M Barber, D Field, P Topping (2000) Grime's Graves, Norfolk. RCHME Survey Report.

49
SUMMARY OF PREVIOUS WORK AT GRIMES GRAVES Introduction As yet, no interpretative synthesis of all the excavation data on Grimes Graves has been produced. The following account summarises the known episodes of antiquarian and archaeological investigation at the site, arranged in chronological order. There are, however, certain omissions where the results of minor excavations have not been published: numerous small exploratory trenches have been cut in various parts of the site (particularly in the years between the First and Second World Wars, when at least one excavation seems to have occurred every year), and in a number of cases detail remains rather elusive. It is also certain that a considerable amount of excavation and surface collection has taken place without record. In addition, the full publication of the most recent campaign, that undertaken by the British Museum in the mid-1970s, is still in preparation. A number of volumes have appeared, however, and as a result this most recent work is dealt with in a more summary manner than the earlier investigations. Early Investigations The first recorded examinations of the site appear to have occurred in February 1852. In addition to the note by Luke quoted earlier, both CR Manning (1855, 1872) and the Rev Pettigrew (1853) reported on excavations undertaken there in the same month. There is obviously a strong possibility that all three are referring to the same episode, although Luke appears to have been merely an observer, while Manning and Pettigrew both refer only to trenches dug under their own direction. Pettigrew reported to the British Archaeological Association on February 25th 1852 that he had ...just returned from visiting a place called Grime's Graves,...where he had opened two of the graves, and found in each some bones of animals, appearing to belong to the ox and the pig. The place was formerly the seat of war between the Saxons and the Danes. No other remains beyond those of the bones, laid in a bed of flint stones, were discovered. Those were in a cavity of coffin-like shape...

Transcript of M Barber, D Field, P Topping (2000) Grime's Graves, Norfolk. RCHME Survey Report.

SUMMARY OF PREVIOUS WORK AT GRIMES GRAVES

Introduction

As yet, no interpretative synthesis of all the excavation dataon Grimes Graves has been produced. The following accountsummarises the known episodes of antiquarian and archaeologicalinvestigation at the site, arranged in chronological order.There are, however, certain omissions where the results ofminor excavations have not been published: numerous smallexploratory trenches have been cut in various parts of the site(particularly in the years between the First and Second WorldWars, when at least one excavation seems to have occurred everyyear), and in a number of cases detail remains rather elusive.It is also certain that a considerable amount of excavation andsurface collection has taken place without record. In addition,the full publication of the most recent campaign, thatundertaken by the British Museum in the mid-1970s, is still inpreparation. A number of volumes have appeared, however, and asa result this most recent work is dealt with in a more summarymanner than the earlier investigations.

Early Investigations

The first recorded examinations of the site appear to haveoccurred in February 1852. In addition to the note by Lukequoted earlier, both CR Manning (1855, 1872) and the RevPettigrew (1853) reported on excavations undertaken there inthe same month. There is obviously a strong possibility thatall three are referring to the same episode, although Lukeappears to have been merely an observer, while Manning andPettigrew both refer only to trenches dug under their owndirection. Pettigrew reported to the British ArchaeologicalAssociation on February 25th 1852 that he had

...just returned from visiting a place called Grime's Graves,...where hehad opened two of the graves, and found in each some bones ofanimals, appearing to belong to the ox and the pig. The place wasformerly the seat of war between the Saxons and the Danes. No otherremains beyond those of the bones, laid in a bed of flint stones, werediscovered. Those were in a cavity of coffin-like shape...

Manning meanwhile reported to Norfolk Archaeological Society on4th March 1852 on

...the result of an examination he had made of the large collection of pitson the heath at Weeting, near Brandon, commonly known as 'Grimmers'or 'Grime's Graves'. This examination appeared to fully confirm that theywere the remains of a British village.

On the 5th July 1866, Manning (1872) provided a little moreinformation when speaking to members of the NorfolkArchaeological Society during a visit to Grimes Graves. Hestill regarded the site as 'a collection of British dwellings', possiblywithin a fortified settlement. Referring to his 1852investigations, he states that trenches were dug throughseveral of the pits. In each case, at approximately 3ft (0.9m)deep 'we came upon an oval wall of flints, evidently a fire-place, containing numerousbones of oxen, but no implements'.

He undertook further investigations in the week prior to thesite visit in order to provide something for members of thesociety to look at. The results this time seem to have beenconsidered disappointing, since 'in the pits now opened the fire-places arenot so distinctly preserved'. A trench was also cut into the Grimshoemound (see below). Manning adds that a visit by the SuffolkArchaeological Society on September 28th 1866 resulted in thediscovery of 'a flint, apparently worked for a celt' in the nearby wood.

Greenwell: 1868-1870

Canon William Greenwell's excavations, undertaken between 1868and 1870, first revealed the true nature of the site (Greenwell1870a, 1870b, 1872; note that Greenwell 1872 is a reprint ofGreenwell 1870a, complete in all respects with the exceptionthat all paragraphs containing the word 'penis' are omitted);he was assisted in 1870 by Lord Rosehill (Rosehill 1871).Greenwell provided no explanation for his decision to excavateat the site. He had been present at Cissbury, West Sussex,during Lane-Fox's first investigations of the pits there (Lane-Fox excavated at Cissbury in September 1867 and January 1868,with Greenwell certainly present on the latter occasion), whenLane-Fox correctly interpreted the pits as resulting from flintextraction, but failed to realise their true depth, having

mistaken the compacted surface of the chalk infill of theshafts for the bottom of the excavated features. However, amore direct stimulus appears to have come from the appearanceof the first detailed accounts of discoveries at the flintmines at Spiennes, Belgium. Some years later, Lane-Foxexplained how

Canon Greenwell happening to be carrying on his excavations nearBrandon, which has always been the great workshop of the gun flintmanufactory, chanced to come upon a collection of pits similar to thoseat Spiennes and Cissbury...and he decided to excavate them, in order todetermine whether they also had shafts and galleries like the Spiennespits. (Pitt Rivers 1884, 70)

Greenwell selected for excavation 'one of the shafts on the eastern side,near the extreme edge, and almost in the south-east angle of the space occupied bythe pits' (referred to hereafter as 'Greenwell's Shaft'). Hedescribed it as being 28ft (8.5m) diameter at the mouth, 12ft(3.6m) wide at the bottom and 39ft (11.88m) deep. The lower18ft (5.4m) of its infill comprised almost entirely chalk.Immediately above this was a 'considerable thickness' of sand withflint nodules and chalk, then a deposit of chalk and 'flintchippings', then chalk rubble, followed by sand and finally chalkrubble at the top. The upper 13ft (3.9m) of the shaft had beencut through a deposit of dark yellow sand. At c. 28ft (8.5m)down on the eastern side of the shaft was a layer of charcoaland wood ash 4ft (1.2m) wide and extending 5ft (1.5m) towardsthe centre of the shaft. The chalk and flint below, andimmediately in contact with, this material was partiallycalcined, suggesting to Greenwell that the burning had occurredin situ. Animal bones were scattered throughout the fill down tothis depth, but below this point they were much less common Atthe base of the shaft was the 'floorstone' layer, which hadbeen exploited by cutting a series of radiating galleries.Greenwell noted that the galleries varied between 4ft (1.2m)and 7ft (2.1m) in width, and the longest of those examinedextended for some 27ft (8.2m), eventually joining up withanother shaft.

The rather impromptu nature of Greenwell's excavation isevident not only from his and Rosehill's accounts, but alsofrom the re-excavation of the same shaft and all of its

associated network of galleries during the British Museum'swork at the site a little over a century later, which not onlydemonstrated the true complexity of the gallery system and itsrelationship with other nearby shafts, but also made clear howmuch had not been excavated by Greenwell and his workmen in thetime he was able to make available for the task (Longworth andVarndell 1996, 9-33). Rosehill and Greenwell were obliged toleave Norfolk a few days after finally reaching the bottom ofthe shaft in 1870, although even then much of the shaft fillhad not been removed - Rosehill notes that 'about one-third of the filledin stuff was left in situ'.

