lable at ScienceDirect

Journal of Archaeological Science 36 (2009) 2493–2501

Contents lists avai

Journal of Archaeological Science

journal homepage: ht tp: / /www.elsevier .com/locate/ jas

Classification and provenance of soapstones and garnet chlorite schistartifacts from Medieval sites of Tuscany (Central Italy): insights intothe Tyrrhenian and Adriatic trade

Patrizia Santi a,*, Alberto Renzulli a, Fabrizio Antonelli b, Antonio Alberti c

a Dipartimento di Scienze Geologiche, Tecnologie Chimiche e Ambientali, Universita degli Studi di Urbino, ‘‘Carlo Bo’’, 61029 Urbino, Italyb Laboratorio di Analisi dei Materiali Antichi, LAMA, Universita Iuav di Venezia, 30125 Venezia, Italyc Dipartimento di Scienze dell’Antichita e del Vicino Oriente, Universita Ca Foscari di Venezia, 30123 Venezia, Italy

a r t i c l e i n f o

Article history:Received 17 February 2009Received in revised form28 April 2009Accepted 17 May 2009

Keywords:Metamorphic rocksSoapstonesSchistsAlpine pietra ollareArchaeometryMiddle ageTuscany

* Corresponding author.E-mail address: patrizia.santi@uniurb.it (P. Santi).

0305-4403/$ – see front matter � 2009 Elsevier Ltd.doi:10.1016/j.jas.2009.05.006

a b s t r a c t

Soapstones (talc-bearing schists) and garnet chlorite schist artifacts found in Medieval archaeologicalsites of Tuscany (Central Italy) were classified, in order to define provenance of the different lithotypes. InItaly and throughout the Central Europe, these greenschist facies metamorphic rocks are generallyknown, among the archaeologists, as the pietra ollare from the Alps. The investigated Tuscan archaeo-logical sites are between 6 and 13th century AD and were strictly linked, in that period, to the welldefined network trade running along Tyrrhenian coast. Samples come from little containers used forcooking and preserving food and showing traces of lathe manufacturing at their sidewalls. According tomodal mineralogy, petrographic texture, XRD, SEM-EDS and whole rock chemistry we recognised, amongthe 18 studied findings, three different petrographic groups of the Alpine pietra ollare. (i) Fine grainedmagnesite talc schists (i.e. soapstones) from outcrops of the Central Alps located in the Valchiavennaarea. (ii) Garnet chlorite schists from the Valle d’Aosta region. (iii) Amphibole talc schists (i.e. soapstones)with a provenance in the Ticino area. It is worth noting that artifacts of pietra ollare lithotypes from theWestern Alps (i.e. garnet chlorite schists and amphibole talc schists) were not detected in the archae-ological sites of the Middle Adriatic coast of the Central Italy, belonging to the same Medieval timeinterval. This emphasises that the petrographic groups of pietra ollare from the Alps spread to the southof the Po Plain according to Western and Eastern trade along the Italian Peninsula, using respectively, theTyrrhenian and the Adriatic Sea commercial routes.

� 2009 Elsevier Ltd. All rights reserved.

1. Introduction

The Italian term pietra ollare (from the Latin olla, i.e. container) isnormally used by the European archaeologists for artifacts of basicand ultrabasic metamorphic rocks, mostly belonging to thegreenschist facies. All of the rocks referred to as pietra ollare aregeologically linked to the Apennine-Alpine ophiolite axis whichspread out for a long distance from Liguria (Italy) to the Stiria(Austria; Dietrich, 1980). These metamorphic rocks were highlyprized for the production of pots and food containers because wellworkable (related to the presence of talc) and their resistance to fire(Antonelli et al., 2006). One of the most common lithotypes ofAlpine pietra ollare is grey to pale-green, coarse to fine grainedrocks consisting of talc, chlorite and carbonate minerals (mostly

All rights reserved.

magnesite) cropping out in the Central and Western Alps (Mannoniet al., 1987). These magnesite-bearing talc schists are soapstones,usually derived from carbonate-free ultramafic rocks by interactionwith an externally derived CO2–H2O fluid phase (Bucher and Frey,2002). In fact, serpentinites represent the usual hydrated low-temperature alteration of mantle peridotites (lherzolites or harz-burgites), but small amounts of CO2 in the fluids are sufficient toconvert serpentine assemblages into carbonate-bearing rocks.According to Bucher and Frey (2002) the three major types ofcarbonate-bearing ultramafic rocks are: ophimagnesites(antigorite þ magnesite), soapstones (talc þ magnesite) and sag-vandites (enstatite þ magnesite). The term soapstone (synony-mous: steatite) is however commonly used for any talc-bearingschist. Pietra ollare groups soft grey-coloured (soapstones) andharder green-coloured lithotypes (chlorite schists, serpentineschists and amphibole schists) both very widespread among theartifacts of the archaeological sites throughout Italy and CentralEurope. A useful mineralogical data-base of the different

P. Santi et al. / Journal of Archaeological Science 36 (2009) 2493–25012494

metamorphic rocks known as pietra ollare and cropping out in theAlps, is represented by the 11 petrographic groups (Table 1)recognised by Mannoni et al. (1987). The Central Alps quarries ofValchiavenna and Valmalenco (magnesite talc schists and chloriteschists), those of the Valle d’Aosta (garnet chlorite schists) and theTicino Valley (various types of schists) in the Western Alps, werethe most exploited sources in the whole Medieval period (Table 1;Mannoni et al., 1987). The use of the green-coloured pietra ollare isdocumented from the pre-Roman Period (Mannoni et al., 1987). Thegrey-coloured soapstones coming from Valchiavenna started to beused in the 1st century AD (Bolla, 1991). Valle d’Aostagreen lithologies (garnet chlorite schists) became very widespreadstarting in the 4th century AD (Mollo Mezzena, 1987) with artifactslargely distributed around the outcropping areas of the WesternAlps (Piemonte, Liguria and Provence). The use of the green-col-oured pietra ollare gradually decreased with time in the Middle Agewhereas the use of grey-coloured rocks (i.e. soapstones) increased(Lusuardi Siena and Sannazaro, 1987). This seems to be confirmedby recent findings of pietra ollare from Medieval archaeologicalstrata of additional excavations in Pisa (works are still in progress)where the proportion of grey-coloured lithotypes strongly increasewith time.

