R adt falls into the Ganges ; v an Twist says it i s a question whether this tributary fromKashmir b elongs to the Indus or the Gange s , b ut finds the latter alternative least probable . K ashmir i s a mountainous country with the K ingdomof Tlieb et (probably misprint for Thieb et) to the east. Then fol lowsthe ordinary account of N agrakut with i ts temple and its pilgrims , and a place of pilgrimage cal ledK alamaka , where flames constantly rise fromthe cold springs and hard rocks . Accord ing to h is de

s cription H ardwar i s s ituated in the province of S iba , which agrees w ith the maps of the time , and

the mounta ins of N agrakot further north . About the source of the Gange s he'

says : »S iba,wiens Hooft

.stadt i s Hardwari , waer v an daen dat gelooft wort, de Ganges haer oorspronck te heb b en nyt een

Rots e , die dit Supersti tieus Volck haer inb ee lden een koeyen Hoft te heb b en e tc.

5 »Ou sa it que I’

As ie est d ivisée par une chaine de montagnes d ’

un bout a l ’autre , dont le strois plus hautes parties ont été nomin ees Taurus , Imans , Caucase . La premiere est la plus avancéedans l

As ie , on appel le toute cette chaine en genera l le mont Taurus . Je d is en généra le , parceq ue chaque partie a son nomparticul ier connu par chaque Nation qu i en es t proche Voyagesd e Mons ieur le Cheval ier CHARD IN , R 11 Pers e , e t autres l ieux de l ’orient, Tome 1 . Amsterdam 1 7 1 1 ,p . 1 1 0 .

1 52 E AST INDIAN TRAVE LLE RS IN THE S IXTE E NTH AND SEVE NTE E NTH CE NTURIE S .

has been the case , both in old and recent times , he reckons on ly as a part of Tartary.

For he s imply says that India borde rs to the north upon Tartary , and that the bes tknown part of India , name ly Indos tan , is bordered to the eas t and wes t by the Gangesand the Indu s , »which have the i r sou rce s in the mounta ins of Zagatay Turquestan .

The s e two countrie s borde r Indos tan to the north , as the Indian S ea s e rve s as its

boundary on the sou th .»

Kab ulis tan also borde rs upon Tartary to the north , on ly s eparated fromi tby mount Caucasus . Two of the rive rs which con tribu te to the Indus have the i rsource s in the s e mounta ins . The Kingdomof Kas hmir has »to the eas t a part ofTibe t» , and to the north , Tartary . It is , howeve r , su rrounded bymoun tains , and one

cannot enter the country except by pas s es and narrow gorge s . The rivers coming downfromthe s e mountains make this coun try one of the mos t fe rtile in India . They jo in and

formthe Tchenas which , through a narrow val ley in the mountains goes to the Indusnear Atoc. The capital of Kashmi r, Syrenaqu er, he place s at 3 5

° N , which is nearlyone degree too far north . He compla ins w i th some reason thatmode rn w ri te rs havegi ven so many di fferent name s to the fiv e rive rs of Panjab , that i t is hard to dis tingu ishthem from each othe r, though Ptolemy had dis tingu ished themsufficiently clearly.

So much i s s u re , that »all thes e rive rs have the ir s ource s in the northe rn mounta ins .»

In THOMAS BOWRE Y ’

s accoun t we on ly find a short re fe re nce to the Ganges,

which he cal ls one of the fou r principal rive rs in the world, »s uppos ed to runne up

Paradis e or the garden of E den» .

2 And he adds : »Forme rly , yea not many yearsagoe , the Inhabitants on the Northerne parts of Bengala tra ined up the i r Children(from the ire infancie) to E at raw fish and flesh , and when growne Up S ent themupon trav aile to discove r the great Ganges

,to find out the garden of E den (by

Orde r of the ire Kings), b u t few or none E ve r returne , ergo now qu i te le ft off as a

thinge Impos s ible to b e accomplished .» 3

Bowrey knows that sou the rn Tartary is mounta inous , and that it rains there»for a quarter of a yeare togeather, and rushe th downe the Ganges and arms the reofwi th Such Violence that all this rive rs afford not is sue for i t, unto the S ea , unti l] acons ide rable t ime b e spent and an abatement made » . S ir R ICHARD TEMPLE is rightin s e e ing in this pas sage a vague re ference to Central As ia , including the Himalayasand the i r heav y ra infal l .

Final ly le t us remembe r JOHN FRYE R , though he has even les s to te l l abou tour regions , than any of h is predece s sors . He has the curious conception that theGhats run north and s ou th >>ti ll they cros s the Imau s » .

4 He also us es the name2 Le s Voyages de Mr. D E THEV E N OT aux Inde s Orientales Trois ieme Partie , Paris 1 689,p . 8 et s eq .

2 A Geograph ical Account of Countries round the Bay of Benga l 1 669 to 1 679 by THOMASBOWR E Y E d ited by S ir R ICHARD TE MPLE , Cambridge , Hakluyt Soc iety , 1 905, p . 1 66 .

Op . c it. p . 2 1 1 . Compare L inscoten ’

s account, p . 1 45 supra , note .

4 A N ew Account of E ast-Ind ia and Pers ia, in E ight Letters be ing N ine Years Trave ls . Begun

1 67 2 . And F in ish ed 1 68 1 . By JOHN FRYE R ,M . D . London 1 698 , p . 1 8 7 . A new ed ition , by

W ILL IAM CROOKE , was publi shed for the H akluyt Soc iety 1 9 1 2 .

CHA PTE R XVI I.

W H EN D ID THE NAME MANASAROVAR BECOME KNOWNTO EUROPE ?

In s eve ral books deal ing with S .W . Tibe t a s tatement is found that the Portugu es e Jesu it Father ANTONIO DE MONSE RR ATE shou ld b e the firs t E u ropean whoeve r heard of and men tioned the lake Manas arov ar. This s tatement s e ems firs t tohave been promu lgated by Capta in .F . WILFORD in 1 808 .

I But Wilford goesmuch farthe r, and what he has to te l l u s of the hi s tory of the s acred lake cannotinspire u s wi th any confidence in his as s e rtion regarding Mons e rrate .

2 For he

pos i tive ly as s e rts that the lake was mentioned by Pliny and Marco Polo , and probably by Ctes ias . We do not ne ed to was te words on Pliny and Ctes ias and the irpre tended know ledge of the Manasarov ar, which is , of course , absurd .

Regarding Marco Po lo ,Wilford probably means Chapter XLVII : 3 »Con cemz

i ng z/ze P rov i n ce of which is said to b e lying towards the we s t . Al lthat Marco Polo s ays is : »There is a lake he re , in which are found pearls whichare white b u t not round.

» If RAMUSIO is right in his vers ion : »a great salt lake » ,the identification wi th Manas arov ar becomes so much the more impos s ible , dis regarding the geographical s i tuation of Ca indu .

4

The s tatement abou t Mons e rrate , on the othe r hand, is s o pos i tive , and so

detai le d that i t cannot s imply b e d ismissed as cons tructed by Wilford ’

s imagina

An E ssay on the Sacred Is les in the West, with other E s says connected with that work . ByCaptain F . W ILFORD . A s iatic R esearches . VIII

,1 808 , p . 3 2 7 et s eq .

2 H is own words should b e observed in this conn ection : »The lake of M an -saraur is menti oned by P lin y, and it i s probably the same that i s mentioned by Ctes ias , who says it was e ighthundred stad ia in c ircumference . M . Polo de scribes it as to the

'

West of T ibet, but does notmentioni ts name . It i s noticed by P . Mons errat, who accompan ied the Emperor Acb ar in his expeditionto Cabul , in the year 1 58 1 . H e ca lls it Mansar-nor, and , fromthe report of pilgrims , places it inthirty-two degrees of lati tude North ; and about three hundred and fi fty mi les to the North-E ast ofSerhind» .

3 YuLE’

s Marco Polo , V ol . I I , p . 53 .

4 Al l the materia l ava ilable for the identification of Marco Polo’s lake is brought together byY ULE and CORD IE R . Op . cit. p . 7 2 .

ANTONIO DE MONSE RRATE . 1 55

tion . The mis s ionary is even reported to have placed the lake at 3 2° N , 1 50 years

be fore d’

Anv i lle placed i t be tween 3 1’/z and 3 2

° N ; in real ity it is s ituated b etwe en and 3 1

° N . Nor is the dis tance of 3 50 mi les fromS erhind to the Ma

nasarov ar much exagge rated, as it is in real ity 300 mi le s . Finally,the lake is said

to b e NE . of S erhind though it is nearly due eas t.As before mentioned, Wilford has been quoted in many books , and even

the great Carl R i tter accepted his authority w i thou t res e rvation .

I E ve rythings eemed to b e au thentic. A Catalon ian Jes u i t , Father Anton io de Monse rrate , wasborn in 1 536 , and died in 1 600. He went out to Goa , and for some years ente redinto the s ervice of Empe ror Akbar in 1 580 , jus t a year be fore the journey to

Kabu l . 2 Mons e rrate and the othe r mis s ionaries who on Akb ar’

s reque s t trave lledto Fathpu r we re rece ived w ith the gre ates t hospitali ty by the Empe ror, and as theywere even invited to accompany the Great Mogu l on some of his jou rneys i t is no .

wonde r that the Mohammedan court became enviou s .3

»Von Augenzeugen, we lche die Routen zwischen K aschmir, Ladakh und Yarkand wiederholtzurtickgelegt hatten , erfuhr A . BURN E s , wie auch n icht anders zu erwarten war, und was schon s ehrfriihze itig demPater MON TSE RRAT , der den Ka iser AKBAR , imJ . 1 58 1 , auf se inemZuge nach Kabulbegle itete , bekannt war, die Bestatigung das s der S trorn an demd ie Capitale Leh ge legen , wirkl ichaus der Nahe des Manasarowar S ees entspringe , e inen “

s ehr lange n Indusarmausmache E rdkundev on A s ien , Band V , Th . VII , Berl in 1 8 3 7 , p . 1 2 .

2 The K ingdomof Kabu l was under Akbar’s brother, M uhammed HakimMirza , who often rebe l led , so that Akbar had to reduce him to submis s ion , which s eems to have happened in 1 58 2

(Whee ler, H istory of Ind ia,V ol. IV , p . Further it i s known that Akbar took a great interest

in Christian ity . H e wrote to the Portugues e V iceroy at Goa , inviting some of the Fathers to comeand instruct h im. The three Fathers Aquavi va , Mons errato , and E nriques were despatched . (Ibidem,p . On ly the most able and learned Fathers at Goa were chosen . A fter a d ifficult journey of4 3 days they arrived in Fathpur, February 1 8 , 1 580 (Kaiser Akbar, E in Versuch fiber d ie GeschichteInd iens ims echs zehn ten Jahrhundert v on Graf F . A . v on Noer, Le iden 1 880, I , p . Noer gi 1 esthe names as Rodolpho Aquaviva , Anton io de Monserrat and Franc isco E nriq iie s . Father Jarrie , theprinc ipa l authority on the matter, says : »Legatus interpre s Goamcumperuenerunt, hasce Patribus Socie tatis in Collegio D . Pau l i commoran tibu s pertulere ; quae soc ios omnes incred ib il i afi’ecerunt laetitia,dums ib i persuade t v nusqu isque , misericordiae d iuinae the sauros ac b onitati s hoc pacto Regi patefac iendos ; hanc s ib i summi s v oti s prou inciam~

qua erat caritate , deposcit. At Prouincialis , v b i negotiumD eo diuturn i s multi sque commendass et, P . Rudolpho Aquauiuae s imu l Anton io Monserrato , quide inde in Aethiop iammis s us , in Turcarummanus inc idit ac Franc isco Fernandio '

hanc prouinciam.

demandauit.» (R . P . Petri Jarrici Tholosan i Soc iet. Je su Thesav rv s RerumIndicarum. Coloniae Agrippinae , Anno MDCXV, I , p . 504 . Monserrate i s also mentioned on p . 51 8 and p . 565 Ibidem) . Or, .

in Father Catron ’

5 vers ion . »When Akbar asked for mis s ionaries fromthe venerable Fathers of St. PaulfromGoa , »the Lot fel l upon the R . P . Rodolphu s Aqua viva , Anthony M anserra t, and Franc is H ene

ric The E ducation of the Second Son of Akebar was committed to Father M ans erat» Lateron Father Rodol phus »sent away the two Companions of his M iss ion , Father Henry for Goa, and

Father Manserat for the Town of Agra wi th the Prince hi s Disciple» . During the Campa ign to Kashmir »Father Mans erat had persuaded h im(Akbar) that ’twere proper to depute h imfor Spain to the

new K ing Don Phil ip (The General H istory of the Mongol Empire London 1 709, p . 1 3 9and Danvers in his »The Portuguese in Ind ia» , London 1 894 . V ol. II , p . 52 , on ly mentions theleader of the mi ss ion , but calls h imFriar Rodulphus de Aquav ina.

3 Abul Fazl expresses these fee l ings thus : »Learned monks also came fromE urope , who goby the name of Padre Thes e monks brought the gospe l , and mentioned to the emperor the irproofs for the Trinity . His Majesty firmly b e lieved in the truth of the Christian rel igion The s eaccurs ed monks appl ied the description of curs ed Satan, and of hi s qual ities , to Muhammed , the best.

1 56 WHE N DID TH E NAME MAN ASAROVAR BE COME KNOWN TO E UROPE ?Wilford the re fore s e ems to b e correct in everything,—except the Manasaro

v ar. For so far as I have be en able to find ou t , Mons e rrate has not le ft any account-of his experience s , and I think that Wilford has s imply made a mis take and con

founded Mons e rrate and .T ieffenthaler, who, nearly 200 years later , calls the lake ,Mansaroar, which i s a lmos t the s ame s pe l l ing as Mons errate

s s upposed Mansaruor.

JOHAN VAN TWIST leads us to comparative ly more sol id ground, though hehas nothing b ut hearsay informa tion to give us .

The borax,he says , is found in the mounta ins of the beau ti fu l province of

Purb et which s tre tches to the front ie rs of Tartaria . There is also musk, nard ,

qu icks i lve r, c0pper and a kind of colour . The borax is found in a rive r cal ledIankenckhaer, runn ing through the mounta ins of Purb e t , coming in a narrow wate rwhich flows away through the middle of the coun try , and which they cal l Mas eroor;

at the bottomof this water the borax grows, jus t as the coral does , and the Gujerats the re fore give i t the name of Iankenckhaer.

Aga in i t is the mu sk that opens the s ecre ts of dis tant Tibe t, for it is in a

de s cript ion of the principa l ware s brought for sale to Guje rat that the unknowncountry of Purb e t happens to b e men tioned . Purb et may b e the northe rnmounta insin general , or Kai las Purb et . The musk and borax , and the fact that it s tre tchesto the ve ry fron tie r of Tartary prove beyond doubt that it is Tibe t . The rive rIankenckhaer mus t b e the uppe rmos t Satlej and Mas eroor is The

confus ed hydrography as gi ven in the text does not inte rfe re w ith this explanation ;much more extraordinary waterways have been fabricated by explorers in our own

t ime , and verbal information , gi ven by nat ive s , may eas i ly b e misunders tood.

V an Twis t’s account is dated 1 63 8 , and I have not been able to find the

lake menti oned by name at any earlie r date , by any E urope an .

The next time I find the name of Manas arov ar is in the introduction to

W’

alte r Schouten ’

s narrat ive . He trave l led from 1 658 to 1 665, b u t the French edi

ti on , at my disposal , was not published un ti l that is to s ay, at any rate ,before the dis coveries of the Lama su rveyors of Emperor Kang Hi , and 25 yearsbe fore d

Anv ille publ ished his maps ; and on the latte r the lake is called MapamaTala i and not Manas arov ar.

Whe the r Schou ten or his trans lator is re spons ible for the fol lowi ng pas sage or

not, w e recogn is e in i t a ce rtain re s emblance w ith v an Twis t, b ut also a new , ve ry

o f all prophets God’

s bles s ings rest on himand his whole house ! a thing which even devi lswou ld not do.» Ain -i -Akbari , H . Blochmann ,

V ol . I , p . 1 8 2 .

The intere sting pas sage run s thus : »Den Borax wort gev ‘

onden in een water, genaemt Iankenc khaer, loopende door het gheb erchte v an Purb et, comende in e en engh water, dat door ’

t middenv an

t lant heenen loopt, dat sy Mas eroor b enamen ; 0p den gront v an d i t water wast den Borax , evenge lijck den Corae l , so dat de Gus aratten haer den naemv an het water Ian kenckhaer ghe v en .» GeneraleBeschrijv inghe v an Ind ien p . 6 1 . E ven Klaproth , quoti ng Chinese sources , says : Le me i lleur boraxs e trouve s ur le s bords du lac Maphamdalai“. II y en a du viol et et da noir. Magas in A s iatique ,T ome II , N o. IV ,

Paris 1 8 2 8 , p . 3 0 7 .

2 V oiage de GAUTIE R SCHOUTE N , I , p . LXX I II .

CHA PTE R XV I I I .

BENED ICT GOES AND ANTON IO DE ANDRADE.

In abou t A . D . 1 600, the learned Fathers at Goa did not know that Cathayand China we re two diffe rent names for one and the same country , and as they hadheard many years before that somewhere in the N .E . of India there exis ted a vas t country, the popu lation of which was nearly en ti re ly Chris tian , they s eriou s ly contemplatedthe plan to s end mis s ionaries thi ther with the object of »d is covering Cathay» .

I An old

Mohammedan merchant who had pas s ed some 1 3 years at Kambalu or Peking, toldFather JEROME XAVIE R at Goa, what he had s e en and heard of Chris tian i ty in thos eregi ons .

2 Such rumou rs had also been heard fromTibe t and China and were b eliev ed to have some kind of foundation . When Father Xavi e r , hims e lf, toge ther wi thBE NE D ICT GOE S , accompan ied Empe ror Akbar on his journey to Kashmi r, he heard>>that the kingdomof Tebat, which fexpanded eas twards fromKashmi r all the way toCheta i or Catay, conta ined great: numbe rs of Chris tians and many chu rches withpries ts and bis hops » .

3

At abou t this t ime Go ’

es happened to b e at Goa . Emperor Akbar had, for thethi rd time , invited Jesu i t Fathers to come and s e ttle at his court , and the three whowent were Jerome Xavie r, Emanue l Pinhe i ro and Benedict Goes , who arrived at

Lahore in May 1 594 . Akbar had great confidence in Goes , and therefore u s ed him

I Sociorum litteri s , qui Mogori s Regiam incolunt, auditumes t in India de ce leberrimo R egnoi l lo, quod Saracen i Cataium v ocant, cuius iamo l imnomen auctore Marco Polo Veneto noti’

i fuerat

E uropaei s , sed a l iquot deinde saeculorummemorize i ta exciderat, vix v t esse credere tur. scrib eb ant

Patres il l i Catai R egnumi llud ad ortumspectare , Borealioremaliqnan to ipso MogorumR egno : in eo

fidei Chri s tianae cu ltores reperiri multos , templa , Sacerdotes , ritus , persuadebfit. De Christiana E x~

peditione apv d Sinas s v scepta, ab Societate Jes v . E x P. M atthee i Ricij eiusdemSocietatis Comentarij s .

L ibri V Auctore P . N icolao Trigav tio Be lga ex eademSoc ietate . Augustee V ind . MDCXV. Lib . V .

C . XI . p . 544 .

2 Xaverius non s ine ratione credidit, i llud ip sume sse regnumCataianum, cu ius Marcus

Pau lus V enetus ali ique Historie i memin ere , cuiusque incolas es se Christianos sed labe in fectos N estorianatradunt R . P . Petri Iarrici Tholosan i Soc iet: Iesv Thesavrv s R erumIndicarum Colon izeAgrippinas Anno MDCXV. Tomus Tertius , p . 2 02 .

3 Benoit de Go '

es M i s s ionaire voyageur dan s l’Asie Centrale 1 603—1 607 par Le R . P . J . Brucker

de la Compagn ie de Jesus . E xtrait des E tude s re l igieuses , Lyon 1 8 79, p . 6 et s eq .

JEROME XAVIE R , MATTE O RICCI AND TRIGAULT . 1 59

in a mis s ion to the vice roy at Goa . Here he was found to b e the right man forthe commis s ion of dis covering Cathay .

The famous Father MATTE O R ICC I who had got new s of the intentions of hisbrothe rs in Goa, wrote to them, before the departu re of Goe s , that Cathay was thes ame as China , and Kambalu was Peking, so called by the Mohammedans . The

s tory abou t innumerable Chris tians was a fable . His arguments , however, we re not

regarded as convincing, and Go’

és got the extraordinary commis s ion to discove rChina .

From information which Fathe r Xavie r had obta ined frommerchants he firs t decidedto s end Go

es v ia Tibe t. Already in a le tter of 1 598 he wrote I that a trave lle rfromLahore shou ld firs t arrive in Cachimir, which s t ill be longs to Akb ar

s empire .

>>From there one trave ls di rectly to the Kingdomof Tebat , 2 the King of which isa great friend of Akbar, and fromwhere one arrives , provi ded with letters from thisp rince , eas i ly to Caygar (Kashgar) . Fromhere i t is on ly a few leagues to the firs ttown in Chatai , which is inhabited by Chris tians .»

The Fathers at Goa had , in this cas e , as BRUCKE R be l ieves , been mis ledby Orte l ius ’ maps and the reports of Jenkinson . Howeve r, the rou te of Kabu l wasfinally chos en , and Goes left Agra on Octobe r 2nd 1 602 , and Lahore on February1 5th , 1 603 , disgu is ed as an Armen ian , and under the name of Abdu llah Isa‘

i'

.

He was accompan ied by two Greeks and an Armen ian , ISA AK .

Refe rring to the Hima laya he s imply us es the expres s ion »the Mountains coveredwith snow» . The Indus was cros s ed at Attock, which he calls Athec. Then he pas s edPassaur (Peshawer), Ghideli and Cabu l . Fromthe road ove r the Hindu-kush three namesare mentioned by TRIGAULT : 3 Ciaracar, Paruam, which is s a id to b e the las t town of

Mogor, and, afte r 20 days through the highes tmounta ins , a regi on ca lled Aingharan .4

Thence he trave l led 1 5 days to Calcia, 1 0 days to G ialalab ath , 1 5 days to Talhan ,

8 days of difficu lt journey to Tengi Badas cian ,5 and final ly Ciarciunar. Fathe r R icci

who afte r Go ’

es’

death firs t tried to s ave as much as poss ible of the informat iongathered on the jou rney , has obvious ly confounded some places , which , in the latte rpart of the itinerary , Is eas ie r to s ee . G ialalabath, for ins tance , is Jallalab ad be tweenPeshawer and Kabu l . From Ciarciunar, perhaps Chahrchinar or the »Four Planetre es » , he had 1 0 days to S erpanil,

6 S er- i-Pamir , and further, afte r pas s ing the s teepmounta in Sacrithma , 20 days to Sarcil, which Brucke r identifies with S ariko l . Two

1 Quoted by Brucker, p . 1 1,fromFather Jean H ay.

2 I . e . Ladak .3 Op . supra cit.4 Brucker identifie s these places w ith the Charekar, Parwan , and I-angheran , vis ited by Wood

in November 1 8 3 7 .

5 Teng-i -Badakshan.

6 Or as Trigault has it : Hin c post d ie s decemin Serpanil deuenerunt, sed desertus locus erat,mortalibus omnique comitatu des titutus ; in arduummon temconscen sum,

cu i Sacri thma nomen est, in

hunc montemeuadere non n is i rob ustiores equ i potuere , re liqui longiore sed faciliore itinere perrexerfit.Op . cit. p . 550 .

1 60 BE NE DICT GOES AND ANTONIO DE ANDRADE .

days more took himalong the mountains of Ciecialith (Brucker : the plateau of Chi "

chiklik), where he had to pas s s ix days in the s nows . Thence he trave lled to Tanghetar, Teng-i-tar, which be longs to Cascar, and whe re there was a great r iver, fromwhich 1 5 days to Iacon ich, perhaps Yaka—arik , as Brucker sugges ts . After anothe r5 days he was in Hiarchan , Yarkand.

From Yarkand Goes wrote a le tter to his Jesu it brothers in India describingthe difficu lt ies of his road s tra ight acros s the Pami r. I

FromYarkand Goes vis i ted Khotan (Cotan or Q uotan), where he was muchinteres ted in a kind of »marble » , cal led Tusce by the Chines e , and of which thereare t wo d ifferent sorts , one »e fiumine Cotan» , the other, of inferior qual ity, froma

mountain at 20 days fromthe capital , and called Cansangu i cascio,»id es t mons

lapideus » .

2

Goes re turned fromKhotan to Yarkand, and left this city defin ite ly on No

v emb er 1 4th , 1 604 , pas s ing Iolci , and furthe r, in the cou rs e of 25 days , the following places : Hancialix, Alceghet, Hagab ateth, E griar, Mes etelec

,Thalec , Horma ,

Thoantac, Mingieda , CapetalcOl Zilan , S arc Gueb edal, Canb asci , Acons ersec, Ciacor,and Acsu.

Brucke r is perfectly jus tified in saying that i t is difficu lt to iden ti fy thes e names ,which have , however, a perfect loca l phys ionomy of geographical names in the jagataiturki language . H e identifies Capetalcol with Chaptal-kul, Zi lan w ith Tchi lan , Sarcwith Sai-arik, and Canb as ci with Kumb ach . The two end s tations of Go '

es’ road placed

i t beyond doubt that he fol lowed the ord inary caravan road along Yarkand-darya toAksu . Chi lan is indeed a s tation on this road.

3 Sarc is not Sa i—arik , for in Trigault 4it s tands toge the r w ith the next name Gueb edal, indicating one name cons is ting of

two words I w i l l show be low that Sarc Gueb edal is probably S arik-abdal .

I Father Iarric has saved this letter : Benedictus a Goes H ircandae etiamcommoratus , iter se

d ifficillimum8: molesti s s imum hab u isse scrib it , per de sertumsci l icet Pamech , in quo qu inque amiseratequos idque ob intens is s imum frigus ; tumquod magna hic Iigni s it penuria, ex quo ign is excitatur,tumvero quod ea fit coe l i inclemen tia v t respirare an imal ia commode vix poss int. h inc equi homines respirare sa pe per vi am demoriuntur. quod homines malum,

a l l io, porro , pomisque ; s iccatis

comestis v itant ; iumenti s vero gi ngiuae al l io perfrican tur. Hoeoe desertums i quando n iuosum, quadraginta emetitur dieb us ; s in minus , paucioribus . Op . cit. III , p. 2 1 7 . This description could have beenwritten by any modern trave l ler as we l l . E ven the onions are still used by all native trave l lers in thehigh parts of the Pamirs and the Kara-korum.

2 In hi s above-quoted work , De Christiana E xpeditions , Trigault writes Tusce, but in anoth erwork , Regn i Chinens is descriptio E x V arij s Authori b us , n d . Batav . CIDIQ CXXXIX , p . 3 45, the sameauthor writes Yusce , which makes it eas ier to recogn is e the derivation of the word , namely yii -shi , or

j ade . A lready in 1 8 2 0 Abe l -Rémusat wrote a learned artic le : R echerches sur la substance minéral eappe lée par les Chinois P ierre de in , e t sur le jaspe des anc iens . As to Cansangui cascio , a nameto which we shall have to return later on , Brucker shows that it means Kan-s ang

-i-kas h , or the

>Mine of jade-stone» .

3 Iolci,Yo lchi , means guide, and may we l l have been the name of a vil lage fromwhich

caravan-men were taken for the road . A vil lage , Yolchak , i s sti l l s ituated just outs ide of Yarkand ,though on the road to Khotan . Pet. M it. I. c . p . 5.

4 In both his above quoted works .

1 62 BENE DICT GOE S AND ANTONIO DE ANDRADE .

can b e found almos t anywhere along the roads in E as tern Turkes tan . One oas iseas t of Keriya and a vi llage be twe en Kargalik and Gume have this name , and a

fores t-tract on the Kara-kash-darya i s cal led O i-tograk .

I Thes e places are , of cou rs e ,not meant, b u t eas t of Lop I have pas s ed a Tograk

-bu lak. Cialis is probably s ituateds omewhe re near Korla and Kara-shahr , and w e shal l have to cons ider it in con

necti on with the maps of the s eventeenth century on which it is nearly a lways shown .

In Aramuth I be l ieve we have to suspect Urumchi , which , by an Armen ian mayhave eas i ly been heard as Aramutchi and half forgott en .

The same confus ion mu s t b e su spected amongs t the names of the earlier partof the road. The tac, tag, mountain , in Thoantac, and Canb asci , refer to a mountainous regi on and not to the leve l road fromYarkand to Aksu . Thoantac and

Tanghe tar probably be long to eas tern Pamir ; I have pas s ed both a Tong and a

Tengi-tar in thes e regions .

2 Tong-tag, or the mountains of Tong , is an expres s ionthat may b e us ed at any moment by a »caruan basa» or »comitatus praefectus » , 3 i . e .

caravan-bashi .The chief points of Goes

’ itinerary are therefore : Lahore , Kabu l , Badakshan ,

Pamir, Tengi-tar, Yaka—arik, Yarkand , Khotan , Yarkand, Aksu ,Kucha , Ugen , Korla ,

Kara-shahr, Urumchi , Tu rfan , Pijang, Hami , the des ert Cara-catb ai (Kara-kita i), andSu-chou .

In the his tory of exploration in that part of Tibet , where the great Indianrivers have the i r sources , the Portugues e membe r of the Jes u it mis s ion in India ,

AN TONIO DE AN DRADE,is the very firs t of whomwe have any re l iable knowledge .

He was the firs t E u ropean to cros s the Himalaya fromthe Indian s ide , and he

accomplished a brill iant journey fromS rinagar in Garhwal ove r the Pas s of Mana toTsaparang or Chaprang on the Satlej . H e was the mos t s ucces s fu l mis s ionary whoeve r entered and preached in Tibet , b ut he made no fu rther u s e of his succes s .

Anton io de Andrade was born in 1 580 in O le i ros in Portugal . In 1 600 he

was s ent to Goa , and here the rumour reached himthat Chris tians exis ted in Tibe t .He says hims e lf that his journey to the unknown country was unde rtaken for theglory of God , and that the Portugues e s ince long had been looking forward to the discov ery of Tibet . Go

'

és’ journey, and the identi fication of Cathay w i th” China, as shown

by R icci , do not s eemto have made any impres s ion upon him, though he cannot havebeen ignorant of thes e facts . In 1 624 he got an occas ion to s tart , and in companywith Father MAN UE L MAR QUE S and two Chris tian s ervants he left Agra, on March3oth , and joined

,at De lhi, a caravan of Hindu pilgrims who were on the i r way

to the holy places at the source of the Ganges . The journey probably went overHardwar, the Gate of Vishnu , through the domin ions of the Raja of S rinagar, a

region that had never been s een by a E uropean . The road along the Gange s was

I Pet. M it. l . C .,p . 3 3 , 1 98 e t s eq .

2 Through A s ia , I p . 2 64 e t seq .

,II p . 707 et s eq .

3 Trigault.

ANDRADE’

s JOURNE Y To THE K INGDOM or TIBET. 1 63

di fficu lt and rocky . The caravan proceeded towards the heart and the great he ightsof the mountains , the same road which

,1 80 years later, was us ed by WE BB and

RAPE R , who did not know that the priori ty be longed to a Po rtugues e . In his ex

cellent narrative Raper has not a word for Andrade , b ut he confirms his obs e rvat ions . In a very cleve r es say, C . WE SSE LS has jus t drawn the paralle ls between thetwo trav ellers .

I VVes s els’ work on Andrade is a very we lcome addition to the

his tory of exploration in Tibet, and he has put an end to the be l ief that Andradeshou ld have trave l led through Kashmi r and the whole of Tibet and that he shou ldhave dis cove red the Manas arov ar.

2

Howeve r, the mis s ionaries continued and reached the holy temples of Badrinath . Webb and Raper proceeded at a later pe riod so far as to the vi l lage of Manaon the Saraswati , the uppe r cou rs e of the Vishnu-Ganga . But Andrade continu ed farbeyond the village and s truggled for succes s . Marques had been left behind ; on lythe two s e rvants and a gu ide fromMana accompan ied the mis s ionary . Afte r threedays some men fromMana came up and advi s ed himto re turn , unles s he wanted dieon the pas s . S til l he continu ed through heaps of s now and in blinding snow-

s torms .

He graphically des cribes all the i r di fficu lties in this inhospitable region , whe re manytrave llers succumb frompo isonous gas es .

3 Finally the l ittle party reached the pas s ,and Andrade des cribes in the fol low ing important s entence what he saw : »Nes taforma fomos caminhando atee o a lto de todas as s e rras , onde nas ce 0 Rio Gangade hum grande tanque , 8 1 do mesmo nasce tamb emoutra , que rega as te rras doTibe t .» 4 Or, in othe r words , he reached the highest point of the rocks , where the

rive r Ganga takes i ts ris e froma great pool , fromwhich also anothe r rive r begi nswhich irrigates the countries :

of Tibet . 5 We shal l have to retu rn to this pas sagelate r on .

From the Mana pas s Andrade saw the Kingdomof Tibe t. But now the irfatigues and hardships we re also at the i r cu lmination . They were snow-blind and

I Anton io de Andrade S . J . E en ontdekkingsreiz iger in de Himalaya en in Tibet ( 1 624De Stud ien, T ijdschrift voor godsd iens t, wetenschap en letteren . Dee l LXXVII , N zo 4 . 1 9 1 2 .

2 It i s not rare to find the statement that Andrade should have trave l led through Kashmir.T hus in J . P . Parraud et J . B . B i llecoq : Voyages au Thibet, Paris L ’

an IV,p . VI : »E n 1 6 24 , le P .

d’

Andrada, jésu ite Portugais , pénétra dans le Thibet par Cachemir.»

3 »Perhap s mofettes» , as R itter suggests ! It was the same story which was told by almost alllater mi ss ionaries in T ibet, last by Father Huc .

4 Novo Descobrimento do gramCathayo, ou Re inos de T ibet, pe l lo Padre Anton io de Andraded a Companhia de Iesu

,Portugue z , no anno de 1 624 . L isboa 1 626 , p . 7 . This l ittle narrative was

tran s lated 1 62 7 into Span ish in M adrid , and into German in Augsburg. See R ichthofen : China I p .67 1 . The German ed ition was used by R itter : Die E rdkunde v on A s ien . Berl in 1 83 3 . Bd . 1 1

,p .

440 e t seq . The Ital ian ed ition , R e latione del novo scoprimento del Gran Cataio, overe Regno delT ibet, Roma 1 62 7 , i s regarded as be ing »olto oscura e incerta» .

5 In the French vers ion at my d isposal the word »tanque» is trans lated with »lac» : »Nous che

minames de cette fagon jusqu ’

a ce que nous arrivfimes au sommet de toutes ces montagnes OI‘

I se voitle lac d

oii sortent la riviere de Ganga e t une autre qui arrose les terres du Thibet» , and : »Le lac

d ont j ’ai parlé, m’

auroit fourn i amplement de l ’eau , e t j’

av ai encore assez de provis ions pour s ix ahuit jours Parraud et B illecoq : Voyages au Thibet, p . 1 6 and 1 7 .

1 64 BENE DICT GOES AN D ANTONIO DE ANDRADE .

cou ld s ee no road . The gu ide had already retu rned to Mana , and now the two

s ervants wou ld also re turn . Andrade accompan ied them, les t they shou ld die inthe s now . After a few days they fortunate ly met Bothia-s couts fromMana and afte ranother thre e days they camped in a grotto whe re they we re jo ined by Marqueswho came up w ith provis ions . After a month ’s res t they returned to the pas s , whichwas then eas ie r s ince a good deal of the s now had me lted . On the other s ide theywe re rece ived by emissaries fromthe King of Tsaparang, and in the beginn ing of

Augus t 1 624 they had reached the »C idade Real» .

In Tsaparang Andrade was rece ived with roya l hospitality by the king and

queen , b u t he s tayed only one month and has nothing to say of geographical interes t , except that the city is s ituated on a river. In the begi nning of Novemberhe is again at Agra, where he wrote his l ittle book , which is dated Novembe r 8th , 1 624.

But already the next year Andrade s tarted for his s econd journey to Tsaparang. How long he remained this time is not known . He reached Ts aparang Augu s t 2 8th , 1 6 25. Apri l l i th, 1 626 , he laid the foundation s tone to the firs t Chris tianchurch in Tibe t . On Augus t 1 5th the s ame year his s econd narrative is dated at

Ts aparang.

I From lette rs he has w ri tten it is clear that, in S eptember 1 627 , he

was s till in Tsaparang. But in 1 630 he was at Goa , and in 1 63 1 he s en t fou rmis s ionaries to Tsaparang . In the beginn ing of 1 634 he prepared hims e lf for a thirdexpedi tion , togethe r w ith s ix compan ions , b ut d ied , March 1 9th , and was buriedat Goa .

C . Wess e ls has found that at leas t e ighte en other mis s ionaries fo llowed Andrade ’s example and continued his work . Marqu es was s ti l l at Tsaparang in 1 642 .

On ly vague echoes fromthos e days have reached our time . It is to b e hoped thatC . Wes s e ls will have eve ry success in his in tention to w ri te an es s ay on thes e un

known mis s ionaries .

Andrade ’s s econd narrative , though extreme ly interes ting to read, containseven les s geography than the firs t . He on ly mentions »the grand Lama of Utsang»,

Dalai Lama in Lhasa ; and when he speaks of a city half a day fromTsaparang,

we suspect that he means Totl ing.

To return to the pool on the Mana pas s , which for centuries has been the

innocent caus e of misunde rs tandings , i t shou ld b e noticed that Andrade speaks of a

»tanque » and not of a »lago». He means , as Wes s e ls clearly proves , the 3 70m. longglacier lake

,which is s ituated on the pas s , and fromwhich the S araswati takes its

ori gin and goes down to the Ganges . The other rive r which irrigates the countries of

Tibet i s a tribu tary to the Satlej . But modern geographe rs have misunders tood him.

MARKHAM , for ins tance , says : »He cl imbed the te rrific pas s es to the source of the

Ganges , and eventual ly , afte r fearfu l su ffe rings , reached the shores of the sacred lakeof Mansarowar, the source of the Sutlej . Thence the undaunted mis s ionary found

I The French edi tion of the same , printed in Pari s 1 62 9 , has the title : Histoire de cc qui s’

estpasse au Royaume du T ibet.

1 66 BE NE DICT GoE s AND ANTONIO D E ANDRADE .

s ou rces he concludes that Chaparangue was s i tuated in Tebeth and thus in Bolor,and that in this ne ighbourhood Andrade was fortunate enough to make s eve ral admi rable discove ri es , amongs t othe r things the sou rces and the fountains of the Gangesand the Indus , about which Kircher had got his notion fromJOSE PH MOGULE N SEwho was a Chris tian and who accompan ied Andrade on all his jou rneys , and furthe rfrom Father HE NR I ROTH , who, on his arrival in Rome , told all about Andrade ’sjou rney . Amongs t other things Kirche r learnt was the fol lowing s tatement, which ,on account of its importance , I give in the original text : I

»ll y a un grand Lac sur les plus hau tes Montagnes de Tebeth les que lles sonttoujours couvertes de nege , duque l prenent nais sance les plus grands Fleuves del’

Inde ; pu isque l’lndus , le Gange , le Ravi , I’Atech 2sortent de ce gouffre . Le Gange

prent son cours vers des precip ites , venant a tombe r il fai t umbru it effroyable ,apres quoy i l arrous e une agreab le v alée , 8: continue de rouller s es fiots ve rs la Mer,

ou il s e v a rendre .3 Pou r ce qu i es t de l

Indus des au tres fleuv es , ils cou lentdoucement le long de la Montague , comme la Carthe nou s le fai t voi r. »

Kircher indeed s eems to b e respons ible for the be l ief that Andrade dis coveredthe Manasarov ar, for late r on in his work he s ays again : 4 » i1 vint ensu ite (afte r Chaparangue) a travers er une hautemontagne au somme t de laque l le il y a un grand lac leque les t (a ce qu ’ i l dit s e lon ce qu ’i l reconnut) la Source du Gange , de l

Indus des

au tres plus grands Fleuves de l’Inde ; de la 11 prit la route vers Radoc qu i es t une

R egion extremement fro ide s epten triona le tres difficille a pas s er ; c’es t pourquoyapres avoi r res té long temps a trave rs e r ce pai s , il arriva a la Vi lle qu i porte lemesme nom» ,Andrade has neve r pre tended to have d is cove red anything except what he de

s cribes in his narratives . Kirche r did not know the exis tence of the Manasarov ar

and never mentions i ts name . And s ti ll , when he talks of a lake , which is the sourceo f the Ganges , the Indus and the othe r greates t rive rs of India ,

he means , withou tknow ing it, the Manasarov ar. The be l ief that the great rive rs began fromthe

Manas arov ar was common in India, and i t is extreme ly like ly that both Jos eph and

Roth had been told by natives that such was the case . Withou t having heard or,

perhaps , after forgetting the name of the lake , they have told the s ame s tory toFather Kirche r, who made his combinations at home and found that the s tory of thelake feeding all the great rivers of India , agreed ve ry w e ll wi th Andrade ’s : »onde

nas ce 0 R io Ganga de humgrande tanque » , etc. It is hardly pos s ible to think thatAndrade shou ld have l ived on the banks of the S atlej for years , withou t knowingthe exis tence of the Manas arov ar. Bu t that is anothe r ques tion . He does not say

a word of the lake . He knew it, b u t he .does not speak of it. Kircher does not

I Ibid . p . 67 .

2 Athee on the map, i . e . Attock .3 Compare Pl in ius , Natural is H istoriae Liber VI , Cap . XVIII .4 Op . c it. p . 8 8 .

RN TIBET ,VOL 1 .

LLA R E

C A S P I U

KIRCHE R ’

S ACCOUNT or ANDRADE’

s JOURNE Y . 1 67

know i t, b ut he speaks of it. In this particu lar point, therefore , I cannot shareWes s e ls ’ opin ion .

Kircher goes on to say that Andrade continued to China or Cathaie fromRudok ; he was on ly two months on the way, and trave l led through the kingdomsof Maranga and Tanchu t of the Tartars . This is of cou rs e impos s ible , and dependsupon a misunders tanding.

The s econd map in Kirche r ’s work I does not qu i te agree with the description in the text, b ut s ti l l , this map is an importan t and cu rious document , be ing thefinal resu lt of a learned man ’

s desperate s truggle with great di fficu lti es in ge tting all

the s eemingly contradictory information he had gathe red, to agree . The map isalso a hope less confus ion , whe re on ly the principal l ines are recogn isable , b ut all thedetai ls ups ide down . Leaving alone the res t of this wonderfu l map for the pres entI wi l l on ly try to expla in its hydrography (s ee Plate XI). The map is so u tte rlydi fferent fromall other maps of the time that i ts compi latormus t he re have had at his

disposal a mate rial not ava i lable to othe r cartographers .

2 Kirche r’s map , publishedin 1 667 , and 1 670, be longs to a type of its own , so far as Centra l As ia and Northe rnHindus tan are concerned. I t is an anomaly in the deve lopment of Central As ia ticcartography . And this mus t depend upon the fact that the compilator has tried to

make an i l lus tration to Andrade ’

s discoveries , as s is ted by the information given byFathe r Henri Roth , and the Chris tian nat ive Jos eph .

To the north the Indian pen insu la is bounded by the Montes Teb e t ici . As .

there are on ly a few de tached moun ta ins south of this range , nobody will doub t thati t i s mean t to b e the Himalaya . The ve ry highes t part of this range is a name les smountain and north of it is Radoc (Rudoc). So the moun tain in qu es tion is meantto repres ent Kai las . Near the s ummi t, b ut s ti l l on the southe rn s ide of the

culmination , there is a rather big lake , and fromthis lake the Ganges takes its beginn ing,and thence flows to the S .E . This is the »tanque» or poo l of Andrade . On the

mounta in and jus t sou th of the lake are wri tten the few words which gi ve the key

to the unders tanding of the whole problem: Origo Gangis e t Indi . In spite of this ,and in spi te of the text , which a lso says that the Indus begins fromthe lake , themap repres ents the three Indus branches as s ituated wes t of the mountain and jo ining at Athec (Attock), which is indeed more correct. Thus the Indus comes downfrom mountains s ituated north of Casmir. At the uppe rmos t part of the wes ternIndus branch we read Pallaur, Remembering the Indian view of the

hydrography we ought not to b e surpri s ed to find the Ganges coming di rectly out

of the Manasarov ar. We find the same mis take on D’

ANVILLE’

s map , nay even on

1 Op . cit . p . 65.

2 It shou ld b e compared w ith V is scher’s map of about 1 680 , N icolos i’

s of 1 660,and Cante l l i ’s .

of 1 683 , described and reproduced hereafter.3 The map has a lso a Belor mon s north of Lassa Regnum.

1 68 BE NE DICT GOES AND ANTONIO DE ANDRADE .

RE NNE LL’

S of 1 78 2 , more than hundred years later. On Kircher’s map the sou rcesof the Indus are more correct than those of the Ganges .

At some dis tance sou th of this big lake , the Manas arov ar, there is another fromwhich two rive rs take the ir origi n . The eas te rn one i s obvious ly meant to b e the

Jumna, as Agra is s i tuated on i t and »De l li» not far fromi t . I t does not join the

Ganges b ut goes its own way to the s ea . The re can hardly b e a doubt thatthe wes tern rive r is meant to b e the Satlej . Both the S atlej and the Jumna havetwo head branches on the map . The lake fromwhich the two rivers iss ue has

no name , b u t on An drade ’s rou te w es t of it is Caparangue , as the cartographe r didnot know that this town was s i tuated on the ve ry river. The name of the river isKauc flu .

, i . e . the rive r of Kauc or Guge (i ). That the ri ver is the Satlej is obviou sfromi ts runn ing S IM and i ts join ing the Indus far be low Attock . The hydrographyof Panjab is of cou rs e far more absu rd than '

on any contemporary map. The rive rcomes froma lake which is s i tuated s outh of the Manas arov ar ins tead of wes t, jus ta s in the cas e of the two lakes at the o rigi n of the Hwang-ho on the same map .

I

This s econd lake is the Rakas -tal. It is no wonde r if the cartographer was a l i ttlebew i ldered by the information in the narratives of Andrade and by thos e broughthome by Henri Roth and Jos eph . Andrade ’s rou te is marked on the map , thoughi t does not harmon ize ve ry we l l w i th the phys ical geography of the same map . But

the map shows that , in Kirche r ’s opin ion , Andrade pas s ed through »S i negar» in

Garhwa l and Caparangue , that he cros s ed the S atlej and went clos e wes t of theRakas—tal and the Manasarov ar and of Kai las , that he cros s ed the Transhimalayanmountains to Radoc (Rudok), by which the cartographe r has made the mis take to

le t the road turn eas t ins tead of w es t. FromRudok he has returned to Ind ia the

s ame way.

- This is what Kircher be l ieved ! But in real ity Andrade neve r wentbeyond Tsaparang.