As far as artefacts are concerned, a polished axe recoveredfrom the first gallery explored turned out to be the mostdebated of Greenwell's discoveries for the next few decades. Atthe time, Greenwell described it as being of basalt, althoughpetrological examination many years later revealed to beepidiorite, probably from south west England (and assigned topetrological Group IIIa - McK Clough & Cummins 1988, 47). Ananonymous letter of 1873 (Norfolk CRO h), possibly written byNorwich antiquarian and headmaster Augustus Jessop, contradictsGreenwell's own first account of the discovery. Thus the letterseeks to correct Greenwell's 'statement that the basalt hatchet was foundfour feet from the entrance to Gallery No. 1. On the contrary it was discovered lyingbetween two deer's-horn picks, at the farther end of this gallery'. Greenwell notedthat the marks of the axe's cutting edge could clearly be seenupon the gallery walls (the presence of these marks wasconfirmed by the RCHME investigation, and they were recordedphotographically - see Appendix iv).

The axe and the evidence for its use in the gallery becameimportant dating evidence for the mines, both to Greenwell andlater authorities. Some years afterwards, Greenwell provided aconsiderably more detailed account of the discovery of the axeand his recognition of the marks made with it on the gallerywall, after doubts had been cast on the authenticity of the axe(see below):

I...saw among the pick marks a clean cut, and I immediately said 'Theyare using a stone axe'. As the work went on I observed that the cutbecame less sharp, when I said 'The axe is becoming blunt', and shortlyafter, seeing that the cut had become shorter at one end, I told the

workman that a piece had been broken off of the edge of the axe. A dayor two afterwards...I saw through the broken chalk a dark-lookingobject, the nature of which was not apparent...I then had the interveningbroken chalk removed, when to my great delight a ground axe of basaltor some similar rock made its appearance, and on examining it I found itwas the identical axe with which the marks were made, with a bluntedcutting edge, and with a small piece broken off one corner. It fittedexactly into the marks on the gallery sides, and was without doubt thetool on which some of the work of excavation had been done.(Greenwell, in Sturge 1908)

A second intriguing find occurred at the end of this firstgallery, at a distance of 20ft 8in (6.2m) from its mouth.According to Greenwell (1870a, 427)

...when the end (of the gallery) came in view, it was seen that the flinthad been worked out in three places at the end, forming three hollowsextending beyond the chalk face of the gallery. In front of two of thesehollows were laid two picks, the handle of each towards the mouth of thegallery, the tines pointing towards each other...Between the picks wasthe skull of a bird, but none of the other bones.

The bird skull was later identified as a phalarope (Clarke1915, 17), a relatively rare migratory species of wading bird.

Other finds from various levels of the fill included numerousanimal bones, charcoal, burnt sand, flint flakes and cores, and'tools of deer's horn' (ie antler picks), the latter identified byGreenwell as the principal mining tool. Also apparentlyuncovered were four 'rudely-made cup-shaped vessel of chalk', concerningwhich the aforementioned anonymous letter of 1873 (Norfolk CROh) states that

One was found in the pit, & in the position he states, but two of the fourwere manufactured on the spot by a person present on the day, & thediscovery of the first, found on a ledge in the gallery...

Greenwell suggested that these chalk vessels had been used aslamps, although they bore no traces of burning. He alsodescribed some other items apparently of worked chalk,including one

...not unlike part of a human leg or arm. the marks of cutting, probablywith flint flakes, are distinctly seen upon it; and the broken ends showthat it formed part of a larger article; the present length is 10 inches,and it is 14 inches in circumference. [Another] may have been part of afinger; it is 1½ inch long, 2½ inches in circumference, and is only afragment. The most remarkable piece is a representation of the glans ofa human penis, which has evidently been broken off from the wholemember...

Greenwell speculated that as these three items had been foundin close proximity to each other (at a depth of 31ft (9.4m)),they may have formed part of a whole figure. However, hehimself appears to have been unconvinced by this, adding thatthe penis may well have been the product

...of some ancient workman who had no further intention, when hecarved it, than the artists who, in a very inferior style, depict the samemember upon our walls and palings.

Greenwell considered that the main period of working at GrimesGraves must have been 'the age when stone was the material used in thefabrication of weapons and cutting-implements', preferring the Neolithicto the Palaeolithic (the two subdivisions of the stone age hadonly been defined as recently as 1865 by Sir John Lubbock), andadding that 'this period, no doubt, was to a certain extent contemporary with theage when bronze was also in use for certain articles'.

Rosehill also conducted his own excavations in the area duringthe 1870 season. he refers to 'opening up a good many barrows in thearea' before taking on a shaft of his own, the location of whichis unknown:

[On] the last day or two of our stay in working at a second pit, at theopposite side of the group from the one Mr Greenwell had so long toiledat. like all those on the west of the series, this pit was much smaller thanNo. 1, and probably not so deep, as the slope of the ground on that sidebrings the layer of 'floor-stone' flints nearer the surface. However, I onlyhad time to get down some 8 or 10 feet, and found, besides quantities ofsplit bones, two very rude adze-shaped tools of flint, of a very ancienttype, almost resembling the flints of the drift, and some hammer-stones;but as in No. 1, no trace of pottery. (Rosehill 1871, 427)

Rosehill's comments on the flints echo some of Greenwell'suncertainty over the dating of a few items, and foreshadow thelater controversy over flint mine chronology (see below).

After Greenwell

There is no doubt that the site was visited by antiquarians andarchaeologists between Greenwell's departure in 1870 and thenext known excavation in 1914. Warburton (1914), for example,refers to visits to grimes Graves specifically to collectflints:

My first two visits to the 'Graves' were practically fruitless as far asimplements were concerned. However, I stumbled across the pitexcavated by Canon Greenwell and brought back a large bone from thedebris surrounding it. Subsequent visits showed the best place to searchwas on the arable land. Every year since then,...I have been enabled tospend several days searching, and have brought away some heavy loadsof spoil.

Clarke, who led the 1914 excavations), was also numbered amongthose who visited the site looking for implements (Clarke1914).

However, after Greenwell, there is no record of excavation atGrimes Graves until 1914, when a long period of investigationbegan, which included some excavation almost every year until1939. This renewal of interest had been stimulated by thepublication of a paper by Reginald Smith (1912) of the BritishMuseum, in which he argued that the main period of flint miningat both Grimes Graves and Cissbury lay not in the Neolithic butin the Palaeolithic. This belief was a direct contradiction ofgeneral opinion - for example, Sandars (1910, 102) had statedthat

There is no proof that flint nodules were mined in palaeolithic times, andit is only during the later neolithic that distinct and incontrovertibleevidence of systematic mining operations by primitive man can befound.

However, he was prepared to accept the idea of earlier opencast extraction, as had been claimed for Obourg in Belgium

(Sandars 1910, 105). Smith's views represented a minority viewthat had been circulating for some time, but had been gainingground in recent years, resulting notably in doubts being caston the authenticity of the stone axe found by Greenwell. Thosedoubts appeared to have been removed by Sturge (1908), who hadsolicited the detailed account of the axe's discovery fromGreenwell (quoted from earlier) as well as undertakingenquiries in the Grimes Graves area. By the time Smith wrotehis paper, the question of the axe was more or less settled,although he took the opportunity to remind readers of thedoubts which had been expressed. Axe aside, the problem centredon a misunderstanding of prehistoric lithic technology at atime when the Neolithic was poorly understood, and had fewwidely accepted diagnostic features. Much of the excavation inthe years up to 1939 were concerned with testing Smith'shypothesis of Palaeolithic mining at Grimes Graves, althoughnew discoveries ensured that the debate was more or lessconcluded in favour of the Neolithic by the early 1930s (Clark1932; Clark and Piggott 1933).

March 2nd to May 9th 1914: Prehistoric Society of East Anglia

This campaign was undertaken primarily in order 'to throw light on thedate of the industry' (Clark 1915), reflecting the debate begun bySmith. However, it was also concerned with fauna, climate,environment, and the nature of exploitation - seasonal orpermanent - and the length of time the mines were in use.

i) 'Pit 1'

The first shaft chosen for excavation is located in the north-western part of the site, deliberately as far as possible fromGreenwell's shaft. Pit 1 was described as being oval on thesurface, 32ft (9.7m) north-south and 29ft (8.8m) east-west, andsurrounded by a 'mound' of upcast material of irregular height.The upper fill included sand and chalk rubble deposits. One ofthese deposits, layer 3, comprised large chalk blocks said tohave entered from the west; some pieces of 'floorstone' werepresent, along with waste flakes, implements and animal bones,and it was suggested that the material represented spoil fromanother shaft. Beneath this, some 10-15ft (3-4.5m) below thesurface, layer 4 (referred to as the 'Black Band') included

chalk rubble and a mass of charcoal. Also noted was a hearthconsisting of six large fired flint blocks, adjacent to someburnt antler, burnt flakes and a lump of wood charcoal,suggested to be a habitation layer.