Provenance of pietra ollare from Alpine quarries and productioncentres was established for numerous findings in Roman andMiddle Age archaeological sites of (i) Northern Italy (Mannoni andMessiga, 1980; Mannoni, 1986; Mannoni et al., 1987; Alberti, 1997;Bonazza et al., 1999; Malaguti and Zane, 1999) and (ii) Adriaticregions of the Central-Southern Italy (Mannoni, 1986; Alberti, 1997;Santi et al., 2005). No comparable detailed mineralogical, petro-graphic and geochemical data exist for the pietra ollare artifactsfound in Medieval (6–13th century AD) archaeological sites ofTuscany. In this work we present a petrographic classification andrelative provenance of 18 pietra ollare fragments coming fromarchaeological excavations located in the provinces of Pisa andLivorno and the cities of Pisa and Florence (Fig. 1). Although themost useful parameter to unravel source areas of the pietra ollaregroups is the modal mineralogy, we also compared the whole rockchemical data of the Tuscan findings with the available literaturedata of the Alpine outcrops such as those of the Valle d’Aosta(Mugnaini, 1995), Valmalenco (Bonazza et al., 1999), and Valchia-venna (Santi et al., 2005). Due to the chemical heterogeneity of thesingle quarries, statistical processing of the whole rock chemicaldata was not carried out. Whole rock chemical comparisons will beused to support a provenance however well constrained by modalmineralogy and texture. These new mineralogical, petrographic

Table 1Summary on the macroscopic colour, modal mineralogy, classification, petrographic gro(according to Mannoni et al., 1987).

Macroscopiccolour

Modalmineralogy

Classi

Light grey-green Talc, carbonates (magnesite),chlorite, opaques

Magnschist

Light grey-green Talc, carbonates, chlorite,amphibole, opaques

Amphschist

Green Chlorite, epidote, garnet,chloritoid, opaques � talc

Chlor

Light green Serpentine, chlorite, talc,opaques � carbonates � olivine

Serpe

Dark green Pyroxene, amphibole, talc � olivine� mica � plagioclase

Meta

Dark green-grey Amphibole, chlorite, opaques � talc� mica � quartz � epidote

Amph

Grey-white Olivine, talc, chlorite, opaques� carbonates � serpentine

Olivin

Green Chlorite, epidote, opaques +/-; albite � mica� quartz

Prasin

and whole rock chemical data will be finally compared with theavailable provenances of the pietra ollare artifacts from otherarchaeological sites of Central Italy, in particular of those con-cerning Medieval findings of the same time interval found in theregions of the Middle Adriatic coast (Marche and Abruzzo; Fig. 1).A framework of the pietra ollare trade networks during the MiddleAge for the Central Italian regions will be finally outlined.

2. Exploitation technology and trade of the pietraollare in antiquity

The Alpine pietra ollare lithotypes were exploited since theprehistoric age and widely used from the Roman Empire to theentire Middle Age when the artifacts became very widespreadthroughout Italy and Central Europe (Mannoni and Messiga, 1980;Mannoni et al., 1987). The most ancient recorded artifacts as oillamps and crucibles, dating back to the Iron Age, come from somevalleys of the Central Alps (Mannoni and Messiga, 1980) in areaslocated very close to the outcrops. The production of small artifacts(containers for cooking and preserving food) with a local networkdistribution, continued until the 1st century AD. Manufacturing ofthe pietra ollare using the lathe is known since the Iron Age. Theintroduction of big hydraulic lathes in the 4th century AD couldhave produced enough pietra ollare to reach areas outside theproduction zones (extra-regional trade). According to historicdocuments (e.g. Naturalis Historiae, Pliny the Elder, 1st century AD;Enciclopedie, Diderot, 1782) the use of pietra ollare for makingartifacts in the Alpine valleys persisted for almost 18 centuries. Theproduction of the pietra ollare containers attained a peak during the5–7th century AD as shown by findings in archaeological sites ofItaly, Switzerland, France, Austria and Germany (Mannoni andMessiga, 1980).

Two type of excavations for pietra ollare were generally recog-nised, correlated with the nature of raw material outcrops:(i) surface exploitation working erratic blocks transported alongfluvial or glacial valleys, far from the primary outcrops; (ii) opencastquarries and underground galleries developed in lenses or veins(Donati, 1986a). The exploitation of the pietra ollare in the Alpsremoved entire outcrops leaving relatively few traces behind.Numerous traces of ‘‘drillings’’ tests are represented by 20–25 cmlong (the average height necessary to make an artefact using thelathe) holes which were made on the surface of the pietra ollareoutcrops to detect the rock suitability (Donati, 1986a). In someAlpine artifacts, Mannoni (1986 and reference therein) and Donati(1986b) recognised four different production techniques on the

ups and provenance of the different pietra ollare lithotypes from the Alps

fication Petrographic groups Alpine Provenance

esite talcs

C (coarse grained)D (fine grained)