Comparing Kirche r’s text w ith his map , and comparing both with Andrade ’snarrative , the whole s i tuation becomes perfectly clear. On the map as in the textthe two lakes dominate the hydrographical sys tem, jus t as in the mythica l poe tryand the be l ief of the Hindus . Andrade cannot b e made res pons ible for the hydrography of the map. For he on ly s aw on e pool , fromwhich the Ganges went tothe s ou th and another rive r, i rrigati ng Tibe t, to the north . And this does not at

a ll agree with Kirche r’s map. This is the firs t time a E uropean has made the

Ganges take its origi n fromthe Manas arov ar. The informan t , whosoever he mayhave been , probably a native who expla ined ma tters to Jos eph or Roth , has certain lys een a river leaving the Manasarov ar and be l ieved i t was the Ganges , and not

s imply the channe l to Rakas -tal. H e , or some othe r informant, has s een the S atlejleaving the Rakas -tal, jus t as the map repre s ents it. He was , so far as we know ,

On K ircher’s map China is taken fromM artini 1 655, except the K ia Lacus , which has dis

1 70 BENE DICT GOE S AND ANTONIO D E ANDRADE .

havi ng accomplished one of the mos t bri ll iant journeys in As ia in a surpris ingly shortspace of time .

It is a great pity that so very few de tai ls fromthis jou rney have remainedto our time . The two Jesu i ts s eemnot to have unders tood the great geographicalimportance of the i r achievement , and they have not even troubled to keep a roughdiary. We are only able to fol low the great featu res of the i r jou rn ey. In somecas es , as for ins tance , when Gruebe r mentions the rivers sou th of Koko-nor, we can

not , w i th any degree of ce rtainty, ident ify them. On some old maps , howeve r, thisjourney has left its traces , and that i s the on ly reas on why I men tion G ruebe r and Dorville together with Go

'

es and Andrade . For in a later part of this work, deal ingwith exploration in the Transhimalaya, w e shal l have to return to this mos t interes ting journey .

TIBET D THE SOURCES

OF THE GREAT INDIAN R IVE

ON ANC IENT MAPS

1 74 MAPS OF THE FIFTE E NTH CENTURY.the Ganges . So far there can b e no talk of a »greate r Tibet» as a phys icogeographical or pol i tical un ity ; at the ve ry mos t the mere name appears on

some of thes e early maps . On ly wi th Del is le is a real ly new type inaugurated ,showing Tibe t as a defin i te ly bordered country north of the Himalaya . And w i thhimwe enter upon more s ol id ground so far as a cartographical clas s ification in typesi s concerned. The next type , the mos t revolutionary of all, is introduced by D ’

AN

VILLE , who totally changes the map of Tibe t, and whos e influence s tretches far intorecen t times , so much so , that i t can s ti l l b e traced on RE CLUS ’ map of 1 88 2 . To

d’

Anv i lle’

s type we mus t reckon KLAPROTH , and all his fol lowers amongs t the greatGe rman cartographers , R ITTE R , GRIMM , MAHLMANN , HUMBOLDT , BE RGHAUS ; andeven KIE PE RT . Afte r d ’

Anv i lle we can hardly speak of any new origi nal type b efore the journeys of MONTGOME RI E

S Pundi ts , a type which has cu lminated in the

actual repres entation of Tibe t, the res u lt ofmode rn E u ropean exploration . HODG

SON , SCHLAGINTWE IT , S AUND E RS and R ICHTHOFE N may also b e regarded as re

pres entati v es of di ffe rent types , b u t as the i r maps are cons tructed on ly on hypothes es , they are of no great importance .

While Dr. Dahlgren ’

s sys temof cartographical types for Japan embraces on lythe one centu ry from 1 550 (Gas taldi) to 1 655 (M ARTIN I), for Tibe t w e may Speakof types fal ling within two centu ries , fromabou t 1 700 (De l is le) to abou t 1 900 (the

exploration which is s ti l l go ing on). This d i ffe rence s imply depends upon the fact thatTibet has made res is tance agains t E uropean exploration much longer than Japan and

any othe r country in As ia . Be fore 1 700 Tibet can hardly b e said to play any cartographical part at all, and our attempts to extract any information fromearl ie rmapswill therefore prove to b e rather negative . But our obj ect be ing also to trace the

local isati on ‘

of the sources of the great rive rs w e have to go back to the middle of

the fifteenth cen tury. In a much late r time E u ropean inves tigation S low ly begi ns toconque r the ramparts round Tib e t .

In the middle of the 1 5th centu ry cartographical and cosmographical s tudiesw e re in a particu larly flourishing s tate at Ven ice and Genoa , and fromthat perioddates the firs t map we are going to cons ider , name ly the Genoe s e Map of the

World of 1 447 , which is pres erved at the National Library of Florence (Pl .Prof. THE OBALD FISCHE R regards i t as the mos t important map of the 1 5th century, and indeed it supers edes the famous map of FRA MAURO

,1 459 , in the way i t

shows both the mountains and rive rs of sou thern As ia .

I Ptolemy’s influence is ve ryclearly to b e s een ; the great featu res of the orography are the same as his , and the

Indus with its fiv e s ource branches is of a perfectly ‘

Ptolemman type . The eas ternmos t of the branches is the Zaradrus of Ptolemy, the S atlej of our days . On the

othe r hand, Fis che r regards the Ganges of this map as drawn frommore recent

Sammlung M ittelaltlicher We lt und Seekarten Ita l ien i schen Ursprungs und aus Ital ienischenB ibliotheken und Archiven . Dr. Theobald Fischer. Venedig 1 8 86 , p . 1 55 e t seq .

THE GE NOE SE MAP or TH E WORLD .—FRA MAURO . 1 75

notions , es pecially fromthe dis coveri es of CONT I , though I cannot s ee any greaterdi fference fromPtolemy than in the cas e of the Indus . The three feeders , Jumna ,

Ganges and Gaghra (i ), come down fromthe great water-parting range ; a righttributary from the S .W . may b e the Son , and the two great tribu taries fromthe

eas t may b e in tended for the Brahmaputra and Barak, if not , as Fis che r s ugges ts , thenorthern one shou ld repres ent the Irrawaddi , for Av a is s ituated on its banks and a

legend, »maius gange am(n is) al i te r daua dicitur» , i s taken fromConti . So far as

I can s ee , the northe rn tribu tary can the refore unde r no condi tions b e the Brahmapu tra . In spite of this mis take by the cosmographe r his map is in thes e regions fars uperior to that of Fra Mauro , which Shows a hope les s confus ion in the hydrography.

I

Between the Indus and the Ganges is the lake, of which Conti , as recorded in RA

MUSIO’

S collection , has some s tories to te l l and of which he says that the whole re

gion drinks of its water. Fis che r identifies i t w i th the lakes of Udaipur and Debar .At the mou th of the Gange s is a city called Berngalia . The eas tern part of thegreat range Fis cher sugges ts to b e the mountains of eas tern Tibe t, though they aremore l ike ly meant to correspond to the eas tern ranges of the Himalayan sys tem.

On our reproduction , Pl . XII , the legends are i llegi ble , b u t on ly a few of themmay b e of s ome inte res t. A long the northe rn range with its s eve ra l towe rs w e read :»has turre s con(s truxit) presb iter Johannes rex

'

ne inclus is ad eumpate t acces sus » ?

The range i s also adorned with a gate , from which the Ganges is sues . The

fiv e feeders of the Indu s come down fromthe »porta ferrea : ub i Alexander TratarosOn the part of the map which corresponds to the pres ent Tibet and at

the s ide of the golden image of a kne e l ing king the follow ing legend was s ti ll vis iblein the 1 8th cen tury : »rex Camb alech hic (P) es t magnu s can is » .

3

Imaus montes inacces s ib iles is a great range which s tarts fromthe wes tern end

of the principal range and runs to the NE . To the eas t of i t is Scythia u ltraImaummontemand s ti l l fu rther eas t two other ranges called montes inacces s ib iles .

In this region , which in late r years went under the denomination Great Tartaria , a

legend reads : »Hic adeo hab itantur ex ebreorumg t me (generat ione)tribus decem qu i l eges suze gener. » Thes e degenerated Hebrew tribes ,as we l l as the Magog in the eas tern part of the regi on , were the Tartars , who hadbeen excluded by Alexander.

There are no traces of Marco Polo who , some 1 70 years before , had travelled through the whole of »Scythia u ltra Imaummon tem».

In 1 457 to 1 459 FRA MAURO , the learned Camaldolens ian fromMurano, drewa map of the world (Pl . XIII), which , though famous , was , so far as As ia is con

I Compare a lso : Zurla : Di Marco Polo e degli altri Viaggiatori Venez ian i Vol. II , p . 345,Venez1a 1 8 1 9.

E pilogue de la Geographic de moyen age e’

tudiée par JoachimLelewel, Bruxel les , 1 857 , p . 1 69 .

3 H e inrich Wuttke : Zur Ge schichte der E rdkunde in der letzten Halfte des M ittelaltersVI . und VII . Jahresbericht des Vere ins fiIr E rdkunde zu Dresden . Dresden 1 8 70, p . 4 7 .

1 76 MAPS or TH E FIFTE ENTH CENTURY .cerned , no improvement upon Ptolemy’s map .

I The re is , however, and probably forthe firs t time on any map, ,to b e found the name of Tebet north of M ihen , Cina ,and India Te rza . This notion is de rived fromMarco Polo. Ganges and Indus comedown fromthe southern s ide of M . Imaus . Indus has fiv e feeders , of which the one

in the middle comes froma smal l lake . The eas ternmos t feeder, corresponding to

the Satlej , is the longes t , and, beyond i t, on the northe rn s ide of the Imaus, is a

lake called Mare Breunto . The Ganges has two principal feeders , the wes tern comingfromM . Imaus , and the eas tern fromTebet. To the wes t of the Ganges we readthe name Phison , indicating that the Ganges was one of the four rivers of paradis e ?

The Indus flows to the Mare Indicumwhere Bangala is s i tu ated , while the Gangesgoes to S inus Gangeticus . Betwe en the Indus and the Ganges and paral le l to themflows the rive r Mandus coming froma name less lake , s ituated in a l ittle range southof and paralle l to the Imaus . The fact that the city of Av a is placed on the eas ternbank of the Mandus proves that this rive r s tands for Irrawaddi , and the informationis drawn fromConti . In this respect the confus ion on Fra Mauro’s map i s muchgreate r than that on the map of 1 447 , where the river of Av a is at leas t to the

eas t of the Ganges .

It i s inte res ting to note that the river of Av a is shown as taking its originfroma lake , jus t as BARROS reported les s than a hundred years later. Shou ld FraMauro’

s name les s lake b e an embryo of the mys terious Chiamay Lacus which caus ed

so much mis chief during later centuries ? I t is not impos s ible that the same Burmes emyths and legends have been the root fromwhich both lakes have derived the i rappearance on early E uropean maps . Whe the r this interpre tation b e correct or not

it shou ld b e remembered that on a map of 1 459 no fewer than three lakes are placednear the s ou rces of the Ganges and Indus

,though two of themare , jus t as the

sources , placed at the s outhe rn s ide of the Himalaya .3

I Prof. Theobald Fischer has published a »Fac-s imile de l Mappamondo di Fra Mauro de l lanno 1 457 , Venez ia Another Mappamondo d i Fra Mauro i s publ ished in Stud i l tal ian i di Fi lologia Indo-Iran ica d iretti da Frances co L . Pullé , Vol. IV . Atlante de l la Parte 1 . Firenze 1 90 1 . In Holdich

s

T ibet the Mysterious is a reproduction of the part of the map where the name Tebet is ins cribed , p . 8 .

For our purposes the sketch in Zurla’

s work , s upra cit. , w il l b e suffic ient. It has the title : »Abbozzodel M appamondo di F.

Mauro Cama ldole se , Cosmografo Incomparabile a l la M eta del Sec . XV. (Gia’

es is tante n el la biblioteca d i 8 . M iche le d i Murano ora ne l la2 The hydrography of Genes is 2 : 1 0—1 4 has dominated the geographica l world formany cen

turies , as pointed out in earli er chapters of this work . In a manuscript co l lection found by MarioLonghena at Parma there i s , amongst other things , an anonymous Orbis De scriptio, probably fromthebeginn ing of the 1 4th century , where the hydrography of Genes is i s explained as fol lows : »At fluv i us

egred iebatur de loco v oluptatis ad irr igandumparadi sumqu i inde di v iditur in quattuor capita nomenun i Phison id es t Ganges ipse est qu i circu it omnemterramE iulat id es t Ind ie ub i nascitur aurumi llius terre optimumes t. Ib ique inv en i tur bde ll iume t lapis onich inus . E t nomen fluv io s ecundo Geon

id e s t N i lus , ipse est qui c ircuit omnemterramE thiopic. Nomen vero flumin is terti i T igris . ipse v adit

per as syrios . Fluv ius autemquartus ipse es t‘

E ufrates : ipse v adit per terramchaldeorum.» Stud i Ital ian i Anno V, Vol. V . Append,p . 2 2 .

3 Fra Mauro’

s map is influenced by Marco Polo . There i s a lake Lop , and a Deserto Lop ;there are Samargant, Balch , and Jerchan . Tangut i s placed at Lake Lop. Chatajo and C ina are

CHA PTE R XX.

MAPS OF TH E S IXTEENTH CENTURY.

On a map called Taou la S uper/207 2's [na’z’z oi Tartar i e.? M aj orz’

s fromPtolemasus Argentorati 1 52 2 , we find at leas t the name of Tebe t north of India , as a

su rviva l from Fra Mauro ’

s map . Baron A . E . NordenskiOld also says that thismap is almos t exclus ive ly founded on Marco Polo traditions .

I But Tibe t soondisappears aga in .

Such is a lso the cas e with the map of the world by NICOLAS DE SLIE N S , of

1 54 1 , PI. XIV .

2 It s eems to b e made es pecial ly for the us e of sa i lors , for along thecoas ts and is lands there are nume rous names , b ut in the inte rior of the continentonly a very few . The Indus and Ganges are the re , as usual , b ut they are les s correct than on Ptolemy ’s map , be ing me ridiona l and paralle l with each other. Anothe rshortcoming of the map is that the un inte rrupted range all through As ia has beendis solved into many different and qu ite isolated mountain groups , each giving ris e toone of the fe ede rs of the two great rive rs . Only the source of the ma in branch of

the Gange s comes froma group cal led Montes Iman i . Fu rther north are , fromeas tto wes t , Can de Tartari e , Cathay , Sogdiana and Bactriana.

On Pl . XV we s ee an example of the bas tard maps , whe re the inte rior ofAs ia is of pure Ptolemaean type , whe reas the ou tl ines of the coas ts are founded on

late r obs ervations . It is an anonymou s map of the world , which , upon insufficientevidence , has been ascribed to GIACOMO GASTALD I . C . H . COOTE has proved thisto b e wrong , and found that i t mus t b e bas ed on contemporary Span ish and Vene tianmaps , for ins tance , thos e us ed by DIE GO HOMEM in 1 558 , and DIE GO GUTIE RRE Zin 1 562 , which are derivations fromDIE GO R IBE RO of 1 529 , and others . The map isin fou r she ets of the two hemisphe res , and dated Venice , Tramiz ini , It is of

I Facs imi le-Atlas , Stockholm 1 8 89 , p . 7 2 .

2 K artographische Denkm’

aler zur E ntdeckungsgesch ichte v on Amerika , As ien, Austral ien und

A frika , aus dem Bes itz der Komglichen 0flentl ichen B ib l iothek zu Dresden herausgegeben v on

Viktor H antzsch und Ludwig Schmidt, Le ipz ig 1 903 .

3 Remarkable Maps of the XV th , XV Ith XV IIth centuries reproduced in the ir original S i ze ,1, the Bodel Nyenhuis Collection at Leyden , E d ited by C . H . Coote , Amsterdam1 894 .

1 80 MAPS OF TH E SLXTE E NTH CE NTURY .If, therefore , both in the coas t-l ines of the Indian peninsu la and in the oro

graphical arrangement of the inte rior, the influence of Ptolemy has comple te ly disappeared ou Gas taldi

s map , we find s ti l l a las t trace of his hydrography in the

topographica l s i tu ation of the Indus and the Ganges , as we l l as the i r principal tributaries . All the feede rs come fromthe s outhern s ide of the Himalaya . Ches imur,Kashmir, i s s ituated on the upper cours e of the main Indus which comes fromNangracot, in the ne ighbou rhood of which is a lso the source of the Ganges . It i s difficu lt to identify the Satlej , unles s i t is the great tribu tary on the bank of which is a

town cal led Capelang. That Moltan is placed on the lowe r cou rs e of the same rive rand Lahor on one of i ts tributari es s ign ifies nothing, for such mis takes are verycommon in much later years .

In the regions north of the Himalaya the greate s t confus ion prevai ls . Thereis no s ign of a Tibetan highland , nor of a country Tibet, though Fra Mau ro hadentered i t on his map a hundred years earlier . Tangv t Pro . is placed north of

R egno de Gamv l, and D is e rto de Lop north of D is e rto de Gamv l, on which a legendspeaks of the horrors of the des e rt . The »Stagno» be tween the two des erts may b eLop

-nor. But everything is ups ide down , and to find s ome more traces of Marco

Polo we have to leave the bas in of Lop and ente r the bas in of S ir-darya . This rive ris cal led Tahosca, or, on the Prima Pars As iae , Tachosca . It takes its ris e in a

meridional range of mountains , at the foot of which is Marco Polo ’

s » large town»Lop. Proceed ing we s twards we find Ciarcian , Poin , Cas char, Cotan , Caschar, all we l lknown fromMarco Polo . Kashgar has thu s been marked tw ice . Acsu is also the re ,on the same tribu tary as a s econd Lahor. Al l thes e towns are s i tuated on the S ir

darya and its tribu taries . But we have no right to blame Gas taldi . H e has ce rtainlydone the bes t he cou ld w ith the material exis ting at his time . Regarding Lop, forins tance , his map agrees with Marco Polo ’

s text : »Lop is a large town at the edgeof the De se rt , which is cal led the Des ert Lop On qu itt ing this city they ente rthe Dese rt .» I And w e mus t not forget that Marco Polo has hardly anything to say

of the moun ta ins and rivers of thes e regions .

In oppos ition to Fra Mau ro , Gas tald i has no lakes in or near the Himalaya .

On ly Abia f. , Ab-i-amu fluv iu s , comes froma lake furthe r wes t . On one of the

Chines e rive rs flow ing eas twards we find the towns Tu rfan and Gamu l .Taken as a whole Gas taldis map does not teach us anything new of the

countries north of the Himalaya , and it coa i a’ not , for no new materia l had beencollected s ince Marco Polo . And s ti l l the genera l appearance and arrangement ofthough he means the other place of th e same name , and not the N agrakot which has given Himalayaits temporary nam

Speaking of Tavern ier Levi has the fol lowing note : »Les monts de N augrocot sont l’H imalaya.

N augrocot, sous la forme moderne : Nagarkot, es t un temple et un pelerinage célébre du pays deKangra , qui e s t s itué a l’Oues t de S imla , an Sud-E s t du Cachemire . Au XVII° s iec le , on étend cc

ri oma toute Ia chaine qui sépare l ’Inde du T ibet.» (Le Nepal , Vol . I . p .I Yule ’s Marco Polo , Vol . I , p . 1 96.

GRISE LLIN I AND THE WALL-MAPS IN SALA DE LLO SCUDO . 1 8 1

his map marks an enormous improvemen t i f compared w i th thos e of his pre

decessors .

In this connection the famou s Wal l—Maps in the Sala de llo S cudo , in the Doge ’sPa lace at Ven ice cannot b e pas s ed by with S i lence . At leas t one of them

, emb racingthe whole eas tern half of As ia , and part of America , i s of special inte res t to u s .

I

The fou r maps we re drawn by G iacomo Gas taldi in the middle of the S ixteenthcentu ry in place of the olde r maps which had been des troyed by fire in 1 48 3 . The

map of E as t As ia is , beyond doubt, a work of Gas taldi . Al l fou rmaps were res toredby FRANC ISCO GRISE LLIN I in

Comparing the two maps , Gas taldi 1 56 1 , and Gris ellini’

s res toration of 1 76 2 ,

with each othe r,we mus t confes s that the latte r is by no means an improvement ,

so far as our regions are concerned . The coas t- l ines of the Indian pen insu las havenot unde rgone any change in 200 years .

3 The two great rivers and the northernmoun ta ins are practically the s ame as before . Bu t Gamu l , Hami , which in 1 56 1 at

leas t was placed north of the Hima laya,i s now removed to the s ou th of i t, amongs t

the feede rs of the Ganges . North of eas te rn Himalaya is Cataio Provincia, and

north of i t Tangu t. S t ill further north is the Des erto di Lop . E .S—E . of the sou rcesof the Gange s , and N .N -E . of the mou ths of the same rive r, i s Theb eth , which therefore s eems to b e placed somewhat sou th of the eas tern continuation of the Himalaya .

Gris ellin i,mus t b e s aid to have been success fu l in pres erving all the charac

teris tic features of Gastaldi’

s map, and he has been extreme ly cons cien tiou s in avo iding all the improvemen ts which were so eas i ly ava i lable at his time . The on lyfact that he has saved such an absurdity as the Chayamay Lago is su fficien t proofthat he wou ld not alter Gas taldi ’s geography in the leas t . For a lready on D E LISLE

s

map of 1 7 2 3 this lake had disappeared altoge ther, and at Gris ellin i’

s time nobodybe l ieved in its exis tence .

4

I In 1 909 I got a photograph taken by the kind in itiative of Prof. M ittag-Leffler in Stockholmand Prof. Volterra in R ome . Later on I had another photograph taken , of which P l . XVIII , i s a reproduction . It should b e compared ‘

with the de l la tavola de l Palazzo Ducale di Venezia» , inStud i Ital ian i

,Anno V, V ol . V , p . 58 .

2 Compare Dr. E . W. Dahlgren in my »Tran s-Hima laya» , V ol. II , p . 406 . The s ecretary ofthe Societa R amus iana in Ven ice , P . L . Ramb aldi , wrote , in 1 908 a very able memoir on the fourmaps , which was s ent to Prof. Volterra , and by himto Prof. M ittag-Lefll er. I have the manuscriptat my d isposal , for so far as I know it has not been printed . In it we read : »A umrestauro avevadetto i l Gri s e ll ini , non s i puo pen sare : non era ll caso di ripul ire , riparare , accomodare 11 soloed un ico Che me rimane é que l lo di rifarle di pianta .» Gris ellin i

s proposal was accepted by theSenate and it was dec ided that the maps should b e restored »as a mark of honour to the PatricianM arco Polo, whose d iscoveries and voyages had been de l ineated in many places of the said maps» .

Ramb ald i shows that Gri sellini ’s intention was to copy Gas taldi ’s maps , and that the Government agreedto that plan . Generally speaking the four central geographical decorations in . the Sala de l lo Scudo

may there fore b e said to give the same repres entation as the maps fromthe midd le of the s ixteenthcentury . Compare also Zurla : Di M arco Polo V ol . I I

,p . 3 74 , n .

3 »N on v’

b a dubbio ; l ’attuale d i segno de l l ’ A s ia su l la parete de l la Sala de llo Scudo arieggi a intutto i l tipo Gastald iano , come fac i lmente s i puo vedere per le due pen isole de l l ’ India e de l l ’ IndoC ina», (Pu lle’ in Stud i Ital ian i Vol . V , p .

4 Compare , for instance , E u ler’s map of 1 760.

1 82 MAPS OF TH E SIxTE E NTH CE NTURY.The s econd As iati c map , which was , at the s ame time , re s tored by Gris ellini ,

embraces wes tern As ia , including the Black S ea and the Caspian , and contains all

the geographical knowledge of his epoch regard ing this part of As ia . As a s u rvi valfrom the Vene tian map of 1 447 i t has s ti l l an unknown regi on with >>S cytae non

subjecti », otherw is e , at its eas te rnmos t edge , i t gi ves us such modern denominationsas Cote l i H indokusj , i . e . Kote l- i-Hindu-kush . Jus t s ou th of the Fontes Ghjon or

sources of the Amu -darya is ParvumTibe t with its capital E ske rud s eu Tibe t, perhaps Iskardu . The north-eas tern part of Hindu -kush is s ti l l cal led Caucasus , and

sou th of it is R egn umCachemirae . This map, therefore , does not at all touch the

regi on w e are s tudying, and the information it gives be longs to a time , 200 yearslater than Gas taldi .

On D iogo Homem’

s map of the world , Pl . XIX, the Indus is almos t as us ual ,while the Ganges comes fromthe NE . ins tead of N .W . The length of both rive rsis very much exaggerated , depending on the fact that the Himalaya has been re

moved northwards to the centre of As ia , halfway be twe en its north and sou th coas ts .

North of the mountains is India des e rta , and sou th of themDes ertumIndie . The

date of th is map s e ems not to b e known , and on ly a ve ry few dates fromD iogoHomem’

s l ife have been pre s e rved to our time . Mos t of his maps remained as

manus cripts ; thos e known we re made in abou t 1 530 to 1 576 .

I

We now come to GE RHARD ME RCATOR ’

s famous Worldmap of 1 569, of whichI have a copy on PI. XX , repres enting our regions of As ia ? I t is the mos t importan t cartographicalmonument of the s ixte enth century and as such itmay b e regardedas the foundation of the more recent s cientific geography . It shows how difficu lt itwas even for the mos t erudite s cholars to abandon Ptolemy, for Mercator does hisbes t to s ave the Ptolemman image of As ia , and to reconci le it with the mode rnd is coveries . The refore he has , with real ly touching pers eve rance and indus try createda mons trous repres entation of S .E . As ia , at a time when thes e parts we re muchbetter known .

3

The coas t-l ines of the Indian Pen insu la are ve ry good indeed , and even be tte rthan thos e of Gas tald i a few years before . In the general orographical arrangemen t werecogn is e the ranges of Ptolemy , the »spina dorsal is » of all the maps , and from itthe s eve ral ramifications in diffe rent di rections ; the me ridional Imaus , the meridionalrange wes t of i t , which is a wate r-parting of the S ir-darya, the Cas i i montes and

Asmirae i montes s eparating the Oechardes from the Baut isus , the Bepyrrus mons , e tc.

On Ptolemy ’s map, the Oechardes and Bautisus run eas twards , on Me rcator’sthey run fromsou th to north . On the Oechardes and its tri bu taries we recogn is e

I Hantzsch and Schmidt : Kartographische Denkm'

aler. Our P l . XIX i s a reproduction fromapart of the ir Tafe l XVII .2 Dre i Karten v on Gerhard M ercator : E uropa—Britis che Inseln—We ltkarte . Facs imile-Licht

druck nach den Originalen der Stadtbib l iothek zu Bres lau , herausgegeben v on der Ges e l l schaft fli r E rdkunde zu Berlin . Berl in 1 89 1 .

3 R ichthofen : China I , p . 643 .

SOUTHE RN TIBET, VOL. 1 .

Bri t ! ” J!

Part of Mercator’s Map of the World , 1 569.

1 84 MAPS OF TH E SIXTE E NTH CE N TURY.The Indus sys tem is nearly exactly the s ame as Ptolemy’s . That the ma in

Indus is shown as coming down from the mountains which borde r Pami r on the

s ou th may b e regarded as an improvement . The Zaradrus , S atlej , has not changed itsappearance s ince Ptolemy ; the on ly diffe rence is that, on Mercator’s map, i t comes froma meridional range s eparati ng the catchment '

area of the Indu s fromthat of the Gange s .

The mos t extraordinary thing is , however, the way in which' Mercator has

drawn the Ganges . He iden tifies Golfo de Bengala w ith Ptolemy’s S inu s argaricus ,

and makes two rive rs jo in in the Ganges de lta . The on e fromthe w es t is cal ledGu enga flu .

, and corre sponds to a ce rta in extent wit h Gas taldi’

s Ganga f. But on

the latte r i s Oris sa and i t is therefore the same river as the presen t Mahanadi whichindeed flows through Oris sa and has i ts own independent mou th in the Gu lf ofBengal . The lowe r cou rs e of Me rcator’s Gu enga , on the othe r hand , repres ents , asi t w e re , a las t fragment of the clas s ical Ganges . The s econd rive r comes fromthe

north w ith two branches , one fromamountain range s ou th of Mien regnum, the othe rfrom V indius mons ; this rive r, afte r the junction

,i s cal led Chab eris , and identical

w i th Ptolemy ’s Chab erus , which fal ls in the S inu s argaricus . As the Chab eris flowsthrough Aracam, a country which Gas taldi had placed correctly , Me rcator has takena rive r fromthe wes t and a country fromthe eas t and le t themme et in the ne ighb ourhood of Bengal . In 1 569 the coas t of Arakan was ve ry w e l l known by all

Portugues e s a i lors , s o one cannot unders tand why Mercator has placed this coas tland in the interior of India .

The wors t i s , however, that he removes the whole Gange s sys temfar to the

e as t , places i t in southern China and confounds the Gange s w ith the Hs i-chiang .

There fore Ptolemy ’s India in tra GangemFluv ium‘

has to fol low on the way eas twards ,and is to

'

b e found in uppe r Bu rma,Yunnan , and Kwang-s i where w e read thes e

words : Indos tan qu e v eterib us Ind ia in tra Gangem. The rive r its e lf he cal ls Cantanflu . ol imGanges , and the city Cantan ol imGange

,which is s i tuated in the Hs i-chiang

de lta e xactly at the place of Can ton . The feeders of the Ganges are exactly the

s ame as on Ptolemy ’s map : D iamuna , Ganges , S arabis and Bepyrus . On the uppe rD iamuna , Jumna , is a town Cabol ; b u t on a wes tern tribu tary of the Indus the re are ,

a ls o correctly, both Cabu l and Ches imur. The mounta ins fromwhich the feeders of

the Ganges come be long to the Imaus , and there i s a legend : Formicae hic aurumeffodientes homines s unt

,the old s tory which firs t was told by He rodotus .

Dr . Dahlgren has expla ined the cau s e of this s ingu lar mis take . On Mercator ’smap of the world w e read, as an addi tion to the name of S umatra : ol imTaprobana .

Ins tead of iden tifying Pto lemy ’s Taprob ana insu la Wi th Ceylon ,he be l ieves i t to b e

S umatra . And as Ptolemy’s Ganges was s i tuated e as t of Taprob ana , the real Gangesmus t come ou t to the s ea eas t of S umatra . He finds even a res emblance betweenthe words Can ton and Ganges .

I He confus es two is lands sou th of As ia , and thereI Dahlgren : Les Debuts p . 3 4 .

1 86 MAPS OF TH E SIXTE E NTH CE NTURY .The location of Theb et reminds us of the way in which it was placed on

the map in S ala de l lo Scudo . But i t i s s urpris ing to find Hami , Turfan and Tibe tplaced on the banks of one and the s ame river, namely, the Mekong ! Only the

extreme ly meagre reports of Odoric and Andrade explain that such mis takes cou ldb e made after the i r journeys , though , of cours e , very likely Orte l ius did not evenknow the i r reports .

For his map [mi zé Orz'

en ta l z'

s l nsu larumgu e adz'

acz'

entz’

zmz typus , Pl . XXII,Orte l ius has consu lted Gas taldi in orographical matters , as is s een both by the namesand the formof Dalanguer mons , N augracot moms and V s sonte mons , and Mer

cator in hydrographical matters . Mercator’s Indus is of Ptolemman type , and s uchi t remains w i th Orte l ius . The Ganges is combined with the Canton river, and s ti l li t is much shorte r than the Indus , which gives u s an idea of the great deformat ionthe formof the continent has had to suffer. E ven in deta i ls Orte l ius has followedMe rcator ; so , for ins tance , Mien , or Burma , is placed wes t of the Jumna : the coun

tries are removed to the wes t , the rivers to the eas t . It is a re l ief to re turn to

Gas taldi afte r such cartographical orgi es !Pl. XXIII , final ly, is a reproduction of a part of O rte l ius ’ map : P ers i ci s i v e

Sop/zorzwz regn i typos , showing the sources of the Indus and the Oxus . The formof the Indus is , to a cons iderable degree changed fromthe sheet in the same atlaswhich I have repres ented as P] . XXII . But the type is practical ly Ptolema an .

Chesmur, Kas hmi r, and other names on the banks of the fe eders are the s ame as

on Gas taldi’

s map of 1 56 1 . The mountains Dalanguer and Naugracot are artis tical lydrawn , in horizontal perspective , the ve rtical perspective be longing to a much late rtime . They form the watershed between the sources of the Indus and the Amudarya , a vi ew that is generally correct , and had been adopted by Gas taldi , after theoriginal example of Ptolemy. The lake fromwhich Gas taldi makes the Amu-daryas tart, is much exaggerated in s ize on Orte l ius ’ map , whe re it is cal led Abie fons .

E ven in our days the view that the Amu-darya s tarts froma lake , Chakmak, hashad defenders , a View that has been defin i te ly proved to b e wrong by Lord CURZON .

I

Amongs t the maps fromthe s ixteenth centu ry which I have s een , one , enti tledC/zz

'

ne , o/z'mS z

n arzmz reg i on i s , noua descripti o. Auctore Ludom’

co Georgi a, 1 584 ,

(PI. XXIV) is the firs t in which Ptolemy ’s Ganges has been comple te ly, and in amos trevolutionary way abandoned . We shal l have to cons ider this extraordinary type inthe next chapter, deal ing withmore famous draughtsmen , who adopted this type of the

Ganges . For the pres ent b e i t sufficient to say, that the two principal sou rce branchesThe Pamirs and the Source of the Oxus . London 1 896, p . 3 6 e t seq . Reprint fromthe

Geographical Journal for July , August, and September 1 896.

When Captain John Wood , in 1 83 8 , discovered the Sor-kul (Lake V ictoria) , he was sure hehad found the source of the Oxus . H e says : »be fore us lay stretched a noble b ut frozen s heet ofwater, fromwhose western end issued the in fant river of the Oxus» . And again : »This , then , i s thepos ition of the sources of thi s ce lebrated river H e cal led the lake S ir-i-kol. A Journey to the

Source of the R iver Oxus . N ew E di tion. London 1 8 7 2 , p . 2 3 2 .

SOUTHE RN TIBE T,

CHA PTE R XX I.

MAPS OF TH E SEVENTEENTH CENTURY .

On our wanderings through the centu rie s we have to proceed beyond the

middle of the s eventeenth before w e can talk of any defin i te change in the rightdi rection in the fie ld of Central As iatic cartography . Properly speaking it is a dis

appoint ing task to s earch through the maps of thre e hundred years w i thout findingany re l iable information about the country we are s tudying in this work

, and s ti ll i tis a task of abs orbing interes t , at leas t for thos e who have s e en thes e regions w i ththe i r own eyes , and the refore w i l l unders tand how and why this terra , surroundedwi th its gi rdle of montes cou ld remain z

ncogm'mfor s uch a long time .

The maps w i l l te l l us , far be tter than any words , how human know ledge , s lowly b uts ure ly , conquered the coas ts , and , s tep by s tep , proceeded fromthemtowards the

interior ; how China , India , Turkes tan and S iberia l ittle by little assumed more rea

s onable forms , whi le the nucleus of High As ia , Tibet , s ti l l remained unconquerable .

In the s eventeenth cen tury we shal l s ee how the s cientific explorers at home and in

the fie ld w i l l approach the borderland of Tibe t w i thou t reaching the country i ts e lf.On some maps fromthis period even the name of Tibe t w i l l b e mis s ing, while othe rdraughtsmen are very unce rtain of its whereabou ts . S ti l l the journeys of Odoric and

Andrade remain forgott en and have no influence at all on the maps , with only one

exception , Kircher’s map , Pl . XI . But afte r the jou rney of Gruebe r and Dorv i lle

in 1 66 1 and 1 662 the cartographe rs cou ld no longer ignore the mys te rious country,which so far had been a fata morgana to them, and had, towards the ev e of the

next century, to as s ign for i t a place somewhe re north of India .

The firs t map we are going to dis cus s is the »N ov z'

s s z'ma ac exactz

s s z'ma toti u s

07 62'

s termrnmdes cript i omagnet » by JODOCUS HONDIUS , Ams terdam 1 6 1 1 , PI. Xxv .

l

He conscientiou s ly u s ed the mate rial exis ting at his time and as no new informati onhad been won fromthe countries north of India, he has nothing to add beyond the

Map of the World by Jodocus Hondius 1 6 1 1 . E d ited by E . L. Stevenson and J . F i scher.N ew York 1 907 .

SOUTHE RN TIBE T, VOL . 1 .

Ind ia and Central As ia on Jodocus Hond ius’

Map , 1 6 1 1 .

XXV .

HONDIUS AND B OE IUS . 1 89

s tandpo int of Mercator and Orte l iu s . Me rcator had spoi lt the Ganges and had beenas s is ted by his personal friend Orte l ius . Unde r the we ight of s uch high authorityHondius cou ld not retu rn to sound clas s ica l l ines , and did not dare to take the

Ganges back from its unfortunate vis i t to China . The Indus ke eps its ground, and

has the same type as on the maps of Gas ta ldi , Me rcator and Orte l ius , a typewhich was inaugurated by Ptolemy . Orte l ius and Hondius have made the river ,afte r the junction of the Panjab feeders , too long , which i s a s tep in the w rong direction fromPtolemy , Gas taldi and Me rcator. The Chiamay Lacus is also there w ithits fan of is su ing rive rs .

I have mentioned above Thomas He rbe rt ’s book , ‘ which brings u s nothingnew fromour rive rs and mountains . Bu t he had trave lled in India and knew thatthe Ganges was not s ituated in China . . So on the l i ttle map (Pl . X) accompanying his book , the Ganges has re turned to its old place , and i t ev en shows how the

sacred rive r comes down froma rock in shape of a cow’

s head, s i tuated n ear Hardware in S iba . The Lacus Chiamay he has left a lone . The gene ral features of the

Indus are somewhat be tte r than on the maps of the great cartographe rs , thoughAttock i s placed be low Mu ltan . The name Pen iab is adopted, .b ut S atlej , whichhad come down fromPtolemy to Gas taldi , Mercator and Orte l ius , has not be en ente red .

Q u ite a new type both in hydrography and orography of Central As ia is to

b e found on a map w i th the fol lowing title : N ov a 07 52'

s Tow/ aof zmz Googof ap/zz'

oa

ac Hydrograpaz'

oa D es cr iptio, ex optz'mz's ga z

'

oa sya o opi z'morami n 1205 ofiero A a c

z‘oram, Taoa lz

'

s a’os amfifa a Fra a cz

'

s oa s H ooz

'

a s (Pl . XXVI). 2 In the northe rn half ofAs ia the Oechardus fluv ius and Bautisus are s ti l l to b e found as in Mercator ’s time ,b ut Marco Polo ’

s s tations are transplaced fromOechardus to anothe r gigantic and

more we s terly rive r, on both s ides of which Pe im,Cotam,

Lop, D es ertumLop and

Ciartim are s i tuated . This rive r is partly in connection with the Chines e hydrographical sys tem,

partly i t comes froma ve ry great oblong lake , which i s also to

b e s een on Hondius’ world map (Pl . XXV). Fu rthe r wes t Tasken t is placed near

Kithai Lacus , through which the uppe r Ob flows . On the Oechardu s is Turfon , on

the Baut isu s Camu l. Caindu and Theb et which Me rcator had allotted to the Bautisus ,have aga in disappeared .

The Indu s i s unchangeable as always and exactly the same as on Hondiu s’

world map . Regarding the Ganges the d i fference is somuch the greater. The rive rhas re turned defin i te ly to the Gu lf of Benga l , its cou rs e is me ridiona l , and i t is inlength nearly twice the Indu s . The sou rce , near the Ptolemae an Turri s lapideamons ,i s therefore s i tuated almos t s traight north of that of the Indus . It is su rpris ing that

Some Years Trave ls into A frica A s ia London 1 63 8 .

2 Remarkable Maps E dited by C . H . Coote , Amsterdam 1 894 . The ed itor informs us .

that it is the earl iest large map on M ercator’s projection issued in Hol land in the begi nning of the

seventeenth century . The original p lates were first engraved for B laeu in 1 605. The name of Jansson ius has been replaced by that of Hoe ius

,the name of Allard i (Coster) by that of A l lardt whoflourished about 1 640.

1 90 MAPS OF THE SEVE NTE E NTH CE NTURY .s o much unce rta inty cou ld preva i l in E urope regarding the Ganges ,

'

which had beendrawn far more exactly by the Alexandrine cartographe r 1 500 years before .

As a cons equence of the exagge rated length of the Ganges , the chain ofmountains , runn ing wes t to eas t, which had been accepted by nearly all geographe rs andmaps , has disappeared almos t entire ly . On ly the meridional Imaus is left in the

middle of a ne t of i rregu lar and fantas ti c range s .

One clearly s ees how Ptolemy’s au thori ty begins to shake . But this was no

gain for geography. Whe re Ptolemy has been respected, as for ins tance in the

cas e of the Indus , the Oxu s and the Jaxartes , the map is tole rable , but to the eas tand N E . of that regi on nothing corresponds to real ity.

We now have to cons ide r two maps that are of the same type as that onPl . XXVI . S o far as rive rs , lakes and mountains are conce rned it may b e cal ledJanssonius

’ type . The firs t , Pl . XXVII , has the ti tle : I ndi a ga a' Orz

'mla lz’sot [mo/e aa

’z

aoen tos . I t is firs t known to us in GUILJE LMUS BLAEU’

s »L o Tlzoa lre

o’u mona’o, ou N oz wol A tlas » , 1 640.

I The s econd, Pl . XXVIII , is fromIOHANNE S

IANSSON IUS’

N ewer Atla s , Weltoes okrez'

ozmg , Ams te rdam 1 64 1 , and has the

ti tle : A s i a rooms s amma ow ' a dolzh oa ta A a ot: l ad : H ondz'

o. It was prepared inJANSSON

s offices in 1 63 2 .

Comparing thes e three maps , Pl . XXV, XXVII and XXVIII , we at once s ee

that they are founded on one and the same original draw ing. The Indus is of Ptolemecan type . On themap bearing HOErUS

name , Pl . XXVI , the rive r flows S .S -W ; on

the two others S .W . As can b e s een on Iansson ius’

_

map of As ia, of which Pl . XXVII Iis a part, the s ource of the Ganges is much farthe r fromthe Indian Ocean thanfrom the G lacial , and shou ld on a mode rn map have to b e placed in Dsungaria ,4 8

° N . lat . ins tead of If the draughtsman had real ly fe lt confidence in the s ituation he has given to Marco Polo ’

s Cotan , Poin , Ciarci , and Lop, he wou ld havebeen a larmed at the thought that the Venetian trave l le r on his way eas twards fromthe las t mentioned place to S achin in the province of Tangu t , s hou ld have beenobliged to cros s the uppe r cours e of the s acred Ganges . But he has not been s ure at

all, and fol lowing the example of Me rcator he has ente red the s ame places once more ,name ly, on the great rive r, which we remembe r fromPl . XXVI , and which obvious lyi s a forerunne r to Hwangho .

On Pl . XXVI the lower Ganges had a S .S -E . cours e , on PI. XXVII and XXVII Ii t runs to the S .S -VV. No res t is gi ven the sacred rive r. In the cours e of a few yearsi t has been sw inging like a pendu lumbetween Canton and Bengal . And when at

las t the mouth ge ts s ett led in the Gu lf of Bengal , the“whole cours e of the rive r b e

gins to swing the othe r way, while the Indus , the river of Alexander the G reat, wastoo firmly fixed by his torical events .

Of the three rive rs we shal l have to work out through the following chapters ,the Indus , S atlej and Brahmapu tra, unti l I have the honour of leading the reade r

I Remarkable maps E d ited by C . H . Coote .

SOUTHE RN TIBET ,VOL . I .

PL. XXVIII .

Iohannes Ians son iu s , Ams terdam1 64 1 .

1 92 MAPS OF THE SEVE NTE E NTH CE NTURY .S irinaker. The rivers of Panjab begin to abandon the Ptolemaean type . As Lahori s correctly on the Ravi , the next rive r to the S .E . , wi th Sultanpore , mus t b e the

Bias , and the next , Cou l fl ., the S atlej , which has its sources in s ou thern ramifica

ti ons fromHima laya .

The great northern wate r-parting of the Indus and the Ganges is called Montde Caucas e . North of it are the sources of Amu -darya , Tartari e , Tibet, and Tu rques tan . Tibet is shown as a l i ttle part of Tartaric . South of Mont de Caucas e is a

n e t work of ranges in all d i rections . One is cal led Dalangu er Mont, anothe r Dow

Lager M .

, Dhau lagi ri (P) . N augracut is not us ed as the name of amountain ,though

as a ru le , on maps of the s evente enth centu ry this denomination s tands for nearlythe whole Himalaya . On S an son

s map i t on ly s ign ifies the province and the townin Kangra

,and is comparative ly we l l placed.

The upper Ganges goes through a lake at Hardwar in S iba . S erenegar, known .

from Andrade ’s jou rney , is placed be low ins tead of above Hardwar. The Gangeshas s t i l l a nearly merid ional cours e . The chief impres s ion one ge ts fromS anson

s

map is that it i s founded on fresh and more re liable information in the country i ts e lf.Another map of the same draughtsman has the ti tle : D escr ipt i on do [a Tar

z‘a rz

o, is engraved in Paris 1 654 , and published in S anson ’

s atlas : Carlos g en era lesdo totn‘es les parti es a

’a moa a’o, 1 658 and I 677 .

I Wi th a few excepti on s this map(Pl . XXX) shows all the characte ris tic feature s of thos e by Hoe ius (Pl . XXVI) and Janss onius (Pl . XXVIII). The O echardes and Bau ti sus are there under somewhat changedfi rms ; on the latter Marco Polo ’

s S achiou is ente red . The Caramoran or Hwanghohas dropped the branch which came fromthe great oblong lake , which , its e lf, hasd is appeared. Th e re i s no s ign yet of Koko-

nor which will s oon begin to wanderabou t in these regi ons as all the othe r lake s . The source of the Hwangho is s i tuated at a short dis tance fromthe coas t of the G lacial Ocean , jus t as before . The

Chines e have through centuri e s , though w i thou t su fficient proof, be en suppos ed to

regard the Tarim as the uppe r cours e of the Hwangho . If this view has beenkno wn to Ians sonius and othe r cartographe rs of the s evente enth cen tury, one cou ldunde rs tand why they have placed the s tations of Marco Polo on or near the uppe rcou rs e of the Hwangho . The rive r (Pl . XXX), on which we read Calachitae (Kara

and lowe r down Lop , s hou ld the re fore b e the Tarim. Cotam, Pc im, and

Ciarti am (Che rchen) are placed s ou th of the main rive r, as is indeed correct. As

the las t fragments of Ptolemy ’s Oechardes and Bautis us,with Tu rfan and Camul, are

als o , beyond doubt , the Tarim, we shou ld b e able to trace this rive r thrice on the

maps fromthe middle of the s evente enth century .

The clas s ical Imaus , that is to say the northe rn half of its me rid ional part is s ti l lthere , near the s ource of the Hwangho . It s eems to us s trange tomis s such geographical

V . Kord t : Materialiy po I s toriy Rus skoy Kartografiy, Series I I , I : XX . K iew 1 906 . In Nordensk iOld

s Periplus,p . 1 93 , i s a reproduction of Sanson

s map A s ia, fromL’

As ie en plus ieurs cartes ,Paris 1 652 .

MARTINI’S MAP or CH INA . 1 93 .

regi ons as Dsungaria , Tian-shan ,E as te rn Turke s tan , and Kwen-lun , though they had

been cros s ed and s e en by s eve ra l trave l le rs . How utterly they are mis s ing is be s ts een by the fact that the S ibe rian Ob and the Indian Gange s have the i r sou rce s inone and the same range . In a corne r be t we en two range s N .W . of the s ource of

the Ganges sufficient roomhas , howeve r, be en found for Tibre t forsan e t Tob rot .

I

On ly a year afte r S anson d’

Ab b ev ille’

s map of Tartaria had appeared , or in

1 655, Fathe r MARTIN I published his Atlas of China whichmarks a perfect revolu tionin the E uropean conception of the far eas tern empire .