A further series of large chalk blocks, again 'shot in from thewest', layer 5, was recorded but with a hard, thin layer ofmixed sand and rubble running through it, suggested to indicatean interval in the infilling process. The usual flint finds,together with part of a human skull came from this layer. SirArthur Keith suggested that it represented an adult male underthirty years old. The lower part of the shaft fill consistedalmost entirely of chalk. The published section also shows anindentation in the shaft profile suggestive of someexploitation of the 'wallstone' flint, at layer 8. A number ofpotsherds were also recovered from this level. At the base ofthe shaft was a deposit of chalk blocks and rubble, layer 10. 'Init were the remains of three fireplaces at different levels...There were no large flints asin the black bands, but a mass of charcoal, which included a piece of charred log'which was 1ft (0.3m) long and 6in (0.15m) thick, together withmore pottery (later identified as Grooved Ware: Longworth et al1988, pp12 & 16) and antler plus two further 'fireplaces'.

The galleries at the base of the pit are described as radiatingout like the spokes of a wheel, although the published plan ofthe area excavated reveals rather less regularity than thisobservation implies. It appeared that in order to extract asmuch of the floorstone flint as possible, the amount of chalkleft unexcavated in situ to support the roof was minimal. theresult was a complex system of interconnecting galleries andtunnels, not all exploited from Pit 1. Not all of the gallerieswere cleared out during the 1914 excavations, but at least fiveneighbouring shafts, the nearest only c.20ft (6m) away at thebase, were shown to be connected to the same network ofgalleries. Three different types of interlinking betweenindividual galleries were noted: wide open 'doorways' of aboutthe same height and width as the galleries themselves; 'creepholes', or constricted openings through which it was stillpossible to crawl; and smaller openings interpreted asproviding ventilation.

Gallery heights were generally about 2½-3ft (0.7-0.9m). The

report numbers them and describes them in sequence. Antlerpicks were a common find, some having the appearance of beingcarefully placed rather than casually disposed of. Scatters ofcharcoal and some small potsherds are also referred to.Longworth et al (1988, 12 & 16) refer to 67 sherds of GroovedWare from Pit 1, mostly from the galleries. Other findsincluded a carved chalk cylinder from gallery 6, a chalk 'lamp'from gallery 13, scatters of flint flakes with the occasionalimplement, and what were interpreted as rope marks in the chalkabove the entrance to gallery 3. In addition to clear evidencefor antler picks having been used in excavating the galleriesand mining flint, marks created by stone (presumably flint)axes were also observed in places.

The excavators attempted to work out the sequence in whichgalleries were dug and backfilled, primarily on the basis ofthe direction from which the backfill came, and concluded thattwo or more shafts must have been open simultaneously, at leastin the area of Pit 1, with 'free communication' existing belowground between such shafts.

ii) 'Pit 2'

Pit 2 was described by the excavators as being 'more in the centre ofthe whole area and nearer to the SW corner'. It was virtually circular atthe surface, c.42ft (12.8m) x 43ft (13.1m), and about 31ft(9.44m) deep from the modern ground surface. The report notesthat 'beside it, to the NE, was perhaps the largest pit in the whole area, which wasused by us to receive the material thrown out of the shaft'. The upper levels offill were similar to those in Pit 1. Layer 2, in particular,consisted of two distinct tips, designated 2A and 2B. Layer 2Aconsisted of a fine dry rubble, and included two 'fireplaces',one predominantly charcoal, the other including burnt flintsand animal bones. Layer 2B featured chalk blocks with some'floorstone' flint. Layer 3 is referred to as the 'Black Band'and described as a 'definite occupation level of two distinct periods',although the difficulty in distinguishing between the twolayers of charcoal is noted. Finds included flint flakes andimplements, animal bones (including dog), some burnt, and 'twoblocks which appear to be ochre'. The top of layer 3 also containedpart of a human skull, including a lower jaw and part of theskull, thought by Sir Arthur Keith to belong to a young girl

aged c.13.

A quantity of unworked 'floorstone' nodules together with someflint flakes and implements were incorporated in the sand andboulder clay of layer 4, while layer 5, described as largechalk blocks 'from the lower part of a pit' contained patches ofcharcoal, burnt flints, flint flakes and implements, andpottery (since identified as Grooved Ware by Longworth et al1988, 12 & 16). In fact, most of the pottery excavated from Pit2 came from this layer.

As with Pit 1, numerous interconnecting galleries radiated outfrom the bottom of the shaft, exploiting as much of the'floorstone' as possible. Once again, the network of passagesconnected up with a number of neighbouring shafts. Unusedflint, together with occasional flakes and implements,including a flint axe at the entrance to gallery 5 and antlerpicks, were among the finds. In total, 118 Grooved Ware sherdscame from Pit 2 (Longworth et al 1988, 12 & 16). In addition, achalk ball and a carved chalk cylinder were found in theentrance to gallery 4. 'Rope marks' were observed above theentrances to a number of galleries. Two instances of linesincised into the chalk were encountered (Peake 1915, 73-4). Thefirst, referred to as the 'Sundial', was found 7.5ft (2.2m)'above the floor of the shaft, over the south side of the spring of the arch at theentrance to gallery 6, immediately above the upper level of the wallstone'. Itcomprised twelve roughly vertical lines, some curving but mostnearly straight and more-or-less parallel. The longest measured9.5in (o.2m). As for the name, 'It was noted that the sun at 12 o' clock noonshone on this face, and the form of the lines suggested a sundial' (Peake 1915,74). The second, referred to as 'tally marks', were cut ontothe buttress between galleries 4 and 7. The marks comprisedfive roughly parallel lines, with another cross-cutting themdiagonally. It was suggested that they may have been cut intothe chalk with a flint flake.

Pit 2 was among those re-excavated by the British Museum duringthe 1970s (Longworth and Varndell 1996, 35-38). As a result itbecame clear that the 1914 excavation had not excavated thegallery deposits down to the original floor levels, althoughsufficient disturbance had occurred to prevent the recovery offurther information. The British Museum survey of the shaft and

its gallery system also highlighted inaccuracies in theoriginal plan produced in 1914 by Leslie Armstrong and featuredin the PSEA report (Clarke 1915)

iii) The 'Floors'

A number of 'floors' were also excavated on the surface. Thesewere regarded as having been 'the sites upon which the miners manufacturedtheir implements...Most of the floors were discovered by noticing the numbers ofchipped pieces congregated in particular spots, which led to further investigation bydigging'. Fourteen such sites were examined in 1914, and aredescribed briefly below; their locations are indicated on the1914 site plan. The fourteen floors were divided into twocategories: group A, consisting of those containing 'massivechipped pieces with primary work only' (2, 3A, 3B, 4, 7, 8, 9, 10, 11,and 14); and group B, those described as 'finishing floors',featuring 'massive primary work' and the 'refuse of finer secondary workplus the finer implements' (1, 3C, 5, 6, 12, and 13).

Floor 1 - located on the north-west face of a pit in the north-east corner of the earthwork complex, close to Section 9 (seebelow). An area 10ft (3m) x 8ft (2.4m) was exposed, revealing'hundreds of flakes...The whole of this floor lay within a foot (0.3m) of the surfacebelow which was 2ft (0.6m) of chalk rubble with sand beneath it'.

Floor 2 - located on the eastern side of the mined area, notfar from the Grimshoe tumulus, and close to Floors 3 and 13.Described as being 'on the western slope on top of a bank', presumablymeaning a spoil heap adjacent to one of the pits. An area 12ft(3.6m) x 4ft (1.2m) was examined to a depth of 1ft (0.3m),below which was chalk rubble and boulder clay interpreted asmining spoil. Again, numerous flakes and implements wererecovered.

Floor 3 - a few metres west of Floor 2 and immediately east ofthe Beaver Pit (see below), this floor was described as themost productive. It was stated that 'the flakes and debris from this floorwould fill several carts'. It was interpreted as representing threesuccessive floors, separated by layers of chalk blocks andrubble. The size of the area examined is not clear, but it wasat least 21ft (6.4m) long, and of varying width. Floor 3A wasinterpreted as accumulated waste resting on the old land

surface, while 3B was described as 'spread on one layer of rubble andcovered by another' and featured the usual flakes and implements,often occurring in pockets in the chalk rubble. Floor 3Cconsisted of a dense mass of flakes and implements, includingfourteen hammer stones. Also, 'in the centre of the area the flakes wereextensively burnt and embedded in wood ashes'. Floor 3C also containedsheep bones, red deer antler, and pig teeth.