Valtellina, Valbregaglia,Valchiavenna

ibole talcs

B (coarse grained)E (fine grained)

Ticino and Toce Valleys

ite schists F (fine grained)G (coarse grained)

Valle d’Aosta, Valtellina,Valmalenco, Val di Lanzo (F)

ntine schists A Ticino and Toce Valleys

-gabbros H Ticino and Toce Valleys (H)

ibole schists I Ticino Valley

e schists K Ticino Valley

ites L Rare outcrops

Fig. 1. Pietra ollare findings from Medieval archaeological sites of Central Italy. Location of the Tuscan 6–13th century AD archaeological sites from the present work. 1: Pisa; 2:Vicopisano (PI); 3: San Miniato (PI); 4: Florence; 5: Collesalvetti (LI); 6: Monteverdi (PI); 7: Populonia (LI); 8: Piombino (LI).

P. Santi et al. / Journal of Archaeological Science 36 (2009) 2493–2501 2495

basis of the stylistic and technological analyses: (i) carving withmetallic tools, (ii) surface finishing by manual lathes (since the 1stcentury AD), (iii) using horizontal manual and/or hydraulic lathes(from 3 to 6th century AD); (iv) making several pots starting froma single rock block (‘‘onion technique’’) using hydraulic horizontallathes (since the 17th century AD).

Table 2Archaeological sites of the pietra ollare samples from the present work. Modal mineralogpietra ollare according to the groups of Mannoni et al. (1987) are also given. AbbreviatioOpq ¼ Opaque minerals, Grt ¼ Garnet.

Samples Location of the archaeological sites

Province of PISASMV ’98 US2038 Monastero di S.Michele alla Verruca, Vicopisano

(11th cen.)PI-VF ’01 US1471 Via Facchini, Pisa (11th cen.)PIPD I 173 Piazza Dante, Pisa (middle 10–middle 13th cen.)PIPD II 19 (scattered) Piazza Dante, PisaPM 92 85 Piazza dei Miracoli, Pisa (6th cen.)MV 2004 US1023 Monastero di S.Pietro di MonteverdiSGO1 (scattered) San Genesio, San MiniatoSGO2 (scattered) San Genesio, San MiniatoSGO3 US 1001 San Genesio, San MiniatoSGO4 ‘‘area 22 000’’ San Genesio, San Miniato

Province of LIVORNOPB ’00 S.1, UT1, S.4 Collesalvetti (11–12th cen.)SQ 2005 US1179 Monastero di S.Quirico, PiombinoPCS 02, Sg.XX, US329 Populonia (11–13th cen.)PCS 05, Sg.XX, US480 Populonia (11–13th cen.)PCS 06, Sg.XXXIVb Populonia (probably 11–13th cen.)PCS (scattered) Populonia (11–13th cen.)

FLORENCEPO1 phase 12 Via De’Castellani, Firenze (8–10th cen.)PO4 phase 19 Via De’Castellani, Firenze (13th cen.)

A standardisation of the pietra ollare artifacts, characterised bythe progressive thinning of the walls, was developed since the 4thcentury AD as the result of using heavy lathes (Donati, 1986b;Alberti, 1997). Most of the main production centres of pietra ollarecontainers were located close to the quarries or in the nearestareas located along the main waterways, rivers and/or lakes

y, petrographic classification and possible lithotypes of the Alpinens: Mgs ¼ magnesite, Tlc ¼ Talc, Chl ¼ Chlorite, Amph ¼ Amphibole,

Modal mineralogy Petrographicclassification

Mannoni et al.(1987)

Tlc, Chl, Amph, Opq, � Mgs Amphibole talc schist E

Tlc, Chl, Amph, Opq, � Mgs Amphibole talc schist ETlc, Chl, Amph, Opq, � Mgs Amphibole talc schist ETlc, Chl, Amph, Opq, � Mgs Amphibole talc schist EChl, Opq, Grt Garnet chlorite schist F–GTlc, Mgs, Chl, Opq Magnesite talc schist DTlc, Mgs, Chl, Opq Magnesite talc schist DTlc, Mgs, Chl, Opq Magnesite talc schist DTlc, Chl, Amph, Opq, � Mgs Amphibole talc schist ETlc, Mgs, Chl, Opq Magnesite talc schist D

Tlc, Mgs, Chl, Opq Magnesite talc schist DChl, Opq, Grt Garnet chlorite schist F–GTlc, Mgs, Chl, Opq Magnesite talc schist DTlc, Mgs, Chl, Opq Magnesite talc schist DChl, Opq, Grt Garnet chlorite schist F–GChl, Opq, Grt Garnet chlorite schist F–G

Chl, Opq, Grt Garnet chlorite schist F–GTlc, Mgs, Chl, Opq Magnesite talc schist D

Fig. 2. Representative fragment of the pietra ollare studied artifacts, showing thefeatures of the external (a) and internal (b) walls.

P. Santi et al. / Journal of Archaeological Science 36 (2009) 2493–25012496

(Gaggi and Leoni, 1984). Stones were carved near the quarriesbecause of the difficulty in the transport of big blocks due to theirweight (Donati, 1986a): the average density of the pietra ollare isbetween 2500 and 3300 kg/m3 depending on the different litotypes(Robertson, 1982; Antonelli et al., 2006).