2 We have to di rect our attention on ly to the fron ti spiece map : [mfiom

z

S z’

na ramn ov a des cr iptio, ofwhich Pl . XXXIi s a part . To begi n with , the cours e of Hwangho

,or Croceus flu .

,3 has been much

corrected,and the rive r take s its ris e in two lakes , Sos ing and S ings ieu , as is inde ed

the cas e .

4 Wes t of Hwangho the Koko-nor makes its appearance under the name

Mare n igrumS in is Cinghai . The Chines e name of the Blue Lake is Ts ing-hai , so

oa ra loam or oa s z

zmz wou ld pe rhaps have been more adequate than mgr/ am,in a

country wi th many Khara-nor and on ly one Koko-nor. How eve r, the exis tence of

the lake on the map prove s that Martin i has neve r confounded the Koko-nor and

the Chiamay lacu s , which somet imes occu rred in late r years . H e does not rejectthe latte r, which he calls Kia L.

, for he had_

no reason to disbe l ieve its exis tence .

Xamo de s ertumhe identifies with what the E u ropeans u s ed to cal l De s ertumLop . To the north of the s and des e rt he has Samahan Tartariae s ive S amarcandaePars , which is su rpri s ing. And s outh of the s ame des ert is Tibe t R egnvm. Thenfol low Kiang Regnvm and s gang R egnum all of which be long to the wes te rncountrie s of S i-fan .

5

As regards the s ituation of Tibet, Marti n i has given a new impu ls e to the

deve lopment of E u ropean know ledge . And s ti l l he has a ve ry unce rta in conceptionof whe re it really is . On his map he places i t in the Koko-

nor d is trict which weknow is inhabited by Tangut and Mongol tribe s . But in re lati on to Martin i ’s Kokonor i t i s to the S .W . as is correct. Speaking of the country Laos , or the

p art of S iam,which he correctly places eas t of Arakan , Father Martin i quotes the

manu s cript of Fathe r LE R I A , in whomhe has great confidence , and who,du ring his

mis s ion trave ls , had vis i ted Laos , and found i t inundated by i ts rive rs in summe r.>>For whe n the snows of the mountains of Tibe t,—I (Martin i) be l ieve that he (Leria)

I Prob ably misprint for T ibb et and Tob b ot.2 Novus Atlas S inens is a M artino Martin io soc . Ies v descriptv s , 1 655.

3 »fluvrus Hoang nomine, Croceumaut luteumd ico, id en imS in ice somat, a colore» (Mar

t i n i , p .4 »Ortu s H oang fluv i i e s t inter Australes montes Q uenlun ,

quos incolae O tunlao v oc i tant.» Otunlao must b e the Odon tala of the Mongols , spec ia lly as Martin i re lates that the water comes up inmorethan hundred springs .

5 In an artic le : R elation de la Tartaric Orientale . Father Martin i says : »Les Chinois d is ent quec e Royaume (S i -fan) e s t borné par les montagnes de M in ,

8 : par la R iv iere Jaune qui y pas s e . Ces

montagnes ont beaucoup d ’étendue, se jo ign ent enfin a ce l le s de Q uenlun , qu ’

on nomme autrementl es montagnes Amasée s , d ’ou la R ivi ere Safl’ran e’ e (Hwangho) tire son origine .» R ecue i l de Voiagesau Nord , Tome III , Amsterdam1 7 1 5, p . 1 63 . This i s an extract fromthe Latin text in h is atlas .

25—1 3 1 3 8 7 I

1 94 MAPS OF THE SEVE NTE ENTH CE NTURY .means Yunnan , which borde rs on Tibe t ,—begi n to me lt , the rivers swe l l to such adegree that the whole country becomes inundated of i t , as E gypte fromthe watersof the N i le .

» In this cas e Le ria probably means S alwen and Mekong which havethe i r s ou rces in Tibe t , and on ly flow through Yunnan .

Final ly Martin i has heard , from Chine s e informants , that nearly all of the

greates t rivers which flow to the sou th , amongs t themthe Ganges of Bengal , and

the rive rs of Pegu and S iam, have the i r sou rces in the mountains of Kwen-lun . We

mus t remembe r that Martin i had at his disposal for the coun trie s wes t of China on lyvery poor and fals e maps , and that he had no easy task in locating the Kwen—lun ;which he s uppos es is not ve ry far fromLahor and the kingdomof Tibe t. 2 Thus inthis cas e the Gange s comes fromthe Kwen-lun . But at anothe r place 3 he s ays :»Jus t there (note : fromWhe re the Ganges take s its sou rce), towards the wes t, is a

ve ry great lake which is cal led Kia , fromwhich come the Gange s 8: the othe r rive rsI have marked on the map.

» It wou ld therefore s e emas i f Martin i imagined the

Kwen-lun to b e s ituated somewhe re between the Tibe t Regnvmand the Kia L . of

his map , as the Ganges comes both fromthe Kwen-Iun and the K ia Lacus . And

he re we have the poor Ganges s tarting upon a new vis it to China .

I t s eems to have taken a few years before the great northe rn branch of the

Hwangho , coming from i ts lake near the coas t of Mare Tartaricumwas final ly ex

tirpated fromthe maps . While F. DE WITT on his map of abou t 1 660 4 has acceptedthe Hwangho of Marti n i , he has re ta ined the long northe rn branch , which is to b e foundon Ludovicus Georgi us ’ map of 1 584 (PI. XX IV), and then on Hoe iu s

of 1 600 (PI.XXVI), and on Ians soniu s

1 64 1 (PI. XXVIII). THE VE NOT , who makes a reprint ofMartin i ’s map, has ente red a rive r coming fromthe wes t and flow ing north of Kokonor, s and not exis ting on the origi nal map. Kirche r fa ithfu lly fol lows Martin i . 6

Return ing to the S .W ., we find that also the example gi ven by Sanson d

Abb e

vi lle in 1 654 was followed in the next years . S uch is the cas e with the Indianmapof de Witt , abou t 1 660, on which the Cincu i Hay Lacus s tands as an anachron ism,

7

and w i th a map of NICOLOSI , To the s ame category be longs the map i llustrati ng the work of FRAN QOIS BERNIE R , and certainly innume rable othe rs .

Novus Atlas , p . 2 4 ; Recue i l , p . 1 64 .

2»Montes Q uenlun, qu ibus ortum(of Hwangho) debet, omn ino Amasaeos esse exi stimo, eosquehaud i ta procu l di ssi tos ab a ltera magn i Mogori s R egia, quamLaor v ocant, aut a regno Tebet, imo

iisdemin montibus scaturire GangemBengalas . M eson Laorum, H i s toremque Camb ogiae , aliosque non

ignob iles fluv ios , qu i S ion Pegu regiones irrigant, ipse locorumregionumque s itus suadet. plurimaenimmaximaque flumina , quae Austrumversus manant, inde suampetere originem, demonstrant S inas .»

N ovus Atlas , p . 1 4 .

3 Novus Atlas , p . 2 3 ; R ecue i l , p . 1 63 .

Tabula Tartariae et majoris partis Regn i Chime .

R e lations de divers voyages curieux,Paris 1 666.

Ch ina I ll ustrata , 1 667 .

Magn i Mogoli s Imperi vmde novo Correctume t Div i sumPer F . de Witt Ams telodami .Del l ’ Hercole e s tv dio geografico d i G io. Battista N icolos i Vol. 2 . Roma 1 660.

Voyages de Francois Bern ier, Amsterdam1 699. The first edition appeared in 1 670.O

M

Q

O

M

&

1 96 MAPS or THE SEVE NTE E NTH CENTURY.t ion from a common source is evident, though they have been spe l t in a more or

le s s carefu l way. Sanson , 1 654 , rejects the lake, s o far as can b e judged fromhis

maps Pl . XXIX and XXX. On Martin i ’s map , 1 655, the oblong lake has d isappeareddefin ite ly , b ut the name has , in a pu re r ve rs ion Cinghai , be en transfe rred to the

real Koko-nor.

\Ve have thus the s erie s Cirhcuai , Cu icu i Hay, Cincu i Hay, and Cinghai , all of

which are the Chines e Ts ing-hai or Koko-nor, and we are jus tified in surmis ing that

the mys te rious lake is nothing e ls e than the Koko-nor. For some 60 ye ars Hwangho has be en suppos ed to come , partly, fromthis lake , b ut Fathe r Martin i has gotmore re l iable news , and shows that it has no connection whateve r w ith the rive r.

Pl . XXXI I is a reproduction of a mos t inte re s ting map : I nd i a Ori en ta l i s ma

n on f u s a lammz adi a oei i ti am N ov a D osom'

pti o P o? N i oolaam V i s so/zer .

I The

ed itor, C . H . COOTE , s ays of it : >>It w i l l be conven ient to note he re the date of the

production of this map ; i t w i l l b e found ins e rted in J . Jansson’

s N ov a s A tlas , das

i s t Welt B os o/zroi én 5tli v ol . , Gros s o A tlas (8 tli part) Was s or Welt, Ams terdam, 1 657

—58 fol i o. (B r i t . M a s . 9 Tao. Also in othe r edi tions of abou t

the same date .» It wou ld the refore seemto b e ce rtain that the mapmus t have been

drawn before 1 658 . But SYLVAIN LEVI proves in amos t convincing way that this cannotpos s ibly b e the cas e .

2 For NICOLAUS V ISSCHER has , probably as the firs t amongs tcartographers

,made us e of Grueb er

s discov erie s , and G ruebe r fin ished his journeyin 1 662 . According to Levi our map,

Pl. XXXII , was ins erted in A tlas M i n or s i v e

toti a s oroi s torraramt on tracta del i neati o, ox ooa atious N i t . Vi sso/zor (Ams terdamB at.apud N i oola amVi s s o/ter). The A tla s M i nor has no date , b ut the map, Pl . XXXII ,can unde r no conditions have appeared be fore 1 663 . Al l the name s mentioned inGrueb er

s narrative , as re lated inKi rche r’s Ch i n a [ll zzs t i f a ta , appear on the map, evenwith the characte ris tic spe l l ing which comes directly fromGruebe r. Coote

s as s e rtionthat the map shou ld b e taken from Jansson ’

s N ov a s A tla s is the refore , according to Lév i , inacceptable , and even a has ty examination shows that Levi is right . 3V is s cher place s N ecb al w ith Cutlu , a misprint for Cuthi S .W . of Las sa ; then fol lowCadmenda , Hedonda , Mutgari , Battana , Benares , Cadampor, and Agra . Patna hadappeared already on He rbert’s map of 1 63 8 (PI. X), and on S anson

s of 1 654 (PI.

XXIX), though placed far above Benares . When V is s cher hears the name Battana of

Gru ebe r, he be l ieve s i t is another city and has both Patna and Battana on his map .

Remarkable maps etc. Part III , N o. 4 .

2 Le Ne’ pal,Tome I , p . 90 e t seq .

3 Grueb er’

s stations were : fromLhasa 4 days to mount“ Langur, which K ircher identifie s withParOpami sus and Marco Polo’

s Be lor, b ut which in real ity i s H ima laya . Thence Grueber trave l led one

month to the first city of N ecb al (N epal) , Cu thi . The next station is Neste , and then the capitalCadmendu , fromwhere he had 5 days to H etonda (Hetaura) and 8 days to the first c ity of Mogu l ,Mutgari (Moti hari) . Further 1 0 days to Battana (Patna) , 8 days to Benares , 1 1 days to Cadamporand , final ly, 7 days to Agra . Compare al so : R ichard Tronn ier: Die Durchquerung T ibets se itens derJe su iten Johannes Grueber und A lbert de Dorv ille imJahr 1 66 1 . Z e itschr. d . Gese l l . f. E rdkunde zu

Berl in , 1 904 , p . 3 2 8 et seq .

1 98 MAPS OF THE SEVE NTE E NTH CE NTURY.narrative w e read : I »The firs t town they pas s ed in return ing was Gourtche , the las ttown dependent upon Kachemire , and fou r days ’ jou rney fromthe city of Kachemi re : from Gourtche

, they were e ight days in reaching E skerdou , the capital ofLittle Tibe t ; and in two days more they came to a smal l town called Cheker, alsowithin the te rritory of Li ttle Tibe t In fifteen days they came to a large fores t

,

on the confines of Li ttle Tibe t , and in fifteen days more they arrived at Kacheguer, asmal l town which was formerly the royal res idence , though now the King of Kacheguer

res ides at Iourkend, a l ittle more to the north , and ten days ’ jou rney fromKacheguer.»

There is no roomfor a doubt ; all the names mentioned by Bern ier are en teredon V is scher

s map . V is scher’

s Kachaure is s imply a care les snes s for Kachemire , andin Gourtche the c has been read as e . If V is scher had recogn is ed Bern ier’s Kachegue r and Iourkend as Marco Polo ’

s Cas car and Yarkan , he wou ld not have placedthes e citi es on the Ganges , though, afte r all, he is les s to b e blamed than S ansond’

Abb ev ille who , on another of his maps (Pl . XXX), has themon the S ir-darya .

And V is scher was jus tified in fol low ing Be rn ie r ’s text : »They s ay that Kachguer l iesto the eas t of Kachemire incl in ing s omewhat to the northward .» The rou te in questi on is not the ordinary Kara-korum road, which s tarts fromLeh and not fromSkardo or Iskardo, and which firs t reaches Yarkand and then Kashgar ; it is amorew es terly road Bern ier has heard of, which does not touch Yarkand at all.

As Bern ier ’s narrative was published for the firs t time in V is scher’

s mapcannot pos s ibly have been published before that year. 3

N ow , as Kircher’s map is from 1 667 , and V is scher’

s fromat leas t three yearslater, V is s cher has not needed to read C/zi i za I llu s tra ta , for he has found all

the material of G rueber ready on Kirche r’s map. Al l the names he u ses appearon the latter : Las sa , Cuthi , Cadmendu , Hedonda , Mutgari , Battana , Benares , Cadampor, and Agra . H is Be lor Mons and Cons agni Mons Lapideus are d i rectly copiedfromKircher, even the horizontal pe rspective and the shadows , as w e ll as the Mon

~

tes T ib et ici . H e has followed Kircher ’s example in adopting Marti n i ’s source of the

Hwangho , origo flu . Croce i , b ut retained, be tween i t and Lassa , themys terious lakeof Chiamay. The Mongol name for Las sa , Burantola , which probably appears forthe firs t time on Kirche r’s map , he has regarded as superfluous .

'

Kircher is a lso thefirs t to mention and to repres ent the name of Nepal on a map .

Pl . XXXIII shows u s a reproduction fromCANTE LL I ’S map of 1 68 3 , publishedby V . KORDT .

4 The s ources of the Indus and Ganges are as usual , the Satlej is rudiI E ngl ish E dition of A . Con stable , p . 4 2 7. Bern ier got his information frommerchants from

Kashgar whomhe met in K ashmir.2 H istoire de la dern iere Revolution des E tats du Grand Mogol , Dediée Av Roy, Par le S ieur

F . Bern ier Medecin de la Faculté de Montp ell ier. A Paris MDCLXX .

3 Dr. E . W . Dahlgren te l ls me that, accord ing to the best sources , V isscher’s Atlas M inordates fromabout 1 680. Under such cond itions Sylvain Levi ’s and my own efforts to prove that themap in question cannot have been publ ished before 1 663 , and 1 670 , should practically b e s uperfluous .

4 In M aterialiy po Istoriy Russkoy Kartografiy. The complete title is : La Gran Tartaria diu isane l le sue parti principal i da G iacomo Cante l l i, da V ignola, conforme le re laz ion i , Che s

'hanno da

200 MAPS OF THE SEVE NTE E NTH CE NTURY.narrative , and has , on his map , not a s ingle detai l for the road from the Ye l lowR ive r to Las sa . The Ital ian vers ion ,

which has s eve ral mis takes , b ut also s everalin teres t ing deta i ls , was the resu lt of an inte rview which Grueber gave t wo gentlemenin Florence in 1 666 , and which was published in later edi tions of Thevenot ’s work .

I

From there ASTLE Y made the following extract regarding the road to Tibe t, afterthey had cros s ed D es art Tartary : »Afterwards he came to the Banks of the KokoN or, which s ign ifies the Great S ea ; be ing a great Lake , or S ea , l ike the Caspian ,where the Ye l low R ive r has its Sou rce . Leaving the S ea behind him

,he entered

into the Country of Toktokay, which is a lmos t desart , and so barren , that i t ne edfear no Invas ion . One mee ts w i th nothing b ut s ome Ten ts of Tartars . It is wateredby the Toktokay, a ve ry fine R iver, whence i t takes the Name . It is as large as

the Danube , b ut so shal low , that i t may b e forded every-where . Thence havingcros s ed the Country of Tangu t, he came to R et ink , a very popu lou s Province , b elonging to the Kingdomof Barantola ; whe re at length he arrived, whos e capitalcity is cal led Las sa .»

2 Thence i t i s easy to s ee fromwhere Cante ll i has got hismate rial for the e n ti re ly new formhe gives Koko-

nor as w e l l as for his Toktokay,R etink (now R et ing

-gompa), Barantola , and Las sa , so much the more as he quotesboth the Jes u i t Mis s ionarie s and Thevenot . But i f Can te l l i had re ad his Thevenotmore atten tive ly, he wou ld have s e en that Gruebe r and Dorv ille , trave ll ing fromKoko-

nor to R etink , (the s ame way as 1 80 years later Huc and Gabet), had to crossthe country of Tangut . This has been placed by Cante l l i in Ala-shan and Mongol ia ,while hi s Re tink is on the S .W . shore of Koko-

nor.

The ve ry poor, b ut ve ry importan t le tters and inte rviews which , fromGru ebe rhims e lf, have come down to our time , have the refore fructified the map of Tibe t ina ve ry high degre e , and the hitherto almos t unknown country begins s low ly to emergefromunder the impene trable mis t which so long had covered its s ecre ts .

\VITSE N’

s map of northe rn and eas tern As ia, made in Ams te rdam 1 68 7 , is

regarded as one of the mos t remarkable of its time .

3 Pl . XXXIV shows a part ofi t . F. G . KRAMP s ays in his notes to this map that for the regions eas t of theCaspian S ea VV its en had trans lations made fromthe Arabian geographers while , forChina , he u s ed Martin i as w e ll as Chines e maps procu red and trans lated for himbyFathe r COUPLE T . H e has obvious ly also u s ed Rus s ian material , as can b e s een in

s eve ral names , f. i . Teb etzami , Tiob etz Ansky (Tib etanskiy), and Tangoetska . The

las t is a town in Tangu th , the t wo firs t are the Tibe tans , in both cas es called Mon

1 In the copy of the Roya l Library of Stockholm(Relations de d ivers voyages cv rievx, etc .

Paris MDCLXVI) , which i s at my d isposal , this narrati ve i s , naturally enough , mis s ing. In the copyat the L ibrary of Gottingen ,

which i s a lso from1 666 , it has been inserted , bu t has the year 1 673 .

Tronn ier, op . ci t. p . 3 3 0 .

2 As tley’

s Co llection , V ol. IV ,London 1 747 , p . 652 . The Italian origina l i s found in Tron

n ier’s article .

3 N ieuwe Lan tkaarte V an he t Noorder en Ooster dee l v an A s ia en E uropa door N icolaesWitsen , anno 1 68 7 . In R emarkable maps

,Part IV ,

Amsterdam 1 897 .

202 MAPS OF THE SEVE NTE ENTH CE NTUTY.

E as t of the upper Ganges Wits en accepts the orography of Kircher and

V is scher w i th Consangu i mons lapideus , and Be lor mons ; wes t of the latter theclas s ica l Caucas i montes are s ti ll to b e s een . In the text he s ays they are not far

fromImau s .

For the Croceus s ive Hoamko P luvius he chiefly fol lows Martin i . An improvement i s that the uppe rmos t cou rs e runs W .N -W . to E .S -E .

,ins tead of s ou th to north .

On the other hand Martin i ’s local isation of Koko-nor was muchmore correct. Wits en

identifies i t with Chimo i , or Chiamai lacus .

As usua l the greates t confus ion prevai ls in the repres entation of Ce n tral As ia,incl'. Tibe t . Be twe en the Caucasus and the Alta i W its en has on ly a narrow s pacefor Tibe t, Tu rche stan and Tanguth . Tibe t which had be en be tte r placed by Cante ll ithan on any previous map , now cove rs a very smal l region in the far wes t, whilethe mos t famou s of all Tibe tan plants , the rhubarb , is placed far to the eas t, on theuppe r Hwangho. This notion comes fromMarco P010, and therefore we find the

legend : »Hier valt vee l Rhabarbe r» , near S oeks i s ive Sochu, and not far from

Tangu th .

I

Wits en has a special chapte r , ma or Tooet, in which he places this countryat 42

° N . lat .

,»though others have it at H e has heard that some people make it

20 days ’ journey round. I t is s ituated to the N .W . of the Chines e province of Xens i ,

a notion which agrees with Martin i ’s map, b ut not with h is own . According to FatherMAGE LLAE N

s »D es cr ipti on of Caina » Tibet shou ld b e to the eas t of the Mogol’

s

Empire , and its capital shou ld b e cal led Kaparange . FromChines e sources he re

lates that to the E as t of Tibe t is a city and province cal led Utsang, Utsana , Uss angue or S ifan , and he quotes Fathe r Andrade , who te lls that Us s ang or Us sangue

is at a dis tance of 20 days ’ journey fromChina .

2 This Utsang is the combinedprovince of U and Tsang, b u t in Andrade ’s description i t s imply s ign ifies Lhas a .

Gas taldi’

s Monte V s sonte may b e the same as Monte Us s angue or Moun tains of

Lhas a .

2 A lready the Arab geographers knew the rhubarb of T ibet. E dris i says the mounta ins roundButhink abound in this plant. In his descri ption of the province of Sukchur with the ch ief c ity Sukchu which be longs to the General Province of Tangut, Marco Polo says : »Over all the mountains of

this province rhubarb i s found in great abundance , and thither merchants come to b uy it, and carryit thence all over the world . (Yu le , Book I , p . Speaking of Suju , the present Su-chau nearShanghai , Marco Polo makes the mistake to write : »In the mountains be longing to this c ity, rhubarband ginger grow in great abundance .» For this rhubarb was brought fromT ibet for sale in Su-chau .

(Yule , Book II , p . 1 8 1,

Father du Halde says of the rhubarb : »La me i l leure e st cel le de Setchuan : ce l le qui vient dans la Province de Chen s i dan s le R oyaume de Thibe(t) , lui es t fort inférieure : i l en croit aus s i ail leurs , ma is qui n ’

e st nu llement estimée , dont on ne fa it ici nul usage .Description de la Chine , Tome III, p . 492 . Compare also Tome I , p . 4 2 .

2 Andrade ’s pas sage runs : »Le T ibet comprend le Royaume de Cogue’ qui est celuy auque lnous demeurons a present (Caparangue , Aoust celuy de Ladac , de Moriul , Rudoc , V tsanad eux autres qui sont vers l ’orient, tous lesque ls auec le grand Royaume de Sopo, qui confine d ’

v n

costé la Chine , de l ’autre la Moscouie , font la grande Tartaric.» Sopo,Sogpo in T ibetan , i s Mon

gol ia . The description i s very good .

SOUTHERN TIBE T ,VOL . I .

F . de W itt ’s Map ,middle of 1 7

thcentury .

204 MAPS OF TH E SEVE NTE ENTH CENTURY.According to V is scher H iarchan or Ya rkand, remains on the upper Ganges ,

while Cas car s ti l l be longs to the uppe r S ir-darya . If Marco Polo , who v is i ted the

three corners in the Tarim triangle , Kashgar, Yarkand, and Lop , had on ly madean annota tion of three or four words abou t the Tarimbas in and its main river, theclass ical Oechardes , the chief mis takes of 400 years in Central As ian cartographycould have been avoided. As on Wits en ’

s map Turfan appears in two editions ,one , Toerfana, north of lake Bervan , the othe r, Tu rpan , on the upper Hwangho .

More interes ting than anything e ls e on de Witt ’s map are the three wande ringlakes , the traces of which w e have to follow up un ti l they d isappear or lead to a

real goal . Lake S iba , Wits en’

s inven tion , i s the re , and near it , though s eparated bya high mountain range , i s lake Bervan , which we firs t saw on Cante l l i ’s map , Pl .XXXII I . The third is our old Chiamay lacus , this phantomlake which has defendedits ground agains t human know ledge ever s ince 1 550, and remains on the maps of

1 50 years as a monument of human credu l ity . And here also Wits en ’

s example isfol lowed and the lake called Coconor. Of Grueb er’s route fromLhasa to Benaresonly the word N ecb al remains .

The las t map of the s eventeen th cen tu ry of which I give a reproduction , Pl .XXXVI , is the one by CORONE LLI 1 695.

I Here the Ganges , though merid ional , hasbeen brought down to a more reas onable length . He has any amount of names forthe Himalayas : Monti Cocas , o Cochias , Alb sor, Adazar, e Salatto ; Caucasus S trab .

et Plin . Caucas ias Rupes Ptol. S trobilus Arriano . Cascar or Kashgar, Pe imand

Chotan are better placed than before , though the re is also a Cotan S .E . of Lago diBe ruan . NE . of Cas car is Lop , and east of i t Acsu . Yarkand has disappearedfrom the upper Ganges , and so has »Chaparangute » . But the city which on Can

te l l i ’s map (Pl . XXXIII), was s imply called Beruanfhas now also the name Chaurana ,which may b e a new ve rs ion of the Caparangue of Andrade , the Kaparange of

W its en , the Chaprang, Ts aprang, or Ts aparang as i t is now cal led by di fferent trave l lers . I do not know what information Corone lli has got . If his Chaurana i sreally Tsaparang , and if he has heard, perhap s , froms ome mis s ionary in Goa , thatthe town was s ituated wes t of a lake , his map indicates that the Lago di Beruan isManas arov ar. When deal ing w i th E dris i we came to the s ame conclus ion regardinghis lake Berwan . On the other hand, the S iba lacu s has been rej ected .

North of the Himalaya , which is indicated on ly by its s eve ra l names , we haveThibe t, Tob b ot , or Tob b at , ve ry correctly placed in re lation to India . Turques tan

and V achun are parts of Tibet . The town Thib et i s again Leh and Ladak .

The eas tern parts are w is e ly taken fromMartin i , b u t there is no Koko-nor at

all. Yet Chiamay lacus is there . The great des e rt has , as on some previous maps ,

1 A s ia Dini sa ne l le s ue Parti s econdo lo stato presente , Descri tta ,dal P . M . Coronelli

It is publ ished in Atlante Veneto, N e l quale s i contiene la descrittione geografica , storica , sacra, profana , e pol itica , Degl ’ Imperij , Regn i , De l l ’ Un iverso, Opera e studio del Padre MaestroCoronelli Tomo 1 . Venetia 1 695.

SOUTHE RN TIBE T. VOL . I.

A b a a z su '

D

M A R E

71 4'

fl v I nmhm’

é

L a la

fi li a l/tam.

CORON E LLI’

S MAP . 205

several names , Xamo, Kalmuk Karacathay, and Des e rto di Lupo, which is an

cellent trans lation of Lop !

Herewith w e have fin ished the reviews of maps fromthe s eventeenth century,and mus t confess that , so far as Tibe t and the sources of our rive rs are conce rned

,

we are not much w is er at its clos e than at its begi nn ing. It wou ld b e the taskof the firs t thi rd of the e ighteenth century to dis s ipate all doubts of the s i tuation of

this country and to draw out s ome of the principal features of i ts hydrography .

CHA PTE R XX I I.

T IBET lN EUROPEAN BOOKS AN D NARRAT IVES OF TH ESEVENTEENTH CENTURY .

It wou ld b e a ve ry u s e les s and unneces sary task to try and pick out fromall

geographical works of the s eventeenth centu ry the pass ages which dea l w i th Tibe t.The res u l t of so much trouble wou ld b e ve ry disappo inting. And afte r the examination we shou ld fee l obliged to admi t that the knowledge abou t this country had

gained very litt le in the cou rs e of one hundred years . The bri ll iant trave ls of

Andrade and G ruebe r, which are now s o easy to fol low, cou ld not b e placed on

the maps exis ting, as clearly proved by the attempt of Fathe r Kircher. O the r information had been gathered by Jes u it mis s ionaries , but was e ither not published, orpublished at a much later date .

The pres ent chapte r has the refore only one object : to give an idea of the

s tore of knowledge poss es s ed by E u ropean geographers , and for this purpos e a few

examples w i ll b e qu ite su fficient. The ordinary way in which the geographers savethems e lves fromall difficu lti es is to copy the words of the i r predeces sors fromthe

clas s ics and Marco Polo down to Ramus io , Roe and Te rry. One of the pas

s ages which are cop ied ove r and ove r aga in is the one in L ioo’o Odoa rdo B aroos a

P ortog/zoso, whe re the s acrednes s of the Ganges , in the re l igious opin ion of the

Hindus , as w e l l as the s ource of the rive r as s i tuated in the terre s trial paradis e , isdescrib ed.

I Such is also the case with the pas sage in NICOLO D I CoNTI ’s narrativewhe re he speaks of the marve l lous lake be twe en the Indu s and the Ganges .

2 FromBARROS ’ as s e rtion regarding the Chiamay lake, firs t published in Ramus io ’

s work a

traditiona l mis take of 1 50 years took its origin.

In some books one also recogn is es the e loquent des cription Fathe r JARRICI A questo fiume Gange vano tutti gl’ Indian i in peregrinaggi o con gran di uotione a lauars i,perche hanno firmis s ima fede , che da poi lanati , s iano netti d i tutti l i lor peccati , per questo salu i ,

conc io s ia cosa,Che i l detto fiume vien da v n fonte , i l qual ha i l suo principio nel paradiso terrestre .

Ramus io, Primo vo lume De l le N av igationi et V iaggi , Venetia MDLIIII , p . 3 49 , c . The fol lowingpassage i s no doubt taken fromConti : »There i s al so a lake lying between the Indus and the Ganges ,the water of which possess es a remarkable flavour, and is drunk w ith great pleasure . A l l the inhabi tants of that d istrict, and even those l iv ing at a great d istance , flock to this lake for the purpose ofprocuring the water. By means of re lays of carriers mounted on horse—back , they draw the water freshevery day.» In d ia in the F i fteenth Century , ed ited by R . H . Major. London , Hakluyt Society , 1 857 ,p . 2 2 .

2 Op . c it. p . 3 78 , C . See above p . 1 75.

NICOLO D I CONTI ; TRIGAULT . 207

gives of Kashmi r, a luxuriant country , surrounded by snow-cove red mounta ins , and

abundantly watered. To the eas t is R ebat , as he calls Tibet , with its colde r cl imate .

He has heard that s ome me rchants trave l fromIndia through Bengal to China , whilea shorter road goes through Kashmir and Tibet directly to Caigar or Kashgar, fromwhere it is easy to reach the firs t Chines e city.

I

In his work on Great Mogu l ’s empire , 1 63 1 , DE LAE T quotes the ancients ,specially Pliny, and calls the northern mounta ins Emodus .

2 FromBe lgian sourceshe has heard that the Ravi , Behat and S ind take the i r origin in the mountains of

Cas s imer and join near Mu l tan . The province of Kabu l borde rs , to the north , inthe di rection of Corumupon Tartaria, whe re the rive r N i lab has its origin . On the

banks of this rive r is Attack , and be low i t the rive r jo ins the Indus .3

In what he says of the rive r Behat w e remembe r Thomas Roe 1 6 1 7 , thoughde Lae t does not find it like ly that the rive r shou ld jo in the Ganges . The climaticcomparison between Kashmi r and Tibe t he has taken fromIarric. At Hardv v a

'

i'

r is

the sou rce of the Ganges fromthe rock with the cow ’

s head. In describ ing the

Kara-korum road from Kashmir v ia Tibb on (Ladak) to Cascar, he has more con

fidence in Finch , 1 6 1 0, than in Xavier,When Fathe r TRIGAULT ,

1 639, in his compilation on China , te l ls us that theYe llow R ive r and the Gange s both ris e from“

the Kwen-lun, he mus t have us ed the

s ame sou rce as Father Mart in i , who, in 1 655, te l ls the same thing.5

I The passage runs : Regnumhoc nu l l i Ind icae region i pulchritudine 8: aéris clementia ced it,etiammedia ae state . Alti ss imis s iqu idemcircumquaque c ingi tur montibus , n iue maxima ann i parte co

opertis ; reliquus ager planus es t, e legans , v iridantibus herb i s variegatus ; s i lu is , horti s , v iridari is , fontibus ,amnibus ad stuporemv sque abundans . Frigid iuscula omn is regio est, idquemaxime ob monte s n iuales ;temperatior tamen es t, qui regnumRebat, quod ab oriente i l l i adiacet. Mense Maio é frigid is R ebatimontibus infinita prope ans erums ilues triumaduolan t agmina , in flumina , quae iuxta Cascimirum, regnicaput, decurrfit, v elut ali is calidiora , s e se immergunt. Sunt qui iter in Cataiamper BengalamGaragati (Ca lcutta) regnum, in quo Magn i Mogor imperiumterminatur, patere dicant. Sed mercatoresqui . v iarumnorfit compend ia , Lahore Caximirampetunt, ac de inde per Rebati regnum(cujus Rex Magn iMogor socius amicu s) recta Caigarenumtendun t ; atque hinc faci li s breui s ad primumCataiae

emporium v ia patet. R . P . Petri Iarrici Thesav rv s R erumIndicarum. Col. Agr. MDCXV, T . I ,p . 558 et 579. Father Iarric h imse l f has got this information fromFather Jerome Xavier, in a letterwritten 1 598 , and publ ished in Father Jean Hay ’s : D e rebus Japon ici s etc . , Antv erpiae 1 605. See a lsoBrucker 0p . ci t. p . 1 1 .

2 De imperi o Magn i Mogoli s s ive Indi a Vera Commentarius . E v ari j s auctoribu s congestus . ndvn i Batav orum

,Anno CIDIDCXXXI, p . 1 et seq .

3 Kabul versus Corums ive arctumfines jungens cumTartaria . In hac fluv ius N i lab oritur,qui versu s meridiemde scendit, donec cumIndo se conjungat Attack a metropol i s ic appel lata ,ad ripamfluv i i N i lab

, qui a Coro descendens ab Indo excipitur Corumis therefore amountainoustract. Shou ld it b e the Kara-komm?

4 Regio (Cass imere) vicina qu idemCaseari , sed tamasperis montibus ab i lla d ivisa, ut nul lusaditus Carav anis pateat ; rari pedi tes in terdumsumma cumdifficultate asperos hos montes penetrant.In montanis bisce degit Regulus , T ib bon nomine E ademregio fines jungi t cumRegno utriusqueTheb et The Rah ia Raja, Regu lus) T ib bon on Sanson

s map 1 654 comes probably fromthisvers ion of F inch ’s T ibbot.5 O f the Ye l low R iver Trigault says : E t hoc est R egn i S inens is a lterummagn itud ine ce lebri

tate flumen , quod extra regnumad Occasumnascitur,é monte , qu i Cunlun appellatur, quemmontemvero s imi l ibus conjecturi s colligitur, cumip sumes se , v el a l iummin ime remoturn ,

é quo Ganges credi tur

scaturire . R egn i Chinens is Descriptio, n d . Batav . 1 63 9, p . 3 1 8 .

208 TIBE T IN E UROPE AN BOOKS AND NARRATIVE S OF THE SEVE NTE E NTH CE NTURY .As the magn ificent atlas of JE AN BLAEU, of which the e leventh volume I

ap

peared in 1 663 , was an »encycloped ia cartographica and geographica» , we cannotexpect to find any new information abou t our regi ons in it, no more than in the

compilations of S anson , V is s cher, Coronelli , and othe rs . The maps are broughttoge the r from d i fferen t au thors and epochs ; maps from Iansson

s and Sanson’

s

offices are s een at the s ide of such fromO rte l ius , and othe rs . The s ame i s thecas e wi th the text where s everal paragraphs are more than a hundred years old, anduncri tical ly accepted as true . India is s aid to b e bordered to the north by branchesfromMount Tau ru s , or Imae , a part of Caucasus . The mos t famou s rivers of India,Indus and Ganges , take the i r ori gin »fromMount Imae , cal led by its inhabitantsDalanguer N augracot , are born , according to the opin ion of the nat ives , fromone

and the same source ; in spite of the d is tance be twe en the i rmouths be ing 300 l ieu es » .

2

A l l the rivers of »Pang Ab» , so far as they were known in 1 663 , are enu

merated from di fferent trave l lers . The part of the Imae where Ganges has itssource , is known as Montagnes Us sontes . Caximir has the kingdomof Rebat to thee as t . Then fol lows

,from Iarric , the notion of the s ource of Chenab as s i tuated in

Kashmir, and from au thors quoted above , the comparison betwe en the climate of

Kashmi r and Tibe t. Blaeu ’

s volume was published the year afte r Grueb er’sre tu rn , but the unce rta in ty abou t the s ituation of Tibet is s ti l l ve ry great . Havingplaced the Kingdomof R ebat eas t of Kashmir , the compilator of this text says of

the province of Theb eth , the capital of which is also cal led Theb eth , that it confines ,on one s ide with the countries of S indinfu and Caindu , and on the other wi th Mangior China , whe re Marco Polo is his source . The province is s aid to b e great b utmuch devas tated by the Tartars . A few pages fu rther on we me et a third Tibe t,the Kingdomof Thib ethe , bordering upon the domin ions of Mogor and special ly Kashmir

, and be ing s eparated fromi t by ve ry high mountains , which , on account of thesnow , cannot b e cros s ed except at ce rta in s easons of the year. »Beyond it is the

Little Kingdomof Th ib ethe unde r the power of the Moors of the s ect of Sofi. The

res idence of the King 15 cal led Babgo .» 3 One fe e ls qu i te at a los s when the compilatorte lls us that the inhabitan ts of this las t edition of Tibet are all Chris tians , and have manychu rches , and a bishop cal led Lamhao ,—all rumours cu rrent amongs t the Jes u its at Goa,and gi ving an impu ls e and a reas on to the journeys of Goes and Andrade . Al ready beforeGo

'

es’

s tart , thes e rumou rs had been proved to b e fables by the famou s Father M . Ricci .\Vhen Blaeu ’

s compilator gi ves the boundaries of China he s ays that it has»Tibet» to the wes t. We have therefore in one and the s ame work a KingdomRebat , a provi nce Theb eth , a KingdomThib ethe , a Little KingdomThib ethe , and

I Onzieme volume de la Geographie Blav iane , contenant l’As ie . Amsterdam, Jean B laeu 1 663 ,p . 1 63 , 2 1 9, 2 50 e t seq .

2 This i s taken fromBarros in R amus io, p . 4 2 7 , c. : E t s econdo 1a fama de l le genti CI l’CODu icine , s i crede che ambi du i maschino da v n mede s imo fonte .

3 By L ittle T ibet Baltistan ought to b e meant. But Bab go sounds more l ike Badgao, a c ityin Nepal , later on known fromthe journeys of Des ideri and the Capuchins .

JE AN BLAE U , VARE N IUS, OGILBY . 209

a country Tibe t, all more or les s vagu e ly de termined on the map . The reason is ,of cours e , that the author has drawn his information fromd i fferent works and ages ,and thereby created an unneces sary confus ion .

BE RNHARDUS VARE N IUS i s regarded as the mos t important geographe r of thes eventeenth centu ry. In his Geograp/zi a g enera l i s , 1 664 , he s tands , howeve r, on

clas s ic ground so far as our regi ons are concerned. He knows the Imau s as two

ranges forming a cros s , the me ridional branch d ivid ing Scythia in the two partsfamou s s ince antiqu i ty . The sou rce of the Ganges is placed , he s ays , by some at 43

°

N. lat .

, by others at 3 and flows fromnorth to sou th .

I

Another compilation , ve ry w e l l got up, was publ ished in 1 673 by JOHN OGILBY .

2

H e does not help us any further, howeve r, and, as all othe r authors of the time , he quotesthe clas s ics , and Barros , Gas taldi , Linscouten , He rbe rt, Jarrie , Be rn ier, and the res t.The Indus ris es fromthe Mountain Parapomisus , »Ganges (now Guengam) aris eth out ofthe S cythian Mountains » . Later on he says of the two rivers that they »come fromthe

Northward out of the Mountains Imaus and Caucas u s , by the Inhab itants (according toCas taldus) call

d Dalanguer and Nangracot , and both (as the Inhabitants affirm) spri ngfromone Head, though some Geographers make the d is tance betwe en themto b e a

hundred and e ighty Leagues , and others b ut a hundred and thi rty The confu s ionis great about the northern mountains ; probably no geographer of thos e days everkn ew what the ancien t writers meant by the ir Paropamisus , Imaus and Caucasu s , not

to mention the S cythian mounta ins of which nobody has ever had any know ledge .

The river R awy (Ravi) is said to spring out of Mount Caximir. The ad

min is trative d ivis ion of Hindus tan is the traditional fromS ir Thomas Roe . So is

also the map, the same which in P a roé as H i s P i lgrims Vol. IV , p . 43 2 is calledS ir Thomas Roe

s map of E as t India . It is of no special interes t, though I ins ertthe part of i t which conta ins the sources of the Indus and the Ganges , Pl . XXXVII .E ven the Cow ’

s Head is conscient ious ly drawn at Hardware .3

I In the original text the two pas sages run thus : »Imai i s mons , crucis forma duab us vi is progreditur tam versus ortum81 occasum, quamversus Septentrionem81 Austrum. Septentrional is parsnunc A lka i perh ib e tur. Protendi tur versus Austrumusque ad Indiae fines 81 Gangi s fluv i i fontesD iv id it Scythiam As iaticam in duas partes , quarum i l la , quae occasurn spectat, dicitur Scythia intraImaummontem; quae orturn dicitur Scythia extra Imaummontem.» And of the Ganges : >Fons ejusremoti s s imus ponitur in latitudine Septentrionali 43 graduumin Tartaria (s ed qu idamretrahunt ad 3 3

grad ) . Ostiumin latitudine eadem2 2 gr. Flu it a Septentrione in Austrum. Tractus es t 3 oo c irc itermill iariumGerman icorum. E xundat s inguli s ann is .» Geographia general is , In qua aflectiones gene

rales Telluri s explicantur Autore Bernh : V aren io M ed : D . Ams te lodami , 1 664, p . 94 81 2 70 .

2 A s ia, the F irst Part. Be ing an Accurate Description of Pers ia , the Vast Empire of the GreatMogol , And other Parts of Ind ia, col lected and trans lated frommost Authentick Authors , and

Augmented wi th later Observations London 1 673 , p . 1 04 , 1 99, 2 4 2 et seq .

3 The title of the map i s Magn i Mogoli s Imperium. It i s a lso to b e found in Mandels lo’

s

Voyages Ce lebre s 81 remarquables , Amsterdam1 72 7 , where it i s called : Royaume du Grand Mogol ,avec tous le s Pays qu i en dependent suivant les Re lations des plus fideles Voyageurs , nouvel lementdonne

au Publ ic par P ierre v an der Aa , a Le ide . The map of 1 7 2 7 was s imply a reprint of the onefrom 1 673 , wh ich was antiquated s ince S ir Thomas Roe ’s journ ey in 1 6 1 7 . It had also been republished by Thevenot in 1 666 .

27—1 3 1 3 8 7 1

CHA PTE R XX I I I.

VERB IEST AND GE RB ILLON .

Father VE RB IE ST , famous for his journeys of 1 68 2 and 1 683 through E as ternTartary , re lates what a Jesu it in Pers ia heard froman Armen ian pries t who hadtrave lled in Tibe t, and fromanother Armen ian trave lle r, a wis e and re l iable man ,

who had pas s ed four years in the country , and whos e narrative is somuch the moretrus tworthy , as i t agre es perfectly with Grueb er’s account .

Thes e informants had re lated that in As ia were two kingdoms w ith the nameTibe t, the Little and the Great . Little Tibe t borde rs on Cachemi r of which Bern ie rhad given such a favourable des cription . But Little Tibet is s terile , and has a coldc l imate , and ve ry poor inhabitan ts . Great Tibe t which some cal l Tebat and othersBoutan , borders on Chines e Tartary . It is not more agreeable nor more ferti lethan Little Tibe t . There are exce l lent fishes in the rive rs

, and there is any amoun tof mi lk. Tibe t abounds in musk , and in gold, though the natives do not unders tand how to work in the mines . The a ir is exce llent, and the inhabitants are s trong.

Las sa is the capital and on ly city of Tibet ; a mandarin s ent by the Emperor re

s ides the re . Ve rb ies t says that it is three months ’ jou rney fromTibe t to this c ity(Las sa), wh ich is s i tuated at the foot of the great wall ; the informant mus t haves a id that it is three months ’ journey fromthe great wa l l to Lhas a . The whole wayone cros s es an u tterly des olate country where on ly w i ld an imals are to b e found, b u ts everal caravans pas s be twe en Tibet and China. The greates t man in Tibet is the

Grand Lama , »Pontife des Lamas », who ce rtain ly is the famous Pries ter John .

I

The great Fathe r GE RB ILLON who , from 1 688 to 1 698 made e ight journeysthrough eas tern Mongol ia , e ither a lone or w i th Father PE RE IRA , or even w ithEmpe ror KANG HI , gi ves u s s ome exce l lent news abou t Tibe t and s u rround ing countries ,far s uperior to anything exis ting before this time . Though Gerb illon had neve rbeen in Tibe t his des cription of the road fromKoko-

nor to Lhasa is much be tterthan that of Fathe r Gru eber. The ve rtical plas tic which is not mentioned in Grue

I Recue il de Volages au Nord . Tome Trois ieme , AmsterdamMDCCXV, p . 3 3 8 .

2 1 2 VERBIE ST AND GE RBE LON .

Gerb illon conqu e rs the new deta ils quoted above . In his three lakes one of the sur

rounding smal l lakes has been included. Ts ing-s u-hai is the Chines e name for the

Mongol Odon-tala or S tar S ea , through which the uppe rmos t Hwangho, Altan-gol,flows , and empties i ts e lf in the wes tern lake .

I

Fu rthe r Gerb illon was told by the s ame mandarin that fromS in ing to th e veryfronti ers of the Kingdomof Tibet the ground cons tantly and at a s ens ible rate b ecomes highe r, and that the great number of mountains one has to cl imb when trav

I The above description should b e compared w ith the one trans lated by Bitchurin, fromChineses ources in 1 83 3 : »The Yellow R iver i s in Chinese called Khuan—khe , in Tangut Morchu, in MongolShara-muren . It takes i ts origin in the western ne ighbourhood of Koko-nor under the name of A ltangol, and run s to Odontala ; after going through the lakes Ts ariyn -nor and Noriyn

-nor it flows to the

S .E .

,and returns to the N .W., and then takes i ts course to the N .E . B itchurin knows four Chi

nese descriptions of the source of the Ye llow R iver, b ut specially refers to the one by Amida who in

1 78 2 was sent to make a deta iled research of the place . This explorer found at 3 00 li west of Odontala a hil l , A ltan-

gas u-ts ilu , on the top of which a lake i s fed by innumerable springs ; here is the

real source of the Ye llow R iver, for the water goes to A ltan-gol, wh ich , itsel f, has i ts sources in Bayankhara-ula . H ari ng rece ived many tributaries A ltan-gol flows to Odon-tala , where innumerable springsdi p up fromthe ground , which , regarded froma he ight, resemble stars ,—therefore Odon-tala, and inChi nese Sin -su-khai , the Star Sea. A ltan-gol rece ives the water of all these springs , as wel l as that oftwo tributari es , after which it fal ls into Tsariyn-nor. A fter runn ing about 50 li to th e S .E . it emptiesitself into the N oriyn

-nor; on the way between the lakes it rece ive s four rivulets . Is su ing fromthe

Nori ya -nor A ltan -

gol take s the Mongol denomination Kh atun-gol (Princess R iver) . Istoriya T ib e ta iKhukhunora s 2 2 8 2 goda do R . Ch . do 1 2 2 7 goda po R . Ch . Perev edena s K itayskago MonakhomIakinfomBitchuriniymS t . Peterburg 1 83 3 , II p . 1 80 et seq .