Floor 4 - known as the 'Pine-Tree Floor' because it lay partlybeneath the roots of a pine tree, this was centrally locatedwithin the mined area, a short distance east of Pit 2. An area14ft (4.2m) x 17ft (5.1m) was examined. Among the flakes andimplements was 'a piece of carved chalk which seemed to have been hollowed outto form a lamp'.

Floor 5 - located towards the south-eastern corner of thevisibly mined area, it lay in a depression 7ft (2.1m) indiameter situated between three pits. Described as 'quitesuperficial', it consisted of flakes and implements lying in and onchalk rubble.

Floor 6 - located a short distance south-east of Floor 5, thiswas the smallest of those examined. A scraper and hammerstonewere apparently the only items of note among the usual 'roughflakes'.

Floor 7 - known as the 'Birch Tree Floor' after a group ofsilver birches growing out of it, this was located in thenorth-west corner of the earthwork complex. It was described asbeing 'quite a surface floor, many of the pieces lying actually exposed, but a mass ofchipped pieces lay on rubble to a depth of one foot (0.3m)'.

Floor 8 - according to the report, north of Pit 1, although thesite plan suggests that 'north-east' would be more appropriate.Described as a collection of 'roughly chipped pieces lying on the surface inthe rubble between the pits'. There seems to have been little workundertaken apart from some surface collection due to thepresence of tree roots.

Floor 9 - this was similar in character to Floor 8 and likewise'north' (ie north-east) of Pit 1. Both floors are described as'quite superficial, resting on the rubble with no sand or chalk below, and only a scanty

covering of moss'.

Floor 10 - another 'superficial floor', 8ft (2.4m) x 4ft(1.2m), roughly central within the earthwork complex and closeto Floor 4. It was discovered by digging, c.1ft (0.3m) below thesurface, there being no surface traces. Part of a red deerantler was among the finds.

Floor 11 - also superficial, and c.4ft (1.2m) x 9ft (2.7m),located a short distance north of Pit 2.

Floor 12 - located between Floor 11 and Pit 2, and just east ofFloor 7. An area 10ft (3m) x 5ft (1.5m) was examined, 'occupyingthe top of a ridge between two pits' and continuing down at least one ofthe slopes.

Floor 13 - located in the north-east part of the site, betweenthe Grimshoe mound and Floor 2. An area 10ft (3m) x 8ft (2.4m)was inspected after the main excavations had ended. Ofparticular note were quantities of tiny flakes, something alsoobserved at Floor 3C, a short distance to the east.

Floor 14 - located a short distance south-west of Floor 3, thislay on the rubble/spoil between two pits and measured c.6ft(1.8m) x 4ft (1.2m). It consisted 'mainly of chipped pieces'.

iv) The 'Sections'

In addition to the floors, a total of eighteen 'sections' werecut in the hope of finding out where the earliest miningactivity occurred, ie to find out where the flint outcropped,or lay closest to the surface with the least overburden ofsand. In this aim they were notably unsuccessful, with sections1, 2, 5, 6, 8 and 17 giving what were described as negativeresults.

Section 1 - a trench 2ft 6in (0.7m) x 3ft 6in (1m) dug in WestField. Its exact location is given as 567ft (172.8m) from thefence on the south-west side of 'the Graves'. Apart from a fewflakes in the topsoil, nothing was found.

Section 2 - a trench 2ft 6in (0.7m) x 3ft 6in (1m), again in

West Field, 282ft (85.9m) south-west of the fence. Again, justa few flakes were found in the topsoil.

Section 3 - a trench 2ft 6in (0.7m) x 3ft (1m), and closerstill to 'the Graves'. The spot was chosen because a 'faint ringwith a depression in the centre' was suspected to be a pit. The trenchproved to have been dug through the 'encircling mound', ie thesurrounding spoil, which had been damaged by ploughing.

Section 4 - a trench 7ft (2.1m) x 3.5ft (1m) dug a shortdistance north of the earthworks, in the area described as openheath sloping down into the valley. Some 3ft (0.9m) of chalkrubble containing some animal bone and antler lay on top of a1ft (0.3m) layer of charcoal, which itself lay on top of chalkblocks. This was interpreted as a pit which had been 'smoothedover'. The quantity of flint on the surface of this part of thefield was taken to indicate the presence of further pits in thearea.

Section 5 - a trench 3ft 6in (1m) x 5ft (1.5m) some distancenorth-east of Section 4, again north of the earthworks.According to the report, 'this section gave negative evidence'.

Section 6 - a trench 3ft (0.9ft) x 5ft (1.5m) dug 100ft (30.4m)east of the north-eastern corner of the earthwork complex.Again, nothing was found.

Section 7 - situated in an area of flat ground beside the grasstrack that separated the Grimshoe mound and adjacent pit fromthe remainder of the earthworks. A trench 4ft (1.2m) x 6ft(1.8m) was dug on what turned out to be the 'edge of a cup-shapedliving-place...8ft (2.4m) in diameter' on the upper fill of a shaft. Thisso-called 'hut-circle' contained a large quantity of animalbones, burnt flints and quartzite pebbles, and charcoal,together with a 'considerable amount' of pottery (Longworth et al1988, 27 refer to 147 sherds of Deverel-Rimbury style bucket-shaped vessels, 113 undecorated body sherds, and a single EarlyIron Age flint-tempered sherd as surviving from theseexcavations) and three bone pins, as well as a few flintflakes, but no implements. This 'section' was considered asproof that some of the disused pits 'were utilised as dwelling-places at alater date than the working of the mines'. Note that this section was close

to the site later excavated and known as the 'Black Hole' (seebelow).

Section 8 - dug in the flat area between a group of shafts justeast of Pit 1. The trench, 3ft (0.3m) x 4ft (1.2m), revealed anarea which was undisturbed apart from the presence of somerubble from the neighbouring pits.

Section 9 - dug in the north-eastern corner to test the theorythat a 'serpentine dyke' represented the surface traces of threeshafts dug in a line. The trench, 18ft (5.4m) x 2ft 6in (0.7m),suggested that three pits had indeed been dug in a line, withonly a narrow wall of chalk left between them.

Section 10 - a trench 14ft (4.2m) x 3ft (0.3m) dug in the westface of the large pit immediately adjacent to, and north-eastof, Pit 2, in order to check on stratigraphy in the upperlevels of the fill, particularly with regard to the sand andthe boulder clay.

Section 11 - known as the 'Beaver Pit' for reasons explainedbelow. A trench of unknown size was dug by Dr Allison in thecentre of the shaft immediately adjacent to Floor 3 'in order to seeif any implements had found their way into the cup'. Animal bones,particularly sheep and ox, were found in 'a definite layer' in thecentre of the 'cup', along with 'a fair number of chipped flints' and aperforated chalk disc, the latter suggested to be a bead(despite reservations about the weight of such a necklace).Also present was part of the skeleton of a beaver. Below thelayer containing the bones, at c.2.5ft (0.7m) down, was a layerof fine chalk rubble. The excavators did not investigate anyfurther.

Section 12 - dug into the centre of the large pit adjoining theGrimshoe mound. The trench, 6ft (1.8m) x 2ft 6in (0.7m),appears to have been dug in an area which had already beendisturbed. The excavators suggested that Greenwell or manningwere the most likely culprits.

Section 13 - this was actually dug as a gangway between Pit 2and the adjacent shaft to the north-east (where Section 10 wasalso sited) in order to facilitate wheeling of the excavated

material (ie the dumping of spoil). With regard to the issuesof the day, the section demonstrated that the shaft had beendug through the sand and the boulder clay. It also appearedthat some chalk was pile dup to form a ring around the shaft.

Sections 14 & 15 - a pair of trenches cut into the north andsouth lips of Greenwell's Shaft.

Section 16 - cut into the Grimshoe tumulus - see below.

Section 17 - a trench cut within the wood 200 yards (182.8m)north of Grimshoe. Some flints were found in the topsoil.

Section 18 - a series of shallow holes east of Grimshoe, whichagain produced some flints in the topsoil. The results fromthis and Section 17 were taken as an indication of 'how extremelylocal [the Grimes Graves] culture is'.