3. Sampling and analytical methods

The studied fragments of pietra ollare (18 samples) come fromdifferent Medieval archaeological sites located in Tuscany; inparticular, 10 samples from the province of Pisa, 6 from the prov-ince of Livorno and 2 from Florence (Fig. 1 and Table 2). Most ofthem come from little containers used for cooking and preservingfoods as emphasised by the smoky/burned external portion(Fig. 2a). Thickness varies between 6 and 14 mm. Traces of lathe

Fig. 3. Representative thin section petrography and texture of some pietra ollare samples:Abbreviations: Mgs ¼ magnesite, Tlc ¼ Talc, Chl ¼ Chlorite, Amph ¼ Amphibole, Grt ¼ Gar

manufacturing are frequently present both in the external andinternal sidewalls (Fig. 2a, b).

Modal mineralogy analyses were done by optical polarisingmicroscope (thin sections) and X-Ray Diffraction (XRD) techniques.XRD data come from a Philips PW1840 diffractometer (radiationCu Ka/Ni with 40 Kv and 20 mA; detection limit 5%; University IUAVof Venice) using powdered samples. Scanning Electron Microscope(SEM) qualitative analyses of mineral phases were carried out usinga Philips 515 equipped with a PV9900 Energy Dispersive System(EDS) analyser (University of Urbino). Whole rock major (Induc-tively Coupled Plasma-Optical Emission Spectrometry) and trace(Inductively Coupled Plasma-Mass Spectrometry) elements anal-yses were determined at the Activation Laboratories LTD (Ancaster,Canada) using a ‘‘Thermo Jarrell-Ash ENVIRO II ICP’’ and a ‘‘PerkinElmer SCIEX ELAN 6000 ICP-MS’’ respectively. Whole rock analyseserrors are � 1% for major oxides and �3% for trace elements.

4. Results: mineralogy, petrography and geochemistry

Petrography (Fig. 3), XRD (Table 3), mineral analyses (Fig. 4) andwhole rock chemical compositions (Table 4) recognised threedifferent lithotypes of pietra ollare among the studied samples.

4.1. Magnesite talc schists (soapstones)

Magnesite talc schists samples (Fig. 3a) are the most abundantamong the studied pietra ollare artifacts (44% of the total) andcorrespond to the ‘‘D group’’ of Mannoni et al. (1987). They arecharacterised by the typical mineralogy (see modal mineralogy inTable 2) and texture of soapstones s.s.: fine to medium grainedrocks with idioblastic talc, heteroblastic magnesite and clinochlore(Fig. 4). Rare opaques minerals (ilmenite, magnetite and pyrite) arealso present. The sample coming from ‘‘Collesalvetti’’ (Livornoprovince; Table 2) contains a great amount of magnesite with a well

(a) magnesite talc schist, (b) amphibole talc schist, (c) and (d) garnet chlorite schists.net.

Table 3XRD analyses of the pietra ollare samples. þþþ ¼ very abundant, þþ ¼ average abundance, þ ¼ present, � ¼ traces.

Magnesite talc schists

Sample MV 2004 SG01 SG02 SG04 PB ’00 PCS 02 PCS 05 PO4

Talc þþþ þþþ þþþ þþþ þþþ þþþ þþþ þþþClinochlore þþ þþ þ þþþ þ þþþ þþ þþMagnesite þ þþ þþ þþ þþ þþþ þþþ �Dolomite þ þ þþ þ þBrucite �Magnetite � þ

Amphibole talc schists Garnet chlorite schistsSample SMV ’98 PI-VF ’01 PIPD I 173 PIPD II 19 SG03 PM 92 85 SQ 2005 PCS 06 PCS PO1Talc þþþ þþþ þþþ þþþ þþþ þþþClinochlore þþ þþ þþ þ þþ þþþ þþþ þþþ þþþMagnesite � �Dolomite þCalcite þAmphibole � þ þ þ þ þBrucite �Magnetite � � � �

0

500

1000

1500

2000

2500

Actinolite

Mg Kev

Cou

nts

Sample: PO1Si

AlFeCa

0

500

1000

1500

2000

2500

3000

3500

4000

Almandine

Kev

Cou

nts

Sample: PO1Si

Al

Fe

Ca

0

500

1000

1500

2000

2500

3000

3500

4000

4500

Chloritoid

Kev

Cou

nts

Sample: SQ 2005

Si

Al

Fe

Mg

0

500

1000

1500

2000

2500

3000Epidote

Kev

Cou

nts

Sample: SQ 2005

Si

Al

Fe

Ca

Cr

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

TalcSi

Mg

Kev

Cou

nts

Fe

Sample: PO4

0

500

1000

1500

2000

2500

3000

3500

0 1 2 3 4 5 6 7 80 1 2 3 4 5 6 7 8

0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8

0 1 2 3 4 5 6 7 80 1 2 3 4 5 6 7 8

Clinochlore

FeCr Kev

Sample: SMV '98

Si

Al

Mg

Cou

nts

Fig. 4. Representative EDS spectra of some minerals in the studied samples.

P. Santi et al. / Journal of Archaeological Science 36 (2009) 2493–2501 2497

Table 4Major and trace elements analyses of the pietra ollare artifacts (ICP-OES-MS). PI ¼ samples from Pisa or province of Pisa; LI ¼ samples from the province of Livorno; FI ¼ samples from Florence.