On d’

Anv ill e’

s map of 1 73 3 the two lakes are remarkably well drawn , under the names Tcharin N or and Orin Nor, S .W. of Koko-nor and on 3 5

° N . lat. Such they remained on E uropeanmapsfor 1 50 years . Petermann ’

s map , Indi en 81 Inner-A s ien , in S tielers Hand-Atlas , N zo 44 b , for 1 8 75,i s , for instan ce , a fa ithfu l copy after d ’An vi lle , though Petermann puts the t wo lakes at 3 4

° N . lat.

On hi s fourth journey , 1 8 8 3—1 8 8 5, General Prshev alskiy was the firs t E uropean to reach thelakes . He gives an exce llent description of Odon-tal a, which he found traversed by s evera l rivulets ,

of which two were more con s iderable , and formed the sources of the Yellow R iver. One of themis ,as Prshev alsk iy supposes , identical with the Altan , or Al tiyn

-

gol of Chine se geographers . He found thelatitude to b e nearly 3 His de scription shows how very correct the Chinese observations had been .

»But as they had been incorrectly placed on the geographica l maps , as no E uropean had vis ited thembefore , I called , w ith the right of the first explorer, the western the E xpedi tion Lake and the easternthe Rus s ian Lake .

» Att Kyakhtiy na Istoki Sheltoy Reki , S t. Peterburg, 1 8 8 8 , p . 1 53 and 1 98 , and

my tran s lation fromthe Rus sian original : General Prschev alski j ’s Forskn ingsresor i Centralas ien, Stockholm1 89 1 , p . 3 1 8 and 3 3 4 ; with a preface by A . E . Nordenskiold .

On Koslov’

s beautiful map , Vostochn iy T ibet (Kain ) , May 1 900—June 1 90 1 , where t he lakeshave nearly exactly the same latitude as on d

Anv ille’

s map 1 70 years earl ier, the names gi ven byPrshev alskiy are reta ined together w ith D sb arin-nor and Ori n -nor. Kos lov calls the A ltan-gol Soloma ,

for which Prshev alskiy had had the name Upper K huan-khe or Hwangho. But on his general mapAttchotnaya Karta k opisan iyu pute shestv iya P . K . Kos lov a v Mongol iya i K am,

we again find the

name A ltan-

gol . The Ts igenor in B itchurin’

s trans lation i s probably the Chikey-nor of Koslov’

s map .

The name for Odon -tala , which was written Ts ing sou hai by Gerb illon , S in-su-khai by B itchurin, andPrshev alskiy, i s identical wi th Father M artini’s S ing s ieu , wh ile the latter’s Sos ing may b e Oring. The

names E xpedition and Rus s ian have of course to disappear, as the lake s already were baptized hundreds of years ago .

The two freshwater lakes at the source of the Hwangho resemble in many respects the Manas arov ar and the Rak as -tal

,and their connection with the Satlej . This fact i s the one reason for this

d igres s ion . The other i s that Koko-nor and the source of the Hwangho are s ituated on one of theh ighways leadi ng to T ibet, a road trave lled by Grueber and Dorvi lle , v an de Putte , Huc and Gabet.We shal l have to return to this region later on .

GE RBILLON’

S DE SCRIPTION OF KOKO-NOR . 2 1 3

e ll ing from China to Tibe t, have a much longe r s lope to the eas t or Chines e s ide ,than to the wes t or Tibe tan s ide . Then fol lows this pas sage , which is ve ry cleve rfor be ing 2 20 years old : »The smal l mounta ins , in which the l itt le river of Altan koltakes its source , mus t indeed b e extreme ly e levated ove r the surface of the S ea , as

this rive r which is rathe r rapid falls in the lakes of Ts ing sou ha’

r’

, while the rive rHoang ho which is sues fromthe Lakes , has for abou t hundred leagues a fa irly rapidcours e unti l its mou th in the E as tern S ea of China ; b ut this country is also ve rycold on account of its he ight : as soon as one commences to enter “Tibe t the groundbegins to fal l , and the cl imate is then also much more temperated.»

At a t ime when Koko-nor was s ti ll treated as a suspicious object on E uropean

maps , Gerb illon knew that S i ha i or the Wes te rn S ea, as it was also cal led , had a

length of 20 long leagues , and more than 1 0 leagu es in breadth, and that i t wass ituated between 36 and 3 7

° N. lat . , and betwe en 1 6 and 1 7° W . long. of Peking.

From the wes t frontier of China he reckons 20 days ’ journey to Hami , thence to

Tu rfan 7 days , fromTurfan to Acsou 2 3 days , fromAcsou to Yarkan 1 0 days , andfrom Yarkan to Bochara hardly more than a month , a long the road which pas s esove r Kas kar. He gives short, concis e and exce llent descriptions of roads , countries ,cl imates and natives

, and is far in front of his time .

Compared with the othe r geographers I have quoted above , all of whomruminate the same old worn material , Gerb illon ’

s annotation about Koko-nor, the

s ource of the Ye llow R iver, and the roads to Lhasa and Bokhara comes l ike a

refreshing breeze of sol id information frommountains and des erts in darkes t As ia .

CHA PTE R XX I V.

MAPS FROM TH E FIRST TH IRD OF TH E E IGHTEENTHCENTURY.

We are approaching the rena is sance in the cartography over Tibet, an epochthat is characte rized by the s ingle name d

Anv ille . But before we reach so far

we have to remember some of his forerunners , the mos t important of whomare

De l is le and S trahlenbe rg. However, I amproceed ing chronologi cal ly as hithe rto.

Pl . XXXVIII shows u s the map of E . ISBRANTS IDE S publ ished in 1 704 .

I

Ides was s ent as an ambas sador through S iberia to China in 1 692 , and returned to

Rus s ia in 1 694 . He had a copy of Wits en ’

s map w ith himon the jou rney , and

improved i t whe re i t was wrong. In its corrected formthe map was added to Ides ’narrative . Wits en s ays of it : »Ides ’ map is to a great part taken frommymapb ut he has improved i t.» 2

In his text Ides does not touch our regions , and all he has to say of Ind ia isthat the G reat Mogu l ’s kingdomembraces all the countries and domin ions which ares ituated

'

b etw een the rivers Indus and Ganges down to the Gu lf of Bengal . 3S o far as our regi ons are concerned the l ikenes s between Ides ’ and W its en ’

s

maps is , howeve r, ve ry ins ign ifican t . Ides has obvious ly, as Wits en , u s ed a gooddeal of Rus s ian material ; nearly all the lakes in the northe rn half of the map are

I Nova Tabu la Imperi i Rus s ici, ex omn iumaccurati ss imis , qua hactenus exti terun t. Imprimisviri Ampli s s . N ic: Witzen delineation ibus conflata , quam ipsa locorum lus tratione edoctus , multum

emendav it E v erardus Ysbrants Ides . Material ly po Istoriy etc . N zo XXVI . The same map, thoughnot reaching quite so far south , is to b e found in Recue il de Voiage au Nord , Amsterdam1 7 2 7 , whereit has the ti tle : Carte de la Tartaric A s iatique suivant la R e lation de l

Amb as sadeur de Rus s ie publiéeen 1 692 ,

which is wrong, for the map was first published in Isbran ts Ides narrative : Dri ejaarige re izenaar China te lande gedaan door den Mokov i schen Afgezant E . Isbrants Ides , Amsterdam1 704 . In

the German ed ition wh ich i s at my d isposa l : Dreyj 'ahrige re is e N ach China E . Yssb rants Ides , Franckfurt 1 707 , the map i s wanting. In Recuil de V oiage s au Nord the narrative has the title : »Voyage deMoscou a la Chine par Mr. E verard Isbran ts Ide s Ambassadeur de Moscov ie» .

2 V . Kord t : Material iy po I s toriy e tc . K iev 1 906 , II , Part I , p . 2 7 . And F. G . Kramp inRemarkable Maps , Amsterdam 1 897 . Ides himse l f ass erts : »Il faut que j ’av oue que je n

eus se peutétre pas pensé a faire une carte générale des Regions que j ’ai parcourue s , s i le celebre Mr. N icolasWitz en Bourguemétre d

Ams terdamne m’

en efit fa it concevoir l’ idée la carte qu’ i l a donnée m’

a

servi de gu ide en p lus ieurs endroits de mon voyage , e l le m’

a fourn i en méme tern s un plan , que jen’

ai fait que continuer. (French ed ition ,p .

3 German edition, p . 43 1 .

SOUTH ERN TIBET , VOL. I.

SVVL R V

The Map

ISBRANTS IDE S . 2 1 5

cal led »os ero» , and the Cas pian Gwalinskoi More . Othe rw is e the re is hardly more thanthe S iba lak e that is taken fromW its en . Another feeder of the Ganges comes fromthe Lac G iamai , which has to its north Kabu l , as capital in Zagataykabul Regnum,

a real ly wonde rfu l invention ! And s ti l l there is a s econd Cabu l wes t of Kachemire .

The Indus , as u sual , doe s not come fromfar away ; some of its feede rs hardlytouch the s outhern foot of the tremendous range which s tretches fromabou t thesou rce of the Ganges all the way to Is pahan , b ut not at all eas twards fromthe

Ganges . Attock , on the Indus , is s ituated between two cons ide rable lakes , of whichthe northern one i s the Wu lu r lake , the sou thern a new reve lation . The Satlej is

as always rud imentary and not eas i ly recogn isable .

Tibe t R egnum falls w ithin the catchmen t area of the Gange s , b ut RegnumTangu t is eas t of the uppe r Ganges , and has Barantola be tween long rows of nomads

’ tents and fu rthe r s ou th the legend : Woonplaats v an Dala i Lama of Zeepries ter,

a notion that now begins to become more fami l iar to As iatic s tuden ts .

The old Chiamay lacus , which is now breathing its las t, appears unde r newnames , and eas t of it the draughtsman has found i t more reasonable to combineMartin i ’s sou rce of the Hwangho with Grueb er’s Koko-nor and cal l the resu lt CroceLac, a sort of »Saffron Lake » , after the mode l of Martin i ’s Croceus fluv ius or

Ye llow R ive r . Ides cal ls i t CrocumFlumen .

If Lop-nor is a wandering lake on the face of the earth , the poor city of Lop

and the D es ertum Lop of Marco Polo are neve r al lowed to get a res t on olde rmaps . E ve r s ince Fra Mauro and down to Ides , it is never mis s ing, b ut a lways .

s ituated at a new surpris ing corner of the map, and a lways wrong . On Ides ’ mapwe find the Des e rtamLop at the uppe rmos t cou rs e of the Irt ish , and not far eas tof the Caspian .

I In Zagataykab ul is Chiotan , probably Khotan , and wes t of itAndeghem(Andishan), Occient (Khodjent), and Taskent . 2

I By the way, it shou ld b e noted that all the old cartographers wrote Lop and not Lob . So

did Fra M auro, so did Gastald i , M ercator, H ond ius , Jans son ,and the others , and so did Ides . D

Anv ille

wrote Lop-Nor, Stielers Hand-Atlas of 1 8 26 has Lop See . On ly the same atlas for 1 8 75 has Lob

noor, fol lowing K laproth ’s spell ing. Prshev alskiy who was the first E uropean of recent years to vis itthe lake K ara-koshun cal led it Lob -nor, and many maps fol lowed hi s great authority, s o for instanceSaunder

s,

Atlas of Ind ia, 1 889 , and Stielers Hand-Atlas for 1 90 1 . In 1 906 , when I trave l led in the

country I never heard any other pronunc iation than Lop , though in the Rus s ian tran s lation ofmy book»Through As ia» , V Serds e A s iy , St. Peterburg 1 899 , it was changed to Lob . But in this case MarcoPolo, the rea l d iscoverer of Des ertamLop i s a greater authority than Klaproth and Prshev alskiy. AI

ready in 1 904 S tieler’

s Atlas had returned to the vers ion Lop . A l l re l iable trave llers , as Ste in and

H untington , write Lop . So did Karl H imly in hi s Chin e se trans lations which he carried out for me(see my : D ie Geogr.

-wi s sensch . E rgebn is se me iner R e isen in Zentralas ien 1 894—1 897 , Pet. M it. E rg.

Bd . XXVIII , 1 900 , p . 1 5 1 e t seq .

, and Sc ientific R esults , V ol . II , p . 2 78 et So does also Prof.A . Conrady in h is work on my anc ientmanus cripts fromLou-Ian , not yet published . Curious ly enough Dr.G . Wegener, who col laborated with H imly, has Lob -nur (Ze itschr. d . Ges . f. E rd . zu Berl in , XXVII I Band1 893 , p . 2 0 1 et Dr. A lbert Herrmann has , of cours e , Lop-nor (D ie a lten Se i dens tras sen zwischenChina und Syrien ,

Wittenberg The spe l l ing Lob ought to d isappear for ever frommaps and books !2 Under the T itle N aaukeurige Kaart v an Tartaryen , P ierre vander Aa, Leyde , the map of Ides

was much spread during s everal years . It accompan ies also Bergeron ’

s Voyages faits principa lementdans les XII , XIII , XIV ,

et XV s iecles , La Haye 1 73 5, Tome I and i s reproduced here as Pl . XXXIX .

2 1 6 MAPS FROM THE FIRST TH IRD OF THE E IGHTE E NTH CE NTURY .With his P lan t

'

sp/zoro torros tro of 1 68 2 , CASSIN I had made a new s tart fora more modern and trus tworthy represen ta tion of the earth , b ut the rea l reformationin the his tory of cartography was the work of the members of the French Academyof Science . This pe riod was inaugurated by DE LISLE during the las t quarter of thes eventeenth centu ry, and came to an end abou t 1 7 20. De l is le is the firs t moderncartographer in the pres ent s ens e of the word . H e is the great reformer who brokew i th old dogmatized mis takes and pres s ed the tru th upon his time .

I

If this Opin ion is true of De l is le as a draughtsman in gene ra l, it is not les s correct for the repres en tation he gi ves of Tibe t . 2 De lis le ’s admirable map of S .E . As ia,Ca rto dos I n des ot do la C/zia o dros so

'

o s a r plu s i eu rs R ola t i oas fiartz’

oa l i oros

R oot ifio’

os par gaolgaos Oosorv a ti oi zs , Par Gu illaume De l is le , Paris 1 705, of whichPl . XL i s on ly a part, shows at once that the draughtsman , with deep knowledge and

sharp criticism, has made u s e of all material exis ting at his time , and i t shows a

ve ry s e rious attempt to combine and compare the d ifferent scanty narratives witheach other, an attempt that has of neces s ity resu lted in an a lmos t correct local is ationof Tibet . Ptolemy has for ever fin ished the part he had played for nearly 1 600

years , and the traditional mis takes of cen tu ries have d isappeared almos t comple te ly.

It is therefore a matte r of absorbing in teres t to trace the chief features of thiswonde rfu l map , within the boundaries of the regions occupying our attention .

The northe rn boundary of Cachemire is formed by a very highmountain range ,fromwhich the Indus , S inde , comes down . On its northern s ide is Pe tit Tibet andE skerdou and Cheker, taken fromBern ier . 3 In re lation to E skerdou , Cheker (Shigar)is very we l l placed , b ut the capital traditional mis take , that both are s ituated on the

upper Amu-da rya ins tead of the Indus , is s ti l l there .

Cachemire w ith the Wu lur lake is drawn at the wrong s ide of the Indus , thenorth ins tead of the sou th. Lahor correctly ra is es “

its wal ls on the banks of the Ravi ,the upper cou rs e of which pas s es by N agracut . The next Panjab rive r eas t of Raviis cal led V an or V ia, i . e . the Beas , as is also shown by Sultanpor. The eas ternmos t tribu tary in the Panjab is called R . de Chau l ou de S ietrnegus ; sou th of its

upper cours e is a town S irina or S erinde , now S irhind.4 This river is therefore the

S atlej . The curious way in which the diffe ren t branches of the Panjab are suppos edI Die R eformation der Kartograph ie um1 700 , v on Christian Sandler, Munchen and Berl in 1 905,p . 1 and 2 3 .

2 So far as Central A s ia and T ibet are concerned the ambitious Witsen had no reason whatever to write to Cuper, N ov . 5, 1 7 1 4 , as follows : »The Frenchman De Li s le is indebted to my mapfor all that he gives about those regions .» (F . G . Kramp in Remarkable maps .) It is sufficient tocompare Pl . XL with PI. XXXIV . The d is coveri es of Andrade and Grueber were open to Wi tsen as

we l l as to De l is le .

3 These names had first made their appearance on V is scher’

s map (Pl . XXXII) . But the c ityE scalcand on the source of Amu-darya, which for the first time was seen on Gastaldi (P l . XVII) andafter h im on n early all maps , i s probably noth ing b ut E skerdou , I skardo, Skardo, the capital ofBalti stan , or Little T ibet.

4 In 1 6 1 6 E dward Terry call s the river on which Jengapor i s s ituated, the R i ver Kaul . Del i s lehas both a c ity and a province Jengapor on his R iver Chau l .

DE LISLE . 2 1 7

to join , does not matte r much , nor the fact that , for ins tance , Suckor (Sukkor) ismarked as s ituated on the Satlej , though it is in rea l ity on thema in river, far be lowthe confluence of all the Panjab branches .

The source of the Satlej is s ituated amongs t the mountains of a province S iba .

I amnot aware of the origin of this name , which may have been prominent in the

beginn ing of the s eventeenth century when Te rry and Roe trave l led in India .

I On

»S i r l omas R oo’

s M ap of E as t I ndi a » (Pl . XXXVII), the s ituat ion agrees perfectlyw ith Roe ’

s text, and from it S iba entered many E u ropean maps of Ind ia, for ins tance ,S anson

s 1 654 (P1. XXIX). De l is le , who rejects Hardwar, has removed S iba somewhat to the wes t, and made it rathe r be long to the Satlej dra inage area . On Wits en

s map of 1 68 7 (PI. XXXIV) we have s een one of the feeders of the Gangescoming out of a S iba lake , a view adopted by I des 1 704 (PI. XXXVIII). Somerumou r mus t have reached Wits en ’

s informant that the Ganges originated in a lake ,the lake of S iba . Has now De l is le heard that it was the Satlej and not the Gangesthat came fromthe lake of S iba , - or S iva , Civ a , the Mahadeva of Hindu mythology, for indeed some 60 years later Fathe r T ieffenthaler heard the Manasarov ar

s tyled as the lake of Maha D eo ? De l is le has , however, taken the precau tion to re

j ect the lake and only to re ta in the old name S iba at the sou rce of the Satlej .The mos t in te res ting is De lis le ’s drawing of the upper Ganges . He re he has

fo l lowed, or, at any rate , be l ieves that he has fa ithfu lly followed the narrative of

Andrade , along whos e rou te he has ente red S irinagar, Mana , Badid pagode , Chaparangue or Tibe t. The s e are all the names which are conta ined in Andrade ’s narrat ive .

2 But De l is le ’s Ganges does not come froma »pool» on the pas s of Mana ; 3 i ttakes its origin at a ve ry cons iderable dis tance beyond Mana , and flows on its waythrough a cons iderable lake , north of the mountains . The lake is name les s as

Andrade ’s »tanque » , and i t is meant to b e the same . E as t of Badid (Badrid) and

S .E . of the lake is Chaparangue or Tibe t . North of i t a long river runn ing eas t to we s ts e ems to ente r the lake , b ut a legend te l ls us othe rwis e : R iviere qu i sort du memeLac que le Gange . E lle arros e le R ‘

de Tibe t . De l is le mus t have fe lt somewhatmys tified by this extraordinary bifurcation , b ut he had nothing e ls e to

'

do than to

I Terry, 1 6 1 6 : »Syb a, the chiefe Citie is cal led H ardwair, where the famous R iver Gange ss eemed to begin , is su ing out of a Rocke , which the superstitious Gentiles imagine to b ee l ike a Coweshead (Purchas H is P i lgrims , V ol. IX, p . Roe , 1 6 1 7 : »S iba , the chiefe Citie so cal led , itborders with N akarkut Southerly, it is all Mountainous .» (Op . cit. V ol. IV, p .

2 A fter mention ing Agra , De l i and La6r, Andrade has S iranagar, cal led S inadagar at anotherplace , R io Ganga , Pagode Badrid , also Spe lt Bradid , Mana, T ibet and Chaparangue . Bradid i s of

cours e the temple of Badrinat, vis ited by many pilgrims . »A ss i somos passando algfis d ias , ate que

acabo de mes , 81 meo chegamos ao Pagode Badrid , que esta nos confins das terras do S inadagar, a

e ste ha grande concurso de gente , ainda das partes mui remotas , como de Seyli o, B isnaga , 81 outrasque a e l le v ememromaria .» (Novo de scobrimento etc . p . 4 , b .) Bu t Andrade has also heard the rea lname and even written it down , for he quotes (p . 3 ) the exc lamation »ye Badrynate ye ye » .

3 On ly 1 00 years later this region was revis ited by E uropeans . On Kingsbury ’s , Parbury’s , andA llen ’

s M ap of India 1 8 25 (Add itions 1 8 2 7) the Alaknanda i s shown as coming froma pool , DebTal, at the southern foot of the Mana Ghat.

28—1 3 1 3 8 7 I

2 1 8 MAPS FROM TH E FIRST TH IRD OF THE E IGHTE ENTH CE NTURY .fol low Andrade ’s te x t, which he has , of cou rs e , misunders tood, though his map, whennoth ing e ls e was known fromthes e regions , s eems real ly to b e in fu l l accordancewith Andrade ’s words : »onde nas ce 0 R io Ganga de humgrande tanque , 81 domesmo nas ce tamb emou tra , que rega as te rras do Tibe t» , exactly as the legend on

the map . From Andrade ’s poo l is sue the Ganges and the river watering Tibe t ;the s ame is the cas e w i th De l is le ’s lake . But by Tibe t Andrade on ly mean t Chaparangue . De l is le be l ieves he meant the whole Royaume du Grand Tibe t which iswritten north of and along the rive r ; north of i t is the Kingdomof Ladac or Laduca,

I

and sou th of i t Rudoc or R edoc. Both Andrade and De l is le leave the river wateringTibe t w ithou t a name . To Andrade i t was on ly a smal l tribu tary to the Satlej ; toDe l is le i t was a great rive r. He makes it flow through the Kingdomof Cogue, thepres ent Guge , which in real i ty makes the river identica l with the upper cou rse of

the S atlej , of which , howeve r, De l is le cou ld have no information . I have shownabove that Kircher’s lakes , Pl . XI , are the Manas arov ar and Rakas -tal. De l is le ,on the other hand , who mis s ed the direct information Father Kircher had obta ined ,had no reas on to confus e Andrade ’s pool w ith the Manas arov ar. Therefore his maphas no direct ind ication of this lake , except that h is lake is too great for a pool .

To recapitu late , De l is le ’s hydrography is the fol lowing : the Indu s and the

Ganges ris e clos e to each othe r in the moun ta ins of Cachemire , for the rive r whichenters the name les s lake from the N .W . is the uppermos t Ganges , as shown on

De l is le ’s map , Carto do Tarta ri e 1 706 (PI. XLI). Two rive rs issue fromthe

lake , one runn ing eas twards through Tibe t , the other be ing the Ganges . The Satlejhas no connection with the lake . The topography is not b ad, though the hydrographical combination of the di fferent waters is wrong. The name les s river shou ldb e flow ing wes t ins tead of eas t, and join the S atlej at S iba ins tead of the lake , andthe Ganges shou ld b e cut off fromthe lake , and the map wou ld b e right. As itis the lake cu ts the S atlej in two pieces , of which the uppe r half belongs to Cogu e ,the lowe r s tarts fromS iba .

On ly a ve ry s can ty know ledge of Tibe t had been ga ined in E u rope abou t1 705, and s til l i t was sufficien t to confus e even the mos t cleve r draughtsmen . To

place the d i fferent kingdoms , rivers and moun ta ins approximate ly correctly in re lationto each othe r was , of cours e , impos s ible . De li s le gives us a ve ry fine proof of themaximum of correctnes s to which such attemp ts cou ld reach . He s eparates TibetfromIndia w i th a double range cal led M. de Purb et ou de N augracut , of which thefirs t name oo it s o mean s mountain .

North of thes e mounta ins De l is le has s even kingdoms , or s ix if Grand Tibe tb e cons ide red as including the whole lot . Leaving eas te rn Tibe t, w ith Las sa alone ,we have so far fiv e kingdoms : Ladac, Cogué , Rudoc, Uts ang, and Moriu l . Hereaga in Andrade is his source . For when the Portugues e mis s ionary , accompan ied by

I In hi s Histoire du T ibet, p . 3 9, Andrade uses the formLodacca .

TIBE T 0N DE LISLE ’

S MAPS . 2 1 9

Fathe r GONZALE S DE SOUSA and MANUE L MARQUE S had made his s econd journeyto Chaparangue , which he reached on Augus t 2 8th, 1 625, he wrote a le tter in whichhe s ays : »The kingdomof Tibe t or The Powe rfu l embraces the Kingdomof Cogue ,where we l ive at pres ent, thos e of Ladac, Mariul,

I Rudoc, Utsang, and s til l twomore s ituated farthe r eas t. With the great kingdomof Sopo (Mongol ia), which on

one s ide borders upon China , on the othe r upon Moscov y, they formtoge the r G reatTartary .»

2 In th is le tt e r exactly the s ame names are men tioned as those found on

De l is le ’s map .

From the N .E . s lopes of the N augracut mountains , a rive r flowing N .E . has

the legend : »R(ivie re) qu i s ort des te rres de Raja R ibron et s e rend diton dans laM(er) Raja Rodorow or Raja R ibron i s to its greates t extent s i tuatedsou th of the mounta ins . I have not been able to find out what it is meant to b e .

Rahia Rodorou was also entered on Sanson’

s map 1 654 (PI. XXIX). Shou ld R ibronb e a surviva l of R iboch , Ribok , R ib oth or any othe r of Odoric’s and Mandev ille

s

vers ions for Tibet ? 3 A rive r originating fromthe northe rn s lopes of Himalaya,wes t

of Nepal , and eas t of Chaparangue cou ld not pos s ibly b e anything b ut the Ts angpo .

But this rive r was s til l unknown , and De lis le ’s river flows to the N .E . , and comes ,on the map, to an end in Grueb er

s country of Toktokai . But on the map of

Great Tartary (Pl . XLI) the river continues through Toktokai and empties its e lfpartly into Cinghai , and partly into Coknor. But as Coknor (Koko-

nor) is in con

nection with the Ye l low R ive r, the latte r, in De lis le ’

s opin ion , has its sou rce on the

northern s lopes of the Himalaya . S t il l he has a lso a Hwangho-source after the type

of Martin i , though he calls the double lake Sosama .

De l is le shows such a great and really touching reverence for the memory of

the Jes u it Fathe rs , that he cannot induce hims e lf to recogn is e in Grueb er’s Koknorthe Cinghai or Mare n igrumof Martin i . And the refore he has two copies of Kokonor on his map, one with the Chines e , the othe r w i th the Mongol name . Here i sa new example to the fact a lluded to so often in the preceding chapters , that information from two different sources abou t one and the same geographical objectmay eas ily lead to re ite rations on the maps . We saw it already in the cas e of

Ptolemy ’s Oechardes and Bautisus , and we s aw it w i th E dris i ’s lakes of Berwan

and Tehama , the rive r be ing in both cas es the Tarim, and the lake Manasarov ar.

I t rs surpris ing that De l is le having marked Grueb er’

s rou te on his mapdoes not , in this cas e , follow the text verbal ly, for Gruebe r speaks of »Kokonor

I It shou ld b e noted that Andrade has M arinl , b ut De l is le Moriul . Csoma de Koros trans latesM ar-yul » low country» , H inen Tsang calls it Ma-lo-pho, Marpo, »red» (Cunningham: Ladak, p .In any cas e Mar-yul i s the same as Ladak , wh ich Andrade could not know , and stil l les s De l is le .

2 Lettere annue de l T ibet del MDCXXVI e t de lla C ina del MDCXXIV , scritta al M . R . P .

Mutio V ite l le sch i , Generale de l la Compagn ia de Giesu Roma 1 62 8 . As this work i s not avai lable to

me I have the quotation fromC . Wesse ls ’ article : Antonio de Andrade , Op . Supra cit. p . 29.

3 It has some resemblance also w ith the Rahia Tib b on of several maps , for instance Pl . XXIX ,and PI. XXXIII . In Terry ’s and Roe

s l ists of Ind ian provinces there are no names remind ing of R0doron or R ibron .

2 20 MAPS FROM THE FIRST TH IRD OF THE E IGHTE E NTH CE NTURY .dunque s ign ifica Mar grande dal le rive de l quale succes s iv amenta discos tandos ii l Padre , entro in te rra Toktokai while on both De l is le ’s maps Toktokai remains at a cons iderable dis tance wes t of Grueb er’s rou te . In another point he follows G ru eber clos e ly , as may b e s een from his maps , name ly when the Fathe rsays of Koko-nor: »Q ues to e un mare di done his. l

origine il fiume G iallodi China .»

If Koko-nor appears twice on De l is le ’s map, we find Lhasa thrice , unde r d i ffe rent

names . The firs t is Utsang, capital in the Kingdomof Utsang, the two provincesof U or We i and Ts ang, which w e remember fromthe t itle of the Chines e workWe i-tsang-t ’u-

chih . This Utsang De l is le places sou th of the rive r Maurou s so,

S ince Hue ’s and Prshev alskiy’

s memorable journeys be tte r known as Mur—u s or

Muru s su,

and also as D i-chu , or the uppe r Yangt s e-chiang . North of the river isM . Ba inhara or Bayan-khara—u la , cont inu ing eas twards in the range Altountchi

>>te rroi r d ’

ou l’

on ti re de l’or» .

I

Continu ing s ou thwards , pas t Moriu l , we reach Pou ta la res idence du GrandLama

,which is the s econd Lhasa of the map . The third, Las sa or Barantola

, is

the capita l of the Kingdomof Las sa or Bou tan .

E .N -E . of Pou tala is Grueb er’

s Ret ink , at the sou thern foot of M . Tan la . Is

this Tan la meant to b e Nien-chen-tang-la

,the eas te rn Transhimalaya ? It s eems so

,

for Potala and R eting-gompa are indeed s ituated at the s ou thern foot of the Trans

himalaya . But s uch conclus ions are u s e les s as there are three Lhas a ’

s on the map.

Add to this that on the northern s ide of Del is le ’s Tan la is a rive r Aghdame , now

known as Akdam,a right tribu tary to the Mur

-us su . Dr. TRONN IE R identifies Grue

ber’s Toktoka i w i th Mur-us su .

2 On De l is le ’s map both are ente red as s eparaterive rs , and Toktokai is a lso ca l led Hatounous so. The lattermay perhaps b e identicalw ith the Hatun ~

gol of Prshev alskiy,3 the Mongol name for the Ye l low R iver where i t

leaves O ring-nor; as gol, u s su , and mu ren all mean river in Mongol , the rive r ' in

qu es ti on can as we l l b e cal led Hatun-us su , exactly the s ame as the Hatounou s so of

the map.

To re tu rn to Tan la we read sou th of this range the name Tgoumera , whichsounds somewhat l ike d

Anv ille’

S Tchimouran and Littledale’

s Ch'

aremaru , as the

highes t peak of N ien-chen-tang—la is said to b e cal led. On De l is le ’s map, however,

the name is attached to a tent—camp .

I In Jaggatai Turki altunchi means both go ld-d igger and goldsmith . In the mountains southof Lop

-nor there are many gold-mines , the most important of‘

which is Bokalik. (See my works :Through A s ia , I I , p . 92 3 , 957 , and 960 ; Central A s ia and T ibet, I , p . 470 , I I , p . 1 97 ; Pe t. M it.

E rg. Bd XXVIII , p . 6 and 2 3 , where two villages are mentioned , called Altuntji , one near Posgam,the

other at Khotan : and Sc ienti fic R esu lts , V ol. II , p . 1 6 and 93 . e tc.)2 L . c . p . 3 46. I shal l have to return to this question in conn ection with the journeys of

Grueber and Huc. For the present b e it sufficient to say that Rockhil l crossed the river Toktomai onJune 1 8 , 1 8 92 , and the Murus on the 2 2nd . Diary of a Journey through M ongol ia and T ibet, Washington , 1 8 94 , p . 2 1 2 and 2 1 6.

3 At Kyakhtiy na Istoki Sholtoy Reki, p . 1 54.

222 MAPS FROM TH E FIR ST TH IRD OF THE E IGHTE E NTH CE NTURY .Turn ing our attention to the north we mus t confes s that Tibe t is very we l l

placed as compared wi th E as tern Tu rkes tan . The re may even b e sa id to exis t anindication of the w es tern Kwen-lun . He has Casgar tw ice ,

Iand calls the sou the rn

one »Casgar ou Cacheguer anciene Capita le du On his map of 1 706 (PI.XLI), the northern Casgar has d is appeared, b ut »Yarkan ou Irken 3 Cap . du R .

de Cachgar» i s s ti l l ten days north of Cachgar. On Pl . XL Cotan is s itu atedon a rive r flow ing S .E . and ending in »Des e rts sans eau» . On Pl . XLI the rive rof Cotan flows north to a des ert which is part of Dese rt de Caracatay ou V ieu Cathay , a name that he prefers to the old D es ertumLop. In this des e rt he has placedthe famou s »Araptan Kan Calmuc» , who , a few years later s ent his armies throughTibe t and captured Lhas a .

O the r name s in th is part of Central As ia w i ll eas ily b e recogn is ed. There are

the routes of Goe'

s and the i tinerary of Bern ier, and an Arab itinerary throughnorthern Tibe t . Acsu , Yu lduz , Turfan and Camou l are we l l placed. The Kingdomof C ialis is s ti ll the re and s e ems to refe r to the s urroundings of Bagrash-kOl. The

rive r of Kenker mu s t b e e i ther that of Kara-shahr, Kha’

r‘

du -

gol, or that of Korla ,Konche—darya . There i s indeed a Conche not far w es t of it, b ut i t i s an alternat ivename for Cucia which is Ku cha . There is also a Ugan , our Ugen-darya .

4 But on lyt wo rivers are drawn in E as tern Turkes tan , Khotan-darya and Konche—darya ; the

Tarim, of which Ptolemy had two copies , i s mis s ing, and so is Lop-nor, which is not

su rpris ing , for on ly be tween 1 760 and 1 765 Emperor CHI E N LUNG d ispatched the

Jes u its D’

E SPINHA ,D ’AROCHA and HALLE RSTE IN tomake a map of the surroundings

of Lop-nor.

5

Compa ring the map of 1 705 (PI. XL) with another map (Pl . XLII),1 7 2 3 , by the same au thor, we find cons iderable improvements in s eve ral respects .

The t itle of the map i s so importan t that i t mus t b e given in fu l l : »Ca rto d’

As i o

dros so’

o f our l’

Usag e da R oy . S nr los momoi ros en v oy e‘z par lo Czar a l

Aoa

domi o R oy a le dos S oi onoos S n r t o one los Araoos n ons on t la i s s e’

do filn s exact

dos pay s ori en ta ux S nr a n grand n omoro do R on ti ors do torro ot do mor ot do

Cartos manu s cr i tos dota i l lo'os . Lo toa t a s s nj oti aux oos orv a t i ons do l’

Aoadomi o ot

a oollos dos R . R . P . P . yos n i tos ot a n tros Ill a t/zoma ti oi on s P ar GUILLAUME DEL ISLE fl an The memoi rs which Ts ar Pe te r s ent to the Academy wereto a ve ry great exten t the resu l t of the ass iduous work carried out by the Swedishofficers kept in Rus s ian captivity at Tobolsk and other places of S ibe ria , as shal l

I The northern fal l s outs ide the margi n of PI. XL.

2»Kacheguer was formerly the royal re s idence , though now the K ing of Kacheguer re

s ide s at Jourkend , a l ittle more to the north , and ten days ’ journey fromKacheguer. » Bern ier, as

quoted above .

3 The name for the place comes fromI sb rants Ides ’ map P l . XXXVIII . On h i s map of 1 705Del i s le call s the c ity H iarchamand Yourkend .

4 I have described these rivers in Pet. Mit. l . C . , and Scientific Resu lts Vol. I I .5 Sc ientific Results , Vol . I I , p . 2 73 e t seq . , R ichthofen , China I , p . 690.

224 MAPS FROM TH E FIRST TH IRD OF THE E IGHTE E NTH CE NTURY.On his map : Carte do P ors o, 1 724 , which is superfluous to reproduce , De l is le

shows the sou rce of the Indu s in Mont Caucas e fromwhe re i t flows to the city ofCachemi rand a lake w i th a palace on an is land , and thence to Atok, rece iving the N i lab .

It may b e su i table to ins e rt he re a few extracts frombooks and narratives of

that time , quotat ions which will gi ve a s tronge r background to the high me ri ts of

De l is le .

MANOUCH I’

S JVI omoi rs be long to a somewhat earlier period, for he s e rved

AURAN GZE B who was the ru ler of India in 1 658 to 1 707 . But as thes e M omoi rswere edi ted by CATROU in 1 709 , they may b e mentioned here .

I There is nothingnew in the work , which was behind its t ime , and the au thor re lies more upon the

obs ervations of Te ixe ira , de l la Valle , Roe , de Lae t, Bern ier, Tavern i er, and others ,than his own . Of the Ganges he says that it takes » its source almos t at the s amehe ight w i th the Indus towards the North» . Regarding the source of the Indus weread : Cachemi re is »the

{

mos t Northern Coun try of the Mogols Domin ions , S ci tuateat the foot of Mount Caucasu s . Cachemire is indeed b ut one large Val ley sur

rounded with Mountains . Some pretend that the R iver Indus takes its S ource the re ,b u t

ti s a mis take not follow’

d by any Geographer. Its true that a great manyR ivu lets ris ing in Cachemire run into the Indu s

In an article : »R oo/zoro/lzos afa i ro da n s los E ta ts occupez par l os M a/i omota ns » ,we get an idea of what was regarded as the chief Indian des iderata of phys icalgeography in It is a l i ttle >>Hints to Trave l lers » 200 years ago . In the

Kingdomof Kachemire , i t is s aid, one shou ld try to find out the caus e of the sudden

changes on Mount Pire -penjale , »where one pas s es froms umme r to winter in les sthan an hour» . One fee ls a s uffocating heat at the foot of the mountain and to

wards its he ights everything is covered w ith ice and frozen snow . The refore the

formand he ight etc. of this mountain shou ld b e examined. In thes e words are ex

press ed the unders tanding of the importance in hypsometrical re lations , which was ,however, fami l iar to Fathe r Gerb illon some 20 years be fore . The glacie rs and

the part they played as birth-places of the great rivers had not yet attracted the

a ttention of geographe rs .

Furthe r : »whether it b e true that in a place cal led S ang-safed,'

and which isnot very far froma great Lake near the even in summer frozen Caucasu s , a greatno is e is capable of exci ting a s trong ra in , and , provided the s tory to b e true , one

shou ld inves tigate the caus e of this phenomenon .» Whe re this »White S tone » and

»great Lake» are S ituated is , unfortunate ly , not told.

Another matte r wou ld b e to s tudy the countries s i tu ated in the mountains of

Caucasus and su rrounding Kachemire , and what is to b e found in them.

I The General H istory of the Mogol Empire , fromi t’

s Foundation by Tamerlane , to the LateEmperor Orangzeb . E xtracted fromthe Memoirs of M . Manouchi , a Venetian , and Chie f Phys itianto Orangzeb for above forty Years . By F. F . Catron. London 1 709, p . 1 5 and 1 94 .

2 Recue il de Volages au Nord, Tome I , Amsterdam1 7 1 5, p . 4 7 et seq .

226 MAPS FROM THE FIRST TH IRD OF THE E IGHTE ENTH CE NTURY .In anothe r book which was wri tten by the e lder PE TIS DE LA CROIX who

died in 1 695, b ut which was published in E nglish in 1 72 2 , and accompan ied by a

map of De l is le , we are as ton ished to s ee that no other au thors are consu lted on

matte rs Tibe tan , than the Arab s .

I The au thor be l ieves that the »reader wi l l b e gladto know some part icu lar account of Tebe t» , which is s i tuated be tween 30 and 40

°

of Lat . , and be tween 1 00 and 1 1 0°

of Long. ,»according to the s i tuat ion ,

given i tby the E as te rn Geographers » .

—»It contains part of the Region ca lled Tu rc-Hinde ,because it is a part of Turquestan on the one s ide , and of India on the other. Somedivide i t into two Parts , cal l ing i t the Great and Les s er Tebe t. The Great borderson China , the Les s is s ituate to the E as t of the l i tt le Kingdom of Kas chemire , jus tbehind the Mounta ins ; and they are both b ut s ixWeeks Journey over. This Countryis fu l l of Towns and Vi llages we l l inhabited ; the People are so good humour’d and

chearful, that they breathe nothing b ut Joy and Pleasure . But what is mos t surpris ing, is , that the re is a Mounta in cal led Jabal Assumoum(’the Moun t ofwhich inspires all thos e with Me lancholy who come near enough but to sme l l i t,nay, even tu rns the ir Tongues black , in such a manner, that they remain black all

the res t of the i r Lives .» Then fol lows the ordinary s tory of the musk and rhubarb .

Of the Jaxartes or S ihon he knows that it has two sou rces in Mount Imaus , and

falls into the Caspian S ea .

2

It is in vain that one reads through the many volumes of »R ecu ei l dos Voy ag es

qu i on t sorv i a l’

otaol i s s omen t ot aux progrez do la Compagn i e dos I ndes Ori enta les » , 3 for there is not a word abou t the interior of the country. The musk tradei s des cribed, as i t was carried on in Goa, and i t is said to come fromTartarythrough China . When Mount Caucasus is once mentioned, the quotation is fromPliny . The w hole collection is nothing b ut capes , is lands , ports , ships , coas ts , s eas

,

I The History of Genghizcan the Great, Col lected fromseveral Oriental Authors , and E uropean Trave l lers By the late M . Peti s de la Croix Sen ior faithfully translated into E ngl ish ,London 1 7 2 2 , p . 58 , 1 1 6 , and 1 73 .

2 In connecti on w ith Caschgar and Hyarcan he has a very surpris ing story which sounds l ikea good joke . A »M . Sparv enfeld t, who was Master of the Ceremon ies to Ambassadors at the K ing of

Sweden’

s Court, be ing at Paris in 1 69 1 , as sur’d the author that he had read in the anc ient Annals ofSweden that the Swede s took the ir Origina l fromthe City of Caschgar in Turquestan» .

Probably Peti s de la Croix has mis understood the learned Sparfv enfeldt, though the latter wassa id to have thoroughly mastered fourteen languages . As a matter of fact Sparfv enfeld t was very farfrom the Opin ion that the Swedes originated fromA s ia. On the contrary , he was , —as his famouscountryman O lof Rudbeck , who li ved at the same time ,—persuaded , that Sweden was the origin ofhuman ity . To prove the correctness of this theory he was despatched , in 1 68 8 , by King Charles XI ,upon an exped ition of s everal years , which brought himso far as to Spain and A frica .

3 Rouen 1 725; Tome V ,p . 1 3 , Tome VI , p . 400.

In another work,Histoire de T imur-Bec, Peti s de la Croix, Paris 1 72 2 , there i s a map , Carte

de I’

E xpedi tion de Tamerlan dans les Indes , which shows great d iscrepancies with De l is le , 1 7 yearsearlier. The l ittle map is of no importance whatever, but I have given a reproduction of it P l . IV . Itshows that Delis le was not general ly accepted . Petit Tibet and Grand T ibet are we l l placed ; the westernHimalaya and Kwen-lun are vis ible ; at the upper Satlej is a place Mansar, which reminds one of Manaserovar ; M I Couké is also be low Mansar. The Ganges come s froma lake be low which is the eterna lCow : »Vache de piere adorée Par les Guebres»

DE SCRIPTIONS OF TIBET . 227

and trade , the naval achievements of a s et of ve ry clever and courageous , and—s i t v en i a

v eréo s tupid sai lors , who had no interes t for the country at a few leagu es from the coas t .I wi l l now say a few words abou t AB UL GHAZI ’s genealogi cal work on the Tartars , special ly on account of the accompanying maps .

I On p . 4 1 of the Frenchedition we are told that Ogus -Chan conquered the Empire of Kitay , the City of

Dsurdsut , and the Kingdomof Tangu t together with Cara-Ki tay, the capita l of whichi s a great Town ; >>this coun try is inhabited by people , as black as the Indians ; theyl ive in the su rroundings of Lake M '

Ohill be twe en Kitay and the Indies , a l ittle on

the south» . Behind Kitay and near the S ea Ogu s—Chan came across a Chan cal ledItb urak . Some of the s trange names in this pas sage are to b e found on one of the

two maps in the work , which I have reproduced as Pl . XLI II . 2The source of the Indus is s ituated in the Imaus Mons on the very fron tie r of

Bucharia Minor or E as tern Turkes tan . The source of the Ganges is a name les s lakefu rther eas t in the s ame mounta in . E as t of the source of Ganges Tibe t is shownas s ituated in Dese rt Goby, s urrounded by mounta ins . Between both is a »Campd’

un Chan tribu tai re au Dalay Lama» . South of Tibe t is a rive r Yekegoll fall inginto t wo lakes , probably a su rvi val of Cinha‘

r‘

and Coconor. South of this rive r isTangu t w ith anothe r tributary chi ef, and Barantola , and, in the southe rn part, Las saand Potala , »R es idence du Dala i Lama Grand Pontife des Callmuckes et Mongales » .

Fromthis arrangement one ge ts the impres s ion that Tibet is repres ented twiceon the map, the southe rn one be ing cal led Tangu t . 3 Las sa and Potala Show that

.Tangu t s tands for Tibe t Proper , and in Mongolia Tibe t is inde ed cal led Tangut.Barantola is the Mongol name for Lhasa , and the capita l the refore is marked twice .

If the northern Tibe t '

of the map is also mean t as Tibe t Prope r, Yekegoll may b ethe Tsangpo or Tsangpo

-chimbo, w hich , as Yike—gol, means Tao Great R ive r.

I H istoire Genéalogique des Tartars . Tradu ite du Manuscript Tartare d’

Ab ulgasi-Bayadur

Chan Leyde 1 72 6 . Abu l Ghazi , Khan of Khiva , d ied in 1 663 . How his History, in 9 books ,of Jengis Khan ’

s fami ly , was di s covered, is told in the »Au Lecteur» of the French ed ition : »Le Publice s t ob lige du pres ent Traité a la prison de s Officiers de Suede en S iberie ; car quelques uns de ces

M rs . qui estoient gens de lettres ayant acheté le Manuscript Tartare de cette H istoire d ’

un MarchandBouchare qu i l

apporta a Tobolskoy le firent tradu ire a leurs de spen s dans la Langue Russe , 81 letraduis irent ensuite eux mesme s en d iverses autres Langues .» These »Mrs» were the Swed ish officersSchOnstrOmand Strahlenberg, of whommore later on . SchOn strOmpresented a M S copy of the workto Upp sala in 1 7 2 2 . The work was published in several E uropean languages . The E nglish ed ition i sknown under the title : A Genera l H istory of the Turks , Mogu ls , and Tatars , Vulgarly cal led Tartars

e tc., London 1 73 0 ; and Vol. II , bearing the year 1 7 29 : An Account of the Present state of the

Northern A s ia , R e lating to the Natural H istory of Grand Tatary and S iberia , e tc .