The flint report for the 1914 excavations was written byReginald Smith (in Clarke 1915, 147-207) and despite hisprotestations to the contrary, he was mainly concerned withmarshalling evidence to support his dating of the mines to thePalaeolithic period. With Smith also responsible for thepottery report, it is perhaps not surprising that thepossibility of pottery manufacture in the Palaeolithic shouldbe discussed. He also made a distinction between the potteryfrom the Floors and Sections and the sherds found in the shaftsand galleries, but was unable to suggest a date for any of it,simply concluding that 'the known Neolithic, Bronze Age, and later wares ofBritain have quite different characteristics'. In fact, in 1914, comparativematerial from non-funerary contexts was scarce, and his datingcontinued to rest wholly on the flint assemblage. Othercategories of evidence were generally regarded as of uncertaindate or relating to later activity.

Excavations 1915-1939

The 1914 campaign was undertaken with the aim of resolving anumber of problems relating to the mines, primarily their date.The apparent failure to solve the issue caused the debate tocontinue, and in turn prompted more excavations. In fact,almost every year from 1915 to 1939 saw work of some

description taking place at the site, although the debate overchronology had essentially been settled by the early 1930s.Much of the later work has never reached full publication.

1915 - Peake and Armstrong

Peake and Armstrong appear to have been the first to return, inthe autumn of 1915 (Peake 1916). They examined four more'floors' and re-examined three of those excavated the previousyear. The numbering sequence followed on from the 1914campaign:

Floor 15 - described as being 18ft (5.48m) from the fence and29ft (8.8m) from the east end of Floor 16. An area 8ft (2.4m) x11ft (3.3m) was examined. The floor was 1-3ft (0.3-0.9m) belowthe surface and resting directly on boulder clay. Numerousflints were found, including very small flakes taken asindicative of in situ knapping.

Floor 16 - described as being 29ft (8.8m) south of Floor 15,the area examined was 45ft (13.7m) long and 3-5ft (0.9m-1.5m)wide. Peake described it as a 'living site as well as a chipping floor'. 2ft6in - 3ft 6in (0.7m-1m) below the surface was a layer of blackmaterial, consisting of fine charcoal, some calcined animalbones, and burnt flints. They also observed a possible post-hole filled with sand and dug into the boulder clay. A layer ofchalk featured the remains of what were interpreted as hearths,one consisting of five large burnt flints close to some bonesand teeth of red deer. Along with flint flakes and implementswere two potsherds, one which 'resembled that found in the shafts andanother of Roman date'. The former may be the plain Grooved Waresherd referred to by Longworth et al (1988, 12-13).

Floor 17 - this lay on the edge of a shaft 'not far west' ofGreenwell's Shaft, and measured 6ft (1.8m) x 5ft (1.5m), andcontained a typical flint assemblage.

Floor 18 - on the edge of a pit in the north-east part of theearthwork complex. The area examined was 6ft (1.8m) x 3ft(0.9m). The range of flints was again typical.

Floors 3, 4 and 14 from the previous year's work were re-

examined. The main aim appears to have been to collectartefacts, and those recovered were felt to support thePalaeolithic dating of the site. In addition, a 'new' gallerywas discovered in Pit 2 following a fall of chalk and sand. Itwas found by clearing away the chalk which had blocked up acreephole leading off gallery 9. Numbered 20,it headed 10ft(3m) to the north. A low creephole led to gallery 21, which ran15ft (4.5m) west, ending in an infilled shaft.

1916 - Peake, Armstrong & the Lingwoods

Peake and Armstrong were back again the following year, thistime accompanied by Mr and Mrs Lingwood (Peake 1917). The workfollowed more or less the pattern of the previous year: 'wedevoted our time to exploring some of the old floors and in searching for new ones'.Once more, the aim was to collect artefacts with a view to'throwing a light on the date of the industry'. As with the 1915 excavations,the floors yielded various quantities of flint flakes andimplements. In all ten new floors were examined:

Floor 19 - an area 7ft (2.2m) x 14ft (4.2m), it lay on the edgeof a pit near the south border of the plantation.

Floor 20 - an area 10ft (3m) across, located on the edge of apit near the western border of 'the Graves'.

Floor 21 - an area about 15ft (4.5m) long to the east of Floor15 'in the gap on the south border where the pits do not extend. An enormous massof material was exposed...mostly waste material'. Peake adds that floors 19-21 were discovered by a Lieutenant A Hall and a Mr D Casey 'whohave some implements form them which I have not seen'.

Floor 22 - a little to the south-east of Section 7, the areaexamined was 32ft (9.75m) x 22ft (6.7m). It also contained 'anenormous mass' of waste flakes. Peake also noted that 'in excavatingthis floor we found that a pit had been smoothed over to form the grass road'.

Floor 23 - a little to the north of Floor 22, this had beenexposed by the uprooting of a large tree.

Floors 24-27 - these are not described in any detail, nor aretheir locations given by Peake, who merely noted that 'the smaller

floors yielded a number of implements...and all lay on the sides of pits...'.

Floor 28 - 18ft (5.4m) x 8ft (2.4m), this 'lay close to the grass roadon the north-east of the pits, and about 230ft (70.1m) from the nearest one. Thepeculiarity of the pieces from this floor lay in the fact that they had nearly all beenused'.

In addition to the new floors, Peake also dug a trench 6ft(1.8m) deep in the centre of a shaft to the east of Floor 4. Athick band of charcoal at a depth of 4ft (1.2m) confirmed forPeake the idea that 'the partly filled shafts were used as dwelling places by therace that mined flint and fabricated implements on this spot'. Beneath thecharcoal layer he also found 'part of the skull of Bos longifrons, and the jawof a pig and sheep'.

Floors 15 and 16 were also re-examined because it was foundthat 'we [had] not completely worked [them] out'. They were interpretedas 'living-spots for the knappers' but Peake also argued for occupationat a later period, as large hearths built up of chalk blockshad been built in a hollow cut through the 'chipped zone which formedthe living level at the time when the mining was going on'. These 'hearths'were associated with animal bones and teeth, and potterydescribed as Bronze Age in date. Ellison (in Longworth et al1988, 25) refers to fifteen sherds of Deverel-Rimbury bucketshaped vessels and a single Early Iron Age sherd, all fromFloor 15. Floor 16 again produced mostly Deverel-Rimburymaterial, but a single sherd of rusticated Beaker and a singleGrooved Ware sherd (referred to above) survive among the knownfinds, along with a few Iron Age and Roman sherds. Peake addedthat 'Floor 15 is now found to be 18 (5.4m) x 16ft (4.87m) and floor 16 29 (8.8m) x24ft (7.3m)'. A portion of a polished greenstone axe was alsodiscovered on Floor 15. Peake noted that it supported theauthenticity of Greenwell's earlier find, but felt that thequantity of 'Palaeolithic' types coupled with the disturbedcontext worked against it as indicator of the date of themines.

The finds report is again mainly concerned with establishing aPalaeolithic date for the flints. As well as the Bronze Agesherds already mentioned, reference is also made to Romano-British and 'late Celtic' sherds. In addition, Peake mentionedthe discovery of a tanged bronze spearhead, something which

would generally be placed in the late Early Bronze Age. It wasfound by Peake '20 inches below the surface in the east margin of floor 16. Therewas at this spot an inserted hearth of chalk blocks with a quantity of charcoal, potboilers, and bones of red and roe deer with those of domesticated animals. The usualfloor of chipped pieces...was disturbed and cut through by it...The spear was 31ft(9.4m) due north of the fence'. Peake added that 'as far as I know, this is theonly bronze implement found in situ in the area of the plantation', a statementimplying that this was, perhaps, not the first to be reportedfrom the area.

1917 - Peake and the Lingwoods

Peake and the Lingwoods returned in 1917, again restrictingtheir activities to sections and floors (Peake 1919), includingshallow investigation of two further pits. Floor 11 was re-examined as the 'enthusiastic helpers' employed in 1914 provedto have 'missed a good deal owing to lack of knowledge'. The Floor provedto rest on 'a mass of sand not in situ [and] beneath lay boulder clay resting onchalk rubble'. A small area of charcoal and burnt sand lay in thesandy layer.

Floors 15 and 16 were reinvestigated, the main aim being todefine the limits of each (ie define the limits of the spreadof worked flint). The edges of the excavated area were expandeduntil 'the limits of these zones' were reached. In addition, a seriesof trenches radiating out from Floor 15 were dug. These 'foundthe ground for the most part unproductive'. 'Sections' were also cutbetween Floors 15 and 16, again proving unproductive. However,the re-examination of the Floors themselves appears to haveachieved the second aim, that of collecting large quantities offlints.