Magnesite talc schists Amphibole talc schists Garnet chlorite schists

PI PI PI PI LI LI LI FI PI PI PI PI PI PI LI LI LI FI

MV 2004 SG01 SG02 SG04 PB ’00 PCS 02 PCS 05 PO4 SMV’98 PI-VF ’01 PIPD I 173 PIPD II 19 SG03 PM 92 85 SQ 2005 PCS 06 PCS PO1

SiO2 44.50 35.43 33.83 34.28 36.08 35.98 37.49 48.35 53.54 50.72 47.76 49.59 51.55 25.19 24.30 23.83 24.40 28.33Al2O3 2.32 2.07 0.74 2.04 2.20 2.70 2.06 2.43 3.23 3.69 5.73 1.95 1.13 19.94 21.66 20.28 24.17 16.81TiO2 0.04 0.06 0.01 0.08 0.10 0.05 0.05 0.08 0.10 0.12 0.06 0.07 0.14 0.43 0.16 0.35 0.40 0.52Fe2O3 (t) 5.51 7.01 6.37 6.71 7.48 7.42 7.11 7.18 6.11 12.02 11.29 13.72 15.05 35.79 34.49 36.68 33.83 34.70MnO 0.09 0.13 0.10 0.12 0.12 0.12 0.11 0.06 0.09 0.12 0.20 0.19 0.16 0.20 0.04 0.05 0.06 0.15MgO 29.63 34.03 33.94 33.12 32.48 34.06 33.28 27.68 30.63 27.92 27.61 25.07 26.60 10.41 11.50 10.35 9.93 9.85CaO 3.53 1.15 0.37 1.48 1.62 1.17 1.24 3.41 0.23 0.17 0.49 3.31 0.71 0.42 0.08 0.03 0.16 1.08Na2O 0.05 0.02 0.02 0.06 0.16 0.03 0.19 0.04 0.10 0.05 0.06 0.13 0.24 0.03 0.03 0.01 0.05 0.51K2O <0.01 <0.01 <0.01 <0.01 <0.01 0.02 <0.01 <0.01 0.03 0.02 <0.01 0.05 0.06 <0.01 <0.01 <0.01 0.03 0.03P2O5 0.02 0.04 0.05 0.01 0.08 0.11 0.12 0.05 0.11 0.02 0.08 0.18 0.03 0.03 <0.01 0.02 0.06 0.12LOI 13.17 21.07 23.09 20.33 18.23 18.69 19.24 10.32 5.99 4.95 6.33 4.76 4.12 6.95 7.98 7.94 7.58 6.67

Total 98.79 101.00 98.54 98.21 98.55 100.40 100.90 99.58 100.20 99.79 99.62 99.03 99.79 99.37 100.20 99.49 100.70 98.78

Sc 13 9 7 9 10 12 12 8 10 4 5 5 5 26 23 29 16 41V 40 37 26 29 41 44 43 69 40 21 11 17 43 95 87 136 84 233Ba 11 8 9 10 20 21 19 19 29 4 9 16 12 9 4 4 4 47Sr 19 9 4 11 13 8 9 30 6 <2 6 61 14 3 <2 <2 8 15Y 6 <2 <2 <2 2 <2 <2 <2 6 <2 <2 <2 <2 62 37 84 23 47Zr 5 5 <4 <4 10 <4 4 5 <4 10 7 4 10 393 100 318 494 39Cr 1900 1550 1620 1590 1670 1990 1490 1860 3760 2070 4240 2990 3090 2450 1450 1240 780 <20Co 68 87 70 82 80 61 80 61 75 125 134 128 52 249 353 250 216 271Ni 1220 1540 1250 1240 880 1240 1420 980 1040 960 1390 1160 710 420 300 270 230 140Cu <10 20 <10 20 40 90 30 90 50 70 50 70 280 <10 <10 <10 20 160Zn 50 40 30 40 60 50 40 50 100 110 140 150 100 60 130 40 40 70Ga 3 2 <1 2 3 3 2 4 10 7 9 5 4 25 25 21 23 29Rb <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 2 <2 <2 <2 <2 <2 <2 <2Nb <1 <1 <1 <1 <1 <1 <1 <1 11 1 <1 <1 1 3 3 4 4 2La 0.3 <0.1 0.2 0.1 0.4 0.2 0.1 0.1 0.5 0.1 0.2 0.6 3.9 3.6 0.9 15.9 8.8 8.4Ce 0.8 <0.1 0.3 0.2 0.9 0.4 0.3 0.4 1.1 0.2 0.4 1.8 7.6 10.2 2.1 50.9 25.2 22.2Pr 0.16 <0.05 <0.05 <0.05 0.14 0.05 <0.05 0.07 0.18 <0.05 0.06 0.28 0.76 1.64 0.35 7.27 3.55 3.16Nd 1 0.1 0.2 0.2 0.8 0.3 0.2 0.6 1.1 0.2 0.3 1.6 2.6 8.9 2.3 37.4 19 16.5Sm 0.4 <0.1 <0.1 <0.1 0.2 0.1 <0.1 0.3 0.4 <0.1 <0.1 0.5 0.5 2.8 0.9 11.4 6 5.4Eu 0.14 <0.05 <0.05 <0.05 0.06 <0.05 <0.05 0.06 <0.05 <0.05 <0.05 0.08 0.05 0.2 <0.05 0.18 0.21 0.33Gd 0.8 0.1 <0.1 0.1 0.3 0.2 0.1 0.5 0.7 <0.1 <0.1 0.5 0.4 7.5 3.1 15.7 7.6 7.8Tb 0.2 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 0.2 <0.1 <0.1 <0.1 <0.1 1.6 0.8 2.9 1 1.5Dy 1.1 0.2 <0.1 0.2 0.4 0.3 0.2 0.5 1.2 <0.1 <0.1 0.5 0.3 10.5 5.8 16.9 5 9Ho 0.2 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 0.3 <0.1 <0.1 <0.1 <0.1 2.1 1.3 3.1 0.9 1.8Er 0.7 0.2 <0.1 0.2 0.3 0.2 0.1 0.3 0.8 <0.1 <0.1 0.3 0.2 6.4 4.1 8.9 2.6 5Tm 0.1 <0.05 <0.05 <0.05 0.05 <0.05 <0.05 <0.05 0.12 <0.05 <0.05 <0.05 <0.05 1.02 0.67 1.34 0.43 0.76Yb 0.7 0.2 <0.1 0.2 0.4 0.2 0.1 0.2 0.8 <0.1 <0.1 0.3 0.2 6.8 4.3 8.2 2.9 4.8Lu 0.09 <0.04 <0.04 <0.04 0.07 <0.04 <0.04 <0.04 0.1 <0.04 <0.04 <0.04 <0.04 1.02 0.64 1.11 0.44 0.64Hf <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 0.2 <0.2 <0.2 0.2 9.3 3.4 10.2 12.8 1.4Th <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 0.4 0.4 0.3 0.5 0.6 0.2 0.2 1.2 0.2 0.3U 0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 0.1 2.9 <0.1 <0.1 0.3 0.1 0.1 <0.1 0.5 0.3 0.2