, the whole compiledout of the Notes be longing to the foregoing History , and d iges ted into Method by the Trans lator.

The French edition is published by i . e . the Dutchman Bentinck (Barbier) .2

»Carte Nouve l le de l’As ie Septentrionale dres sée Sur des Observations Authentiques et toutesNouve l les» . On the copy at the Royal Library of Stockholm,

Aug. Strindberg has written a penci lnote : »This first map is the one published by Bentinck in Leyde 1 7 2 6, and based on Strahlenberg’smap.» There i s no doubt a cons iderable resemblance in type between the two, though Strahlenberg’smap was published only in 1 73 0 .

3 In a note , p . 4 2 , the ed itor also says of T ibet : »I l e st partage en deux Parties dont la PartieMerid ionale s’

appelle proprement le Tangut 81 la Septentrionale le T ibet.»

228 MAPS FROM THE FIRST TH IRD OF THE E IGHTE E NTH CENTURY.Sou th of Tangu t is a lake Cara Nor, the same as Ides ’ Cananor, the Chiamay

of older maps . As on mos t maps of the time Indus and Ganges reach too far north,3 7 and 3 8

° N. lat . India the refore becomes , to a great extent, the immediate ne ighbou r of E as te rn Turkes tan , and Tibe t is removed eas twards as if it were s i tuatedbe tween India and China .

R egard ing the s econd map , Pl . XLIV ,

Iwe find that the geography is ex

actly the s ame , only the names are changed . Both Tibets are now joined under thename Tangut, and in the northern half is the town Tangu t. On the upper Yekegollis a place Akas chin .

2 Las sa has been changed into the town Ds urdsut , which in thetext was s aid to have been conquered by Ogus-Chan ,

and Cara Nor is now calledLac MOhill, s i tuated in Cara Kitha i , which practi cal ly coincides with uppe r Burma.

The map therefore , as its title indicates , agrees with Abu l Ghaz i ’s text, but not withreal facts . The edi tor be lieves that the au thor, with his Cara Kithai , means Tonkinor Cochin China , as be ing s i tuated south and S .S -W. of China .

The fol lowing pas sage gives an idea of the conception one had at that timeabou t the natu re of Tibe t . In a note the ed itor s ays that for trave l l ing fromGreatTartary v ia Tangu t to Tonkin , and Pegu , one has to fol low the boundaries of Chinaor Great Mogu l, for it is impos s ib le to pas s through the cen tral parts of Tangu ton account of the vas t s and des e rts which occupy the interior of this kingdom, and

which s tre tch fromthe ve ry frontiers of the Kingdomof Av a, to a cons iderable distance northwards and beyond the fronti e rs of the Kingdomof Tangu t. Al l thatwas known of the inte rior of Tibe t had been brought back by Grueber, and he hadshown that it was poss ible to cros s the country .

Hwangho, unde r the name Cara-muran , is s aid to b e one of the greates trivers in the world ; it has its sou rce at 2 3

° N . lat . on the boundaries of Tangu tand China , »in a great lake , which is enclos ed within the high Mountains s eparatingthes e two S tates » . Fromits s ource the rive r runs north a long the front iers ofXien-s iand Tangu t to 3 7

° N . lat ., after which it continues ou ts ide of the great wa l l and

waters Tibe t, which shows how very vague the knowledge of thes e regions was .

On the two maps , Pl . XLIII and XLIV, Jerke en or Jerkehn is s t il l northof Caschgar, and the re are Axu ,

Ku ts char, Chateen , Lucz in , Tu rfan , and Chamill ;the Tarim is final ly to b e s een , after such a long abs ence , s ince Ptolemy ; itempties i ts e lf in to a great lake , which mus t b e the Lop

-nor though s ituated north

of Tu rfan . The moun tains sou th of this bas in are ca lled on our map Imau s Mons ,on the othe r Mus Tag .

4 Ve ry like ly this more intimate information of E as te rn TurI Carte de l’As ie Septentrionale Dans l’E stat OI

I E lle s’

est trouvée du temps de la grande Inv as ion des Tartares dans l’As ie Meridionale sous la Conduite de Z ingis-Chan .

2 The E ngl ish E d ition , p . 492 , says : »by others cal led Cashin» .

3 French edition ,p . 3 57 , Martin i had which i s at any rate much nearer to the 3 5

°

wherethe source is in real ity .

4 In the E ngli sh edi tion , p . 469, the fol lowing explanation is given of the s ituation of the K ingdomof Kashgar : »

T is bounded on the North by the Country of the Callmaks and Mungals ; on the

CHA PTE R XXV.

LAGO DE CH IAMAY.

Before proceeding further in our res earches in the deve lopmen t of the knowledge of Tibe t during the firs t th ird of the e ighteenth centu ry , we have to go backin time and try to trace the his tory of the e xtraord inary cartographica l phenomenon ,which , at its firs t appearance , was baptized as Lago de Chiamay, and then , withins ign ificant variations of name

,and with admirable pe rt inac ity, remained on nearly

all maps for some 1 60 years .

I will firs t lay be fore the reader the mate rial that has been available to me ,and then draw the conclus ions to which i t may lead . I t makes no pretens ions of

be ing complete , b ut it is s ufficien t for the conclus ions to which I a llude .

Unde r the heading D el la li i s tori a del S ignor G i ov an do B arrosIwe find in

Ramus io’

s N a v iga tion i et Vi agg i , 1 550, a short commun ication , and s ti l l themos t complete in exis tence , conce rn ing the lake Chiamay ; 2 he has been speaking of the

s ixth d ivis ion of As ia and of the town of Malaca, and then goes on to des cribe a

ve ry mighty rive r which flows through the whole of Pegu 3 and which comes fromLago di Chiamay, 200 leagues d is tan t to the north in the interior of the coun try,and in which s ix notable rive rs take the ir origin . Three of thes e jo in to formthe

great rive r which flows s tra ight through the Kingdomof S iam,4 while the othe r th ree

fal l into the Gu lf of Bengal . One of thes e cros s es the Kingdomof 'Caor, fromwhichit derives its name , further through the Kingdomof Comotay and Cirote , after whichit empties its e lf, above Chat igan , in the great branch of the Ganges s tra ight acros sthe is land of Sornagan ; the s econd , the one of Pegu , flows through the Kingdomof

Ana, 5 and the third goes out at Martaban , between Tanay and Pegu , at 1 5° N. lat .

6

I Barros was born in 1 496 at V izeu in Portugal . In 1 53 3 he was nominated, by King JohnIII as a Treasurer and General Agent of India. In 1 54 1 he was ordered to write a history of Ind ia,a work that was continued after h is death (in 1 570) and published at L isbon .

2 I have at my d isposal the ed ition of 1 554 , where the passage i s read on p . 4 3 2 , D .

3 Irrawaddi .4 M enam.

5 Irrawaddi .6 S alwen .

A MYSTE RIOUS LAKE . 23 1

Further on I he aga in remarks that the river of S iamcomes,

.

to the greates t part, fromLago di Chiamay. On accoun t of the great volume of wate r this rive r carries down

,

the S iames e cal l it Menam , that is so much as >>Mothe r of the Waters » . Then itfal ls into the s ea .

2

The text, therefore , leaves no roomfor a doubt, so far as all the rive rs , ex

cept one , are conce rned. As to the Irrawaddi the re can b e no mis take , for it flowsthrough Av a and Pegu . But Gas taldi

s map, Pl . XVI , il lus trating the text , does '

notagree with Barros ’ description . The map has on ly fou r rive rs leaving the lake , thetext s peaks of s ix, of which the three to the eas t jo in and formthe Menam, whichon the map is only one s ingle rive r the whole way. The three wes te rn rivers goto the Gu lf of Bengal . One of them, the one farthes t wes t, travers es the Kingdomof Caor. On its right bank are the citi es of Caor and Comotay. Above Chatiganit ente rs the de lta of the Ganges , which it joins at the Ganges -branch on which thecity of Bengala is placed . Consu lting only Gas tald i ’s map of 1 550, it wou ld b eimpos s ible to te l l whethe r this rive r is mean t to b e the Meghna or the Brahmaputraor anything e lse . The s econd eas twards is the Irrawaddi , and the third is , as in

dicated by Martaban , Salwen , though its mou th in re lation to Tawa and Pegu is misrepres ented on the map.

If we remembe r for a moment the long and hard fight abou t the source and

origin of the Brahmapu tra, and that on ly in our days the transverse va lley throughwhich it pierces the Himalaya has been absolu te ly s ettled, though s ituated so near Calcutta, we shou ld not fee l surprised that 360 years ago mis takes were committed bythos e who firs t heard of thes e rive rs and the coun tries they cros s ed . The geographe rsof gene rations had to accept the hydrography wh ich Barros with the whole pondusof his name had given, and the mys terious lake with its four rive rs crys ta ll iz ed out

on the maps , and was impos s ib le to b e got rid of unti l a late r time came with fres hinformation . And s t il l we have to confes s that certain parts of the cours e of Salwen

Ln-chiang and of the Mekong—Lan-tsan-

chiang are unknown . The latte r river, howeve r, is , on Gas taldi

s map in Ramusi o, the firs t to the eas t, which , unde r the name

I Op . cit. p . 43 3 , A .

2 As this passage has been quoted in many works and be l ieved in as Gospe l during nearly2 00 years , I gi ve it a lso in Ramus io’

s text. H e starts with the Irrawaddi : »che parte tutta la terra diPegu , iI qua l vien dal Iago di Chiamay, che s ta verso tramontana per d istantia di 2 00 leghe ne l la interior parte de lla terra , dal quale procedono sei notabili fiumi , tre che s i congiungono con altri , 81fanno i l gran fiume che passa per mezzo del regno di S iam, 81 gli a ltri tre vengono a sboccar in

que sto colfo di Bengala. Vno Che vien trauersando il regno di Caor, donde i l fiume prese i l nome ,81 per quel le di Comotay, 81 per quel lo d i Cimte , doue s i fanno tutti le eunuchi che sono condotti diLeuante

,81 vien ad v scir di sopra di Chatigan in que l notab il bracchio del Gange per mezzo de l lai s ola Sornagan , l’altro di Pegu passa per i l regno Aua che e dentro fra terra, 81 l ’altro esce in Marta

b an fra Tanay 81 Pegu in latitud ine di 1 5 grad i E t s eguendo p iu innanzi 40 . 1eghe e il capodi S ingapura , doue princip ia al lungo del d ito ind ice la settima diui s ione che e de l i fin al fiume diS iam,

che la maggior parte de que llo procede dal Iago di Chi amay. Al qual fiume per causa de llamolta abundantia de l le acque, che porta seco l i S iamini chiamano Menam, che vuol dir madre de l l ’acque

,81 entra nel mare

232 LAGO DE CHIAMAY .

of Mecon is independent of Lago di Chiamay, and has its own cours e fromfar away ,fromthe unknown mounta ins of Tartari Mogori .

Comparing now the map in Ramus io , Pl . XVI , with Jacopo Gas taldi’

s Te rtiaPars As ia of 1 56 1 , PI. XVII , we find , on the latter, muchmore detail , both in hydrography and nomenclature . The hydrography is in perfect accordance w ith the textof Ramus io. S ix rive rs leave the lake , the three eas te rn join ing to formS ian or

Menan F. The next, the river of Martaban , leaves the lake at a place called Chiamai and has the name Caipumo f. Then fol lows Aua f. w ith the cities Aua and

Pegu , and final ly the Caor F . of the Older map, name les s on the latte r. On its

banks are the Citi es Gor, Camotai , and Chirote, and it joins the eas ternmos t branchof the Ganges de lta much fu rthe r fromthe coas t of Golfo de Bengala than on PI. XVI .Both maps have in the de lta the cities Catigang and Bengala , of which the formermus t b e Chittagong . The Is land Sornagon ,mis s ing on Pl . XVII ,may b e Sundarb ans ?

On Pl . XVI Lago de Chiamay is repres en ted as a real Madre del l ’ acque : itrece ives no affluents from anywhe re , b ut s ti l l emits four enormous rive rs . On Pl . XV I Iit rece ives at leas t two s trong feede rs fromthe moun ta ins between India and China.

‘He re the lake is cal led Cayamay Lago ; it s ends out four rive rs , though the eas te rnhas three heads .

North of the lake (Pl . XVII) w e find the Kingdomof Macin , which probably isMahachin or Grea t China us ed by Pe rs ian write rs as synonymous with Manz i , for inPe rs ian parlance Machin and Manz i we re identi cal. I The re are also two cities ,Amuyin-macin and Toloma , reminding of Prshev alskiy

s Amne-machin and Soloma at

the uppe r Hwangho.

2 But this is only a co incidence , for Gas ta ld i has on ly triedto reconcil iate Barros ’ lake with Marco Polo’

s geography, as the latter did not, nat

urally enough , know of any such lake . We have therefore to think of Marco Polo’

s

Province of An iu , in Ramus io’

s ve rs ion Amu , in Pauthier’

s An iu .3 Toloma is ob

v ious ly Marco Polo’

s Province of Coloman , which in mos t vers ions is called Toloman .4

Barros , on the othe r hand, is respons ible for the two names Caor and Camotay,and he has even a KingdomCaor in the mounta ins north of the lake , leading our

thoughts to the Garo Hills , the wes tern foot of which is washed by the Brahmaputra . But on Pl . XVII the name has been changed into Gor, and the rive r i s a

boundary betwe en the Regno de Camota i and Regno de Ve rma or Bu rma . As the

city of Ve rma is placed on a river which joins the Ganges de lta furthe r sou th , and

as there is no rive r s ituated between the Caor and the Ganges on Pl . XVI , the Caormus t needs b e Brahmapu tra

,which becomes much more evident on Pl . XVII .

As to the province and ci ty of Caor, al ias Go ’

r, w e find i tmentioned by Te rryin 1 6 1 6 : »Gor, the chiefe Cit ie so called, it is fu l l ofMountaynes . The R ive r Pers ilis

I Yu le ’s M arco Polo , II , p . 3 5.

2 At Kyakhti e tc . p . 1 50 .

3 Yu le , op . ci t . p . 1 2 0 .

4 Ibidemp . 1 2 3 .

234 LAGO DE CH IAMAY .

On the Gas taldian map in S al a de l lo S cudo there are on ly three rivers leavingthe lake ; on the w es t rive r is the city of Bengala ; be tween the two eas t rivers isAv a . DIOGO HOMEM ’

S map (Pl . X IX), on the other hand, has no Chiamay lake at

all, which s eems to ind icate that th is part of Homem’

s map may b e previ ous to

Gas ta ld i .The appearance of the lake was of cours e no improvement on Ptolemy’s map.

But far more unfortunate was ME RCATOR ’

S removal of the Ganges . The Chiamaylacus he has , howeve r, re tained , and his Caor rive r, which can no more join the

Ganges de lta , goes out at the eas t coas t of the Gu lf, pas s ing Catigamor Chittagong ;it has therefore a S .W . cours e and shou ld rather b e the Karnaphul than the Brahmapu tra, which is , of cours e , absu rd. Gas taldi

s Aua rive r or Irrawaddi has been changedinto a Cosmimflu . , and there is no Aua on it . The Salwen, Gas taldi

s Caipumo, hecal ls Martab am, and the re he is right again . Menan flu . has not been changed at all.

ORTE LIUS , 1 570, (Pl . XX I I), chiefly copies Me rcator, b ut he w is e ly gives the

four rive rs the names ind icated by Gas tald i . He wri tes the name of the lake Ch '

yamailacus . Gas taldi

s Amuyin macin and Toloma on the north s ide of the lake he haschanged intomay on the north and Toleman on the eas t. Thes e places had beenregarded as suspicious by Mercator, who has not marked themon the Chiamay lake .

But Mercator has placed themon another, name les s lake , far to the N .E . , where wefind Amu and Tholoman . Orte l ius has fol lowed his example , b ut ca lls the lakeCa

'

yamai lacus . Lago de Chiamay therefore appears twice on O rte l ius ’ map of

E as t India .

LUDOUICUS GE ORGIUS s hows on his map of 1 584 , PI. XXIV, the lake underthe name of Chiama lacu s with fiv e effluents , jo in ing, as usual , into fou r rivers , herew ithou t names . The Caor river, which on Gas taldi

s map flowed W.S -W ., flows

he re S .S -W . Gouro , obvious ly the ancient Gaur, is correctly placed on the Ganges ,and . the region of the Gov ros is s ituated be tween the lake and the Ganges .

The firs t map in LIN SCOUTE N’

S work Shows the lake as usual with four ri versb ut withou t name . On the s econd map, which embraces S .E . As ia

,on ly half Chiama

Lacus is s een .

HONDIUS , 1 6 1 1 , PI. XXV, has moved the lake a cons iderable d is tance to the

north . On ly the three eas tern rivers have names , the s ame as thos e of Mercator :Cosmin , Montab an ,

Iand Menan . On the latter is Aracam, though a province and

city of the same name is also placed between the Indus and the fals e Ganges , at

the uppe r cours e of which we find the Gouros , oppos ite to Caor on the Caor river.Cati gan is now changed into S atagam. The city and province of Bengala is nearthe mou th of the Cosmin river, Av a or Irrawaddi . Verma , Bu rma is on the rive rwhich has taken the place of Ganges . Burma and Bengal have therefore changedplaces .

I The sea-port at i ts mouth is more correctly written Martab am.

THE LAKE 15 THE OR IGIN OF FOUR RIVE RS . 235

The l ittle map in HE RBE RT ’

S work, publ ished 1 63 8 , PI. X, has four riversleaving the lake sou thwards , and in the ir »Mesopotamia» the names of Caor, Anva ,Satagam,

Brama , and Ve rma are placed wi thou t any kind of order. Martav an is to

b e found at the place of Bangkok .

I

On two maps , N ov a toti a s Torrarnmoroi s g eograjfiki oa ao nydrograpni ea

taou la , A net: H enr: H ondi o 1 63 0, and Oroi s Torraramtypzes do i n tegro mu lti si n lot i s omenda tns , by N . I . Vi s s ener 1 63 9, which both in real ity are the samemap, on ly two rivers go out of the lake , though one of themdoes s o with two headbranches .

2

Gas taldi show ed the lake at a cons iderable dis tance eas t of the Ganges de lta , sothat the Caor rive r had to flow nearly wes twards to reach it . HOE IUS and ALLARDT ,

1 640, PI. XXVI , have the Chiama i Lacus s tra ight north of the Ganges de lta, so thatthe Caor river has to flow S .S -E . in order to join it. The rivers are fou r and havethe ordinary names . WILLIAM BLAE U, 1 640 , PI. XXVII , accepts Amuy and Tolema ,and has both Caor and Gouro, oppos i te each othe r. The cities on the Caor rive rare : Caor, Comotay, Cirote , Verma , and Satagam. The s econd river eas twards isapocryphic, with the cities of Bengala and Aracamat its double mouth , and Chat igam, a s econd Chittagong, wes t of Bengala. The third rive r, with Av a and Pegu ,is Irrawaddi . Martaban is left withou t a river. In fact the s econd and third river,Irrawaddi and Salwen , have changed places .

On JAN SSON ’S map, 1 64 1 , PI. XXVIII , the lake has not undergone any

notable change from the previous map, and the Caor river flows , as on PI. XXVII,to the S .S -W . In the text to the map of the E as t Ind ies it is sa id : »Der grOs ste

aber unte r den S een dies es Lands is t der Chiamay, we lcher 400 Meyl wegs in s e i

nem b ez irck b egreifft, 500 Meyl 3 v on demMeer ge legen ist , und vie l flies s endeWas ser v on s ich ergeus t» , nothing more , b ut often much les s than in Barros ’ original des crip tion of the lake .

In the narrat ive of his jou rney the Portugues e trave l le r FE RNANDO '

ME NDE ZPINTO , a lso mentions the lake . De scribing diffe rent nat ions he came across , he says :Nou s en vismes d

autres au s s i qu i auoient d’

as sez longue s barbe s , le visage s eméde lenti lle s , le s ore illes 81 les narines pe rcée s , 81 dans le s trous de pe tits fils d ’

or

fa its en agraphes ; ceux-cy s’

appelloiét Gz’

napéog a os , 81 la Prou ince don t ils e s toientI In a German work on Indi a fromabout the same time (E xtract Der Oriental ischen Ind ien .

Das i st Aussftjhrliche und vollkommene H istorische und Geographische Beschre ibung A lles mitsondermF le iss beschrieben Durch Ca saremLonginvmH is toricvmFrankfurt amM aya 1 629, (SecondSupplement, p . the fol lowing passage i s found in the chapter on S iam: »Durch d iss Landt fleustder Flu s s Menan, we lcher gle iche Art hat, wie die andern , ufi entspri ngt auss (IE gros sen See Chiamay,und f 'alt in das M eer.» As the leading map s of the time accepted Barros ’ lake , the compilators of

geographical handbooks had of course to do the same .

2 Remarkable map s , II—III , N o. 6 and N o. 9 .

3 I do not know which kind of miles this is meant to b e . The legoa or old Portuguese league ,cons isting of 3 M ilhas , was m. long. If the »500 M eyl» are meant as M i lhas , the d istance shouldb e km. ,

which , in a north-westerly d irection , and starting fromthe sea at the Ganges de lta shouldtake us nearly to the actual source of the Brahmaputra.

2 36 LAGO DE CHIAMAY .

natifs , S nrooosoy , lesque ls par dedans les montagne s de Lauhos son t bornez du lacde C/zi ammay , 8

1 de ceux-cy le s v n s son t v e s tus de peaux ve lu '

es , 81 le s au tres decu it bronce He eve n te l ls abou t a war undertaken by the King of S iamaga ins t the King of Chiammay : A ins i pas sant pays i l arriua au lac de S i ngapamor ,qu’

on appe l le ordinairement C/zi ammay , O I‘

I il s’

arre s ta vingt-s ix iours , durant le sque lsil prit douze fort be lle s place s enu ironne

es de b ouleuarts 81 de fossez , a la facondes nos tres In thi s cas e , the re fore , the Lago di Chiamay is identified withanothe r lake , unknown to us , name ly S ingapamor.

I

We now come to the Voyag es f amonx da S i eu r Vi n cen t lo B la n e, publishedby PIERRE BERGE RON in 1 649 . There are two pas sages which are too precious tob e trans lated . S peaking of »S ian» the au thor says : 2 »Ce pais confine au iourd

huy

du cos te de l’Occident z‘

r celuy de Pegu , du Nort au pais de Chiamay La villede S ian es t s ituée sur la be l le 81 grande riviere de Menan , qu i vient du renommélac de Chiamay, 81 qu i a de be l les murai lles . Further he s ays of the » l icornes » : 3»Il s

en trouve , a ce qu’

ils d is ent , aux environs du lac Chiamay Ce lac a deuxcens mil de tour, d ’

or‘

r sort un grand nombre de grandes 81 fameus es rivieres , commece l le d ’

Aua , Caypumo , Menan, Cosmin , 81 au tres , qu i ont les mesmes inondations 81desb ordemens que le N i l . Ce lac a du cos té de Leuant de grandes fores teS '81 des

mares cages impene trables 81 dange reux Of Pegu he says that mos t of thiscoun try is cros s ed by the great rive r Amoucharat , or as it is cal led in Pegu Caipoumo 8 1 Martaban »Cependan t ie d iray que nos Geographes s e trompen t, qu ime tten t la riviere qu i arrous e le pa is de Tangu , pour lamesme que ce lle-cy de Pegu,quoy qu ’

e lles so ien t d ifferentes 81 bien es loignées : Car ce l le cy vien t de ce grand lacChiammay, 81 pas s e a Brema ou Brama.

» Fina lly some Peguans told himthat a fewyears be fore his vis it a king of the old dynas ty had s everal l ieu tenan ts in the

countri es of the Brama »toward lake Chiamay» and one of themwas in the kingdomof Tanga , which had revolted aga ins t him.

LE BLANC not on ly knows the s ize of the lake , he a lso gives us some glimps esof its natural his tory . In the great fores ts and dangerous swamps on the eas t shoresof the Chiamay the un icorns have the ir eldora do. He discus s es hydrographicalques tions , and ment ions four rivers : Aua (Irrawaddi), Caypumo (Salwen), Menan(Menam), and Cosmin (Irrawaddi), thus us ing two d iffe rent names for one river, andomitting Caor. But the four are on ly examples out of the »great numbe r of greatand famous rivers » which is sue fromthe lake . He even al ludes to his torical eventswhich have taken place near the lake . He bl indly and uncrit ically be lieves in itsexis tence . He has never s een it, nor has anybody el s e . Ve ry l ike ly he has neverheard of i t e ithe r, and only be en impres s ed by the maps of his time . For one can

I Dr. E . W. Dahlgren has d irected my attention to P into’

s work . The title of the Frenchedition is : Les Voyages adv antv revx de Fernand Mende z Pinto , Paris MDCXLV, p . 655 and 8 1 1 .

2 Op . ci t. p . 1 55 and 1 57 .

3 The un icorn or Rhinoceros of Burma. Compare Yule ’s Marco Polo, Book II , p . 1 07 .

)UR R IVE RS . 237

more than once . From the

enters the lake on his maps .

the country , b ut has no op

lms elf of the exis tence of the

uthority of Gas taldi, and nothhim. N ow , as le Blanc saysaudacious to create the lake ,exis tence . And there i t was ,

n his beau tifu l map of 1 654 ,

trad ition . He has fou r rivers:

ads , and he has complicatedfl . be longs to the Menan fl.

,

Pegu (Caypumo). The city:ay, appears on the Caor, and

is a provi nce Vdes sa , which

XI), and has i t to the S .W .

of which the one is formeddrography is very surpris ing :»uchant, il y a umfort grandR ivie res que j

ai mis e s dansni

s map .ln his R ela ti ons do

added the name Ganges flu .

th'

erto had been called Caor.

.VIart ini and Thevenot reckon1 fromlake Chiamay. They:ls e cou ld Martin i have got theaor rive r. On ly one rive r wasthe crty of Bengala, name ly,it was the Ganges that took

re has compres s ed the wholes t in such a degree , that histhe Hwangho . And s ti ll he

s taldi’

s map of 1 550 , and sti ll betterUdeza he says in his narrative that3 p. c it. p . 43 3 .

the Atlas , p . 2 3 , the pas sage runs

multaque al ia flumina , qua mappa

GANGE S RE PRE SENTE D As ON E OF TH E FOUR R IVE RS . 237

n ot suppose that such a mis take cou ld b e committed more than once . From the

misunders tanding of one E uropean , Barros , Gas ta ld i enters the lake on hi s maps .

Nearly a hundred years later another E uropean vis its the country , b ut has no op

pOrtun ity to proceed towards the inte rior to persuade hims e lf of the exis tence of the

lake . He had no other choice b ut to trus t the great au thority of Gas taldi,and noth

ing superior had occurred in the mean time to replace him. Now , as le Blanc says ,the lake had even become famous , and i f it had been audacious to create the lake ,i t wou ld have been s t ill more audaciou s to deny its exis tence . And there i t was

,

s urrounded by its impene trable swamps and fores ts .

_

And so i t came that S ANSON D’

ABBE VILLE on his beau tifu l map of 1 654 ,

PI. XX IX ,had no reason whatever to break aga ins t the trad ition . He has fou r rivers

from Lac de Chiamay, the eas te rnmos t with two heads , and he has complicatedthe hydrography more than his predece s sors . The Av a fl. be longs to the Menan fl.

,

while the lower cours e of the Irrawaddi is cal led R . de Pegu (Caypumo). The cityGouro is we ll placed on the Ganges . A new city , Totay, appears on the Caor, and

the city Caor is cal led Caorforan . Wes t of the lake is a province Vdes s a , whichLevi , obvious ly incorrectly , identifies with Ori s sa .

I

In 1 655 Martin i cal ls the lake K ia L . (Pl . XXXI), and has it to the S .W .

of Tibe t Regnvm. He has as usual four rivers is su ing, of which the one is formedby two very long head-branches . H is text to the hydrography is very surpris ing :»La meme (d

ou le Gange tire,

sa source), vers le Couchant, il y a un fort grandlac qu i s

appelle K ia , d’

ou vient le Gange 81 les au tres R ivieres que j ’a i mis es dansla carte .

»2 THE VE NOT who has a reprint of Martin i ’s map in his R ela t i ons do

di v ers v oyages curi eux , 1 666 , PI. XLVI , has even added the name Ganges flu .

to the wes ternmos t of the four rive rs , the one which hithe rto had been called Caor.

I t is worth noting that two such able geographers as Martin i and Thevenot reckonthe Ganges amongs t the four rive rs taking the ir origin fromlake Chiamay. Theycannot have ignored the previous maps , fromwhere e ls e cou ld Martin i have got thelake at all ? S ti l l they do not care in the leas t for the Caor river. On ly one rive r wasknown to flow through Bengala and ente r the Gu lf at the c ity of Bengala, name ly,the Ganges . Thus the Caor mus t b e a mis take , and it was the Ganges that tookits ris e in the wes ternmos t part of the lake .

Whatever Martin i ’s conception may have been , he has compres s ed the wholeof As ia enormous ly, and approached the wes t to the eas t in s uch a degree , that hisS amarcanda Pars is jus t ou ts ide the great wal l and the Hwangho . And s ti ll he

I Le Nepal , I , p . 9 1 . Oris sa is we l l placed already on Gas taldi ’s map of 1 550 , and stil l betteron that of 1 56 1 . Sir Thomas Roe has his Orixa correctly, and of Udeza he says in his narrative that» ir is the utmost E ast of the Mogols Territorie beyond the Bay» . Op. c it. p . 43 3 .

2 Recue i l de Voiage s au Nord , Tome 11 1 , p . 1 63 . In the Atlas , p . 2 3 , the pas sage runs

» Inibi versus occasumingens lacus es t K ia d ictus , ex quo Ganges multaque al ia fiumina, qua mappaadscrips i , profluunt.»

238 LAGO DE CHIAMAY .

knows that S amarkand is not far fromthe Caspian S ea .

I This compres s ion of the

continent has , of cou rs e , noth ing to do with the Lago de Chiamay.

As a ru le the trave l lers avoid as much as poss ible mention ing the lake , of

which nobody except Barros had ever heard. And i f they mention i t in connectionwith the Indo-Chines e rivers i t is always w i th the s ame words as in the firs t narrative . The original work of MANDE LSLO ,

1 658 , ke eps s i lence , while the French edition , 1 72 7 , has taken the lake to grace . WALTE R SCHOUTE N , who trave l led 1 658

to 1 665, has found that Arakan , Pegu and Bengal we re cros s ed by a great numbe rof rivers . »Ganges and the great river of Aracan are , by way of branches , jo int ats everal po ints .

» He does not s ay which the latt er rive r is nor fromwhere it comes,

b ut as no other river b ut the Brahmapu tra joins the Ganges , th is river mus t b emeant . Fu rther he says : »The Dutch ships , 81 thos e of Me rchan ts fromother nati ons .

enter usually on the great river that is called Menam , 81 which goes out of a greatlake or an interior S ea cal led Chiammai , which is at more than two hundred leaguesnorthward, 81 that flows sou thwards , travers ing the K ingdomof S iamfor reachingthe s ea of this s ame name .

» Menamis the on ly river for which he takes the re

spons ib il ity of its s tarti ng fromChiamay. He has probably s een the Brahmaputrab ut has noth ing to say of its sou rce : »Afterwards we saw the great river of Calcu la,which comes also from the E as t from the Kingdomof Aracan . Here the rive rGanges begins to get more narrow.

» He is far fromclear, b ut a river join ing the

Ganges , and coming fromthe eas t can on ly b e the Brahmapu tra , the Caor river ofGas ta ldi .NICOLOSI , 1 660, fol lows S anson rather clos e ly, though he has managed to

complicate th e rivers of the Burmes e Mes opotamia a l i ttle more . In the text of thegigantic Geograp/zi o B la v i an e Vol. XI, 1 663 , an honourable attention is pa id to our

lake , b ut there is nothing new ,i t is word for word the s ame s tory which was told

by Barros .

2

I Of Samarcanda he says (Ibidemp . »E s t, cur credamhasce terras non mu ltumamariCaspio remotas ess e , ab ii sque AlexandriamBactri anamab A lexandro Magno in Bactrrs ohmcondrtamhaud longe abesse .»

2»Mais le lac de Chiamay, qui est du coste da Nord vers la Tartaric, est comme le pere de

s ix grande s rivieres , dont trois s’

unis sent avec d ’autres , font le grand fleuv e qui pas se par le mil ieudu Royaume de S ian ; 81 aus s i les trois autres rivieres s e vont rendre dans le Gol fe de Bengala. L

une

traverse le Royaume de Caor, dont e lle porte le r1 0m, 81 celuy de Comotay ; pu is ce luy de Cimte ,se font tous les E unuques qu ’

on emmene du Levant, 81 se v a de‘

charger au des sus de Chatigan , en cc

renomme’ bras du Gange , contre le mi l ieu de l’ISle de Sornagan ; l ’autre qui prend le nomde Pegu ,pource qu ’ i l partage cc Royaume en deux, pas s e par le Royaume d ’

Av a, pu is par le Pegu , 81 finale

ment fa it nu grand Go lfe pres de Tav ay ; 81 le trois ieme se rend .dans la M er 51 Martaban entre Tav ay81 Pegu ; que lques Geographes nomment cette riviere Martaban .

Au reste , passant du Cap de N igraes a Tav ay, il y a nu grand Golfe ple in de plus i eurs Is les81 bancs de sable , que fait a la facon du Gange , umautre grand fleuv e qui partage tout le Pegu ,venant du lac de Ch iamay ou Chi antay, qui est du coste du Nord , au dedans de la terre ferme ,éloign é de d eux cens l ieues . Avant qu ’

arriv er a la M er e lle court l ’espace de cent c inquante l i eues .

Quant a la riv i ere de S ian , que ceux du pays appe l lent M enam,c’

e st a dire mere des eaux, acause de la grande abondance qu’

e lle en a , e l le vient du grand lac de Ch iamay du cos té.

du Nord ,ou tro is rivi eres qu i partent da mesme lac s

as semb lent pour faire ce grand fleuv e Op . Ci t. p . 1 98 .

240 LAGO DE CHIAMAY .

from Chiamay i t s eems to b e the same which had been called Caor by Gas ta ldi.Gas taldi may there fore have known the Brahmapu tra , the famous rive r of Ass am.

Chris tian S andle r has given us a reproduction of CASSINI ’S P lan i spnoro

torres tre of 1 694 ; the firs t edition had appeared in 1 68 2 . On this map the Indu sand Ganges are s ti ll shown as me rid ional rivers coming froma range ofmoun ta insrunn ing E as t to Wes t, and north of which is P. Tibet, or Little Tibe t, while G randTibe t is eas t of the Ganges at abou t the same place whe re Chiamay lacus us ed tob e , for Cas s in i has not entered the lake at all. In th is he has be en wis e , b ut un

fortunate ly he has left on ly one of the four traditiona l Indo-Chines e rivers . This i sno improvement. The exis tence of the fou r great r ivers was the important thing ;whether they came froma lake or not was so far of s econdary importance .

I

CANTE LLI ’S map of 1 68 3 , (Pl . XXX II I), is , whateve r e ls e its value may b e , veryin teres t ing in the his tory of Lago de

'

Chiamay. He calls it Lago Chimai , and has

fou r issu ing rive rs , each wi th on ly one head. The i r names are : . Caor, Cosmin ,Chab eris , and Menan . The S inus Argaricus of Ptolemy had by Mercator beenidentified with the Gu l f of Bengal . The refore Mercator transplanted Ptolemy’s Chabe rus P luvius to the place whe re Ganges had been formerly. Now , by someextraordinary reve lation, Cante ll i has transplanted this res tless river to Lago de

Chiamay where i t occupies the place which by Gas tald i had bee n as s igned to Cai

pumo or the rive r of Martaban , S alwen . The Caor river he drops ha lfway to the

junction wi th the Gange s and places As s en , As sam, on the wes t s ide of the lake .

But there are s ti l l be tter surpris es ! The lake s eems to get unqu iet. It iswandering. I t has approached Tibe t a ve ry great s tep , nay, so much so, that i t hasRegno di Barantola to the eas t, and Las sa at a great d is tance S .E . Vis scher had

Lassa S .W. of the lake ; now it is S .E . , which means that the lake has moved wes twards as compared with its su rround ings . At some dis tance north of the lake isRedoch, our Rudok.

D is regarding the res t of the orientation , e specially the re lation of lake Chiamayto Nupal , Bov tan and As s en , I wou ld ask : wh ich other lake cou ld b e practical lys ituated in Tibe t, at a cons ide rable d is tance south of Rudok , and at a s t i l l greaterdis tance N .W . of Lhas a, than the Manas arov ar! And to make perfectly s ure of the

iden tification , the Manasarov arwas , s ince remote anti qu ity , regarded as themother of fourgreat rivers , jus t as the Lago de Chiamay was cal led a M adre dell

agi l e by old BarrosI t is hardly pos s ib le to thi nk that the qu ite new su rroundings of the lake cou ld

have been brought forward on ly by a caprice or coincidence . Cante l l i has not beeninflu enced at all by Fathe r Kircher whos e map of 1 667, (PI. XI), shows a qu ite d ifferenttype . But we shou ld not forget that Can te ll i ’s map of 1 68 3 was published in Rome ,where Kircher had ques tioned Fathe r Roth and the convertite Jos eph abou t And

I Sandler says : »In Hinterindien hat die bisher gewohnte Darstellung, als kamen d ie vier grossenPlus se des Landes aus e inem See , demChiamay lacus , e inems ehr e infachen B i lde P latz gemacht.»Reformation der Kartograph ie , p . 1 0.

MANASAROVAR CONFUSE D WITH THE KOKO° NOR . 24 1

rade ’s journeys , where Grueber had been inte rviewed, and where ce rtainly a gooddeal of information fromd ifferen t mis s ionaries was brought togethe r, of which not

an echo has been pres erved to our time .

Cante ll i ’s map shows us Lago Chimai and Kokonor 0 Mar grande s ide bys ide , as two great lakes . WITSE N , on his map of 1 68 7 , (PI. XXXIV), makes on lyone lake of both , and calls his new creation Coconor, v el Chimo i Lacus , s ive Zim.

The lake has now , in its adventurous exis tence , come so far as to s e rve as a vicarfor Koko-nor! There :is no Ts ing—hai on Wits en ’

s map . Hwangho does not comefromKoko-nor, as on Cante ll i ’s map. The source of the Hwangho is shown as a

comparative ly smal l lake. As Wits en probably found it indecent to let the r ivers ofCaor, Av a , Martaban and S iamtake the i r origi n in Koko-nor, he has pre fe rred to

deprive the lake of all sorts of e ffluents , and does not care a bit for the origin of

the Indo-Chines e rivers .

Fathe r CORONELLI ’S map of 1 695, (PI. XXXVI), shows us a new re trogrades tep. The Padre Maes tro den ies the exis tence of Koko-nor a ltoge the r, and res toresLago di Chiamay o Cunab etee to its previous dign i ty as a M adre dell

agne.

F. de WITT ’

S map , (Pl . XXXV), bears no date , b ut is bas ed on Wits en and

othe r draughtsmen . Therefore our lake is cal led »Coconor at Chiamay Lacus » .

Again , Las sa or Barantola is to the eas t of the lake . But on the shores of

the lake we find old and new names together , as Tolema , Aczu , Sochen , and othe rs,

while Radoc, our Rudok , has been removed an enormous dis tance to the N .N -W .

The Burmes e Mesopotamia i s a mixtumcompos itum, s urpas s ing eve rything we haves e en hitherto. Al l the rivers and the i r branches have names : Caor, Cosmin , Chabe ris , Pegu , Av a , and Menan . N ecb al, our Nepal, is caught as a fly in this cobwebof rivers , and has Av a as its next ne ighbour. As all the previous draughtsmen the

author of this map is sure of the exis tence of the lake . The rive rs are w e l l knownto exis t. The space be twe en the Ganges and China is narrow , and eve rything hasto b e ente rtained within this narrow s trip of land.

The s ituation is therefore the following : the Coconor, fromwhich Gruebe rwent v ié Las sa to N ecb al, gives birth to the Indo-Chines e rivers . The lake , a ls ocal led Chiamay, is , as . shal l b e shown hereafter, probably the Manasarov ar. The

Koko-nor, far away to the north , has there fore be en confus ed with the Manasa

rov ar, s ti l l further wes t. Or, thes e two lakes , which are s eparated from each othe rby mi les , have been repres ented by a third lake which does not exis t !

Compared w ith de Witt , Martin i was indeed a cleve r man . For he had a

Cinghai or Koko-nor, and a Kia Lacus or Chiamay, qu ite independent of each othe r,and s eve ra l years be fore Grueb er’s discoveries were made known to the world .

W its en s eems to b e the father of S iba lacus . Perhaps he had heard somerumou r that the Ganges came from a lake . Father Martin i had pos itive ly as se rted thatthe rive r came fromKia or Chiamay lacus . For Wits en the S iba lacus was therefore at leas t a surrogate . ISBRANTS IDE S , on his map of 1 704 , (PI. XXXVIII), goe s

3 1—1 3 1 3 8 7 I

242 LAGO DE CHIAMAY .

much fu rther. Here our lake has not only defin ite ly begun , b u t a lso partly accomplished its homeward jou rney. To evacuate and extingu ish the lake bas in altoge the rhas been too much for Ides , and his Caor Flu .

,Cosminus Flu . , Capoumo Flu .

, and

Menamis Flu . s t il l take the ir origin in the old traditional lake of Barros . But he has

changed its name into Lac Kananor et Cunab etee . The latte r name , whateve r itsderivation may b e , we have already found on Coronelli

s map , the former, Kana-nor,may s imply b e a care les s spe l l ing of Koko-nor. For there is no other Koko-nor onthe map, and Croce Lac includes all the lakes of Odon-ta la .

But where is Lago de Chiamay ? It has wandered north-wes twards and is oneof the sou rces of the Ganges , in the middle of the Himalayas . With a s l ight changeof spel l ing the name is the old one fromBarros ’ days , Lac G iamai ! Both the Indusand the Ganges have two lakes at the ir uppe r cours e , and jus t north of both riversKabu l is placed . The whole arrangement has a certa in res emblance with Kircher’smap and his Origo Gang i s ot I ndi . The two lakes of the upper Indus do not , how

ever, awake our s uspicions , for the uppe r one is the Wu lu r lake , and the lowe r oneis s ituated be low Attock. So much the more important is the fact that Ides showsthe Ganges as coming fromtwo lakes , one in the Himalaya , the othe r north of the

mountains . The one in the Himalaya , Lac G iamal , is the Manas arov ar. On Ortel ius ’ map, (Pl . XXI I), we had two copies of the lake , both cal led Chiamay, on Ides ’map we have also two copies , though on ly one of them is cal led Chiamay.

Fromwhere has Ides got his information , and what news has induced himto

undertake such a radical change ? The Jesu its in Peking ! In his narrat ive he des cribes his meeting with Fathers GRIMALD I and THOMAS PERE YRA , and he te lls us

that none les s than the famous Father Gerb illon s e rved as interpre te r at the aud ience which : Emperor Kang Hi accorded to the Rus s ian Amb as sador.

I The Jesu itFathers no doub t knew a good deal abou t Tibet, which they have not des cribed inthe ir le tters and books . On ly a few years afte r the vis it of Ides the Jesu its in Pekinggot a ve ry good native des cripti on of the two lakes , Mapama Tala i and Lanken , or

Manas arov ar and Rakas -tal, of which they we re told and firmly be l ieved that theygave b irth to the Ganges , and to no othe r r iver. Why cou ld not Gerb illon havegot some earl ie r news fromChines e s ources , and told Ides that in his Opin ion the

Chiamay lacus was the real source of the Ganges , while Barros and Martin i hadconfus ed it w ith another lake , the exis tence of which nobody so far had any opport

un ity to deny .

I In R ecue i l de Voiages au Nord , Tome VIII, p . 1 50 a Chapter has the fol lowing title : L’

Amb as sadeur vis ite les Peres Jésuites de Peking. Description de leur maison : de leur E gl ise : d ’

une colation que ces Peres donnent a I’Amb assadeur 81 a s a suite , etc. The audience begins in the fol lowingway, p . 1 3 4 : »L

un de ces trois Re ligieux etoi t Frangois , 81 s’

apeloit Pere Jean-Francois Gerb illon .

Les deux autres , dont l ’un s’

ape lloit Pére Antoine Thomas , étoient Portugais . L’

Empereur comandaau premier de ven ir me parler, lequel aus s i tot s

etant aproché de ma place , me demanda en Itali en dela part de S . M . combien de terns j 'av ois employé ‘

a ven ir de Moscou a Peking ? Ce Prince parlaensu ite un moment avec le Pere Gerb illon

LAGO DE CH IAMAY BE GIN S TO DISAPPE AR . 243

Lac Giamai on Ides ’ map is therefore placed a long way wes t of Barantolaand the res idency of Dala i Lama, as is indeed the cas e with Manasarov ar.

Next year, 1 705, a new change of the s cene takes place . De l is le , (Pl. XL),reta ins the lake and cal ls it Lac de Chaamay, b ut removes it so far eas twards as

pos s ible , obvious ly with the calcu lation that it shou ld not inte rfe re toomuch with thecomparative ly s e ttled phys ical geography of Hindu s tan and Tibe t. The lake is s ituated s tra ight sou th of Lhas a, which is also too far to the eas t . De lis le has droppedthree of the fou r rivers , Irrawaddi , S alwen and Menam,

which on Gas taldi’

s mapis sued from the lake , though the great French cartographe r has not yet dared to

leave the Irrawaddi qu i te withou t contact with Chiamay : a tribu tary to the river ofAv a, our old acqua intance Caipoumo, I s ti ll takes its origin fromthe lake . Otherwis e hehas only one river that gets eve ry drop of its wate r fromLac de Chaamay, namelyR iviere de Laqu ia , the river fromthe torra i neogn i ta of Lakhimpur. If w e compareth is modern repres en tation of the lowe r Brahmapu tra with Gas taldi ’s Caor river of1 56 1 , eve ry doubt d isappears regarding the identification of the las t-mentioned rive r.

De l is le does not accept Ides ’ Giamai lake , b ut he has another name less lakewhich he calls the sou rce of the Ganges , and in his opin ion Ides ’ Giamaimus t havebeen the same as Andrade ’s »tanque». So far he agre es with Ides . Andrade ’s»tanque» was , as shown above , not the Manasarov ar. But De l is le ’s name les s lakeand Ides ’ G iamai are both in real ity , though ignored by the draughtsmen thems e lves ,the Manas arov ar, and Lac de Chaamay, from which the Brahmapu tra takes its source ,is also the Manasarov ar. The name les s lake and Lac de Chaamay are therefore inreal ity one and the same , or, in othe r words , the Manasarov ar has given ris e to twofals e lakes on the map.

But De l is le got time to change his opin ion thoroughly. On his map of 1 72 3 ,

(Pl . XLII), the Chiamay has d isappeared withou t leaving any S ign behind . Here ,on ly ten years before d

Anv ille’

s map was publ ished , we have a map withou tthe s l ightes t trace of the Manasarov ar. For the re is no Chiamay, no Beruan , no

S iba , and no name les s lake . Only the Coconor is left, which at one t ime playedthe part of the Manas arov ar.