Floor 29 - this was a section cut through the side of a shaftnorth of Pit 1, or on the east fall of a pit a little to thesouth of Pit 1, according to the different summaries provided(Peake 1919, 75 & 86). Again, flints were plentiful.

Pits 3-5: note that although the three shafts examined in 1917were so numbered by Peake, when Armstrong (1923) beganexcavating new shafts (see below), he also followed on from the1914 numbering system and began with the number 3.

Pit 3 - a section at least 5ft (1.5m) deep was cut across ashaft a little to the north-east of Floor 15. The first 2ft(0.6m) was full of flint waste similar to that from the Floors,and interpreted as knapping debris. Some animal bones and teethwere also present. The flint was mixed with chalk rubble andshells (snails). Below was more chalk rubble, finer in texture,and containing a thick band of charcoal with red deer bones 'andremains of a large jar of pottery'. Below this layer were chalk blockssuggested to have come from another shaft. Longworth et al (1988,27) refer to 'a base angle and four undecorated wall sherds of a [Deverel-Rimbury style] bucket-shaped vessel...labelled as from 'hearth 3' (0.9m) deep in pit 3',as well as a rim sherd labelled as from 'hearth in pit near midden northof Floor 15'. Longworth et al suggest that this last sherd is almostcertainly that described in Peake (1919, fig 87B) as comingfrom the hearth in Pit 3.

Pit 4 - situated a little south-east of Pit 1, a section dug inthe centre yielded a large collection of 'flint chipped refuse'. Thesection was dug to a depth of 5ft (1.5m). Topsoil, chalk andflints lay above a layer of chalk rubble containing snailshells and some flint. No charcoal was found this time.

Pit 5 - this is the pit adjoining Pit 3 'and situated in the samehollow'. The trench was 'nearly devoid of chipped flints and bones, and nohearth was discovered'.

1918 - Kendall and Armstrong (unpublished)

At this point occurs one of many lacunae in the publishedaccounts of the various Grimes Graves excavations. It ispresumed that there were Floors numbered 30 to 45.Unfortunately, there appear to be no contemporary publishedaccounts of them. The Reverend HGO Kendall (1920a, 1920b),rector of Winterbourne Bassett, Wiltshire (a keen flintcollector and, in 1922, the first person to establish byexcavation the Neolithic date of the earthworks on WindmillHill, Wilts) refers to flints from Floors 35 and 36, bothapparently dug in 1918, but provides no indication of theirlocation. Armstrong (1921a) (see below) includes illustrationsof items from Floor 42, but otherwise provides no detail.Longworth et al (1988, 25-27) provide some information in theirdiscussion of pottery from Grimes Graves. Floor 42 was situated

immediately to the north of Section 7 and probably joined upwith Floor 85. It appears to have been divided into two (A andB); Longworth et al suggest that this may indicate more than onechipping floor horizon, with A being the uppermost. This wouldcertainly be consistent with the numbering of other floors. Anote with the pottery from Floor 42A is quoted as stating 'northside of Floor 42. 1919-20. Pottery. Floor of pot boilers. Bones of red deer, two horns reddeer, jaws small sheep, ox horn core...polished tine, pin, chalk lamp'. Potteryincludes sherds of Deverel-Rimbury bucket-shaped vessels, andsherds from an Early Iron Age globular jar.

1919 - Richardson, Kendall and Peake

To return to the contemporary published accounts, excavationswere undertaken in 1919 by D Richardson (Richardson 1920). Anew Floor, numbered 46, was discovered 'in the open field close to thenorth border of the plantation, about 50 yards (45.7m) north-east of Pit 1'. Thetrench 'revealed the presence of a filled-in pit, of which there was no surfaceindication, the field having been levelled down for cultivation'. The trench wasabout 11ft (3.3m) square and 5ft (1.5m) deep. Below the topsoilwas a thin layer of flints above chalk blocks and rubble, whichin turn covered a 'mass of flints without any matrix' sandwiched between'tips' of chalk blocks and sand. A single small potsherd'similar to that found in the shafts' [Grooved Ware?] and a fewitems of worked chalk were also recovered.

As far as Richardson was concerned, the most striking thingabout the floor was the quantity of 'celts' recovered,representing about 20% of the total number of implements found,hence his description of the site as a 'celt-making floor'. Sixobjects of carved chalk were also found. Finally, in reviewingthe continued debate surrounding the date of the site,Richardson puts forward the view that 'the only thing to be done is to goon digging'.

In June and July 1919, and the same time that Richardson wasworking on the site, Kendall (1920) excavated a further seriesof Floors, numbered 47-59:

Floor 47 - the location given by Kendall is rather complicated.'The centre of Floor 52 was fifty paces north-east of the corner of the West Field, andfourteen from the edge of the cart track north of the wood. Twenty yards (18.2m)

north-west of the said centre was Floor 47, nearer the cart track coming up out of thevalley...'. He regarded this as one of the later floors, primarilybecause of its small size and the poor quality of theassociated flintwork. A 'hearth' lying in the middle of theFloor was c.4ft (1.2m) in diameter and roughly circular. theFloor appears to have been a few feet wider all round, and'...began at 9in (0.2m) below the surface and sloped steeply down to its base in themiddle at 3ft (0.9m)'. The stratigraphy is described simply as 'humus1ft (0.3m); flaked and chipped pieces 2ft (0.6m)'. Quartz hammerstones and afew animal bones were also present.

Floor 48 - 'lay between the second and third pits, east-north-east from the cornerof the West Field'. It was 4ft (1.2m) or 5ft (1.5m) in diameter.

Floor 49 - 'in the North Field, sixty paces north-west of the cart track, along the Nof the wood, and almost in a line therefrom with Mr Richardson's Floor 46'. Kendallcontinues 'the space cleared was 3ft (0.9m) by 3ft (0.9m). Small flakes werenumerous, and little chips numberless. A hearth, with burned sand and a littlecharcoal, lay on the east side, and contained a quantity of burnt flakes'. Thesection is described as 'sandy humus 14ins (0.4m), then red sand'. Theflint mainly came from the lower half of the sandy humus.

Floor 50 - 'a few yards from a floor of Mr Richardson's, about three or four pits SEof Pit no. 1'. The floor was 'on the brink of a pit showing a flat surface'. Theflints recovered were mainly from the top of the chalk rubbleand at the base of the humus, the latter about 1ft (0.3m) deep;it was apparently not much larger in extent than Floor 49.

Floor 51 - 'out in the North Field, roughly fifty paces north of floor 52...At 1ft(0.3m) depth, in the bottom of the sandy humus, were flakes and cores similar in sizeand style to those of Floor 47'.

Floor 52 - 'It lay in the North Field, as already described' (see Floor 47,above). Roughly oval in shape, c.20ft (6m) by 12ft (3.6m),Kendall described it as 'one of the two largest and most prolific discovered inmy researches'. The floor area actually consisted of 'a more or lesscontinuous series of groups of chipped flints', and was deepest in themiddle. Four 'hearths' were noted. Kendall suggested that thedistribution of the flakes, cores and implements found stronglyindicated that knapping had occurred around them. Sections weredescribed as generally 6in (17cms) of sandy humus, above up to18ins (0.45m) of chalk rubble. The majority of the flints were

at the bottom of this chalk rubble, some lying on the red sandbeneath. A few pieces of worked chalk (other than large blocksbearing tool marks) were also found.

Floor 53 - 'lay in the North Field, its edge being about 10ft (3m) from the nearestpart of 52'. Again, it featured a 'hearth'. The flints were oncemore at their thickest in the centre of the Floor. A tine ofred deer antler was also recovered.

Floor 54 - located an unknown distance east of Floor 53, thismeasured just 4ft (1.2m) by 3ft (0.9m). A few flints deemedworthy of comment are mentioned by Kendall but no 'hearths' orany other finds were noted.

Floor 55 - 'was forty-nine paces about west-south-west from the line of the birchesacross the end of the West Field and near the outer cart track...It was...9ft (2.7m) by 7ft(2.1m) in extent, and of oblong shape. Flakes and cores of fair quality, but noparticular interest, were numerous'.