P.Santiet

al./Journal

ofA

rchaeologicalScience

36(2009)

2493–25012498

P. Santi et al. / Journal of Archaeological Science 36 (2009) 2493–2501 2499

developed idioblastic crystals (up to 4 mm). The chemical compo-sition of all the magnesite talc schists samples (Table 4) shows SiO2

33.8–48.3 wt%, low Al2O3 (0.7–2.7 wt%) and high MgO (27.7–34.1 wt%) values (Fig. 5). Magnesite is also responsible for the veryhigh LOI values (10.3–23.1 wt%; Table 4) of this pietra ollare lith-otype. Some other (undetected) accessory silicate minerals arepresent in the two samples showing the highest SiO2 contents (MV2004 and PO4; Table 4) which are also characterised by an elevatedCaO content of 3.4–3.5 wt% due to the presence of dolomite(detected by XRD; Table 3). In all of the samples without dolomite,CaO content is in fact restricted to 0.4–1.6 wt%. Magnesite talcschists show elevated contents of Cr (1490–1990 ppm) andNi (880–1540 ppm). Very low LREE (e.g. Ce � 0.9 ppm) and HREE(e.g. Yb � 0.7 ppm) contents seem to exclude the presence ofaccessory phases in which REE can enter the mineral structures.Rare brucite was detected by XRD analyses (Table 3).

4.2. Amphibole talc schists (soapstones)

Amphibole talc schists (corresponding to the ‘‘E group’’ ofMannoni et al., 1987) represent 28% of the studied pietra ollarefragments. They consist of fine grained rocks with heteroblastic talc(Fig. 3b) and clinochlore as the major minerals and minor amountof actinolitic amphibole (Fig. 4). Opaque minerals (magnetite andpyrite) and magnesite can be present. This modal mineralogy (Table2) is well reflected in the whole rock chemistry (Table 4) as thesesamples show high SiO2 (47.8–53.5 wt%) and MgO (25.1–30.6 wt%)

0

5

10

15

20

25

30

35

40

SiO2

Fe 2

O3

Garnet chloriteschists

Magnesite talc schists

Amphibole talc schists

0

5

10

15

20

25

22 27 32 37 42 47 52

22 27 32 37 42 47 52

SiO2

l 2O

3A

Amphibole talc schists

Garnet chloriteschists

Magnesite talc schists

Findings from the province of Pisa Findings from the

Valchiavenna outcrops (Santi et al. 2005) Valmalenco outcro

Fig. 5. Binary diagrams for the pietra ollare studied artifac

and relatively low Al2O3 (1.1–5.7 wt%) values (Fig. 5). High Crcontents (2070–4240 ppm) reflect the presence of chromianmagnetite, Ni (710–1390 ppm) and HREE (e.g. Yb � 0.8 ppm)contents are comparable to those of the magnesite talc schists. LREEare relatively high (e.g. Ce up to 7.6 ppm) and this may be due to thepresence of some accessory phase (epidote?) although not detectedby XRD or optical microscope.

4.3. Garnet chlorite schists

Garnet chlorite schists correspond to the ‘‘F and G groups’’ ofMannoni et al. (1987) and represent 28% of the total investigatedsamples (see modal mineralogy in Table 2). This group consists ofmedium to fine grained rocks with abundant heteroblastic greenclinochlore, opaque minerals and rounded almandine garnet (Figs.3c, d and 4). Rare crystals of quartz were also detected. Chloritoidand epidote are also present (Fig. 4), the former being characterisedby crystals up to 1.6 mm in size (sample SQ2005) and indigo to bluepleocroism. Rare brucite is detected by XRD analyses (Table 3).All the samples have a well developed schistosity, locally with adecussate texture (Fig. 3d). They are chemically detectable (Table 4)with high Al2O3 (16.8–24.2 wt%) as the result of abundant chlorite.Fe2O3 tot (33.8–36.7 wt%) prevails over MgO (9.8–11.5 wt%) whereasSiO2 (23.8–28.3 wt%) is the lowest among the three lithotypes.Relatively high V (84–233 ppm) and Sc (16–41 ppm) with respect tothe other two petrographic groups, can be attributed to themagnetite-spinel series. High Fe2O3 tot, HREE (Gd up to 15.7 ppm; Yb

9

14

19

24

29

34

39

44

SiO2

MgO

Amphibole talc schists

Garnet chloriteschists

Magnesite talc schists

0

1000

2000

3000

4000

5000

6000

7000

SiO2

Cr

Garnet chloriteschists

Amphibole talc schists

Magnesite talc schists

22 27 32 37 42 47 52

22 27 32 37 42 47 52

province of Livorno Findings from Florence

ps (Bonazza, 1999) Valle d'Aosta outcrops (Mugnaini, 1995)

ts compared with literature data of Alpine outcrops.