He rewith we have reached the las t days of Lago de Chiamay. It took of

cou rs e some years to get i t defin ite ly extingu ished fromE u ropean maps,b u t thos e

who s ti l l protected the lake we re ignoran t fabricators of maps and books . E xamplesof s uch maps are Pl . XLI I I and XLIV , whe re it appears unde r the names of

Cara Nor and Lac Mohill , 2 and PI. XLVII , which in 1 7 2 7 was publ ished by

I De l is le seems to identify Caipoumo with the Chindwin river.2 In Histoire Généalogique des Tartars , it i s said , p . 1 2 2 : »Le Lac MOhiIl dont nostre Auteurparl e en cette occas ion e st Ie mesme que nos Geographes modernes appe l lent Ie Lac Giamma’

r’

ou

Koko-N or, 81 que les Callmoucks du Tangut appe llent encore a l ’heure qu ’ i l est Cara N or: L ’

H istoire

Chinoise es t conforme a cet egard a cc que nostre Auteur avance en cet endroit, d ’autant qu ’

e lle neparle jamai s de ce Pays , que comme d ’un E stat qui a este pendant plus ieurs S i ‘ecles sujet a l ’Empirede la Chine .»

244 LAGO DE CHI AMAY.

\VICQUE FORT in a new edi tion of MANDE L SLO’

S n arrat ive . The map comes fromthe offices of PIE RR E VAN DE R AA , and is ve ry much of the same type as Jansson

s map of 1 64 1 , (PI. XXVIII). I At any rate i t is nearly a hundred years too old,

and shows what the publishers al lowed thems e lves , and what the public had to ac

cept. To the s ame clas s be longs the map reproduced as Pl . XLVIII . I t dates fromabout The Chiamay Lac. has be en captured by the Ganges rive r sys tem,

though the old Chiamay Lake , under the name of Cananor or Cunab ete , s ti l l playsits part as be ing the sou rce of the fou r Indo-Chines e rive rs . On s eriou s maps the

lake does not exis t any more , and d’

Anv ille has the sou rces of the Indo-Chines erivers in or on the borders of Tibe t .

The name of the lake has unde rgone ve ry ins ign ificant changes in the cours eof time . Barros w ri tes Chiamay and his spe l l ing has been adopted by Me rcator,He rbe rt, Jans son , Blaeu , S anson d

Ab b ev ille , V is scher, de Witt , Coronelli , and

ce rta inly by a great many others who have not been d iscu s s ed in the precedingchapters . Gas ta ldi (1 56 1 ) has Cayamay. Orte l ius writes Cayamai , though his s econdcopy is cal led Ch

'

yamai . Hondius and Hoe ius have Chiama i . Martin i w rites K ia,

ins tead of Kiamay. Cante ll i has Chima i , Wits en Chimo i , and De l is le Chaamay;names which have on ly augmented the confus ion are Coconor, Zim, Cunab etee , Kananor, Cara N or, and MOhill . D is regarding thes e late r attempts , the prototype is Chiamay, and such the name has remained from 1 550 to 1 705, or perhaps s omewhatlonger, w i th s ome un importan t variations in spel l ing. This is very natural , for thelake had been created by a misunders tanding, and Barros has heard the name as

Chiamay. As nobody eve r heard of it after his t ime , the name given by himhadto rema in intact. Nobody had any new information to bring, nor any new correetion to add.

How cou ld Barros pos itive ly as s e rt the exis tence of a lake in an unknowncoun try , whe re , as has fu lly been proved by the explorat ion of a later time , no lakeexis ts at all ? This ques tion is imposs ible to answe r with any degree of certa inty .

Natives on the banks of Irrawaddi , Salwen , and Menamwou ld not pre tend that the irrivers come from a great in land s ea . The information is probably not firs t hand.

It may have come fromHindus l iving on the banks of the Brahmaputra , who havebeen convinced that the ir river came froma lake , which gave ris e to three othe rgreat rive rs . As the Indus and the Ganges eve r s ince Ptolemy ’s time were supposedto b e known as n ot coming froma lake , and as the Satlej was almos t unknown ,

I The lake i s mentioned in Wicquefort’

s ed ition , Vol. I , p . 54 , where the province of Udes sa

i s placed between the provinces of Kanduana , Patna , Jesual, and M ewat, and the lake Chi amay.

P . 290 he says of the Kingdomof. Benga l : »Le Gange la traverse presque tout, 81 11 e st encore arrose’par les rivieres du Guenga , da C3 01 , 81 du Cosmin .» And of the river of Menan we learn ,

p . 3 05»el le est s i longue , que jusqu ’

ici on n’

a pas encore pfimonter jusqu ’a sa source» .

2 This map which was kindly shown to me at the Royal L ibrary of Berlin , has the title : As iaRecenti ss ima Delineatio , qua Status e t Imperia totius Orientis unacumOrientalibus Indiis exhib entur

Authore Jo. Bapt. Homanno Norimberga .

LAGO DE CH IAMAY IS IDE NTICAL WITH MANASAROVAR . 245

Barros has mis unders tood his informant and suppos ed that the Indo-Chines e riverswere meant. But the informan t has meant the fou r rivers which were suppos ed to

take the ir ris e fromthe sacred lake of Brahma, and one of which was the Caor river.It is easy to unde rs tand that in a time when even Tibe t was unknown , nobody wou ldb e able to locate this lake , even approximate ly at its right place . We have als o to

remember the form and s i tuation given to Ind ia , the Indus and the Ganges . It wou ldhave s e emed absurd to think that the Indus and Ganges came fromthe same lakeas the Caor rive r. For then the Caor wou ld have be en forced tomake an enormousbend through the D is e rto de Camv l and Mon te N augracot and to enter

'

into ve ry int imate contact with the Chines e rive rs . The more natura l it appeared that the IndoChines e rivers came fromthe lake that Barros or his informants had heard of.

Barros may have got news of a journey fromInd ia to China, or v i ce v ers a ,

by some now unknown native or E uropean trave l le r, or perhaps abou t s everal journeysundertaken on that l ine . H e has be en told that the road cros s es fou r great and almos tparalle l rivers . At the s ame time Barros may have heard that a lake , the sou rce of

four great rivers , exis ted somewhere to the north . He cou ld not know that this lakewas the Manasarov ar, nor that the rivers were the Indus , S atlej , Map

-chu and Brahmapu tra . So , ve ry natural ly , he identified the rivers w ith thos e cros s ed on the road b etween India and China. This is mos t probably the real and fundamenta l cau s e of

the trans formation of the real Manasarov ar into the imaginary Chiamay.

It is su rpris ing that the e lemen ta ry laws of phys ical geography , of eros ion and

b ifu rcation , cou ld b e so complete ly unknown that eve ry geographe r born s ince 1 550

accepted w ithou t the s l ightes t attempt of oppos ition such a mons tros ity as fou r rive rs ,in fiv e branches , flowing ou t in almos t the same d irection fromone and the samelake . This is exactly what innume rable natives of India and Tibe t be l ieved to takeplace , and s t ill be l ieve ,—ln the cas e of the Manasarov ar! But that E uropeans cou ldb e taken in so far is indeed surpris ing

Manasarov ar became known to the Jes u i ts in Peking in the beginn ing of the

e ighteen th century . In 1 73 3 d’

Anv ille made i t known to E u rope . Before that timethe lake had been unknown ,—and s ti l l i t had exis ted for 1 8 3 years on nearly all

maps , b ut at a wrong place , feed ing wrong rive rs , and pos s es s ing a fals e name .

We mus t leave it to philologis ts to s e ttle the ques tion of the derivation of the

name Chiamay. In Hindu mythology the Himalaya mounta ins have many d iffe rentnames , as Himachala, Himadri , Himavat, and others . Alb eruni calls themH imav ant .Moorcroft in 1 8 1 2 wri tes Himachal , Francis Hami lton , 1 8 1 9, Himadra and Himaliya,James Fras e r, 1 8 20, Himala, and so forth , innumerable variations on ly 1 00 years ago .

How eas ily cou ld not the word have been misunders tood 3 60 years ago ! And Chiamay is not s o ve ry far fromHimalay. The informant may not have known the nameof the lake and s imply cal led it the Himalaya lake , and indeed the Manas arov ar is

s till regarded as one of the mos t s acred »tirthas » amongs t the Himalaya Mountains .

CHA PTE R XXVI.

THE MAPS O F STRAH LENBERG AN D RE NAT .

Among the Swedish officers who , after the battle of Poltava were s ent aspri soners to Tobolsk and other places in S iberia , was Lieu tenant-Colone l PHILIP PJOHANN VON STR AHLE NBE RG, born in S tralsund 1 676 . Afte r a s tay of 1 3 years inS iberia , he re tu rn ed to Sweden in 1 7 2 3 , and died there in 1 747. N o more than hiscomrades was he broken down by the long and ted ious years of S ibe rian capt ivity

,

b ut u s ed his time to work ou t a great map of Northern and Cen tral As ia , a workin which he had great as s is tance fromCapta in JOH . ANTON MATERN and Lieute

nan t-Colone l PE TTE R SCHON STROM , both Swedes .

It wou ld take us too far to pay du e attention to S trahlenbe rg’s wonderfu l l ife .

In th is connection w e are on ly conce rned w ith his work. I have a lreadymentioned I

the important part played by S ch ’

OnstrOmand S trahlenbe rg in connection wi th BE NTINCK

s H i s toi re Genea log i gn e dos Tartars , where the au thor of the explanatory notesis S trahlenberg . Thes e notes w ere afte rwards publ ished as a s eparate article : R ela

ti on de la Grando Tartari e, D res s e’

e s n r l es .Mémoi res Origi na ux dos S n e’

doi s

P ri s on n i ers en S ioe'

ri e, pendan t la Gn erro do l a S nodo av ec la R nss i e.

2 In d’

An

ville ’s Atlas , Haag 1 73 7 , an article accompan ies the maps : D es cr ipt ion do l a B on

cé a r i e P a r n n ofi ci er S u édoi s qu i a fa i t quelqu e s e'

j onr dans co P ay s . Accord ingto AUG . STRINDBE RG this officer was aga in S trahlenberg, who d ictated the articleto a friend , probably Bentinck .

3

Then fol lows the mos t important of S trahlenberg’s works , which , according to

S trindberg , was firs t published in Le ipz ic 1 730, unde r the title H i s tor i o der R ei s en

i n R u s s la nd , S i b i r i en a nd dor g ross en Tartarey . The edition at my d isposal waspri nted at S tockholmthe s ame year, and was trans lated into s eve ral other languages .

4

I See above p . 2 2 7 , note .

2 Recueil de Voyages au N ord . Tome X . AmsterdamMDCCXXXVIII , p . 1 e t seq.

3 August Strindberg : Philipp Johann v on Strahlenberg och hans karta Ofver A s ien . Sv enska

sallskapet for Antropologi och Geografi, Geografiska s ektionens T idskrift, 1 8 7 9. Band I , N zo 6 , p . 1

et seq. And Kulturhi s toriska studi er, by the same , Stockholm 1 8 8 1 , p . 1 8 e t seq.

4 The fu l l title runs : Das Nord und Os tliche The i l v on E uropa und A s ia, In so weit solchesDas gantze Russ ische Re ich mit S iberien und der grossen Tatarey in s ich b egreiflet, In e iner Historisch

248 TH E MAPS OF STRAHLE NBERG AND RE NAT.

Comparing the news he got fromhis friends the Tatars in '

S ib eria with the

w ri tings of the clas s ics , S trahlenberg some times arrived at rathe r adventurous con

clus ions the audacity of which we gladly forgive , remembe ring that he travel led and

wrote exactly 200 yea rs ago, when hardly anything was known of Tibe t. He saysthat a lready in Alexande r’s time the name Catba i was used by the Scythians

,for

the Sophitic kingdommentioned by Curt ius is the same as S trabo’

s Catha a . »The

SOphiti c kingdom, howeve r, is nothing e ls e than Thib eth and the country of the

Tangu ts whe re the wis e and high Prie s t Dala i Lama l ives wi th his wou ld-b e holyLamas or Pries ts . For with the appe l lation of Sophis ts the Greeks have indicatedthos e , who gi ve thems e lves up to wisdom And as the Mungals and Calmuckss tate , the i r Dalai Lama or so cal led Pries te r John has already s ince some thousandyears had his res idency in the country of the Tanguhts , he mu s t have res ided therea lready at and before the time of A lexande r. And that the Gre eks have had somereason to cal l thes e Lamas or Pries ts in the country of Thib eth orMats chin Sophis ts ,is confirmed by the manners and cus toms thes e Lama s ti l l have in the i r re ligion,name ly the important part that is played in it by sorceri es , prophezyings and chiromancy.» I

He compares the me ridional Imaus with that of Me rcator, and both with actualthings , and agrees that there is really a me ridional range which goes out northwardsfromthe southern Imau s or »as it now is ca lled Imus -tag».

The very greates t part of S trahlenberg’s work is devoted to the descriptionof the Russ ian empire , its name , its s ituation , boundaries and cl imate , its provinces ,its dynas ties , more e specially the Romanow dynas ty, the government of Tsar Pete r,the ti tles Tsar and Grand Duke , re ligions , income , army, the aris tocracy, the greates tci ties , etc.

, and finally he has a l is t of mine rals , plan ts , curios ities , an tiqu i ties ,manufactures , fabrics , »und de rgle ichen D ingen» .

In this l is t he occas ional ly men tions things or ques tions touching upon Centra lAs ia and Tibe t. Under the word »co tton» he te l ls u s that the Mohammedans l ivi ngin S iberi a never us e s tu ffs manufactured by heathens , b ut on ly such which are manufactured in cities with a Mohammedan popu lation , as Turphan , Chamill, Jerken ,

Kutzai , Axu , Caschkar, etc. The word Be lgian which is found on some maps of thetime is , accord ing to S trahlenberg, derived from »bill» or »be ll» mean ing mountain ,

2

while the name Imaus is pronounced Imus -tag, or Imus sahr by the Tatars , for »mus »or »maus > means ice , and »tag» mountain ; thu s the meaning of Imaus is ice or snowmountain , and Imaus is a corruption of the native word . Caucasus he derives fromthe Pe rs ian Cuho or Coho (Kuh), mountain .

The musk of Tibe t he found to b e three times as s trong in sme l l as the

S ibe rian , the price of which was only one fourth of the Tibetan . R egarding the latte rhe quotes Martin i .

I Op . cit. p . 40 .

2 Bel means indeed »pass» in Jagatai-turki .

248 THE MAPS OF STRAHLE NBE RG AND RE NAT.

Comparing the news he got fromhis fri ends the Tatarswritings of the class ics , S trahlenbe rg some times arrived at rathee lus ions the audacity of which we gladly forgive , remembe ring tf

wrote exactly 200 years ago, when hardly anything was knownthat a lready in Alexander’s time the name Catba i was us ed by

the Sophitic kingdommentioned by Curtius is the s ame as S tral

Sophitic kingdom,howeve r, is nothing e ls e than Thib eth and

Tangu ts where the wis e and high Pries t Dala i Lama l ives withLamas or Pries ts . For with the appe l lation of Sophis ts the Gr

thos e , who give thems e lves up to wis dom And as the Mur

s tate , the i r Dala i Lama or so cal led Pries ter John has a lready s f

years had his re s idency in the country of the Tanguhts , he mus talready at and before the time of A lexande r. And that the Grereason to cal l thes e Lamas or Pries ts in the country of Thib eth o

is confirmed by the manners and cus toms thes e Lama s ti l l haname ly the important part that is played in i t by sorceries , propmancy.» I

He compares the me rid ional Imaus with that of Mercator,things , and agrees that the re is real ly a merid iona l range whichfromthe sou thern Imau s or »as it now is cal led Imus -tag» .

The very greates t part of S trahlenbe rg’s work i s devoterof the Rus s ian empire , its name , its s i tuation ,

boundarie s and cli

its dynas ti es , more especial ly the Romanow dynas ty , the gov ernrthe ti tles Tsar and Grand Duke , re l igions , income , army , the arisci tie s , etc.

, and finally he has a l is t of mine ra ls , plan ts , curios iti e

factures , fabri cs , »und dergle ichen D ingen» .

In this l is t he occas ionally mentions things or ques tions tor

As ia and Tibet. Unde r the word »co tton» he te lls u s that the l\in S ibe ria neve r us e s tuffs manufactured by heathens , b ut on ly S tfactured in c i ties with a Mohammedan popu lation , as Turpha

Kutzai , Axu , Cas chkar, etc. The word Be lgian which is foundtime is , according to S trahlenbe rg, de rived from»bi ll» or »be l l» 1

whi le the name Imaus is pronounced Imus -tag, or Imu s sahr by tl~

or »mau s » means ice , and »tag» moun ta in ; thus the mean ing of Inu

moun tain , and Imaus is a corrup tion of the native word . Cauca .

the Pe rs ian Cubo or Coho (Kuh), mountain .

The musk of Tibe t he found to b e three time s as s tru

S ibe rian, the price of which was only one fourth of the Tibetan .F

he quotes Martin i .I Op . cit. p . 40 .

2 Bel means indeed »pass» in Jagatai-turki .

250 THE MAPS OF STRAH LENBE RG 'AND RE NAT.

there i s practically no S ign of the Tarimbas in on Wits en’

s map , whereas the fundamental featu res of this regi on , e specially its wes te rn part, on S trahlenbe rg’s map verymuch approach real ity and surpas s eve ryth ing execu ted so far. He is evenmuchsupe rior to De l is le ’s maps of 1 705 and 1 7 2 3 , (Pl . XL and XLII), for whileDe l is le knows only the Khotan-darya and makes i t flow the wrong way, S trahlenbergknows the fol low ing rive rs : Axu (Aksu-darya), Ters ik (Taushkan-darya), Kas chkar fl .

(Kashgar-darya),Meles cha (Yarkand-darya), Chote en (Khotan-darya), Orankas ch (Yurunkash), Karakaisch (Kara-kash), Tarimo (Tarim), Kari a fl . (Ke riya-darya), Tschikman(Cherchen-darya P), Kipzak and Chaidu fl . (Khaidu-gol), or, in a word, all

the great rive rs of the Tarimbas in , though , of cours e , much was left to b e des iredregarding the i r relative s ituation . He knows all the ci ties of the same regi on and

some vi llages as we ll , and places themas a ru le in correct re lation to each other,for ins tance , not Kashgar sou th of Yarkand as De l is le had done . Or, to mention on ly the mos t important : Kaschkar, Ierken , Sarikol , which has fal len too far

eas t, Kargalagga (Karga lik), Choteen , Karakaisch, Gurumkas ch, Gumma (Gume),Mods che (Muj i), Karia, Gans (Gas s-kul, which he be l ieves is a city), Tarim, markedas a city, Karas chai (Kara-shahr), Chialis ch al z U luk Iuldus , the Cialis of Goés and

olde r maps , Kitz ik Iuldus (Li ttle Yu ldus), Axu , Baij , Arwat , Kutschai (Kuche , wronglyplaced in re lation to Aksu), Uts chferrnent (Uch—turfan), and , furthe r eas t,many placessuch as Turpan , Chami l , Lukz in (Lukshin , which Grum-Gshimailo found to b e

s ituated be low the su rface of the s ea), and many othe rs . Cu rious ly enough he hasno Lop and no D es ertumLop.

I Three years later d’Anv ille completes S trahlenbe rg’s hydrography eas twards by making the Yarkand-darya fal l into Lop Nor.

Of the greates t interes t to us is the way in which S trahlenbe rg has su rroundedBucharea Minor or E as te rn Turkes tan with moun tain ranges , and nobody w i ll contradi ct my saying that he has known both the Tian-shan and the Kwen—lun . The

eas te rn half of the latte r he repres en ts as a boundary be tween the Tibe tan highlandsand E as tern Tu rkes tan , as indeed is the cas e . Regarding its wes tern half he commi ts exactly the same mis take as Ptolemy, namely , to be l ieve that the s lopes fallingdown to the sou thern margin of E as te rn Turkes tan and the s lopes fa ll ing down to

the plains of Hindu s tan be longed to one and the same range , one s ingle ranges eparating India fromthe Tarimbas in .

S trahlenberg’s eas tern half of the Kwen-lun i s the refore correct, while the

w es tern half includes both Kwen-lun and Himalaya as we l l as everything e ls e s i tuatedbe twe en them, as for ins tance , Kara-kommand Transhimalaya . This wes te rn halfhe cal ls Mus Tagk al ias Imaus Mons .

H is northern range , the Tian-shan , S trahlenberg cal ls Musart , and

, qu ite cor

rectly, he has Uramtza (Urumchi), and Barskol Lac. (Bar—kul) north of this range and

I Where the Lop desert ought to be , he has a Desertum,

the name of which i s il legi ble on

the origina l copy at my di sposal . On PI. XLIX it reads cltus . On Strindberg’s reproduction in our

Geographical Journal 1 8 79, quoted above , it read s Sul tus .

STRAHLENBERG’

S EXCELLENT OROGRAPHY. 251

a t its very foot. The inte rruption caused by the rive rs Karausun and Ulanusun,

the Black and the Red Wate r, does not agree with real i ty. S trahlenberg has praetically dis cove red the exis tence of Tian-shan , though it s eems ve ry un like ly tha t hee v e r proceeded so far sou th pe rsonally . Of this fact ALE XANDER VON HUMBOLDTs ays : »S trahlenberg has the merit to b e the firs t who has repres ented the Tian-

shana s a specia l , independent range , withou t know ing the volcan ic activ ity of the

Humboldt is les s fortunate in his criticism of S trahlenberg’s nomenclatu re .

I cannot he lp quoting two or three pas sages by Humboldt in his own words . H e

s ays : »S trahlenbe rg hat den Thian-shan auf der Karte , we lche s e inem We rkebe igegeben is t , un ter demunrichtigen Namen Musart so darges te l lt, das s er z iem

l ich le icht e rkannt werden kann ; er nennt den Bolor Mus tag und s e tzt dabe i nocho l imParopamisus , was auf e ine S .

—N .

-Ke tte sehr s chlecht pas s t . D ie BenennungenMusart und Mus tag s ind Corruptionen des tartaris chen Wortes Muz

-tagh .»2 In con

nection with S trahlenbe rg ’s me rit of repres enting the Tian-shan on his map Humboldt says : »E r gi eb t ihm den s ehr unbes timmten Namen Mousart : der

,we i l der

Bolor mit demallgeme inen , n ichts indiv idualis irenden , nur S chnee andeutenden NamenMus tag be legt wurde , noch e in Jahrhunde rt lang zu e ine r i rrigen Dars te llung und

a lbernen , sprachwidrigen Nomenclatur der Geb irgsre ihen nordlich v omHimalaya Anlas s gegeben hat , Me ridian und Paral le l Ke tten mit e inande r verwechs e lnd.» 3

What Humboldt s ays of such a name as Mus tag is qu ite right ; it gives ris eto misunde rs tandings and shou ld disappear fromthe maps . In our own days one

s ee s fromtime to time the Wes tern Kara-korumcal led Mus tag fromthe pas s of the

s ame name . I u s ed to cal l the eas te rnmos t range of the Pami r Mus-tag,

4 thoughCo lone l Burrard prefe rs the name Kashgar Range , a proposal I mos t w i ll inglyaccept. 5

S trahlenbe rg’s Musart , on the other hand, is no corruption , as Humboldt b eliev es . He says the correct word is Mous sour tagh orMous sar tagh of which Musart

s hou ld b e the corruption . On the las t edi tion of S l z'

e/er’

s H a n d-A tla s , N o. 62 , the

name is spe lt exactly in the s ame way as on S trahlenbe rg’s map 1 83 years ago ,

name ly Musart . The on ly mis take , ve ry excusable indeed, which S trahlenberg committed in this cas e was to trans fe r a name , which be longs only to a very importantpas s s ituated jus t eas t of Khan-tengri

,to the whole sys tem, the Chines e name of which

was unknown to the Swedish office r. Anothe r mis take was to be l ieve that Tians han cons is ted of on ly one s ingle range , b ut this was also Humboldt’s Opin ion 6

Kosmos . E ntwurf e iner phys ischen Weltbe schre ibung. 1 8 58 , Vol. IV, p . 607 .

2 Central-A s ien , Ub ers etz . Dr. W. Mahlmann . Berl in 1 844 . E d I , p . 3 7 2 .

3 Kosmos , IV, 1 858 , p . 607 . Compare also Fragmens A s iatiques , Tome Premier, Paris 1 83 1 ,p . 48 and 60 .

4 See my l ittle map in Ze itschrift der Gesellsch . f. E rdkunde zu Berl in Bd . XXIX, 1 894, Ta

fel 8 . Through As ia, Vol. I , p . 2 73 and 3 1 1 , Vol. II, p . 705.

5 A Sketch , e tc . Part I . Calcutta 1 90 7 .

6 Fragmens A s iatique s , map .

252 TH E MAPS OF STRAHLE NBE RG AND RE NAT.

and E uropean draughtsmen have made the same mis take regarding Transhimalaya, untilI re tu rned frommy las t jou rney .

S trahlenberg’

s Mus tag is no corrupt ion e ither, for the word can as we l l b es pe l t Mus tag as Muz

-tagh . S i z’

eler’

s H and-A tlas has Mus-tag as S trahlenberg, and

not Muz-tagh as Humboldt . »Olirn ParOpamisus » i s unfortunate . But when Hum

boldt says that S trahlenberg has cal led the Bolor Mus tag, he s eems to have overlooked that S trahlenberg, wes t of his Mus tag has a M . Belur, al ias Bu lu t

,i . e . Mon

tes tenebros i , and “ bet ween both he has a Plan iti es Pame r, or »the Plain of Pamir» .

S trahlenbe rg ’s Mons Be lur the refore corresponds to the range I cal led Sarik-kolRange in 1 894 , a name that has a lso been adopted by Colone l Burrard under theformof Sarikol Range . S trahlenberg ’s repres entation is therefore in pe rfect accordancewi th the actual geography, his northe rn Mus Tag be ing the s ame as the KashgarRange wi th Mu s

-tag-ata , and his M . Be lur be ing the Sarikol Range , and his Plan i

t ies Pame r be ing the Sarikol and Tagdum-bas h val ley. He even knows the TerekDaban , a name which he gives to a range ins tead of a pas s . The refore S trahlenbe rg ’s map is , in this region , far supe rior to Humboldt’s

,which was published

more than a 1 00 years later, for while S trahlenberg has two paral le l me ridionalranges wes t of E as te rn Turkes tan , Humboldt has on ly one , which he calls Bolor, andwhich cros s es the wes te rn part of Tian-shan in a mos t extraordinary way at righ tangles . And this wes te rn part Humboldt calls Mouztagh , which he s uppos es to b e

an improvement upon S trahlenbe rg’s Musart . Humboldt ’s Bolor shou ld soon b e givenup, while S trahlenberg’s Mus tag (Mus

-tag-ata) and Musart and his two me ridional

ranges , which we re 1 1 4 years olde r, shou ld l ive for eve r on the maps of As ia .

S trahlenbe rg was of opin ion that the Hindu-kush , which he calls Hendukes ch ,

was the immed iate wes te rn con tinuation of Mu s Tagk , the sou thern Mus-tag or Imaus

Mons . What e ls e cou ld he do as he did not know the s eve ral other ranges northof Himalaya ! And we shou ld not forge t that the connection between the Hindukush and one or s everal of the ranges to the eas t is not defin ite ly cleared up yet .

The Swedish office r made his obs erv ations and combinations fromS iberia , and got

his information the re , and the fu rther sou th , the greate r, therefore , are his e rrors . H e

makes the same mis takes from th e north that the ancients made “fromthe sou th .

They mee t halfway and arrive , in some respects at leas t, at the s ame conclus ions .

So, for ins tance , the Indus and the Ganges have the i r sources in the same range as

the Keriya—da rya . On S trahlenberg’s map we find, howeve r, an indication of a doublerange even in the wes t . The Khotan-darya and Yu run-kas h of the map are repre

s ented as coming froma short range s i tuated north of the Imau s .

The uppe r Gange s and its lake is taken fromDel is le . E as t of i t we find Tibetet Tangu tia Pars al ias Lasa R egnum, while Lac. Sofing is fromMartini . S trahlenbe rg has two copies of the Koko-

nor, which he cal ls Zinhac Lac: , and Koko Nor.

S trahlenbe rg ’s map is , unde r all ci rcums tances , well'

worthy of be ing drawnfromthe obl ivion into which it has been kept for such a long time .

254 THE MAPS OF STRAHLE NBE RG AN D RE NAT .

s aved, and to the e fforts of S trindberg that it was d is cove red 1 50 years afte r itsappearance . Fromthe original it was copied by I . A . Benzels ti erna in 1 73 8 .

Then follows in the text a chapter abou t the geographical know ledge of the

coun tries shown on Renat’

s map , fromwhich we learn that s evera l Rus s ians hadtrave l led there in the beginning of the e ighteenth cen tury withou t draw ing anymaps .

It was impos s ible to make a map fromthe ir des criptions . Therefore the Rus s ianssay of Renat

s mapL

honneur d’

av oir dres s é ce tte carte , en s e bas ant tan t s ur s es propres ob

s erv ations , qu e s ur des interrogations , appartient donc sans con tredi t a Remat . Il

fut le premie r a nous donner un tableau graphiqu e déta il lé de la Dzoungarie et du

Turkes tan Oriental a ins i qu e des parti es des pays contigus , tableau qu i , s’

il n’

av ait

pas été égaré , au rai t été le me i lleur , du moins quant a la Dzoungarie , avant l ’edit ion de la carte de l’As ie Centrale de Klaproth et des cartes récentes russ es , fondéessur des recherches géographiques immédiates , e ffectuées pendant le s iecle dern ier dela part de la Chine et pendant le s iecle cou rant de la part de la Ru s s ie Maisdans différen ts déta i ls , par exemple , dans les contours des lacs Balkhach et Is sykkou l, la carte de R enat s e rapproche plus des cartes mode rnes , que ce l le de Klaproth .

This is ve ry high and ve ry jus t pra is e of this old map . But it remainedforgotten and neve r came to play the importan t part i t wou ld have done if i t hadbeen known and explo i ted by late r draughtsmen .

Then the Ru s s ian article goes ove r to a very thorough and ve ry able discus s ion of the map its e lf. I t contains 250 geographical names and legends . Out

of this numbe r more than half is s ti l l to b e found on modern maps .

We are not concerned with the greates t part of the map , b ut as its sou thernhalf embraces regions of As ia , which so late as in 1 875 we re practically unknown ,

and where Prshev alskiy made some of his mos t bri ll iant discove ries , and as thes eregi ons have been deal t with in the preceding chapters , I think it worth while to

cons ide r this part of the map. Of E as tern Turkes tan the Rus s ian text s ays : 2Le Turkes tan Orienta l es t dés igné s ur la carte de Renat sous le nomde

Kottoner,Kottons , sans dou te d’apres la vi l le de Khotan , qu i jadis é ta it la plus im

portante de ce pays Q uan t a la difference que R enat fa it dans l’orthographiedu peuple Kottoner, Kottons , e t de la ville

,de Choton , Khotan , e lle es t de peu

d’

importance . Meme des s avants géographes écri ven t différemmen t Khoton , Khotan ,

Koton , Kotan , etc.,e t Renat , qu i n

av ait pas une ins truction bien sol ide , a pu remplacer une le ttre par une autre et meme me ttre un double t sous au tres combina is ons étymologiques .

The word Kottonner which Renat has written ove r a great part of E as ternTu rkes tan has nothing to do with the name of the city Choton , or Khotan . Bythe expres s ion Kotton was meant the inhabitan ts of towns , and especially me rchan ts

1 Op,cit. p . 9 et s eq .

2 Op . cit. p . 2 6 .

RE NAT’

s KNOWLEDGE OF EASTERN TURKE STAN . 255

from Bokhara , in the same manner as nowadays me rchants from Rus s ian Turkes tanwho l ive and trade in the cities of E as te rn Tu rkes tan are called Andishanliks , fromthe city of Andishan . S trahlenberg s eems to re fe r the word to the Mongol wordkhoto, town , as for ins tance in Koko—khoto, the Green City , as Kwe i-hwa-chong inS han-s i is cal led by the Mongol ians .

The formKottonner is Swedish plural , and s tands for the E nglish Kottonians .

The same word enters in the Swedish title of the map : Carla af S ongarsé é

K a/mucé z’t, fiwar u nder K at/07272137 726 fié'

rer, or Map of the Dsungarian Kalmuckyto which the Kotton ians be long .

2

The cities which have given ris e to this name are Kutza (Kuche), Acksu (Aksu),Utz (Uch-turfan), Kas char (Kashgar), Iercken (Yarkand), Choton (Khotan), and Kiria(Keriya). The hydrographical arrangemen t of the Tarimbas in is perfectly exce l lent andve ry nearly the same as on our own maps .

3 I t is far more correct than the repre

s entation on No . 44 b in S ti eler’

s H and-A tla s for 1 875, which shows the s tandpoin tof the knowledge before Prshev alskiy

s s econd jou rney . I will on ly add that on Renat’smap the cities Yarkand, Khotan , and Ke riya have each been placed one s tep too fareas twards . For the river wes t of Iercken is Yarkand-darya , and at the place whe reIercken is entered, Choton ought to have been s i tuated . The Tarimis s imply cal ledDaria . Chaidu , or Khaidu-gol is there , and Schisken-toroij is Cherchen-darya .

South of the latte r is the country of the Koss iauts and the Tangu ts . Two

important geographical names are ente red here : Gas , and Schidamor Tsa idam . Alake in the same val ley as that fromwhich Che rchen-darya comes , may pos s ibly b eAyak-kum-kol. Remat has even heard of the gold in thes e regions , though he hasmisunders tood the information he got , and places the gold-mines near Kiria

,where

Under the word Koton , Choton , Strahlenberg says : Die Ost-Tatern nennen die Handels-LeuteBurger und Bucharen in ihren Stadten Choton ; Daher die Nahmen der Stadte in der Ost-Tatarey al s

K ara-koton, i . e . schwartz, Koko-koton, i . e . gri in , und Naun-koton dasjen ige Koton , das an demN annS trohm l iegt. Wei ] d iese nun vie l mit BaumwollenemZeug hande ln , s ol lte n icht etwa der N ahmeCatun hiermit V erwandtschafft haben ? We lches n icht behaupte , sondern nur e ine Meynung i st. Op .supra cit. p . 3 8 9 .

Pallas expla ins the Dsungarian name of a certain s icknes s , Ge ile Seuche , with the fol lowingwords : D ie Soongaren nannten se lbige Chotton-jarra (bucharische oder Stadt-pochen) we i l s ie durchd ie Bucharen unter s ie gekommen war. P . S . Pallas : Sammlungen histori scher Nachrichten uber dieMongol ischen Volkerschaften . E rster The i l . Frankfurt und Le ipzig 1 7 79, p . 24 2 . In h is chapterabout the admin istration etc. amongst the Kalmuks , Pallas write s : Die Dorfschaften oder Nachbarschaften

,we lche immer auf e inemP latz b eysammen ihre Fi lzhti tten (Garr) aufschlagen , und ihre H eerden

zu sammen we iden , pflegen aus etwan zehn oder zwolf Fami lien zu bestehen, und werden Chotton oder

Chottun genannt, we lches etwan e inen Hof oder Kre i s s bedeutet, we i l s ie s ich aml ieb sten in d ieserF igur lagern : Der altes te imChottun , demdie Au fs icht der Beywohner aufgetragen ist, wird ChottuniAcha genannt und hat e ine Art v on Ansehen unter den ihmuntergegeb enen Geme inen . Ibidemp . 2 85. And finally , in an Order of Galdan Chuntaiji : Auch al le Rechtshande l zwischen demOelot

schen und Chottonschen Volk (st '

adtischen Bucharen) sol len genau geprt'i ft werden Ibidemp . 3 2 2 .

2 The scale of the map is in Swedish mi les , and the explanations of s igns refer only to towns ,the Regent’s tents , and (ord inary) tents .

3 The text of the Geogr. Society in Petersburg runs : Le bas in de la riviere Tarim, qui formele Turkestan Oriental , est representé sur la carte de Remat presque comme sur les cartes de nos temps .

Op . cit, p . 28 .

256 THE MAPS OF STRAHLE NBE RG AND RE NAT.

he has the legend : Har finns gu ld , or He re gold is found ; the origina l informantmeant that many of the gold-d iggers came fromKiria .

Proce eding wes twards we find, sou th of Choton , the Garriker, or Garrikians ,

which I cannot iden tify . Then follow the Kesmirer or Kashmi rians , sou th ofMus tack

or Mus-tag. The rive r which on Renat

s map is the rive r of Yarkand b ut in real ityKhotan-darya , comes from a lake Charchol, obvious ly Kara-kul, so much the moreas it is s ituated in the country of the S iara-Kols er or S iara-Kols ians , the people of .

S arik-kol or Sarikol . In Badakshan again gold is to b e found .

Tian-shan is we ll drawn , and it is cal led Musar ola , or Musart ula with two

Turk i and one Mongol ian word . The correct trans lation , Isb erg, Ice-mounta in , is

ente red . Is s ik-kul is remarkably we l l placed and its form ve ry good. It is cal ledToskal Noor and has the legend : Finnes jarn sand , or Iron s and to b e found .

S trahlenbe rg has the name Iss ech kul for this lake , not nearly so we ll placed byhim. Renat knows even Son-kul and the Narin river

,and has themwonde rfu lly

we l l placed .

The interior of E as te rn Turkes tan is , on . Renat’

s map , shown as rich in fores ts ,though in real ity the fores ts grow only in narrow s trips a long the rive rs . On ly sou thof the eas te rn part of the Tarimis the groundmarked as occupied with s and des erts .

l

The Tarimempties its e lf into the Lop-nor, which on Renat

s map is cal led L '

ap.

In the lake the re are some is lands . The eas tern end of the lake has been rathe runknown to Remat and his informan ts , for a rive r, Bolans iu , which mus t b e the E u

lunds ir, enters this part of the lake . Leaving this inaccuracy alone , one cannot he lpnoticing that the orientation of Lop

-nor on Renat

s map, in re lation to the Tarim,

agrees ve ry we l l w i th my theories abou t the s i tuation of the old, Chines e Lop-nor.

2

For the Tarimgoes s traight eas twards through the b ed of the Kurruk-darya to the

northern , now dried-up lake , and the Che rchen-darya enters the south-wes te rn cornerof the lake , withou t join ing the Tarimat all. The fol lowing is a pas sage fromone

of my narratives : 3 »The old Chiefta in of Abdal , Kunchekkan Beg, a friend of Przhe

v alsky and also my es pecia l friend , is e ighty years old . Both his fathe r, j ehan Beg ,

and his grandfathe r, N umet Beg l ived to b e n ine ty years of age . Kunchekkan Begtold me , that his grandfathe r l ived bes ide a large lake north of the exis ting Lop-norof Ps zhev alsky, and that , where the latte r now is , there was at that time nothingb ut the sandy des e rt . The firs t formation of the sou the rn Lop-

nor dates fromthe

year when N ume t Beg was twen ty-fiv e years old,in cons equence of the Tarims eek

i ng a new channe l for i ts e lf, and the lake bes ide which he dwe l t, and in which hisforefathe rs had fished , dried up . It was he (Numet Beg) who founded Abdal , and

1 In this respect the d ifferent copies of Renat’

s map vary . On the Mogol original, there are noforests and no s and marked , showi ng that Remat has added thes e detai ls to his copy fromverbalin formation . On Benzelstierna

s copy there are no trees , b ut on ly sand deserts .

2 I have d is cuss ed this problemin several of my previous books , more especially in Sc ientificResults , V ol. I I. . Compare also Dr. A lbert H errmann ’

s important d iscus s ion , Op . supra ci t.

3 Through A s ia, V ol . II , p . 8 78 .

258 THE MAPS OF STRAHLE NBE RG AND RE NAT .

the map hims e lf. ’ This s tatement cannot b e taken l i te rally. The Mongol ianru le r 2 can at the mos t have orde red the map to b e made . By whom? It s eems veryunl ike ly that any Mongol ians , E as t or Wes t, wou ld have been able to draw a map,the whole appearance of which i s pe rfectly E uropean . The Chines e have always beenve ry cleve r s urveyors , but the map (Pl . L) has not at all the ordinary Chines ehabitus , i t has not even been influenced in the leas t by Chines e me thods . And as

I s a id before the Mongol original is exactly the same as Pl . L, except the names

which are in Mongol w riting , and a few detai ls added on Renat’

s copy . The Chines ealways us e to show the mountains in horizontal pe rspective

,not in ve rtica l projection .

On De l is le ’

s , S trahlenbe rg ’s and d’

Anv i lle’

s maps the mountains are s ti l l shown inhorizontal projection ; b ut on Renat

s they are drawn as s een froma bird ’

s eye view ,

which gi ves the map a much more mode rn s tamp than heaps of E uropean maps of

a much later date .

Al l we can do is to confes s our ignorance abou t the real origin of this mos tvaluable map , which had been bu ried and forgotten 1 50 years when S trindbe rg discove red it .

3 Anyhow the Rus s ian Geographical Society may fee l sati sfied that RematNagra smfi annotationer giorde wid en kort conference emellan Lieut: Re h at, somifrfin Cal

mukie t fi terkomit, och B isk : Hogw . H zr Dr. E ric Benzelius 1 73 8 . In the E ngestromska Saml . B II ,2 , 2 5, Royal Library , Stockholm,

where Dr. E . W. Dahlgren has kindly d irected my attention to thisdocument.

2 R egarding the title Kontaij i or Khun taij i Pallas has the fol lowing passage : D ie machtigs tenunter den kalmt’i ckischen und mongol ischen Ftirs ten haben s ich sonst the ils v on ihremge istli chen Oberhaupt demDala i—Lama , theils v on ihren machtigen Nachbarn demRus s ischen und Chines ischen Beherrscher den Chanen-T itu l b eylegen las sen , und den T ite l Chuntaidshi (Schwanenfii rs t) welcher vie lenSoongarischen , Choschoti schen und Mongol ischen Ftirs ten eigen gewe sen i st, unter d ie s en Horden , da

s ie noch frey und v on ke iner au swartigen we ltl ichen Macht abhangig waren , al le in v omDalai-Lama er

the ilt, und gab , den Rang fiber d ie gemeinen Fti rs ten und d ie R echtmas s igke it der Macht, we lche s ichso lche ti ber die minder machtige anmas ten . Op . c i t. p. 2 75. Accord ing to a letter which FatherGaub il s ent fromPeking to A . M . Del is le in 1 755 or 1 756 , the power of the Kungtaij i seems not tohave lasted long. H e says : » l l y a que lque s année s que le s Rus s es allerent sur la riviere I li , avec lapermis s ion des princes Kalmuks , qui y éta ient alors ; ils y ont ob serv e sans doute dans le pays entrel’

Irtyche e t Ili ; ici , nous n’

avons eu aucune conna issance de ces observa tions La pu issance desKa lmuks Koungtaichi , dont le principal campement éta it sur la riviere I l i , es t comme détruite . Lesprinces éleuths de cc pays se sont fa it la guerre , les princes mahométans de Yarkand , de K achgar e td'

Aksou s e s ont rendus inde’pendants che z eux. Le prince de Tourphan s'

est donne a l ’empereur avecque lque autre . P lus ieurs ont péri ; d ’autres s e sont soumis a la Chine , d '

auti es se sont retires che z lesR usses . L

empere ur de la Chine s’

est rendu mai tre du pays des Kalmuks Tchongkar ; i i a un corpsd’

armec sur les riviere s d ’

Il i et de Borotala .» Lettres du P . Gaub i l, adressées a plus ieurs savants deParis . Nouveau Journal A s iatique , Tome X. Paris 1 8 3 2 , p . 4 1 1 , 4 1 2 .

3 In the last moment be fore the printing of this vo lume I get fromthe Library of the Univers ity of Uppsala the copy of a letter, dated StockholmApri l 25, 1 743 , written and s igned by JohanGustafw R emat, and adres sed to the Librarian Andr. N orrelius . H aving spoken of the Songars or

Dsungarians , R enat continue s : >>Amongst the se Songars I have been l iving, and have got the greatestmap (of the four which he had , A pril 1 3 , pres ented to the Library of Uppsala) fromthe Chines e ,when they invaded the territory of the Kalmuks at Barcoll (Barkul) or Turphan , a town which also,on some maps , is s ituated on the ir frontier. H alf of this (map) I have copied careful ly with Latinletters , and intended to trans late the other hal f as we l l : b ut the fine writing and my weak eyes haveprevented me fromit. However, I be l ieve that it has been copied in Russ ia , for, 5 years ago , I del ivered it to Profes sor de Lie ll (J . N . De lis le) of the Academy in S t. Petersburg, who returned it tome , promis ing that I shou ld get a printed one later on , b ut so far I have not rece ived it. It has

RE N AT’

S ADVE NTURE S IN KALMUK CAPTIVITY . 2 59

is not a rival of Prshev alskiy in the dis covery of Lop-nor. For if Renat eve r was at

Lop-nor, i t was at the old , Chines e lake , while Prshev alskiy dis cove red the s outhe rn ,

Kara-koshun , 1 1 0 years afte r the jou rneys of d’

E spinha, d’

Arocha and Hal lers te in .

E ven before the journeys of thes e mis s ionaries , DE GUIGNE S had , fromChines es ources , told E urope some important truths abou t the geography of Lop

-nor and

its rivers .

‘ He speaks of two or three rive rs ris ing in the »mountains to the wes t»L ’

une , soit que le terre in n’

a it pas as s ez de pente , soit que les sables accumulés en empéchent le cours , s

arréte s e pe rd au mi l ieu du des s e rt . Les deux au tresvont plus loin , apres s

etre réun ies , e lles s e jettent dans umgrand lac appe lleLop , qui e st s i tué dans la partie la plu s bas s e de tout ce grand terre in .

The firs t rive r is Khotan-darya,the two others Yarkand and Aksu-darya . That

Kashgar-darya i s not meant can b e s een by the fol lowing words : 2 Dans les memesmontagnes Tgung

-l ing vers Yerken i l sort un grand fleuv e qu i v a s e rendre dansce lu i qu i part d’Aksou , l’un l ’autre s e j e ttent dans le lac de Lop .

A new glimps e of the wonde rfu l adventurous life of Lieu tenant J . G . Renat

i s given us in a ve ry able article by S IGR ID LE IJONHUFVUD abou t his w i fe BRIG ITTA S CHE RZE NFE LDT

,who was born in southe rn Sweden in 1 684 , and thus , 26

years younger than her third husband , Renat .

3 She accompan ied her firs t husbandduri ng the Ru s s ian campaign of Charles XII , and became a prisone r. At the fortressof Yamishoff

,where her s econd husband was killed, she came into Kalmuk cap tivity .

Here she made the acquaintance of Renat who became her third husband . She

accompan ied him back to Sweden and died in S tockholmin 1 736 . In her »pe rsonal ia>> 4 we are told that she got a ve ry high and honoured pos ition at the cou rt ofthe Kalmuk princes s es . When one of themwas to get married to a grandson of

Aynka Khan , Mrs Renat was s ent to Litt le Bokharia to b uy all neces sary equ ipment for the wedding, which was to take place three years late r . She pass ed two

years in Ge rken or Yarkand,and as she , a woman

,cou ld trave l abou t in the then

unknown heart of As ia,i t wou ld not b e s urpris ing if her husband

,who had a s ti l l

been copied by a Kalmuk f romt/ze C/zimse orig ina l , and th is as we l l as the other map has , on myrequest, been given to me by the Songar R egent when I left, and this (map) he has h imsel f made of

his country . Though I should have been able to enter many names of mountains and rivers whichare to b e found there , I have , however, preferred to de l iver it such as I rece ived it fromthe R egentpersonally , as I have found, on other maps , names of both places and rivers , of which I , duringmy 1 7 years l iving there , have heard and s een noth ing. ) Upp sala Un iv ers itets B ib lioteks handskri ftssaml ing , G : 2 0 : a . (Bre f ti l l och frfi n G . Benzelstierna). The original map i s thus Chinese and copied by a K almuk . But probably the Jesu its in Peking have given it i ts E uropean habitus , which ita lready had when it came into the hands of Profes sor De l is le . With the topography of the mapR emat had nothing at all to do . Therefore it does not give us any indication about the journeys of

Remat. Histo ire générale des Huns , Paris 1 756 . Tome II , Introduction , p . V .