Floor 56 - 'lay between the north-west corner pit and that to the eastward, ormore strictly the north-east of it...The floor was, at the deepest, 2ft (0.6m) from thesurface of the ground and was not of large extent, as worked by us, at any rate'. Itappears that Peake subsequently excavated a further extensionof this floor which, according to Kendall, 'was surmounted by humusand yellow sand, and sometimes a little chalk rubble'.

Floor 57 - '12 yds (10.9m) south-east of Floor 51, produced only one rough celt,given by myself to Mr Nichols'. Again, the excavated area seems to havebeen extended subsequently by Peake.

Floor 58 - 'of equal importance to Floor 52. It was sixty-three paces south of thecorner of the West Field, close to the track, on the Graves side thereof'. A trialhole 'disclosed only rather large chalk blocks, with flakes interspersed among themsomewhat thinly'. Subsequently a trench 6ft 8in (2m) x 3ft (0.9m)was opened 6ft (1.8m) to the south. This uncovered a 'hearth'and considerable quantities of flint flakes 'above and just within thered sand' which lay beneath the chalky rubble. A few pieces ofantler, most if not all from red deer, were also found.

Floor 59 - 'lay among the birches to the north-east of the wood'. Some flintimplements are described, but the Floor itself is not.

Finally, Kendall (1920, 305) also investigated a shaft 'on thenorth side of the wood, some distance north-east of Floor 52, the two trenches maderunning crosswise, and to a depth of 3ft (0.9m), and at one side, 6ft (1.8m). Noimportant results were obtained, and the trenches were half filled in again'.

July 1920 - Kendall

The next excavations known to have been undertaken appear to bethose of Kendall (notes in the Alexander Keiller Museum,Avebury); however, given that he is describing Floors 73-77, itagain seems probable that some excavations remain unreported.These five Floors are described very briefly in his notes, butit is clear that they do not represent the sum total of hisefforts that year. Kendall (1925) refers to an arrowheaddiscovered by him in 1920 from Floor 48, though whether thiswas excavation rather than surface collection is unclear.

Floor 73 - examined on July 15th, this was located 16 pacesnorth-east of Floor 58. 6ft 4in (1.9m) in extent and 6-18in(0.15-0.45m) deep, it yielded, among other items, two 'celts or celt-like implements'.

Floor 74 - dug into on the following day, July 16th. Located 13paces north-west of Floor 46, excavated the previous year, itwas described as being up to 1ft 10in (0.5m) deep, andconsisting of sand, rubble and flakes. It apparently dippedsteeply to the south. A 'celt' from a 'hearth' in Floor 74 wasapparently given to 'an officer who dug it out'.

Floor 75 - excavated during the same month. Crescent-shaped, itwas '14 paces long in the curve', 4 paces wide, and 12-20in (0.3m-0.5m)deep. It was situated on the edge of the North Field. Armstrong(1925) describes an arrowhead found at this floor by hisdaughter Morwenna.

Floor 76 - examined on July 24th, this yielded 'handsome cores andflakes', and was 3-4in (7-8cms) deep above chalk rubble.

Floor 77 - centre to centre, this was 54ft (16.4m) north-westof Floor 75. Up to 18in (0.45m) deep, it featured a 'hearth' ofc.3ft (0.9m) diameter on its west side.

September 1920 - Armstrong

By the time Armstrong returned, two months after Kendall'svisit, there had apparently been further unreportedexcavations, as his published account mainly concerns Floor 85(Armstrong 1921a, b). Longworth et al (1988, 26) provide someinformation on Floors 79 and 80. Floor 79 apparently lay northof Floor 42 (and east of the 'unknown' Floor 81). A few sherdsof Deverel-Rimbury bucket-shaped vessels were found, along 'withpot boilers and charcoal and bones and small flint flakes. 1920...above Grime's Gravesfactory level'. They also refer to a manuscript list which adds'Bronze Age hearth' to the aforementioned finds. Floor 80 issaid by Longworth et al to have been excavated in 1930, whichsuggests that it was examined out of sequence, unless the 1930episode was a re-excavation. It was close to and just north ofFloor 42. They quote from a surviving letter from Armstrong,which refers to the Floor as a domestic site with pot boilersand Bronze Age pottery, a chalk lamp, a point from a polishedbone pin, bones of red deer, sheep, ox, and the teeth of acarnivorous animal. An associated chipping floor was apparentlypartly buried in sand. Not all the pottery described byArmstrong survives, but among that which does are Deverel-Rimbury sherds.

Armstrong (1921a, b) was primarily concerned with reporting thediscovery of a number of pieces of 'engraved flint crust' fromFloor 85, which is therefore discussed in some detail (theseand other similar items published by Armstrong have figuredvery little in later discussions of the site, and their statusis best described as uncertain. They do not, of course,represent the only examples of prehistoric 'art' of dubiousorigin from Grimes Graves). The other specimens among the 'severalinteresting discoveries' alluded to are omitted from Armstrong'sreports. However, it should be noted that his illustrations (inArmstrong 1921a) include items from Floors 42 and 75, bothotherwise absent from published accounts. Armstrong (1921b, 84)refers to other items bearing 'engravings...of the apparently formlessvariety', as opposed to the animal representations from Floor 85,having been recovered from various Floors by Peake between 1916and 1920, though not published by him. Two such items alsofeatures animal heads - one reportedly found in September 1920on Floor 75 (a Floor otherwise described not by Peake but by

Kendall - see above). No provenance is given for the secondexample.

Perhaps in acknowledgement to earlier flint mine controversies,Armstrong (1921b, 85) finished his account by stating that 'I wasassociated in these excavations with Mr BWJ Kent (FSA (Scot)), and assisted on severaloccasions by Mr JB Sidebotham, both of whom were present when the engraved pieceswere discovered'. In the discussion subsequent to the delivery ofhis paper (reported in Armstrong 1921b, 85-6), Reginald Smithnotes the identification of the most complete animal engravingas an elk, which he noted 'went back as far as the Forest-bed of Cromer, atthe base of the Pleistocene', but otherwise avoids directly commentingon the chronological debate which he had initiated almost adecade earlier. Armstrong had, in any case, argued that theengravings supported the idea of a Palaeolithic presence atGrimes Graves, a point which Kendall found contentious (inArmstrong 1921b, 86), arguing that the evidence to the contrarywas 'overwhelming'.

Floor 85 itself was found near the south-east margin of themining area, and immediately to the west of the so-called'Tumulus Pit', ie the large pit immediately adjacent toGrimshoe. An area of 36 yds² (30.1m²) was examined to a depthof 3ft (0.9m). Three phases of occupation were identified byArmstrong (numbered 85 A, B and C) in addition to Iron Age andRoman potsherds from the topsoil. The three floors appear tohave been stratigraphically distinct, separated by layers ofchalk rubble with varying amounts of sand, and varied inextent, with the uppermost, 85A, only observed on the northernmargin of the excavated area, appearing 'to extend for a considerablearea'.

Floor 85A, immediately below the topsoil, featured black sandand charcoal, and contained animal bones, pot boilers, BronzeAge, Neolithic and Roman potsherds (Longworth et al 1991, 176),bone tools, and some bronze tweezers which may be of LateBronze Age date. Floor 85B was a compacted mass of flint flakesand implements, together with animal bones, antler, and 'smallhearths'. Floor 85C was described as an 'extensive chipping zone'. Atthe west end was a 'hearth' 4ft in diameter consisting offorty-five large flint nodules and blocks, with some smallerflakes, the whole thickly covered with charcoal. Near the

centre (presumably of the 'hearth') was the base of a potteryvessel. Much flint waste and nodules occurred in the generalarea. More bone tools and pottery also came from other parts of85C. The engraved flints from Floors 85C and 75, plus a fewnoted from earlier excavations, provided Armstrong with what heregarded as further evidence of Palaeolithic activity at thesite, although the idea of later occupation was by now clearlya less contentious matter.

Armstrong also re-examined Pit 1, finding an additional gallerywest of Gallery 18, and mapped the courses of three previouslyknown but unexplored galleries (21, 22 and 24). He also notedthat in the six years since Pit 1 was excavated, the shaft hadfilled up to a minimum depth of 7ft 6in (2.2m) with erodedmaterial. This accumulated debris proved to contain a socketedbronze axe not too far below its surface. Armstrong suggestedthat it had probably fallen in with material from the uppersandy layer above the pit.