Fig. 6. The Western (Tyrrhenian Sea) and Eastern (Adriatic Sea) trade of the Alpinepietra ollare during the Middle Age. Arrows indicate different provenance of thepetrographic groups. Filled circles show the pietra ollare findings in the middle Adriatic(Santi et al., 2005) and Tyrrhenian (this work) regions. VA: Valle d’Aosta; TV: TicinoValley; VC: Valchiavenna.

P. Santi et al. / Journal of Archaeological Science 36 (2009) 2493–25012500

up to 8.2 ppm; Lu up to 1.1 ppm), Hf (up to 12.8 ppm), Y (23–84 ppm)and Zr (up to 494 ppm) values are well related to the modalalmandine crystals (Table 4). Relatively high contents of LREE (La upto 15.9 and Ce up to 50.9 ppm) are linked to the presence of modalepidote. The PO1 sample (Table 2) shows a slightly different modalcomposition (Fig. 3d), with clinochlore, actinolitic amphibole,magnetite, ilmenite, almandine garnet � albite. This distinctivemineralogical assemblage is also reflected in the variation diagrams(Fig. 5) where the PO1 sample shows the highest SiO2 (28.3%) andthe lowest Al2O3 (16.8 wt%) contents among the samples of thispetrographic group. The relatively high Ba (47 ppm) with respect tothe other samples of garnet chlorite schists (Ba 4–9 ppm) is certainlydue to the presence of some albite crystals (Na2O 0.5 wt% in thewhole rock chemistry).

5. Discussion

The archaeological sites in the Livorno and Pisa provinces andFlorence all contain magnesite talc schists and garnet chloriteschists; amphibole talc schists are represented only by samplesfrom Pisa province (Table 2).

Soapstones, detected in all the sites, are the most abundant (13total samples consisting of both magnesite talc schists andamphibole talc schists) whereas garnet chlorite schists are subor-dinate (5 samples). According to the archaeometric literature(Mannoni et al., 1987) the different groups of Medieval pietra ollarediscovered in the Italian archaeological sites come from the Alps.

Grey-coloured pietra ollare lithotypes (Table 1) consisting oftalc þ chlorite þ magnesite come from the Central Alps

(Valchiavenna, Valtellina and Valbregaglia), whereas amphiboletalc schists have a provenance from the Western Alps (Ticino andToce valleys; Table 1; Donati, 1986b; Pfeifer and Serneels, 1986;Mannoni et al., 1987). The most abundant green-coloured pietraollare lithotypes (Table 1), where chlorite is the prevailing mineral,come from both the Western Alps (Valle d’Aosta) and the CentralAlps (Valtellina and Valmalenco) according to Pfeifer and Serneels(1986) and Mannoni et al. (1987).

5.1. Provenance of soapstones

Among the investigated samples, the large presence (72%) offine grained, grey-coloured soapstones (‘‘D and E groups’’ ofMannoni et al., 1987; Table 1) is probably due to the fact that theserocks are well workable by means of lathe. Concerning all theMedieval archaeological findings of pietra ollare studied in theMiddle Adriatic coast of the Central-Southern Italy (5–15th centuryAD; Marche and Abruzzo regions; Santi et al., 2005) only soap-stones belonging to the magnesite talc schist ‘‘D group’’ (Mannoniet al., 1987), are present. Findings of pietra ollare from the Adriatic-Ionian regions of Southern Italy (artifacts from Ugento, Lecce)belong to the same group of magnesite talc schists (Mannoni, 1986).

Modal mineralogy of the magnesite talc schists from the‘‘D group’’ of Mannoni et al. (1987) indicates that quarries are mainlylocated in Valchiavenna (Central Alps). Compositions of the eightTuscan findings belonging to this group are clustered in and aroundthe range of the low Al2O3 and Fe2O3 and high MgO contents of theValchiavenna quarries (Fig. 5). We can summarise that the finegrained magnesite talc schists found in all the archaeological sites ofTuscany, come from the Central Alps, in particular from the wellknown quarries located in the Valchiavenna area.

The five Tuscan findings made of fine grained amphibole talcschists probably represent the ‘‘E group’’ (Table 1) of Mannoni et al.(1987). This modal mineralogy suggests provenance from theTicino valley (Western Alps). Comparison with mineralogy andpetrography given by Pfeifer and Serneels (1986) for some outcropsof amphibole talc rocks from the Ticino valley, seem to support thisprovenance. Unfortunately, no chemical data are available for theseoutcrops. The five samples of amphibole talc schists are charac-terised by high SiO2 and Cr, low Al2O3 and intermediate MgO–Fe2O3

contents and do not match any quarries of Valchiavenna, Valma-lenco and Valle d’Aosta (Fig. 5).

5.2. Provenance of garnet chlorite schists

The five artifacts made of garnet chlorite schists closely matchthe ‘‘F group’’ of Mannoni et al. (1987; Table 1) from Valle d’Aosta(Western Alps). The chlorite schists from Valmalenco are, however,not so enriched in garnet. Only the Valle d’Aosta outcrops approachthe Fe2O3-rich, MgO-poor samples of the five Tuscan garnet chlo-rite schists (Fig. 5).