2 Ibidemp . XXXIX .

3 Tvé'

t kv innooden fra°.n karolinska tiden . Nord isk T idskrift, Stockholm1 907 , 2 :dra haftet, p . 45et seq .

Personalier ofwer Lieutenantens af Kong] . Mai j ztz arti llerie Regemente Herr Johan Gus tav Renats i Lifztiden K ier E lske lige Hus Fru och Maka

,Fru Brigitta Schersenfeldt.

260 THE MAPS OF STRAHLE NBE RG AND RE NAT.

more important and honou red pos ition at the cou rt , enjoyed the same opportun ity andl ibe rty. This becomes the more l ike ly as not on ly b ut all the other townsof Little Bokharia are entered on the map . And there fore also the conclus ion of

the Ru s s ian Geographical Society : »Il es t ev ident qu e Remat n ’

a jamais vis ite ces l ieux» ,los es ve ry much of its s trength . So is also the cas e w ith the Russ ian doubts abouthis achievemen ts amongs t the Kalmuks , for in his lett er of res ignati on fromthe

mi l i tary s ervi ce which he s ent in 1 739 to King Fredrik of Sweden ,he says that he

has taught 200 Kalmuks the art of arti lle ry, and that he made a campaign with theKalmuks agains t China .

I The above-mentioned »persona l ia» te l l us , that Renat , whenhe re turned home , was arrested in Mos cow. The Rus s ians »had got a great hatredto him, on account of the as s is tance he had given the i r enemy the Kalmuk to put

thems e lves in a pos s ib il ity to defend thems e lves agains t fu tu re hos ti le attacks » . Throughthe e fforts of the Swedish legation at S t Pe tersburg he

_regained his l iberty .

Finally I wil l say on ly a few words abou t a ques tion which has something to

do w ith Renat ’s map , name ly abou t the wild came l . To the authors of the Chines eannals the wild came l was not the same s trange r as to E uropean naturalis ts . I giveonly one quotation , fromthe Thang epoch :

S e lon le Ths e fou chin kou e i , la cinqu i eme année Kha i -youan , a la s ix i emelune

,il vint des ambas sadeu rs du royaume de Iu-thian , les que ls offri rent quatre che

vaux propres au jeu de paume , un chameau sauvage,au pied léger comme le vent ,

et un na ou renard .

2

In his geography RITTE R mentions the wild came l in s eve ral places , for ins tance :3»ImS ijden v on Tu rfan ist ‘

ub erall kahle S teppe (Kobi), wo wi lde Kameele und wildePfe rde in zahlre ichen Heerden umherz iehen .» S ti l l zoologis ts doubted the exis tence of

this an imal , unti l Prshev alskiy, on his memorable jou rney 1 8 76—77 , bought the skins

of four wild came ls fromnative hunters . He s ays hims e lf : »Ich brauche n icht zu s agen ,w ie froh ich war, Fe lle v on demThiere erworben zu haben, dess en s chon MarcoPolo erwahnt

,das aber b is j etzt ke in E ur0p ’

aer ges ehen hat .»4 And with fu l l rightDr . Brehm adds : »Prshewalskiy hat imvorletzten Jahrzehnte nachgewies en , das s inlnneras ien wirklich wilde , n icht blos s ve rwildert e Trampe ltiere (Came lus b actrianusferus) zahlre ich v orkommen .» 5

I t is therefore interes ting to note , that nearly 200 years ago the exis tence of

the wild came l was known to a Swedish office r. For on Renat’

s map we find,jus t

eas t of Gas chon Noor , a lake now bette r known as E b i-nor, a fores t tract , Carangoj ,under which the legend is written : I Inur Skog finns willa Kameler, or, wi th the

Loc . c i t. p . 51 .

2 Abe l-Rémusat : Histoire de la ville de Khotan , Pari s , 1 8 20, p . 7 1 .

3 Die E rdkunde v on A s ien , VI I, p . 43 5.

Przhewalsky’

s Re ise an den Lob -N or und A ltyn-Tag Pet. Mit. E rganzungsh . Nr. 53 ,p . 1 9 . Marco Polo does not mention the w ild came l .5 Brehm’

s T ierleben . Saugetiere Dritter Band . Le ipzig und Wien 1 89 1 , p . 1 51 . Compare ,concern ing the l ife of the wi ld came l , my Central A s ia and T ibet.

CHA PTE R XXV I I .

TH E J ESU IT CARTOGRAPHERS IN CH INA, AN D D ’

ANV I LLE ’SMAP OF THE LAKES AND TH E IR SURROUND INGS .

R ichthofen has given us the mos t admi rable de scription of the Je s u i t miss ionaries ’ cartographical work in China , and I cannot do be tte r than give a short extract fromhis s tandard work . In 1 58 2 the Ital ian , MATTE O RI CCI inaugurated the

his tory of the Jes u i t mis s ion in China ; he vis i ted many place s and founded s tationsbe fore he finally arrived in Peking in 1 60 1 . S ee ing the great s ucce s s of cathol icism,

in 1 6 1 7 , the Empe ror ordered all the Je su i ts to leave his domin ions . They,how

eve r, took refuge and we re we l l rece ived in the hou s es of conve rts , and afte r s omeyears of inactivi ty we re again al lowed to s tay . N ew men came out fromE urope ,amongs t others the famou s Tyrolian MARTIN MARTINI , whomR ichthofen regardsas the founde r of the geographical know ledge of the inte rior of China , the Ge rmans ,FABE R , and ADAM SCH A AL , and the Be lgian , VE RB IE ST . Du ring the dynas tic fightsthey took the s ide of the Manchu s , and the i r influence with the firs t Manchu emperorsthe refore became rathe r powe rfu l . After the Domin icans and Francis cans had arrivedin China

,in 1 630, a time of d isputes began , b u t the Je s u i ts kept the i r ground

through great d ifficu lties , and in 1 692 Kang Hi is sued the important decre e whichopened the gates to cathol icism.

Only very s cienti fical ly trained members of the orde r were s ent to China . In

1 688 the great Frenchmen , GE RB ILLON and BOUVE T arrived , and Gerb illon accompan ied the Empe ror on all his journeys . In the mids t of new inte rior s truggle swhich nearly des troyed the mis s ionarie s ’ own prospects of s ucces s , the Je su i ts won

the i r greates t s cien tific triumphs , and the firs t s eventeen years of the 1 8 th centurymark a rea l epoch in the i r his tory . Du ring the s e years they accomplished the

work of the i r famous map of China .

When the great Emperor Kang Hi had e ffected the conques t of China whichhad been begun by his predeces sors , he decided , in 1 708 , to have a map drawn of

his countrie s , and ordered the Je su i t mis s ionarie s , who we re learned in mathematicalR ICHTHOFE N : China I

,p . 653 e t seq .

EMPE ROR KANG H I’

S GRE AT MAP or CH INA . 263

s ciences and whos e abilities he admi red , to unde rtake the work . They began roundthe great wall . The Fathe rs Bouve t, REGIS , and JARTOUX de te rmined the s ituations ; the two las t-mentioned continued the work throughout the year 1 708 .

I In

January 1 709 they re turned to Peking with a map 1 5 fee t long which pleas ed the

Emperor s o much , that he des ired to have such maps of the whole of his Empire .

In May 1 709, Régis , Jartoux and FRIDE LLI drew the map of Manchu ria and the

province of Chi-Ii which kept thembusy during 1 7 1 0 . In 1 7 1 1 Régis and CAR

DOSO made the map of Shan-tung. Late r on Régis was as s is ted by MAILLA and

HE NDE RE R for three more province s . In 1 7 1 5 Régis made the map ofYunnan , andafte rwards , togethe r with Fride lli , he carried out the mapping of two othe r province s .

REMUSAT says that Régis did mos t of the work hims e lf. 2 Two memoirs of Régisare published in DU HALDE , name ly abou t Korea and Tibe t. But much of the

othe r mate rial in du Halde is taken fromRégis , a lthough his name is not mentioned. Du Halde te lls us how the mis s ionarie s proceeded

when surveying the greatE mpire , »which unti l now has been so l i ttle known, even to the mos t ab le geographers » . But when the Je su i ts began the ir topographical work they found earlie rChines e maps in eve ry provi nce . No such maps of Tibe t , howeve r, we re known to

exis t, though DE LLA PE NNA mentions some sort of a map in Pota la , which he ca llsa Ta éu la C/zorograpfiz

'm.3

Tibe t was not included'

in the work of the Je s u i ts . But Empe ror Kang Hihad already s ent a specia l embas sy to Tibe t . Although the chie f object of thismis s ion was a po litical one , name ly, to try and gain the Tibe tan princes for the new

Chine s e policy , the ambas sador, who s tayed two years in Tibe t , had be en orde redto have a map drawn of »all the countrie s which are immed iate ly submi tted to the

Grand Lama» .

In 1 7 1 1 the map was ready and pres ented to Fathe r Régis , who was to workit out in the s ame formas the maps which had be en made of the Chine s e province s ;b ut afte r having examined the mate rials and put some ques tions to the s urveyors ,the Fathe r found that he had no fixed points to s tart from , and that dis tances we res imply taken from native information . He the re fore decl ine d to us e the mate rialbrought to him .

4 But it was obvious from the mate ria l that the country was muchI A BE L-REMUSAT : Nouveaux Mé langes A s iatique s . Paris 1 8 29 . Tome I I , p . 2 3 5 et seq .

2e t quand on songe qu

une entreprise géographique , plus vaste qu’aucune de ce lles qu ’

on

a jamais tentées en E urope , fut achev e’

e par quelque s re l igi eux en huit anne’ es , on me peut s ’empécherd'

admirer cet efi’

et d’

un zele qui n ’

etait pas un iquement ce lui de la sc ience , quoiqu ’

il en servit s i bienles intéréts . Ibid . p . 2 3 7 . »Jean Bap tiste Régis , jésu ite franca is , mis s ionaire a la Chine , et habilegéographe , do it etre compte parmi les sav ans re l igieux qui ont fait le plus d ’

honneur a cette mis s ionde la Chine , s i ferti le en hommes d istingués Tout ce qu ’

on sa it de lu i s e borne a ce qu ’ i l a fa itde glorieux pour les sciences et d

honorab le pour son pays .» H e l ived 3 6 years in Peking, and d ied ,in 1 759, 7 1 years old . In the h istory of exploration in A s ia R égis ’ name has a very prominent place .

3 R ITTE R ; Die E rdkunde v on A s ien. Bd I I , p . 466 .

4 Le P . J . B . DU HALDE : Description Géographique , Historique , Chronologique , Politique et

Phys ique de I’

Empire de la Ch ine et de la Tartaric Chinoise . Paris 1 7 70 . Vol. IV ,p . 459 et s eq ,

264 TH E JE SUIT CARTOGRAPH ERS IN CH INA , AND D’

AN V ILLE’

s MAP or THE LAKE S E TC .

greate r and riche r in important geographical matte rs than had been be lieved be fore .

There fore the Empe ror decided to have a re l iable mapmade . He orde red t wo Lamaswho had learnt geome try and arithme tic in a mathematical academy which was unde rthe prote cti on of his thi rd son , to make the map of the whole country fromS i-n ingto Lhas a and thence to the sou rce of the Ganges and also orde red themto bringhim some water from that rive r. The work was carried ou t punctual ly . Not unti lthe year 1 7 1 7 , howeve r, was the new map de live red to the Emperor, who s ent it tothe mis s ionary geographe rs . They found i t by far superior to, andmore re l iable thanthe map of 1 7 1 1 .

I

By means of us ing and checking all re liable iti ne rarie s fromChina to Tibe tthe mis s ionaries did the i r bes t to cons truct a map of the whole of Tibe t, »the

exactitude of which was worthy of the attention of the public», as nowhe re e ls e we reany de ta i ls regarding towns , mountains , and rive rs to b e found. The map of Tibe twas added to the great map of the whole of China which in 1 7 1 8 was pre s ented tothe Empe ror .

This map of China was engraved in Peking and the Jes u it Fathe rs s ent a copy

to Paris , which was pres ented to the King and kept in his private l ibrary at V er

sai lles . D u Halde had a copy of it made in Paris and s ent to d ’

Anv ille , askinghim to reduce i t and prepare it for publication . The Paris copy was , howeve r, a

ve ry incomple te extract of the origi nal maps , and the geographica l names had beentrans lated in Peking by somebody who did not know Chine se s ufficiently.

2

KLAPROTH te lls us , that i t is wrong to be l ieve that the Jesu it map of Tibe thad no othe r foundation than the work of the two Lamas s ent by Kang Hi to thatcountry . The mate rial they brought to Peking was control led and improved by othe rpe rsons whomKang Hi had s ent for the purpos e of s tudying Tibe t, as we ll as bythe i tine raries of the Manchurian armie s which , during his re ign , pacified Tibe t . 3

From du Halde we learn that the Lama surveyors of Kang Hi had somepolitical difficu lties to ove rcome du ring the latte r part of the i r work . He says thatthe grandson of COUCHI HAN , TALAI HA N , was attacked by the great TSE VANGRABTAN . Talai han had some men , and Ts

ev ang Rab tan’

s general onlyor But Talai ban was de feated and killed and the country of

The following pas sage is important, and does not throw an altogether favourable l ight overthe way in which the material was used : »E lle ne leur parut pas néanmoins tout-a-fait exempte dedéfauts : ma is par respect pour l ’ecole d ’

ou ces Lamas étoient sortis,ils s e conten terent pour lors de

corriger les plus sens ibles , qui auroient choqué les yeux de I’Empereur. IIs lai ss erent meme Lasaau des sus du 3 c

c degré de latitude , ou le s Lamas I’av oient mis , ayant plus d ’

égard a la mesure ac

tue lle dont ces Lamas s’

étoient s ervis, qu

a l’

ob serv ation astronomique .» Ibid . p . 460.

» Il ne faut donc pas attribuer a d ’Anv ill e les erreurs que présentent le s cartes d ’

apres ces

calques , et insérées dans la description de la Chine , du P. du Halde . Quand aux feui l les particuliére sdu Tubet que renferme cet ouvrage , d ’

An v ille les a fait copier as sez exactement d ’

apres les calquesdont on vient de parler; mais i l s

est permis , dans la carte générale dans laque lle il le s a réunis ,plus ieurs changemens qui n’

ont pas tous été heureux Mémoires re latifs a l’As ie .

'

Par M . J .KLAPROTH . Paris 1 8 2 8 . Tome I II , p . 3 7 2 .

3 KLAPROTH , loc . c it . p . 3 89.

266 IPPOLITO DE SIDER I .It is pre tty ce rtain that the Lamas have vis i ted the Manas arov ar and Rakas -tal

and the s urroundings of Kai las , as this part of the country is particu larly we l lmappedand , so far as I can s e e , much be tt er than the re s t . But they had no time to de

cide the ques tion of the s ource of the Gange s . Here they have re l ied upon information s upplied by natives .

It is d ifficu l t to say whe the r the su rveying Lamas or the Jesu i t mis s ionarie sor d

Anv ille and du Halde are re spons ible for the mis take abou t the source of the

capture of Lhasa in 1 7 20 . H aenisch has trans lated no less than 3 7 documents on this war. The manwho held the power over T ibet in h is hand was Latsang Khan in Dam, and Tsev ang Rahtan thoughtthe favourable moment had come to beat Lats ang with ass istance of hi s enemies in T ibet. As the dis

tance fromthe D sungarian frontiers to T ibet was long, Tsev ang had to act s i lently and quickly . TheDsungarian general Ts eringdondob (Ss ering Donduk) cros sed the passes of the J ingkurting mounta insand s urpris ed the guardians of Latsang. Only seven months after the Dsungarian invas ion in T ibet,Emperor Kang H i got the first news of what had taken place . Latsang sent word to the Emperor toget h is he lp . The Emperor at once dec ided to proceed and fight the D sungars . In the beginn ing of

1 7 1 8 one of the two imperia l army corps advanced v ia Litang and Batang towards T ibet. But on lyin 1 7 2 0 was the offens ive taken . August 2 2 the joint Chin ese army stood at the gates of Lhasa . The

c ity was conquered without d iffi culty . In a proclamation the T ibetans were told that the ChineseEmperor had s aved themfromthe Dsungarian yoke . Tseringdondob and his army s eemto have hada very hard time in T ibet. As they had no commun ication with the ir own country they had to l iveon robbery , and suffered horribly fromhunger and cold . At the arrival of the Chinese troops theyhad to return northwards , probably persecuted and constantly attacked by the natives .

The following passage on Tseringdondob’

s hardships , tran s lated by H aenisch , i s interesting»Als Ts eringdondob , we lchen Tsewang Rab tan ausges andt hat, mit Mann heranmarsch ierte ,um T ibet zu erobern

, ging ihmLatsang mi t se inemH eere entgegen und nahmden Kampf mi t ihmau f, wobei e s auf be iden Se iten weder S ieg noch N iederlage gab . D ie Leute aus Tseringdondob

s

Heer, we lches v on we it her durch den Schnee gezogen kam,waren verfroren und verhungert, ihre

Pferde und Kame le ge fallen und krepiert, so dass alle Leute zu Fuss marschieren mussten und s ichauf demWege mi t demF le isch v on M ens ch en und Hunden nahrten . V on den Mann , unterwe lchen s ich mehr Uriyanghai

-Leute (e ine Landschaft nordl. Kobdo) als Oleten be fanden , kamen nur

noch M ann an . D ie ub rigen waren zu Tode ers chopft und hatten n icht Schritt halten konnen .Hieraufhin gab der Ka iser fo lgenden E rlass an den Staatsrat Nun befand s ich noch dazuTseringdondob

s Heer b ei se iner Ankunft auf demG ipfe l der E rscht’j pfung. Und wenn es jetzt erst insGefecht gekommen ist, und man den Abgang an Gefal len en und an Krankhe it Gestorbenen in Be

tracht z ieht, ni ochten schwerl ich noch Mann da se in . Ob er damit noch imstande ware , Latsang

s Burg zu nehmen ? Ne in ! Sondern ,wenn Tseringdondob imS inne hat, das Land mit Waffen

gewalt zu erobern , so i st s e ine Stre itmacht n icht mehr ausre ichend . Und wenn er den Ruckzugin die H e imat antreten wil l , so wird er auch keinen Weg zur Rettung mehr finden . Da es aber n ichtausgeschlossen ist , clas s er in se inen Bedrangni s raubend und plundernd umherz ieht, s o is t es fur uns

unbedingt notig die umfassendsten V ors ichtsmassregeln zu trefi'

en.»

Further reports reached the Emperor : »We iter batten die Gross lamas T solcimund Tsangb u

R amjamb a berichtet : Als wir v on G’

angdisa nach Lasa he imkehrten , teilte uns Latsang mit, das sTsewang R ahtan demTs eringdondob e in H eer v on tiber Mann gegeben habe , mit we lchemdies er ti ber das J ingkurting-G ebirge here ingebrochen s ei und s ich in Bes itz der gefahrlichen Pas sagenund strategi sch wichtigen Platze gesetzt habe . M it uns kampfend habe er mitten in der Nacht diePasse ti b ers tiegen und das Land Damgenommen , wo er zur Ze it noch s itze .»

This also appears fromthe fol lowing passage in du H alde : »Ils s e con tenterent de faire une

carte de la source du Gange , 8: des pays qu’ i l enve loppe , sur le rapport des Lamas qu i demeurentdans les Pagodes vois ins , sur les mémoires qu ’

i ls trouv erent a Lasa chez le grand Lama. Mais ilsmanquerent a umpoint e s s entie l , qui éta it de prendre hauteur aupres du mont Kentais se , ou autrement Kan te chan , comme les nomment les Chinois , lesque l s étendent ce noma toute la chaine demontagnes qui v a a I

Occident ; ou du moins dans le Pagode oil ils s’

arréterent, pour s ’informer ducours du Gange qu i s ort a l 'Occiden t de cette montagne , tandis que le Tsanpou , qu

ils ont suivi etmesure, vi ent a I’orient vers Lasa .» Du H alde Ioc . cit. p . 465.

MANASAROVAR AS SOURCE OF THE GANGE S . 267

Ganges , a ques tion which afte rwards was so eage rly dis cus s ed in the A s i a ti c R e

s ean /zes . I cannot be lieve , howeve r, that the s urv eying Lamas or the Lamas of thegompas round the Manasarov ar made the mis take , as even at the pre s ent day all

Tibe tans round the lake s know that the rive r which is s ued fromRakas -tal was the

Satlej , so much the more , as both Tibe tans and Chines e know and u s e the Tibe tanname of Satlej even for that part of the rive r which is s i tuated eas t of the Manas arov ar.

It mus t have been ow ing to the mis inte rpre tation of the E uropeans who obta ined the

mate rial ; or the Lama s urveyors may have been compe l led to s ave thems e lves beforethey had time to ente r upon that que s ti on and to follow the Satlej to the we s t .Howeve r, as the mis s ionarie s were convinced that the latitude of Kai las was un

ce rtain ,they addres s ed thems e lve s to d

Anv ille and found himon ly too will ing to

combine the materia l fromthe l i ttle known parts of As ia which had to b e improvedin the way that had the greate s t probabi lity of be ing correct. The nece s sary material for making corrections was regarded to b e more eas i ly obta inable in E uropethan in China . The mos t surpris ing thing of d

Anv i lle’

s map is , that both the

uppe r Satlej and the upper Indus are dragged into the sys temof the Gange s . And

s till the uppe r Indus , e xcept the ve ry sources of the two branche s , is tole rably correct .We find there : Tchas ir tong (Tashi-gang), Tchics e (Tiks e ), Latae (Leh, Ladak) andPitoe (Spittok). But fromPitoe the river turns s traight southwards to the Gange s .

On the uppe r Satlej the hydrography is correct andmuch be tte r than on manymaps,

published du ring recent years . We find the river is su ing fromMapama Ta lai andente ring Lanken L and continu ing fromthere as the Lanctchou R (Langchen

-kamba)to Cogue (Koge), Dshaprong (Tsaparang) and Tchoumourt i (Chumurt i), and then ,

again , sudden ly turn ing south , jo in ing the Indus and forming the Ganges .

As to the s ources of the Ts angpo , this part i s much more re l iable than eventhe map of NAIN S ING . RYDE R

s map was the firs t to prove how near the Lamaswere here to the rea lity . To this and othe r parts of the Lama map we shall haveto return late r on .

O the rw ise there is not much abou t Tibe t in du Halde ’s compilation . He

has some thing to say of the his tory and re l igion of the people . He knows the

route fromBenga l to Tibe t which had be en trave l led for s everal years . He thinksthat a good deal of the gold brought to China come s fromTibe tan rive rs .

In this chapte r we have to pay special attention to D’

ANV ILLE’

s deta i ledmapof s outh-wes te rn Tibe t , of which Pl . LI is a reproduction . We have s een fromRégis ’ report in du Halde that, regarding the Ganges , the s urveyors contentedthems e lve s w ith the information they obtained fromthe Lamas in the gompas roundthe Manasarov ar. But as to the lakes and the ir rive rs one has only to compared’

Anv ille’

s map w ith the l i tt le map of the lake s which I published in 1 909 .

I Southeas t of L . Mapama we find the mountains of Lantchia Kepou , a region which is

1»Trans-Himalaya» , at the end of Vol. II .

268 TH E JE SUIT CARTOGRAPHE RS I N CH INA , AND D’

ANV 1LLE’

S MAP or TH E LAKE S E TC .

s ti l l called Langchen-kabab ’ or the »mouth of the e lephant ) , that is to s ay, the sourceof the Satlej . Froma l ittle lake at the foot of thes e mounta ins a rive r flows to the

N .\V. and emp tie s i ts e l f in the Manas arov ar. This rive r is the Tage-tsangpoof my map , which comes froma glacie r, Gang-lung, and not froma lake

,though

a mora ine - lake may eas i ly b e s ituated in fron t of the s nou t of the glacie r . A l ittlefarthe r north d ’

Anv ille has anothe r name le s s rive r, which i s the same as the Samots angpo of modern maps . This ri ver also comes froma smal l lake which no longerexis ts . Clos e to the eas t of i t we find L . Conghe which is Gunchu-ts o ofmode rn maps .

The rive r flowing down fromthe valley of Pachung is also to b e found on d’

An

ville ’

s map , wi thou t a name . But the mounta in at its head, which I pas s ed in1 908 , is on d

Anv ille’

s map ente red unde r the name M . Patchon , although it hasbe en placed north of Ka i las ins tead of e as t . The re are two lagunmat the northernshore of L . Mapama , which also exis ted in 1 907 and 1 908 . To the wes te rn lake le tt wo smal l rive rs flow from the mountains eas t of Kai las , which is correct, and

so is the l ittle rive r which ente rs L . Lanken fromthe north . The gene ral arrangement of the mounta ins is pe rfectly correct, proving that the Lamas have unders toodthat the t wo lakes are s i tuated be twe en t wo great sys tems , the Himalaya and the

Transhimalaya .

Farthe r, we find the channe l going out of L . Mapama and ente ring L . Lankenand final ly the Lanctchou R .

, or as now , Langchen-kamba , is su ing fromL . Lanken .

D’

Anv i lle’

s map prove s , that in 1 7 1 7 the channe l bet ween the lakes was fil led w ithwate r , and that the Satlej flowed fromRakas -tal . It also proves that the Samots angpo did not is s ue fromthe Gunchu-tso , which would have been a phys ical impos s ibi l i ty , at leas t during the latte r part of the pos t-glacial pe riod.

On account of the original maps having been copied, the maps of d’

Anv i lle

conta in e rrors probably not to b e found in the origina ls . Klaproth, and wi thhim R i tte r, regre t that d’

Anv ille did not ke ep to the de ta i led she e ts of Tibetwhen joining them to his general map of that country which has there fore los t agood deal in exactne s s . The hydrography of our lakes as we l l as the i r s u rroundingsare the same on both maps , b ut the latit ude has be en changed from29

’/z°on the

specia l map to 3 1’/z

°

on the gene ral one , the correct s i tuation be ing ju s t be tweenthe t wo .

2

Langchen-kamba i s the name of the river.

2 In later chapters we shall often have to refer to th is important map , as we l l as to the gen

era l map of T ib et, published by d ’Anv ille in 1 73 3 . As the latter i s more important in connectionwith the h istory of exploration in the country north of the Tsangpo , I have pre ferred to insert it inthe part of this work wh ich deal s w ith Transhimalaya , Vol . III .

The reader w i ll find that I have h itherto cons idered the world ’

s know ledge of Tibet as a whole .

In the fol lowing chapters I amgoing to separate the two lakes and the sources of the great rivers and

deal w ith thems eparate ly , leaving the Transhimalaya and Kara-korumsystems for spec ial cons ideration .

E xceptions to this rule w i l l b e found only occas ionally .

270 IPPOLITO DE SIDE R I .forteres s e ou rés ide le Roi , s e fa it entre de s montagne s , qu i s ont une vra ie image de la tristes s e Le haut le b as des montagne s sont également imprat icab le s ; on e s t obl igé demarcher ami-cote , le chemin y e s t d ’

ordinaire s i étro it, qu ’

a pe ine y trou ve-t-on as s e z d ’

e spacepour pas s er le p ied.

He give s a ve ry vivid de scription of this horrible journey and the dange rousroads , which make. him tremble when he thinks of them. Then he speaks of the

ve ry hard cl imate and of his s u ffe ring eye s .

N ous fime s a p ied Ie voyage de Kaschemire a Ladak , qu i dura q uarante jours , nous

n’

y arri v ames que le 25 Ju in . Ce Royaume du s econd Th ibet, commence , comme je l’ai de’ ja

remarqué , au mont Kante l , 8: s’

étend du s e ptentr ion vers le levant. II a na s eu l Gh iampo ou

Ro i ab s o lu ; ce lu i qu i regne aujourd ’hu i s e nomme N ima N angial.Then he ta lks of the re ligion of Tibe t, a matte r that of cours e would inte res t

himspecially. He arrived in Leh , June 2oth, and had jus t begun to learn the lan

guage and hoped to s e e some fru i ts agreeable to the D ivine Maje s ty »cropping up

be twee n the rocks of Tibe t» , whe n he was told the re exis ted also a third Tibe t .Aprés p lus ieurs deliberations , il fut conclu ,

contre mon incl inat ion , que nous irions en

faire la d écouverte . Ce voyage e s t d’

env iron 6 a 7 mo is , par de s l ieu x dés erts dépeup le’ s .

Ce trois ieme Th ibet e st plus e x pos é aux incurs ions des Tartare s qu i sont l imitrophe s , que lesdeu x autre s Th ibets . N ous partime s donc de Ladak le 1 7 Aofrt de l

année 1 7 1 5, nous

arri v éme s a Las s a , d’

ou j’

a i l’

honneur de vous écrire , le 1 8 Mars 1 7 1 6 . Je vous lais se a con

jecturer ce que j’

a i e u a s ouffrir durant ce voyage au mi l ieu de ne iges , des glaces , du fro ide xces s i f qu i regne dans les montagnes .

And he re his geography came to an end , excepting that he give s fourmonths ’journey for caravans fromLhasa to Peking . The le tte r was w ri tten on ly 2 2 daysafte r his arriva l in Lhasa , so that his impre s s ions of the jou rney we re s ti l l qu i tefresh . And s ti l l he has nothing to say about the long jou rney in Tibe t . He has

Some thing to te ll of the firs t part of the journey, which was much more difficu l tand demanded much more attention on account of the accentuated landscape , thanthe leve l and Open va l leys of the uppe r Indus and Tsangpo . The journey fromLehto Lhasa took himexactly s even months , which i s s low trave lling.

2

R i tte r, us ing this s ame le tte r, says i t is a pity we pos ses s so l ittle information about this memorable journey . He says De s ideri ’s rou te was the s ame con

ce rn ing which BE RNIE R had got s ome information and that Leh was rediscove red byMOORCROFT in 1 8 20. But De s ideri was the firs t E uropean who vis ited Leh ,

which , before his time , was comple te ly unknown in E urope .3

2 Re v . A . H . FRAN CKE writes his name Nyima namgyal. H e re igned about 1 680—1 7 20. AH istory of Western Tibet, p . 1 1 6 e t s eq .

2 I have trave lled between Shigats e and Ladak in e ightmonths , includ ing excurs ions to the s ide s .

3 »So kurz auch le ider der mitge the ilte Bericht ist , so ist er doch der erste , ja fruherh in dere inzige , der auf d iesen hochst beschwerl ichen Wegen das ganze Himalaya-Gebirge ti b ers teigt und uns

zum ersten Male in d iesen Bergreviere e twas orientierte .» R itter, d ie E rdkunde v on A s ien , Bd II.Berl in 1 8 33 , p . 43 4 .

THE LETTE R OF DE SIDE RI . 27 1

In all, some ten lette rs fromD es ideri’

s hand we re in exis tence , although the

one quoted above s e ems to b e the on ly one of geographica l importance .

‘ In 1 8 75,

howeve r, a manu script of 630 page s in smal l fol io, was discove red in Pis to ia , De s ide ri ’s nati ve town in Ita ly . It has the ti tle : B ret/e e s urcim‘

o ragg u ag l i o del v i agg i o

a ll e I ndi e or i en ta l i del P adre Ippol i to D es i der i del la Comfiag i zi a. de 6 2522. Thistitle only refe rs to a smal l part of the contents , as the author is al so deal ing withPe rs ia , Kashmir, India and the mis s ion of the Je su its ; and more than half is aboutTibe t . The manu s cript , which is not in Des ideri

s own handwriting , was publishedin 1 904 by the Italian Geographica l Socie ty.

2

The Congregatio di Propaganda confe rred by a specia l decre e of 1 703 the

exclu s ive rights to mis s ionary work in Tibe t upon the Capuchin Fathe rs . The rew i th began a very inte re s ting chapte r in the his tory of Cathol ic mis s ion in As ia , a

chap ter which came to an end in 1 745, and to which we shal l have to re turnlate r on .

In the meantime the Jesu i t Ippol ito De s ide ri obta ined the pe rmis s ion of the

chie f of his orde r and the ble s s ing of the Pope to go out and conve rt the Tibe tans .

And i t is abou t the expe rience s on his long and adve nturous journey he te lls in the

precious and valuable manu s cript.In the firs t chapte r Pu in i give s a D es cr i z i on e s ommar i a del v iagg i o fa l lo

da l P . D es i der i . The mis s ionary le ft Rome on S eptembe r 2 7th 1 7 1 2 in companyw i th P . I ldebrando Gras s i and w e nt to India . Fromthe re he took P . MANUE LFRE YRE as a compan ion .

3

Afte r having provided themse lve s with an inte rpre te r, De s ide ri and Freyrele ft Kashmi r on May i 7 th 1 7 1 5 and went ove r a ve ry high and s nowy mountain , Kante l , 4 the Z0j 1

-la of our days , and on 3oth of May, ente red the Primo

2 N . DE LISLE has collected some »N otes sur le Tubet par le P . Hippolyte De s ideri»,published byKlaproth . Nouveau Journal A s iatique , Tome VIII , Paris 1 8 3 1 , p . 1 1 7 . It does not contain any important geographica l facts . For, travel l ing fromYarkand to T ibet, it i s sa id , one has to pas s the desert

of Ngari Jongar, at which one arrives through a pas sage between two mountains , w ith the exceptionof which the whole country i s impracticable . At Karthou (Gartok) and in Ngari Jongar there are

T ibetan garrisons to de fend the pas sages . Barantola i s the name the Chi nese Tatars give to Lhasaand the southern part of T ibet. The second or Great T ibet i s cal led Lata yul . The town of Lata isthe capital of Great T ibet. Two road s are sa id to exist between Lhasa and S ining. FromLata fora two months ’ journey there are no d ifficu lt mountains . In the great de sert between Lata and Lhasathere are very high mounta ins , which are the summits of Imaus . Tochoa, R e thOa or R edok » in mongol» are places on the road . The first T ibet i s under Chin a, the s econd is independent, the third i sunder the Great Mogul .

2 CAR LO PU1N 1 : l l T ibet (Geografia , Storia, R e ligione , Costumi) Secondo 1a re lazione de l viaggiodel P . Ippol ito Des ideri ( 1 7 1 5—1 7 2 1 ) Roma , 1 904 . In h is introducti on , Pu in i has a note regard ingthe h istory of the geographical knowledge of T ibet unti l the beginn ing of the 1 9th century.

3 I amhere going to fol low the s ummary of the d iary , on ly fi l l ing in thos e parts of the laterchapters which are of spec ia l interest to us froma geographical point of view.

4 The ed itor tells us that V IGN E identified Kante l with Zoji-la , but does not quote the place .

H e gets hi s authority fromMARKHAM (Narratives of the M iss ion of GE ORGE BOGLE etc . p . 3 04)where it i s sa id : »Accord ing to V IGN E , the summit of Bultul (Zoj i la pass) is the Mount Kantul of the

old maps . » Ne ither does Markhamgive the words of V igne , which run as fol lows : »The elevation of

272 IPPOLITO DE SIDE RI .or Piccolo Thibe t , which in the native language was called Baltis tan . Fromthe rehe ente red the Secondo or Gran Thibe t, which is cal led Lhata-yul by the nat ives .

The capital is Lhe or Lhata, whe re they arrived on June 26th . He re they we rerece ived w i th great honours and hospital i ty by King N ima Nangial, the min is te rs ,Lamas and o the rs .

FromLeh the ir way goes e as twards ove r an un inhabited plain , called Cianghthangh , »the north plain»

,and final ly they arrive in Tre s cij-khang , or Tashi-gang,

as we call i t. They le ft Leh on 1 7 th of Augus t and reached Tashi-gang on 7th of

S eptember . A part of this jou rney took themthrough mountainous and inhabitedcountry , b u t late r on they arrived at »a wide plain , Ciangh-thang, a great part ofwhich was occupied by dead and rotten wate r, ve ine s of su lphur and pools of s ul

phurous wate r» . This »rOttem wa te r and the pes tilential a ir was ve ry dange rou sboth to men and an imals . For some days they had to abs tain , as far as pos s ible ,fromdrinking the b ad wate r .

From this description i t is not e asy to s ay which road Des ideri has real lytaken between Leh and Tashi-gang. For on the ordinary road the re is no b ad

wate r nor are the re su lphurous pools . The whole way long one is clos e upon the

Indus , and has no need whateve r to drink b ad wate r , so much the more as the

rive r is at its highes t leve l at the end of Augu s t and beginn ing of S eptembe r,and the re fore large enough even for ve ry thirs ty Je su its .

So far as I can s e e , Des ide ri has not taken the high-road along the Indus ,but has gone v ia Rudok . On this northern route he may eas i ly have come acros ssalt wate r pools . Thu s he wou ld have had to cros s the wes te rn part of Transhimalaya in orde r to descend fromthe Chang-tang to Tashi-gang . It mu s t, howeve r, b e remembe red that the name Chang- tang doe s not only be long to the plateauIand north of Transhimalaya . For at the pres ent day, when Ladaki me rchants gofrom Leh to Gartok , they always say that they are cros s ing or go ing to Changtang. S ti l l , De s ide ri wou ld neve r have complained of the wate r i f he had fol lowedthe Indu s .

Trescrj-khang »is the las t inhabited place in this direction» with a fortres s and

a su rrounding wal l and a moat w i th bridge s . He re the domin ion of the S econd

Tibe t comes to an e nd and the Third Tibe t begins . As the place was a fron tie rfortres s there was a governor and a garrison in Tashi-gang . The fortre s s had at

some pe riods be longed to Ladak , and the D sungarians , and the Tartars of high independent Tartary , who were known to b e ve ry ambitious , unre l iable and treacherous ,we re not far off. The King of Second Tibe t had given the mis s ionarie s le tte rs of

the pass of Paien-i -Kotal or Bal~Tal (above—be low) , otherwi se Shut-j i -La (pronounce Zoj-i-La) , the hil lof S iva , i s , by

'

thermometer, about feet. The pas s I have just descended , i s that which ismarked as mount K an tal in the old maps . Kantal means a lofty hil l or pass , and as it happens tob e that by wh ich the low land of K ashmir i s quitted on the highway to Duras , it was , of cours e , noticedby the Jesuit miss ionary Des ideri in hi s way to Great T ibet and Butan .» G . T . Vigne : Travels inK ashmir, Ladak , Iskardo e tc. London 1 844, Vol. I I , p . 3 95.

274 IPPOL1T0 DE SIDE R I.

In the more de tai led part of the narrati ve we read , amongs t othe r things»This p lace is called N gari G iongar and is approached b y th is people w ith great respect

and venerat ion on account of a certa in Urgh ien Padma S ambava), who is the founder ofthe re l igion wh ich is pro fes s ed in T i bet. In the s ame Ngari Giongar there i s a mounta in,tremendous ly h igh , very large in circumfe rence , w ith the summit cov ered b y clouds and w ithete rnal s now and ice and also v ery horri ble , rock y and d ifficu lt on account of the mos t b ittercold that is re ign ing there .»

In a grotto in this mountain , Urghien was s a id to have l ived in total re ti rement and con ti nua l contempla tion . But now i t was us ed as a temple , and in an

annexed monas te ry the re were a Lama and some re l igiou s men . The Tibe tans vis itthe grotto whe re they leave some offeri ngs , and then w ith much z eal , they makethe wandering round the whole mountain which takes some days , and by which theybe l ieve that they will rece ive the greate s t indu lgence for thems e lve s .

Thi s description of the Ka i las and of the Tibe tans ’ vene ration for the sacredmounta in i s a lso admirable .

»Three weeks a fter Ngari , go ing towards the eas t, you reach anothe r valley calledTos cioa ; and after some more days of journe y another ’rig ida campagna’ where the lake of

Re toa i s to b e found .»

Or more de ta i led : »The I st of Decembe r we arrived at a plain place morefre e from s now , which is ca l led Tos cioa» . To let the men and an imals ge t someres t the prince ss s tayed he re for two days . He re some shephe rds ’ tents were found.

The shephe rds we re graz ing great flocks of pon ie s , mu le s and yaks , be longing to

the Gran Lama and to the king.

If he did not say that he had some more days fromTos cioa to the lake , Ishou ld have identified Toscioa with Tokchen . A lthough, of cours e , and as he has

had particu lar opportun ity to s tudy the pilgrims ’ wande rings round Ka i las and the

Manas arov ar, one may b e al lowed to conjecture that he wen t down fromKa i las vielDarchen to Tokchen , and thence round the lake . This becomes the more plaus iblewhen we read the next pas sage , the mos t important and inte re s ting in the wholenarrative of this able and inte l ligen t Father 2

»Furthe r on we pas sed a p lain called R etoé , where there is a great lak e wh ich has a

circumfere nce of s ome day s walking, and fromwh ich the Gange s i s suppos ed to originate .

However, as a cons equence of w hat I cou ld ob s erve on my way and wh ich I heard e x perts ofthe p lace s as we ll as of the Mogol to agree in , it s eems tome that the abovementionedmountof Ngar i G iongar s hou ld b e recogn i zed as the true origin and source of the Gange s , as we l las of the r i ver Indus . That mount be ing the h ighes t, fromwh ich the land s lopes on boths ides , the one as we ll as the other, the waters , e ither fromrain or fromme lting snow ,

w h ichdescends fromthere on the wes tern s ide

,flow into the s econd Ti b et, as practically s hown b yfacts ; and a fter hav ing gone through it , cros s the L ittle T i bet. Then rus h ing down through

all the mountains of Kas hmir,the y finally approach L ittle Guzaratte to formthe Indus , very

2 Op . c i t. p . 44 et seq .

2 Op . Cit,p . 45.

DE SIDE RI’

S DE SCRIPTION or MANASAROVAR . 275

w ide and nav igab le . In the same way the waters descend ing fromN gari Giongar on the eas tern s ide flow firs t into the s a id lake Re toé , then , follow ing the ir way down , they formthe ri verGange s .

A proof of that i s to b e found in the follow ing fact. The golden sand of the Gange s

is greatly renowned in the writings of our ance s tors ; therefore , i f we wou ld as sume that theorigin and source of that river w ere in a d ifferent p lace , we s hou ld s ta in our ances tors as l iars ,because in no other place of the ri ve r cours e (e x cept in Mount Ngari G iongar and in lakeRetoa), e ven the s l ightes t appearance of such s ands i s to b e found . On the contrary , i f whatI jus t have sa id b e accepted , that is to s ay that the s ource of the Ganges is s ituated on the

s a id mount and summit of Ngari Giongar and in lake Re toa, the statement of the anc ient writersw il l b e found to b e really in accordance w ith my op in ion, b ecaus e , as it is we ll known , I daresay,throughout the world , on the s hore s and in the s ands of that lake much loos e gold is to

b e found , wh ich fromthe s urface of Mount Ngari Giongar and other intermed iate mounta ins ,is carried down b y the waters flow ing fromthence . To s earch for such gold and to collectit, T i betans and s ome merchants mee t fromtime to t ime at the said lake , and draw muchprofit from it . Moreover that lake is the object of a great veneration b y those supers t itiouspeop le ; there fore they meet s omet imes there in a p ilgrimage , and make the tour all roundthe lake w ith great devotion believ ing to acqu ire many indulgences , and in some way to w in

many part icular jub ilees .»

The lake of Re toa is of cours e the Manas arov ar. It cannot pos s ibly b e any othe rlake . Re tozl or Redoc is Rudok , a name which he has heard and e i the r misunders tood or misplaced when w riting his narrative . He has heard that the Gange s wass uppos ed to originate fromthe lake , exactly as the Lama surveyors we re informedand as they represented i t on the i r map. And they vis ited the lake at the s amet ime as Des ide ri . But Des ide ri is more critica l than the Lamas on this part icu larpoin t, because he has obs e rved personal ly and heard from othe rs , that both the

Indus and the Gange s have the i r sources on Mount Kai las . Regarding the Indushe has not the s l ightes t doubt of this be ing the cas e ,

) as is practica l ly shown byfacts » . But as to the Ganges , he agree s that fromits sou rce in the Kailas i t rea llyruns through the lake . And later on he re iterate s that the source of the Gangesi s s ituated on the s ummit of the Kai las and in the lake . Thus he criticis e s the

gene ral s uppos ition of his t ime on ly so far that he s ays the lake is only a secondarysource of the Ganges , Mount Ka i las be ing the rea l and primary sou rce . He de

scribe s the cours e of the Indus all the way down . It is a pity that he . does not say

whe re his Gange s goes to. For the re can b e no doubt that it is the same as the

Gange s of the Lamas , i . e . the Satlej .

De s ideri confirms the hydrography of the Lamas indirectly . »Parmi pe ro,da que l che ho nel pas saggio os se rvato . l» He re fers to his own obs e rvations onhis way. He has s een that the source of the Gange s is s ituated on the Kai las , bywhich he means one of the smal l rivers go ing down to the Manasarov ar. And, if hehas made the wandering round the lake , he has s een the branch , Ganga , going out

of the lake , and the re fore gives two s ources : Kai las and Manasarov ar. He is not

to b e blamed in placing the s ou rce of the Indus on the wes tern s ide of the Kai las .

This has been clone 1 90 years late r. He does not know Langak-tso at all.

276 IPPOLITO DE S1DE R1 .Then he talks aga in of the flocks and shephe rds and the great amount of

bu tte r that goes fromth is regi on to Lhas a and othe r place s . On the 2 and of De

cemb er he came to anothe r pla in whe re he a lso found s everal ten ts and shepherdswande ring abou t w i th the i r flocks . In s pite of the des ert be ing arid enough and

tota lly devo id of s edentary popu lation , s ti l l the Grand Lama has a great profit fromits gold and bu tte r as we l l as fromthe tol ls levied fromthe me rchants of Gartok.

»January the 4th , 1 7 1 6 , hav ing finally terminated to des cend the w hole of th is greatde s ert we arri ved to the firs t popu lation in the th ird Th ibet.»

In S erchia,xwhich was a big fortified place and the capital of the province of

Zang-to and whe re the re was a great gove rnor, they had to s top on account of thes icknes s of the pri nces s . On January the 2 8th they left Sereb ia and trave lled throughmany villages and inhabited place s , b ut they had to trave l s low ly for the princes s

sake .

Fifteen days from S erchi a they reached anothe r big fortified place , which isthe begi nn ing of the kingdomof S echia.

2 Fromhe re they had 4 or 5 days to the

great city of Giegazze,3 »the capital of the old kingdomof Zang-to , which nowadays

is d ivided into 2 provinces : Zang-to , the capital of which is S e rebia, and Zang-me ,the capital of which is G iegazze» .

4 Ove r Chiangze (Gyangt s e) the party fina llyarrived in Lhas a on the 1 8 th of March 1 7 1 6.

A special chapte r in the valuable publication of De s ideri’

s narrative has re

ce iv ed the ti tle : D zfi eol l c‘

z del v iagg i o, and conta ins a perfectly admirable des criptionof all the hardships Des ideri and his fe l low trave l le rs met w ith on the i r way, whe re ,during 3 months , they did not s e e a s ingle vi llage and whe re they had to bringwi th them all the ne ces sary provi s ions , tea , bu tte r, roas ted corn flour and mu tton ,which was exce llently pres e rved by the cold. The hors es had to b e fed with flou rand corn as no grass cou ld b e found on account of the s now . They obtained wate rby me lting ice and snow, which, howeve r , can only have been for a certa in part ofthe road ; as for the res t they had the Tsangpo near at hand. No firewood worthmention ing cou ld b e found and they had to collect droppings of hors es and yaks .