6. Conclusions

The presence in the archaeological sites of Tuscany dating to6–13th century AD of some pietra ollare lithotypes from Valled’Aosta and Ticino valley can be strictly associated with a Westernpietra ollare trade along the Tyrrhenian Sea (Alberti, 1997). Thecirculation of these groups of pietra ollare towards the MiddleAdriatic coast was precluded by the presence of the NorthernAppenine barrier (Fig. 6). In the archaeological sites of 5–15thcentury AD of the Middle Adriatic coast (Alberti, 1997; Santi et al.,2005) only the soapstones coming from the Central Alps, werefound, supporting an Eastern pietra ollare trade, from the

P. Santi et al. / Journal of Archaeological Science 36 (2009) 2493–2501 2501

Valchiavenna area to the Adriatic Sea along the main waterways:Mera River–Como Lake and finally the rivers Adda and Po. Never-theless, the widespread talc magnesite schists from Valchiavennaalso reached Tuscany through the Western trade. We can conclude,as shown in Fig. 6, that the existence of the Tyrrhenian and Adriatictrade during the Middle Age determined the distribution ofdifferent Alpine pietra ollare lithotypes in the regions of the Italianpeninsula to the south of the Po Plain.

Acknowledgements

We wish to thank C. D’Amico (University of Bologna) whoprovided us many thin sections of pietra ollare samples from Valled’Aosta. We are also grateful to S. Swanson (University of Georgia)and an anonymous reviewer for their useful suggestions.

References

Alberti, A., 1997. Produzione e commercializzazione della pietra ollare in Italiasettentrionale tra tardoantico e altomedioevo. In: 1� Congresso Nazionale diArcheologia Medievale. Pisa, 1997. Proceedings, pp. 335–339.

Antonelli, F., Santi, P., Renzulli, A., Bonazza, A., 2006. Petrographic features andthermal behaviour of the historically known ‘‘pietra ollare’’ from the ItalianCentral Alps (Valchiavenna and Valmalenco). In: Maggetti, M., Messiga, B.(Eds.), Geomaterials in Cultural Heritage, 257. Geological Society of London, pp.229–239. Special Publication.

Bolla, M., 1991. Recipienti in pietra ollare. In: Caporusso, D. (Ed.), Scavi MM3.Ricerche di archeologia urbana a Milano durante la costruzione della linea 3della metropolitana 1982–1990, 3.2. I reperti, pp. 11–37. Milano.

Bonazza, A., Lazzarini, L., Vaccaro, C., 1999. Nuovo contributo archeometrico allostudio della ‘‘pietra ollare’’ padana. In: 6� Giornata ‘Le scienze e l’archeometria’,Este, pp. 171–180.

Bucher, K., Frey, M., 2002. Petrogenesis of Metamorphic Rocks. Springer-Verlag,Berlin, 341 pp.

Diderot, D., 1782. Encyclopedie, ou Dictionaire raisonne des sciences, des arts et desmetiers. Lausanne-Berne Ed.

Dietrich, V.J., 1980. The distribution of ophiolites in the Alps. Ofioliti, Special issueon Thetyan ophiolites, 1, Western Area, pp. 7–51.

Donati, B., 1986a. La pietra ollare in Valmaggia dalla cava al laboratorio. In: 2000anni di pietra ollare. Quaderni d’informazione, Bellinzona, pp. 17–42.

Donati, P., 1986b. Archeologia e pietra ollare nell’area Ticinese. In: 2000 anni dipietra ollare. Quaderni d’informazione, Bellinzona, pp. 71–141.

Gaggi, S., Leoni, B., 1984. La pietra ollare. In: Quaderni della Provincia, Amminis-trazione Provinciale di Sondrio.

Lusuardi Siena, S., Sannazaro, M., 1987. Pietra ollare di Luni, La pietra ollare inLiguria. In: Atti della Giornata di Studio in ricordo di Lella Massari, Finale Ligure,1985, pp. 165–198.

Malaguti, C., Zane, A., 1999. La pietra ollare nell’Italia nord-orientale. ArcheologiaMedievale 26, 463–479.

Mannoni, T., Messiga, B., 1980. La produzione e la diffusione dei recipienti di pietraollare nell’Alto Medioevo. In: 6� Congresso Internazionale di studi sull’AltoMedioevo, Milano, 1978. Proceedings, pp. 501–522.

Mannoni, T., 1986. Materia prima e scarti di produzione dei recipienti in pietraollare. Rivista di Studi Liguri 1–4 (LII), 155–164.

Mannoni, T., Pfeifer, H.R., Serneels, V., 1987. Giacimenti e cave di pietra ollare nelleAlpi. In: La Pietra Ollare dalla Preistoria all’Eta Moderna, Como, 1982.Proceedings, pp. 7–45.

Mollo Mezzena, R., 1987. Primi elementi per lo studio della pietra ollare in valled’Aosta. In: La Pietra Ollare dalla Preistoria all’Eta Moderna, Como, 1982.Proceedings, pp. 59–114.

Mugnaini, D., 1995. Studio petroarcheometrico dei manufatti strumentali e orna-mentali in serpentinite e clorito scisto di siti neolitici friulani e trentini. Thesys,University of Bologna, 110 pp.

Pfeifer, H.R., Serneels, V., 1986. Inventaire des gisements de pierre ollaireau Tessin et dans les regions voisines: aspects mineralogiques et miniers. In:2000 anni di pietra ollare. Quaderni d’informazione, Bellinzona, pp. 147–235.

Robertson, E.G., 1982. Physical properties of building stone. In: Conservation ofHistoric Stone Buildings and Monuments, Report of the Committee onConservation of Historic Stone Buildings and Monuments. National AcademyPress, Washington, D.C., pp. 62–86.

Santi, P., Antonelli, F., Renzulli, A., 2005. Provenance of Medieval pietra ollareartefacts found in archaeological sites of Central-Eastern Italy: insights into theAlpine soapstone trade. Archaeometry 47 (2), 253–264. Special issue: thelinking role of the Alps in past cultures: an archaeometric approach.