Camping was cold and uncomfortable on the s now-cove red ground. They us ed roundTartar ten ts . But whe ther thes e we re pitched on s and or on frozen ground , theywere ve ry much expos ed to the furious w ind . Des ide ri te lls u s how they we redres sed and how di fficu lt i t was to save one

s nos e , hands , and feet frombe ingfros t-bitten .

The journey is de s cribed as s pecial ly hard, as one has to b e in the saddlebe fore sunrise and ri de the whole day ti l l s uns e t. Arnongs t innume rable mounta insone has to proce ed in ice and s nowfall , and froms carcene s s of gras s the hors e s

2 Probably Saka-dsong.

2 Sakya—gompa .

3 Shigats e .

4 I . e . Tsang-tod and Tsang

-med , the upper and lower Tsang.

273 1PP0L1To DE s 1DE R 1 .

and the Ki-chu we re one and the s ame rive r. But Des ide ri had fol lowed the

Tsangpo the whole way fromnear its s ou rce , and we shal l s ee in a late r chapterthat he knew pe rfectly we l l which was the head rive r and which the tributary.

By far the greates t part of D es ideri’

s narrative is a de s cription of the cu ltand re l igion of the Tibe tans , which is not les s admirable than the geographical part .I have quoted above only thos e parts of his obs e rvations which are of importancefor th is his torica l sketch . They w i l l b e sufficient to prove that Des ideri

s work isone of the be s t and one of the mos t re l iable eve r wri tt en on Tibe t and that it leavesfar behind it the contemporary narrative s of Beligatti and de lla Penna .

Des ide ri mus t b e regarded as the firs t E uropean trave ller who has vis ited and

des cribed the Manas arov ar, and at all events he is the E u ropean d is cove rer of theKai las . He is the firs t explore r in recent time s to s tart the ques tion and controv ersy abou t the s ituation of the source of the Indus , a problemwhich shou ld b ecome finally solved on ly 200 years late r. It mus t, howeve r, b e confes s ed that hehas conjectured the s ituation of the source of the Indus ve ry near its rea l place .

He was the firs t to s tart the problem of the source of the Ganges , which shou ld b esolvedmore than a hundred years late r on . He was told that the sacred rive r originatedfromthe Manasarov ar, b ut fromhis own obs ervations arrived at the conclus ion that therea l source was s i tuated on the Kai las . By a curious co incidence he make s the samemis take as the Lamas , confus ing the Satlej w ith the Ganges . E ven on Kirche r’smap we have s een the Ganges s tart froma lake which mus t b e the Manasarov ar

He re i t is a pity that Des ide ri mentions no name at all. If he had us ed the

expres s ion » il fiume» as he“

doe s for the Tsangpo , he wou ld have le ft the ques tionope n whe the r the rive r en te ring the Manasarov ar and leavi ng the lake aga in were the

Gange s or any othe r rive r. But already the fact that he knows a rive r is su ing fromthe Manas arov ar is important enough . This prove s that the superfluous wate r of theManasarov ar in the winte r of 1 7 1 5

—1 6 ran ove r to Rakas-tal through the channe l .And as the rive r is su ing from the s acred lake was known to proceed to India , wege t through De s ide ri an indirect, though re l iable proof, that the S atlej also wentout of Rakas -tal in 1 7 1 5

—1 6 . He has gi ven a ve ry corre ct and clear account ofthe pilgrimages to the sacred mount and lake . Beyond doubt he is the firs t E urOpean who eve r cross ed the famou s and important pas s ofMaryum-la . If his Toscioais iden tica l with the pres ent Tokchen , and his Sereb ia iden tical with Nain S ing’sSarka and Ryde r’s and my Saka-dsong, Des ide ri can hardly have us ed any othe rroad than the one cros s ing Maryum-la. This is the more like ly as he trave lled incompany with a native princes s and all her fol lowers , Who certainly only wou ld us e

the mos t comfortable road , and even this road s e ems to have been toomuch for theprince s s , as she became i l l on the way. Profe s sor N. KUE HNE R pos itive ly saysDes ide ri went ove r Maryum-la, although this is only conjecture .

‘ With the new

2 Opisanie Tib eta , Vlad ivostok 1 907 , Part I , p . 3 6.

DE SIDE R I THE FIRST E UROPE AN OF MAN ASAROVAR . 279

material we now pos s es s one may b e a llowed to as s e rt that Des ideri , in Maryum-la ,

has discove red the wate r-parting be tween the Satlej and the Brahmapu tra . Des ide riis , finally, the firs t E uropean who has fol lowed the cou rs e of the Tsangpo the wholeway fromMaryum-la to Che tang and at the s ame time he has dis cove red the sou rc eof the tribu tary to the Tsangpo , which is called Maryum-

chu .

Add to this the gene ra l merit of his narrative , the abs ence of fantas tica lspecu lation, the qu ie t matte r-of-fact way in which he give s his obs ervations and no

body will cal l i t an exaggeration i f I regard Ippol ito De s ideri as one of the mos tbrill iant trave l le rs who ever visi ted Tibe t, and, amongs t the old one s , by far the

mos t prominent and the mos t inte l l igent of all.

CHA PTE R XX IX.

FATH ER T IE FFE NTHALE R, FATH ER GAUB IL AN D

AN Q UETIL DU PERRON .

JOSE PH TIE FFE NTHALE R was a Jesu i t Fathe r, born at Botzen about 1 7 1 5.

I In

1 743 he arrived in Goa , and l ived in India for 42 years . In 1 759 he entered intocorrespondence with the famous s cholar ANQUETIL DU PE RRON , who was then at

Surat . Tieffenthaler occupied himse lf with s tudies on natura l his tory , re l igionand geography , and in 1 773 his geographical mate rial firs t came to E urope . In

1 7 76 his three maps arrived through the as s is tance of M . BERTIN , in the hands of

Anqueti l du Pe rron who was then in Paris , whe re he published an art icle on

thes e maps , which was reprinted in In 1 784 Anqueti l du Pe rron published ,

on one shee t, the maps of T ieffenthaler unde r the title : Ca rte g e'

ne'ml e da cow '

s da

Ga ng e el da Gag ra . The parts of this map which are of specia l inte res t to us havea short time a go be en re

—published by M . CH ARLE S E UDE S BONIN , 3 who says thatthe original map or maps we re probably drawn by a Hindu and pres ented to one of

the great Mogu ls , probably Akbar (1 556 The s e maps or copies of themcame into the hands of Tieffenthaler and afterwards to Anquet il du Pe rron .

In the Ge rman ed ition of T ieffenthaler’

s , Anquetil du Pe rron’

s and RE NNE LL’

S

work , 4 the ed itor, BE RNOULL I , says , that T ieffenthaler’s work to a great extentis taken from Ain-i-Akbari , to which important work Profe s sor Sprenge l hadd irected the attention of the Ge rman s cholars .

5 T ieffenthaler’

s work is there fore2 He wrote his name T ieffentaller, but I use the spell ing of his commentators .

2 Journal des sgav ans , 1 77 7 , Janvier. E d . de Hollande .

3 Une ancienne carte des source s du Gange . Annales de Géographie N zo 1 1 2 , XX° année ,1 5 Ju illet 1 9 1 1 , p . 3 3 8 et seq . , and in his book : Le s Royaumes des Neiges , Paris , 1 9 1 1 , p . 256

2 79 : Les sources du Gange. BON 1N gives a very able d i scuss ion on T1 E FPE N THALE R’

S standpoint.4 De s Pater Joseph Tieflenthaler’s his torish-geograph ische Beschre ibung v on H industan , etc. ,

Berl in and Gotha 1 785, V ol. I , Preface , p . x .5 BE RN OULL1 continue s : »Ausser d iesemhat der Pater Tieff enthaler vie l andere der b ewahrtes ten

Geographen und Geschichtschre ib er v on Ind ien benutzt und se ine e igenen auf . Rei sen gesamme ltenBemerkungen in se ine Ausztige e ingewebt. Aus dema l len i s t nun e in flir E urop ’

aische E rdb eschre ib er

und Geschichtforscher hochst brauchbares und clas s isches Werk entstanden , das aber an s ich , ich leugnedies n icht, noch e ine z ieml ich rud is et indigesta moles ist.»

282 FATHE R TIE FFE NTHALE R , FATHE R GABUIL AN D AN QUE TIL DU PE RRON .

Anque t il du Pe rron is the commentator, I amgo ing to re fe r to the French editionin review ing and cri ticis ing his v i ews .

I

In Vol. II Anque ti l du Pe rron d is cus s es T ieffenthaler’

s maps and fu rn ishesthem wi th his own criticisms frommate ria l a lready e x is ti ng at that date .

2 WhenT ieffenthaler pre tends that the real source of the Gange s is unknown and that itw il l for ever remain undis cove red as the roads beyond the >>mou th of the Cow»

are impracti cable , Anqueti l du Pe rron s ays that he is of a d iffe rent Opin ion »as

the re is nowhere in the world a road absolu te ly impracticable for thos e who havefe e t» , and inde ed the source of the famou s rive r had been dis covered by Andrade ,and was redis cove red late r on by Webb and his comrade s carrying out Coleb rooke

s

plans for that pu rpose . He reminds us of the fact that in Bengal and Tibe t the ideaof the impos s ibi l ity of reaching the sou rce of the Ganges had grown immovable as

the origin of the rive r was suppos ed to b e in heaven . Then he quotes the viewsof Linscouten , and of Abbé TOS I who says some people be l ieve the rive r begi nsat Mount Nagracot , whe reas othe rs think that i t come s fromfar away

,fromthe

Scythian mounta ins and pas se s through the gorge , at the s ide of which a rock isl ike the head of a cow . Tieffenthaler be l ieve s the Gange s is sue s froma rock inthe Tibe tan moun ta ins at abou t 3 3° N . lat . and 73

°E . long. fromPari s and then

enters the gorge of Gangotri »s eu Cataracta Gangis , quame tiamOs Vaccze ap

pe lant» .

At the end of the 1 6th century the great Akbar s en t his expedi tion to the

s ource of the Ganges .3 D is cus s ing the resu l ts of this exped ition Anqu eti l du

Pe rron a rrives at the conclus ion that the »cow ’

s head» is not the rea l sou rce , b utthat this mus t b e looked for somewhe re in the inte rior of gre at Tartary, an explanation that can b e unders tood on ly if we cons ide r his want of confidence in the Lamas ’

map , and his negligence of Andrade .

As to the river Anqueti l du Pe rron s ays Tieffenthaler is the firs t whohas eve r mapped i t and made the whole of its cours e known in E urope . But he

thinks natives have furn ished the information abou t the upper part of the rive r. 4The Gagra of Tieffenthaler is obviou s ly the Map

-chu , which lowe r down is

ca lled Kauriala and Gogra . He found that the uppe r part of his Gagra was cal ledSardjou , which is the river known unde r the name s of Kal i , Sardu and Chauka .

2 Description historique e t géographique de l’

Inde, etc. 1 . La Géographie de l’Indous tan ,

e tc ., par le Pere Jos eph Tiefienthaler, Jésuite M is s ionaire apostoli que dans l ’Inde . 2 . Des Rechercheshi stori ques chrono logique s sur l

Inde , 8: la D escription du Cours (in Gange da Gagra , avec une

tres grande Carte , par M . Anquetil du Perron . 3 . La Carte generale de l’Inde , etc.

, par M . Jaque sR enne ll . Le tout, augmenté de remarques d

autres add iti ons , rédigé 81 publié en Frangois , par

M . Jean Bernouilli . Berlin 1 786 .

2 Op . c it . Tome II , IIe Partie : Le Déve loppement du Cours du Gange de ce lu i du Gagra,tire de s Cartes M anus crites fa ites sur les l ieux, par le P . T ieffenthaler 1 78 7 , p . 2 66 et s eq .

3 V ide note p . 7 2 .

4 Thi s als o is obv ious fromT ieffenthaler’

s own words : »Neque solumips us , hasce regionesperlustravi , sed hominem arte geographica in s tructum ab legav i , ut locorumintervalla, mundiquePlagas rite exploraret.»

AN QUE TIL DU PE RRON AND THE OR IGIN OF THE GRE AT INDIAN RIVE RS . 2 83

As to the origi n of the rive rs fromthe lakes , we find in Anquetil du Perron ’

s work the fol lowing inte res ting pas sages .

‘ He talks of two great lakes s ituatedin Tibe t.

Celu i de l’E st,nommé Mansaroar (ou Mansara) , tres ce

' le b re dans le pays , a 60 cos s es

d e tour, s e lon le P . Tieffenthaler dans sa notice . Mesu ré s ur la carte , i l n’

en ofl'

re que 33 a

3 5. Le Lac de l’

Oues t , nommé Lanka , mesuré de meme , a p lus de on ze coss es de tour .

S e lon le s avant M is s ionaire , on dit que le fleuv e Brahma poutre (ou poutren), qu i v a aAs cham8: a Rangamati , s ort du lac Mansaroar. Du méme lac Mansaroar coule au Sud 81 au Sud

Sud -E st,une ri v iere dont i l n ’

y a de trace que douze Cos s e s Du haut du lac Mansaroar,

ou N ord-N ord-Oues t, sort nu fleuv e sur leque l on l it en Persan : grande ri v iere de Satloudj ,q u i v a du cOté du Pendjab et par cons équant a l

Oues t.

To this Pe rs ian text Anquetil du Perron has added the fol low ing remark»On dit que le Satloudj , qu i v a a Belaspour 81 a Lodiane , sort de ce lac ;mais ce tteas s e rtion ne mérite aucune croyance : car il e s t plus v ra is emblable qu ’ i l (ce lu i qu isort de ce lac) s e je te dans l’Allaknandara , qu i arrose Badrinat 8L S irinagar .»

Regarding the Satledj , Anqueti l du Pe rron has the fol lowing pas sage :Ma intenant que lle d ifli culté, que le Satloudj du lac Mansaroar, parti da 36°

1 5 a

s e rende au 3 1 , cou lant a l’Oues t, depu is le 77 °

1 5 ou 20’

jus qu’

au 73°

40'

(Ie 76 de M . Renne ll) ;ce qu i ne fait

,a , ces hauteurs qu

env iron 1 50 l ieues en d iagonale : ou meme que , s u i vant laCarte du Mis s ionaire , s e soutenant a 36°

35'

de lat itude ju s que pas sé 70°

de longitude , il ba is sea 69

°

30'

de longitude , jusqu ’

au 3 12 de latitude ; e s pace de 235 a 240 l ieue s ; 8: paro is se alors

avo ir s a s ource dans le s montagne s ? d ’

un cOté c’

e s t une mer, C ’e st a d ire nu grand fleuv e ; cc

q u i annonce un cours tres étendu , comme nous verrons p lus b as a l’E s t, ce lu i du Brahmapou tren : 8

: il cou le vers le Pendjab : C ’e s t donc le Satloudj de cette contrée . S i l’

on supposeq ue Ie Satloudj , sorti du lac Mansaroar, se jete dans I’A llaknandara , appe lé Gange , on s e réu n is se au Gange meme , alors ce dern ier fleuv e v iendra médiatement du meme lac Mansaroar;

ce qu i rentre dans l’

op in ion des Lamas C’

e s t encore de 1a que le P . Tie ffenthaler a puprendre le sentiment qu ’ il propos e sur le cours du Satloudj s ort i du lac Mansaroar: mais las ource du Gange , Gangotri , s ero it toujours d ifferente de la précédente .

With this view of the hydrography Anquetil du Pe rron compares the mapof De l is le , 1 705, (PI. XL) and finds b oth views tolerably l ike each othe r. He identi

fie s De lis le ’s lake with Tieffenthaler’s Mansaroar and the rive r of Cogue with the

B rahmapu tra .

Regarding Langak-tso or Rakas -tal we read 2

A c6té du grand lac Mansaroar, a l’

Oue st , e st le lac Lanka que le Miss ionaire allemand écrit Lanka Dhe, c

es t a d ire Lanka (lac) d ’

abondance : de deh, qu i donne ; ou lac Dew ,

le lac d iv in . Ce lac , d’

ou sort,a l

Ouest, le Sardjou est beaucoup plus petit que le Mansaroar.

So much s e ems probable , that before 1 776 , when the Fathe r s ent his mapsto Anqueti l du Perron , probably some time abou t 1 760 or 1 765 as we ll as in the

re ign of Shah Akbar, there was a rive r is su ing fromRakas -tal. It is difficu ltto find out the correct date s . It is qu ite clear that Tieffenthaler has not got

Op . cit. p . 3 47 et seq .

2 Op . ci t. p . 3 50 .

284 FATH E R TIEFFE NTH ALE R , FATH E R GAUBIL -AND AN QUETIL DU PE RRON“anything fromthe Lamas ’ map, b ut has gathe red his information fromIndian sourcesand frompilgrims who have been at the place and re lated to himwhat they haves e en . But he also thinks the matter worthy of fu rthe r exploration , for he says :»Fonte s hujus fluminis , ex narratu v iatorum, qu i ad hunc lacumperegrinantur, comperti sunt . Ce rti oria a l ias exploranda .» H is w ish was to b e fu lfilled some 80 yearslate r, when H . STRACHE Y went to the place .

In the chapte r on »the ide ntity of the lake s Mansaroar 81 Lanka dhe w iththe lake s Mapama 81 Lanken» , Anqu et il du Perron compare s the Lamas ’ mapw i th thos e provided by T ieffenthaler. He reminds his reade rs of d

Anv ille’

s

words that Kang H i ’s cu rios ity had given the world the know ledge of the s i tuationof the source of the Gange s . He orde red some men ,

» ins tructed in mathematics » ,to pene trate to the ve ry place , fromwhere the Gange s originate s . Fromthe irmap ,»the Lamas ’ map» , d

Anv ille had learn t that Mount Kenta’

r‘

s sé was the water-shedof two great rive rs . One of them, the Gange s , was formed froms evera l sources ,and went through firs t one and then the s econd of two great lake s and took itscou rs e towards the we s t unti l i t me t w ith a mounta in that forced i t to turn to the

south or to India and only by force cou ld it s ecure for its e lf a pas sage through themounta ins .

We have s e en be fore that the explore rs of Kang Hi we re not le s s succe s s fu lthan thos e that Akbar s ent ou t for the same pu rpos e . In the cas e of Kang H i ’sexplore rs the mos t s e riou s misunde rs tanding was the identification of the Satlej w iththe Ganges . And the same confus ion aros e from the information brought byT ieffenthaler

,who mixed up the Satlej and the Gagra and be l ieved his Gagra to

begin from Langak-tso. D’

Anv ille had to tru s t to the mate ria l brought to him,although he found the Ganges to b e rathe r long.

Comparing the Lamas ’ map w ith the Indian one of T ieffenthaler, Anquetil

du Pe rron , to begi n w i th, conclude s that the lakes are the same , as in both case s

they are s urrounded by mountains and give ris e to s eve ra l big rive rs go ing e as tand wes t. He a lso compares the names ; Chines e : Lanken and Mapama , Pere Gaubi l : Lanka and Lapama Ta lai

, the Indian map of T ieffenthaler: Lanka Dhé and

Mansaroar.

Then he approache s the que s tion of the rive rs .

‘ The Chinese map (Pl . LI)shows a gre at rive r ca lled Latchou coming out fromthe mounta ins , at the foot ofwhich are the lake s ; i t flows to the wes t. On the Indian map, Pl . LII, the Satlej is sue s from Lake Mansaroar, and flows firs t to the N .W .

, then to the Panjab .

But he forge ts that on d’

Anv ille’

s Lama map there is a lso a Lanctchou , whichis sue s from Lake Lanken and come s fromLake Mapama . So Anqueti l du Per

ron ought rathe r to have compared the »Satloudj» of the Indian map with the

Lanctchou ins tead of the Lats chou , which doe s not at all touch the lake s and in

2 Op . cit. p . 3 54 .

SOUTHE RN T IBE T ,V OL I .

D’

Anv ille'

s Map of the Lakes and of the Sources of the three great R ivers , 1 73 3 .

.ICATE D HYDROGRAPH ICAL PROBLE M . 285

»urce of the Indus , though not nearly so care fu lly and

ource s of the Satlej and the Tsangpo . But in spite of

the identity of the names , the sou rce s , and the cours e sow ing to the very meagre know ledge of thes e regionsrs tand his mis take .

te opin ion of the native s the Brahmaputra begi ns from.p of the Lamas (d ’

Anv ille) he finds that a great rive r.u ated near lake Mapama , turns to the eas t , S .E . and

7a . It is cal led Ts anpou . On the Indian map , again ,of a grea t rive r running to the eas t, north of Nepa l .

hinois e , 85 le Brahmapoutren de la Carte Indienneeuve .»

far,Anquetil du Pe rron goes on to examine >>whe the r

Lt is sues from Lake Lanken as the Chines e map re

he will prove that the Chines e Gange s isi s , he proves that i t is the Gagra, amis take for which ,is respons ible .

shows how and why the Lama explore rs are incorrect .rfficient to re late the way in which the map was madeh e conclus ion : »that the discovery of the source of the

1 8 is nothing less than certa in» . The impe rial explorerse, or what they w ished shou ld b e the sou rce of the

of Tseringdondob cut short the i r work and hinde redtry obs e rvations on the place . They asked the natives.me e rrors that made the map of 1 7 1 1 insufficient re1 du Perron ’

s opin ion is , that a map such as d’

An

no othe r au thori ty than that of the Lamas cannot b emap made on the place by the pe ople of the country.

n is take of Anquet il du Pe rron : that he has greate rt to him by T ieffenthaler, than in the Lamas ’ map,:su its . And s t ill he finds a good deal of accordanceian maps . The mos t important exception is that whi leGange s is sue fromthe Lanken , the Indian map makeslake . Ins tead of one mis take another is introducedthat the exploring Lamas obtained the i r informations tery at the s ou the rn shore of Mansaroar (i . e . Tuguknew that pilgrims came from Hindus tan to Man

amas d i sparoi s sent . La Carte du T ibet n ’

est plus qu ’un travail faittpres l

év aluation des mesures itinéraires des Chinois , sur le rapportveux , sera plus critique : ma is ceux qu i le font n ’

out pas été a lals la présentent comme connue . Op . cit. p . 3 62 .

SOUTHE RN TIBE T ,V OL . I .

D’

Anv ille'

s Map of the Lakes and of the f

286 FATHE R TIE FFE NTHALE R , FATH E R GAUBIL AND AN QUE TIL DU PE RRON .

saroar and Lanka , b ut w ithou t regarding the latter lake as the sou rce of the

Ganges .

Anqueti l du Pe rron does jus ti ce to the Lamas of the monas te ries of La

pama . The Lama s u rveyors ask them the name of the rive r which , N .W. of the

Manas arov ar runs to the wes t, and the Lamas living on the lake answe r : Lats chou ,the same rive r as the S atloudj of the Indian map. Then they ask the name of the

rive r which is sues fromlake Lanken and goes to the we s t, and the Lamas of the

place answe r : Lanctchou or the rive r of Lanken , the s ame rive r as the SardjouGagra of the Indian map . The explore rs ask i f the Ganges is not a continuati onof thes e rive rs . And the Lamas answe r that the Ganges is farthe r wes t pas s ingTs chaprang, the place vis ited by d ’

Andrade in 1 6 24 , who be l ieved that he had

d iscove red the sou rce of the Gange s , while , probably , he has on ly s een lake Lanken .

‘ But , u nfortunate ly, Anqueti l du Pe rron confounds Ts aparang w ith Deupragon the Gange s , at the confluence of the Alaknanda . Then he remarks that theLama explore rs give two sou rces to the Ganges , one from Kenta

r’

s s é and the othe rfrom Lanken , and be low the junction of the two they put in the name Ganga or

Gange s . But as thus , nolens vo lens , the Gagra was taken for the Ganges , »the firs tand real source of the Gange s res ts unknown , as it was before the pre tended discov ery by the Chine s e Lamas . »

We mus t remembe r how very little real ly was known at the time whenAnquetil du Pe rron wrote . No Moorcroft had been at the place , and the

only E uropeans who had s een the lake s , De s ide ri and Freyre , had not even menti oned the ir name s . The more de ta iled report of De s ideri was unknown at the

time . Anqu etil du Pe rron forgot that the Lama map had been execu ted withthe s pecia l and expre s s view of finding the sou rce of the Ganges . And , of cours e ,i t was not eas ie r for himor even for Renne l l than it had be en for d

Anv i lle to

make out the complicated hydrography of thes e places .

Unde r the head ing : »L e Ts anpou (52 le B ra/imapon tren s ont l e méi n efl eune» ,

Anque til du Pe rron give s a ve ry cleve r and inte res ting e s say in his work . The

native s had told Fathe r Tieffenthaler that the Brahmapu tra take s its ori gin fromthe Manas arov ar. Fathe r Régis had le ft the qu e s tion as to whe re the rive r reallygoes to uns e ttled , b ut he had thought i t mos t probable that the rive r fina lly tu rnedS .W . to the ne ighbourhood of the mou th of the Gange s . And Anqueti l du Per

ron addsc’

e s t a d ire , que cc fleuv e sera cette Mer de la Carte Ind ienne , qu i , al lant a l’E s t ou

Sud-E s t, pas se aude s s us de N e ipal, de Ts choukra, travers e une grande partie du T i bet, 81 versle s l imites de Ce t E tat, tourne au S ud-Q ue s t, 81 cou le pres de l’emb ouchure du Gange .

He trie s to find from the narratives of Bernie r and Tavern ie r some thingto support his view , b u t they on ly know that the rive r come s fromthe fronti ersof Tibe t.

2 Op . cit. p . 3 64 .

THE TSANGPO-BRAHMAPUTRA ORIGINATING FROM TH E MANASAROVAR . 287

D’

un autre COté le P . R ég is nous apprend que le Ts anpou coule de s env irons du LacMapama, 81 que pas sé le s memes frontieres du T i b et il tourne au Sud -Oue st, ti rant du cOté du

Gange ; 81 l’

opinion du pays est que le Brahmapoutren , qu i v a a As cham81 a Rangamati , sortdu lac Mansaroar, le meme que le lac Mapama : Le Ts anpou 81 le Brahmapoutren sont doncun s eul 81 meme fleuv e .

He also po ints to the di ffe rence be tween the view of R égis and that of theChine s e map , which may b e e xpla ined by the fact that the Lama explore rs neve rwe nt beyond Lhasa . As we know ,

the ques tion of the lowe r cou rs e of the Brahmapu tra gave ris e to a s truggle that had to continue unti l not very long ago , when i twas finally s e ttled by the Pundi ts .

Anquetil du Pe rron sums up the »geographical tru ths » that are repre s ented on

T ieffenthaler’

s Indian maps , in the following words :»La premiere Part ie de la Carte du Gagra, fa ite sur le s l ieu x par des Ind iens , en pré

s entant les deux Lacs Lanka 81 Mansaroar, nou s donne la source , jusque ’ ic i inconnue , destro is p lus grands fleuv e s de cette contrée ; le Sardjou , qu i s ort du lac Lanka , 81 dont le coursne s e trouve s ur au cune Carte E uropéenne ; le Satloudj , qu i sort du lac Mans aroar au N ordOues t, 81 cou le vers le Pendjab ; 81 le Brahmapoutren , le meme que le Ts anpou , qu i a sa s ourcedans le meme lac Mans aroar, a l

E s t, 81 qu i , apres avo ir trav erse une grande part ie du T ib et,tourne au Sud-Oues t, se jette dans le Gange au des sous de Daka

He a lso come s to the conclu s ion that'

the lake s as we l l as Mount Kenta‘

r’

s sé

mus t b e removed some fiv e degree s farthe r north than on d’

Anv i lle’

s gene ra l mapof Tibe t . And this unfortunate idea he calls : »Corrections importantes en Geogra

phie , Découve rte s meme , s’

il e s t pe rmis de le dire , qu i donnent une nouve lle facea la vas te étendue de pays que je viens de nomme r.» 2

T ieffenthaler’

s maps give indeed a new as pect to the country round the

lake s ! Bu t has geography gained by thes e change s ? By no means ! S tartingfrom d

Anv ille’

s map , the map of Tieffenthaler is a great s tep backwards . One

cannot even compare the two . Tieffenthaler’

s map is very e rroneous , whe reasthe Lamas ’ map is admi rable

,although it is 200 years old. The Lamas we re

qu ite independent of eve ry kind of re l igious prejudice , whe reas the pilgrims fromInd ia under all condi tions lead to be l ieve even what they cou ld not s ee w i th the i rown eyes , name ly , that the s acred rive r Brahmapu tra began from the s acred lake ,the Manasarov ar

,and the Gogra fromits ne ighbou r , the Rakas -tal. One can eas i ly

unde rs tand the pleasu re i t mu s t have been for such a learned and able s cholar asAn quetil du Pe rron to ge t hold of Tieffenthaler

s documents and maps in which hehad unl imi ted confidence

,and which he cal ls dis cove rie s .

Then Anque til du Pe rron give s a de tai led description of the cours e of the

G agra . He finds it ve ry probable that the uppe rmos t part of the rive r, fromthe

la ke s and down to the cataracts , is bas ed on the description of a native fromIndia .

The res t, fromthe cataracts and downwards , shou ld b e T ieffenthaler’

s own work .

2 Op . cit. p . 3 70.

288 FATHE R TIE FFE NTHALE R , FATHE R GAUBIL AN D AN QUETIL DU PE RRON .

The whole as pect of the rive r, and its me rid ional runn ing in long bends is rathe rgood . We find Taklakot on the le ft bank , not as now on the right one . The

Gagra of T ieffenthaler is in its uppe r parts cal led Map-chu , lower down Karnal i ,

Kauriala .

R i tte r has no high opin ion of the maps drawn by natives .

But when he says that now as we know the real hydrography of the

place , i t is easy to explain the mis takes of the Lama map and the Hindu mapof Tieffenthaler, he is wrong and too much influenced by Moorcroft. R i tte r cou ldnot know that although the Lama map was qu i te a d ifferent thing than Moorcroft ’smap, both we re perfe ctly corre ct, depending on the pe riodical change s taking placein precipitation . And s ti ll he s uspects the pos s ibil ity of a pe riod ici ty .

The big map of the Jes u it Fathe r has the follow ing t itle : Ca rte Ge’n e’ra l e'

a'

u

Cam's da Gang e el del Gagra , dres s ée s eer les Ca rl ey parti cu l i er“da P . Tz

efieninn/er Par M . Anquet i l da P e i / i ron , P a r i s 1 784 .

Anquetil du Pe rron has on ly diges ted the mate rial brought home by Tieffenthaler, and as the mate ria l is provided by ordinary pilgrims , and the two E uropean s cholars have neve r been to the place , the resu lt cannot b e expected to

b e particu larly good. On Fig. III (Pl . LII) we read the legend : »Calqué sur

l’

Origi nal fa i t par umIndien» . I will now try to show that this map, made by a

Hindu , in spite of its te rrible errors , is s ti l l an important and valuable document .The Hindu has w ri tten s ome explanations in Pe rs ian , and Anquet il has added someobs ervanda of his own .

Beginn ing fromthe eas t , we find a big rive r is s u ing fromthe Mansaroar, and

the followi ng Pe rs ian words w ri tten along it : Dariar taraf N e ipal rafteh , or the rive rthat goes ‘

in the d irection of Nepal . And Anquetil du Pe rron adds : »Ir is sa idthat the Brahmapu tra , which goes to As chamand Rangamat i , is sues fromthis lake » .

The Lama explore rs wou ld neve r have made thems e lves gu i lty of such a gravee rror. The Indian explore r of T ieffenthaler or his predece s s or in Akbar

s timehas wande red round the lake clos e to the water’s edge , for the pilgrims keep as

near as pos s ible to the brimof the sacred lake ; so he has s een and cros s ed all the

2 He s ays : »A lles , was ii b er die Vergle ichung d ieser sche inbar officie llen und doch in jederH ins icht zwar nicht ganz fal schen , aber doch nur halbwahren Chines ischen Daten der Tt‘i b etischen Lama ’s , mi t eben so halbwahren durch H indu-Pi lger verfertigten Original-Kartenze ichnungen , v on den

G angesque llen , mit E rlauterungen in Pers ischer Sprache , we lche be ide der Pater Tieffenthaler aus

H indostan nach E uropa gesandt hatte , s ich nur sagen lasst we l l all that i s to b e said of thesemaps , has already been pointed out by Anqueti l du Perron .

The following golden words which R itter adds are as true nowadays as they have ever been»Moge j ener s charfs inn ig und mi t vie ler Ge lehrsamke it gefii hrte Stre it ii b er be iderle i Darste l lungen derGanges

-Que l len , in Behauptung der Hypothesen auf geographischemFe lde behutsammachen , umdasan fruchtbaren Thats achen so ub erre iche Fe ld der E rdkunde n icht noch durch unfruchtbare H ypothesenund Consequenzen , daraus fib er Dinge , d ie man noch n icht genau wi ssen kann , zu ii b erschwemmen ;

e ine e in zige pos itive Beobachtung an Ort und Stel le klart al le Zwe ife l schne l ler und bess er auf,als Prob ab ilit

'

aten,die wir darub er aufstellen konnten .» D ie E rdkunde v on A s ien , Band I I, p . 476

e t seq .

SATLE J ORIGINATING FROM TH E MANASAROVAR . 289

great wate r-cours es entering the lake or leaving i t, and all the smal l one s as we l l .The Rakas -tal doe s not ente r into the ci rcle of pilgrimage . This is the cau s e why

the Mansaroar is , comparative ly , so we l l laid down , whereas the Rakas -tal is ve rywrong both in formand s ize . The Indian pilgrimhas s e en a rive r at the eas ternshore , in direct commun ication w ith the lake , and he has cross ed i t , ce rtainly not

w ithout difficu lty . He may have forgotten in which direction the wate r was runn ingb ut it is more l ike ly that he wished to make the Brahmapu tra begin fromthe sacredlake . The natives , as a ru le , are ve ry good obs e rve rs in the fie ld , b ut when re l igi oncome s in they ha v e to become blind . The year of his wandering has be en ve ry rich inprecipitation , as can eas i ly b e s e en fromthe res t of the map . And hardly any othe rriver can have made the pilgrimthink of the Brahmapu tra except the Tage-tsangpo , ‘which always and under all conditions carrie s a much greate r volume of wate r thanthe othe r rivers at the eas te rn s ide . So, what the pilgrimcal ls Brahmapu tra is inreal ity Tage-tsangpo .

At the S .W . shore of the lake w e find anothe r rive r, which obvious ly is one of

thos e which enter the lake fromGu rla Mandata . There is no indication on the mapas to the direction of its cours e , but we know that i t can only b e an ente ring rive r .Near its mouth the re is a temple and a vi llage , the Tugu -gompa .

The information that the lake shou ld b e 60 Indian mi le s in c i rcumfe re nce doesnot agree with the s ca le be low , but i t does not matte r, as nati ves a lways e xaggerates uch things . The mountains s outh and north of the lake , Gu rla and Kai las , are

drawn in ve ry thin ou tl ine s and repre s ented as s e en fromthe s ide . They have no

names .

The mos t inte re s ting information given on the map is , howeve r, the exis tenceof a big rive r is su ing fromthe N .W . shore and runn ing to the N .W . The Pe rs ianle gend is ve ry short and clear and doe s not leave any roomfor doubt : Daria-iSatloudj taraf Pendjab rafteh , or, the rive r Satlej going in the direction of Panjab .

The firs t part of the remarks on this rive r is correct : 2 »It is s a id that the Satloudj ,which goes to Be laspour and Ludiana , is su es fromthis lake » ; that is to s ay, he has

obtained this information from native sources , and i t shou ld b e not iced that thenative informant does not s ay that the rive r take s its beginn ing in or fromthe lake ,b ut only that i t is su es from the lake , which leaves roomfor the as sump tion thatthe river Satlej en fers the lake s omewhe re e ls e . The information Tieffenthaler ob

ta ined fromhis Indian pilgrim,who had be en at the place and who , on this po int ,

was perfectly correct and trus tworthy , was , howeve r, s poi lt by hims e lf or Anque til

du Perron when he adds : »b u t this as s e rtion is not at all worthy of be ing be l ieved ,as i t i s more probable that it jo ins the Alaknanda , which irrigates Badr inath and

S rinagar , or some othe r rive r.»2 Th is i s al so Bou in ’

s Opinion . V ide l . c . p . 3 46 , and hi s Royaumes des Ne iges , p . 2 70 .

,

2 Bon in says that these legends can only b e attributed to Anquetil du Perron . But probablythe original information has been gathered by T ieflenthal er.

3 7—1 3 1 3 8 7 1 .

290 FATH E R TIE FFE NTHALE R FATH E R GAUBIL AND ANQUE TIL DU PE RRON .

T ieffenthaler’

s pilgrim, who knew that this rive r was the uppermos t partof the Satlej of Panjab , had no idea of its go ing through the Rakas -tal or LankaDhe which he had s o near to the we s t . But he never eve n went so far as to asce rta inwhe re the rive r went to , and thus on the map the neck of land be tween the two

lake s is not pie rced by any channe l .

But s o much was gathered by the pilgrimthat a river is s ued fromthe LankaDhe. It is a pity that ju s t he re a Pe rs ian legend is mis s ing. The re is on ly a legendof Tieffenthaler, who s ays that i t is really the Gagra that goes out from the

lake , and as to the lake , he informs u s that it is through the reports of »trave l lers »the sou rce of the rive r is known . But he mus t have fe lt the unre l iability of the

s tatement as he adds the really Charming words : »certiora al ias exploranda» .

D is regarding the Brahmapu tra , the mos t curious featu re of Tieffenthaler’

s

hydrography is that the eas te rn lake is regarded as be longi ng to the Indus -sys temand the we s te rn lake to the Ganges -sys temand that the two lakes are pe rfe ctlyindependent of each othe r . The Lama map made both lakes be long to the Gange ssys tem, but in this cas e we have s e en that the hydrography was perfectly correct andthe on ly thing unknown was where the river w ent to . E ven i f Ti effenthaler be l ievedin the tale of the Brahmapu tra go ing out of the lake to the eas t , and the Satlej tothe we s t, i t is surpris ing that Anque ti l du Pe rron cou ld ever accept such an extraordinary bifurcation , a cas e that, at leas t to such an e x tent , wou ld b e pe rfectly un ique .

And s til l i t is eas y to explain the mis take of Tie ffenthaler’

s pilgrim. We

only need to remembe r that the re is one rive r leaving the Manas arov ar to the wes t,

and one river leaving the Rakas -tal to the wes t . H e did not know that the e ffluents ,which he repres ents as two di ffe rent rivers , we re one and the same , or, in othe rwords , that his Sardjou was in real i ty the continuation of the Satlej . When he saw

the uppe r Map-chu above Taklakot, he found it probable that that rive r came ou t

fromthe Rakas -tal.Thus we are able to trace the ou tl ines of tru th even in this confus ion .

In the R ecu e i l published by P . SOUCIE T ,

2 P . GAUBIL give s some informationof the sou rces of the Gange s . The title of Gaub il

s article runs : »S i l nn l i on de

2 Boni n has a qu ite correct view of this problem, I. c . p . 3 46 : w e s erait-ce pas la l’

indice

qu ’ i l existait alors dans cette d irection na canal de jonction entre le s deux lacs , cana l dont le vis iteurind igene a seu lement v u et des s iné l’amorce au Nord-Q uest du Mansarov ar, sans la su ivre jusqu ’

a son

débouche dans le Rakas -tal ? Les con statati ons de Sven Hedin ont établi que le s eaux de cc dernierlac se déversaient dans le Sutledje lorsqu ’

e lle s atteign aient nu certain niveau, comme i l e s t marque sur

les cartes antérieure s Si donc il y a eu commun ication e ntre le s deux lacs , comme tout tend a

l’

etab lir, le fai t de nommer Sutledje l ’e’mi ssaire Nord-Ouest du Man sarov ar n’

étai t pas au fond absurde ,bien que l’

auteur indigene n’

ait pas s u ou conduire e t fa ire aboutir le canal dont il n’

av ai t v u que lajonction avec ce dern ier lac.

2 Observati ons M athématiques , A stronomiques , Chronologiques , e t P hys iques Tire’ es de s anciensl ivres chinois , ou faite s nouve llement aux Indes et a la Chine , Par les Peres de la Compagn ie de Jesus .

Redigees et publiées Par lo P . E . Soucie t, de la meme Compagnie . Paris 1 7 2 9 .

GAUBIL AND TH E SOURCE S OF TH E GANGE S . 29 1

P ou ta la , demeure dcc Gra nd Lama , (les sources dzc Gang e é’

a’

des pay s circon

v oi s i n s , le tou t l z'

re’

des Car i es C/zz’

n oi s es (5°

Ta rtares , pa r le P . Ga zcéi l , de l a

Compagn i e de yes u s , av ec des R emargn es dzc i n é‘

in e

Father Gaub il is by no means su re of the correctne s s of the Lamas ’ map,as he says

Je ne saurois b ien répondre du détour 81 de la figure du Gange d ’

ab ord apres s a source .

Le P . R égis cro it aus s i que ce la do it étre corrigé . Je s u is b ien sfir qu e le s pos itions des deuxCarte s Chino is e 81 Tartare que j ’a i vue s , ne s ont pas e xactes dans ce tte carte de la source duGange . L

entredeux de toute s le s ri v iere s marquee s dans cette Carte , le s environs tout lepa ys e s t montagne u x .

Gaub il gi ves a l is t of the pos itions of all the place s ente red on the map .

For Lake Lapama he has 29°

50' N . lat .

, 3 5°

50' W . long . of Peking, for Lake

Lanka 29°

50'

and 3 6°

30' resp .

, and for Mont Cantes (Kai las ) 30°

30'

and 35°

50

res p .

, and he adds :

Ces pos it ions s ont fort approchante s de s Cartes Ch ino is es 81 Tartares que j ’a i vue s .

E l le s me parois s ent fautive s . E lle s n’

ont été pris e s que s ur le rapport de s gens du pay s . La

mes ure actue lle , fa ite par de s Lamas , a donné la pos ition du Mont Cantes 81 de s lac Lanka 8 :

Lapama. Le s Lamas y allérent de Poutala en mesurant.On d

Anv ille’

s map, the shee t of wes te rn Tibe t (Pl . LI), the lakes and

Ka i las are on the same lati tude as on Gaub il’

s . Gaub il’

s mate rial is drawn fromChine s e and Tartar maps . The whole s i tuation and the hydrographical arrangement is the same as on d

Anv ille’

s Lama map . The name s on both maps are

also ve ry much the s ame , although s pe lt in diffe rent ways . Otherw is e the habitusof the maps is s omewhat di ffe rent . Gaub il

s map was published in 1 7 29 in Soucie t’

s

Oéscrv a t i on s , and d’

Anv ille’

s in 1 7 3 3 . I am not aware whe the r Gaub il has gothis materia l fromthe firs t Lama map which was rejected in 1 7 1 1 or whe the r he hasdrawn his s ke tch fromthe same mate rial as d ’

Anv ille , that is to say the later and

more re l iable Lama map.

2 But i t may b e that Gaub il has us ed earlie r nativemate rial . 3 It is a ls o in tere s ting to learn fromGaub il that the Lamas went fromLhas a to the Ka i las and the two lake s , su rveying on the i r way up . Tha t is whythe map is be tt e r in the we s t than in the eas t , whe re dange r aros e , and be tter alongthe Tsangpo than north of this rive r. Pe rhaps they did not go at all to Bongbaand othe r places north of the Ts angpo and on ly had to trus t the ve rba l informationthey obta ined about the s e regions . At le as t one ge ts that impres s ion when comparing the country round Ka i las and the lakes w ith othe r parts of Tibe t on the i rmap .

2 Gaub il has a little map of the sources of the Gange s , which I reproduce here as P l . LIII .2 In h is explan ation to the maps Anque til du Perron says : »E nfin la Figure V Ie (Pl . LIII)presente la faus s e Source du Gange , te l le qu ’

el le s e trouve dans la carte des Lamas Chinois , dans ce llesde MM . d

Anv ille , R enne ll 81 C .

3 Bon in i s right in suppos ing it to b e Chinese , on account of the characteristic way in whichthe mountains and lakes have been drawn , the almost Square outl ines of the river-course s , and the

Chinese or Mongolian orthography of the names .

292 FATHE R TIE FFE NTHALE R , FATH E R GAURIL AND AN QUE T IL DU PE RRON .

Reproducing the important map commun icated by fathe r Gaub il (Pl . LIII),Anqueti l du Pe rron gi ves i t the title : F a u s s e S ca rce dzc Ga ng e . And s ti ll this mapcon tains the prototype of the real ity . The re is , howeve r , the s ame great mis takeas in the Lamas ’ map of d

Anv ille and on Rennell’

s map : the Gange s ins tead of

the Satlej . Or rathe r : two big rive rs , para l le l w ith each othe r, flow ing towardsthe wes t . The northe rn one i s the Indus , the sou thern one the Satlej . But the

furthe r fate of thes e two rivers was enve loped in darkne s s , for the authors of the

map ,who be l ieved themto b e the fe ede rs of the Gange s , had themto join we s t of

Ladak and then le t the joint rive r turn eas t again and appear as the Ganges . The

northern branch has its source on the north-we s te rn s ide of Mont Cantes (Kai las ,Kentais s e , Kang Tis e). So this mu s t b e the north-eas te rn branch of the Indus .

Further wes t we find t wo we ll-known name s on this rive r : Ts chasr’rk ing and Latac ,

or Tashigang and Ladak , a lthough the firs t mentioned place is on the right ins teadof the le ft bank of the rive r .

The southe rn rive r i s regarded as the real s ou rce of the Gange s and at its

head our s pecial inte res t is concentrated . The eas te rn lake is cal led Lac Lapamaand i t is s ituated sou th of Mont Cantes and is s u rrounded by mountains to the eas tand s ou th as we l l . Three smal l and s hort rive rs e nte r the lake fromthe eas t . Thenorthern of the s e is obvious ly meant to b e the joint Pa-

chung and Pa—chen or one

or the othe r of them; the middle one is the Samo-tsangpo of Tokchen ; and the

s outhe rn one is Tage -tsangpo . On Gaub il’

s map all three are cal led Source s duGange , s o as to indicate that the river has real ly three heads .

The wate r of the s e thre e rive rs is s treaming fromLac Lapama as a short ands trong ri ve r to Lac Lanka , s ituated d i rect wes t ofLapama . And fromLanka (Langak-ts o)the rive r continues to the wes t, pas s ing Kouke (Guge), Ts eprong (Ts aparang), and

Ts choumourt i (Chumurt i), all places or dis tricts we l l known to b e s i tuated on the Satlej .I f we regard only the uppe r cours e of this rive r to a po in t a l ittle wes t of

Tschoumourti , we mus t confe s s that, i f on ly the name »Gange » were changed intoS atlej , the uppe r part of the hydrographica l sys temwou ld b e repres ented in a per

fectly corre ct way. The explore rs from China who drew this map had a cleare rview of phys ica l geography than many geographe rs of our own day. The onlypoint , except the name , in which the native explorers have be en uncerta in , is :

which of the three feeders to the Manasarov ar shou ld b e regarded as the rea lsource , and there fore they have ca l led all three the sou rces of the Gange s . To

s e ttle this que s tion one has only to measure the vo lume and care fu l ly s tudy the

o the r characte ris tics of the three feede rs , as I d id in 1 907 .

The channe l of commun ica tion be twe en the lake s , which was redis cove red bythe S tracheys in 1 846 and 1 848 , is marked on Gaub il

s map as clearly as evercan b e de s ired.

‘ So 200 years ago the re was not a s hadow of doubt as to i ts

2 BON 1 N says , 1. c. p . 3 48 : »Cc que cc croqu is ofi'

re de plus remarquable , c’

e st que , pour lapremi ere fois , il indique tres nettement la communication entre le Mansarov ar e t le Rakas -tal, qu i