August 1921 - Armstrong

Armstrong returned and 'carefully and systematically explored' anadditional 40yds² (33.4m²) of Floor 85 over a period of sixdays. Immediately beneath the turf was found a block of flintwith what Armstrong described as iron scoriae attached to it.Floor 85A produced more bone tools, flints, charcoal, animalbones and potsherds, as well as forty-two small quartz pebbles.Floors B and C were interpreted as representing a series of 'one-man chipping sites grouped around a small central hearth...'. Finds weresimilar to the previous year, and included further examples offlints bearing engravings or incised lines.

Continuation of the trench to the north-east suggested that thelowest floor, 85C, had originally extended across the area nowoccupied by the 'Tumulus Pit' adjacent to Grimshoe. Armstrongsuggested that the sandy material above 85C was also cutthrough by the shaft. He offered this as a possible explanationfor the apparent mixing of flint assemblages in the shafts and'chipping floors'.

Armstrong attempted to reconcile the apparently conflictingdating evidence from the site as a whole by concluding that 'the

Grime's Graves culture is not representative of only one period, its mines andworkshops are the cumulative results of many generations of work spread over severalcultural periods...'. He suggested the following sequence:

'Lower Grime's Graves' - this was represented by, eg, Floor85C, immediately overlying the boulder clay and apparently cutthrough by one of the mine shafts. Although this phase saw theuse of mined flint, Armstrong felt that no shafts attributableto this phase had yet been found. He suggested that they maylie 'blanketed in sand on the north-west side of the plantation', a suggestionwhich disregarded the problem of the excavated shafts havingclearly been cut through the sandy layer.

'Middle Grime's Graves' - represented by Floor 85B.

'Upper Grime's Graves' - represented 'probably' by Richardson's'celt-making floor'.

Interestingly, Armstrong now appears uncertain as to the datewhen mining commenced, noting the 'Palaeolithic form' of manyimplements compared with the Neolithic or later date implied byother forms of evidence. His conclusion, echoing that ofRichardson a few years previously, was that 'the key...will certainly befound if we just keep on digging'.

July 1922 - Armstrong

Armstrong returned to Floor 85 (Armstrong 1924b) and increasedthe area of the excavation to 109 yds² (99.6m²), concentratingon the south, south-east, and east sides of the previouslyexcavated area. The new excavations showed Floor A to be absenton the south side, but still 'well-developed' on the east andnorth-east, containing charcoal, animal bones, pot boilers andsherds of pottery in the latter areas, as well as flintimplements and flakes, quartz pebbles, hammer stones andwhetstones. Floors B and C appear to have been encountered inall areas, but varying in thickness, while on the south sidethe sterile layer separating the two was very thin making thetwo floors almost indistinguishable in places. Armstrong'sassistant, Dr Favell, also dug a trial section some 50 yards(45.7m) south-east of Floor 85 in order to see if Floor 85Ccontinued in that direction. Armstrong noted that 'various

implements and waste material characteristic of Floor 85 were found, many of them incontact with the red sand, but as the overlying material showed signs of considerabledisturbance and the area was a network of tree roots it was abandoned'.

Longworth et al (1988, 26) describe the surviving pottery fromFloor 85 as follows: 'a rim sherd of plain Middle Neolithic bowl marked Floor85 and a probable wall sherd from Floor 85A. Ten sherds of bucket-shaped vessels,together with four undecorated wall sherds, survive from Floor A and an undecoratedwall sherd from Floor C as well as two unlocated wall sherds. There are also two sherdsof Early Iron Age pottery labelled as from Floor C, a sherd of Roman jar from Floor Aand sherds of Samian...presumably from the topsoil'.

A Floor 86 is also referred to by Longworth et al (1988, 26). Itis said, presumably by Armstrong, to have been one of the'earliest factory sites'. A single potsherd survives. The Floor waslocated a short distance to the north-west of Floors 79 and 80.Also in July 1922 Armstrong undertook some preliminary workaimed at putting his theory concerning the earliest mines tothe test (Armstrong 1923). A series of sections were dug 'on thesouthern slope of the valley, north of the Graves...'. Armstrong noted thatvaluable geological information was recovered (ie no flints orfeatures).

September 1922 - Kendall

Notes written by the Rev Kendall (held at the Alexander KeillerMuseum, Avebury) refer to an extension of Floor 75, which hehad first investigated two years previously. He appears also tohave examined (or re-examined) Floors 58, 89, 90, 91 and a newsite, Floor 92, for which he provides no clear indication oflocation, but describing it as

...a small Bronze Age (?) floor with many rather small flakes &c had beenfound by us previously and abandoned. On the last afternoon I put thespade in deeper and got out a good knife. Only an hour or two was left,but as far as could be seen, we cleared out the earlier (Neo) floor, whichwas under the later hearth. We got a beautifully made little celt, and asplendid ovate. The floor was covered with chalk rubble. These threeimplements and some flakes were in it; other flakes were on the redsand, under the chalk blocks'.

His notes are less detailed as far as the other Floors are

concerned. Floor 89 may well be the one referred to (September2nd) as being 'right in the path, near the little sycamores, leading to Pit 1'. OnSeptember 4th, '...as it was drizzling, [I] dug two holes on a slope of groundleading from a pit in the wood down to a very deep pit; apparently a floor of not verygood flakes...', while the next day he dug more holes alongside the'West track'. He found a Floor and a 'hearth', adding 'this floor notnumbered; no doubt Bronze Age'.

1923 - Armstrong and Lingwood

In 1923 Armstrong began a long series of excavations financedby the Percy Sladen Memorial Fund, initially aimed at locating'primitive flint mines' along the suggested line of a possibleformer outcrop of floorstone on the valley slopes to the north-east of Grimes Graves (Armstrong 1923). In all, seventeen trialholes were dug, and although Armstrong described the results as'most interesting', most of the trial holes are not discussed inhis report.

The first major discovery occurred in trial-hole no. 5, 'at theextremity of a line of sections extending from the lap of the valley southwards'. Thetrench revealed the existence of a shaft of which there were nosurface indications. However, as Armstrong noted, 'as pits wereknown to exist on the area formerly cultivated immediately to the south of the trialhole, the presence of a pit at this point was not altogether surprising'. Armstrong(as noted above) decided to follow on from the 1914 numberingsystem, calling this Pit 3, a number previously used by Peakein 1917 (Peake 1919). However, others appear to have followedArmstrong's numbering, although some confusion has occasionallyarisen (eg Mercer 1981a, 5-6).

The shaft was fully excavated. It proved to be about 12ft(3.6m) in diameter at the mouth, and 13ft 4in (4m) deep,although the 'floorstone' was reached at c.12ft 4in (3.7m). Nearthe bottom it was 6ft (1.8m) in diameter before undercuttingexpanded it to 8ft (2.4m) x 5ft 6in (1.6m). Armstrong regardedit as early because the top was sealed by 18in - 2ft (0.6m) oftopsoil, chalk rubble and sand 'without the suggestion of the slightestsaucer-like depression'. No galleries were observed, although theshaft appeared to have been widened at its base in order toretrieve as much flint as possible, according to Armstrong.Access to the shaft appears to have been by means of some

'rudely cut' steps on the south-west side, terminating in alevel platform c.4ft 4in (1.3m) square about 4ft 6in (1.3m)above the bottom of the shaft. The main features to note in theshaft fill were a 'chipping floor' up to 6in (0.1m) thickcontaining 'waste material [which] was rudely flaked and mis-shapen and ingeneral facies did not resemble the waste of the ordinary Grimes Graves floors'. Thisfloor was immediately below the sandy layer beneath thetopsoil. Another floor, more limited in extent, was found c.4ft(1.2m) below the first one. in addition, waste flakes and lumpsof 'wallstone' were encountered throughout the fill, but fewimplements were recovered.

This shaft was re-excavated in 1976 (Longworth and Varndell1996, 35ff) during the British Museum's programme of excavationat Grimes Graves. This re-examination showed that the shaft hadclearly been dug in two stages, the first down to c.2m to enableremoval of nodules in a zone of cryoturbated chalk.Subsequently half was deepened to the level of the'floorstone'. In addition in 1976, a second shaft similar toand 1.5m south-east of Pit 3 was located and partiallyexamined. This shaft, numbered 3a, proved to be very similar incharacter to Pit 3, with working clearly having occurred at twolevels. A trampled platform observed between 3a and an adjacentunexcavated shaft (3b) is suggested to demonstrate that bothwere open at the same time (Longworth and Varndell 1996, 45).

Another shaft, numbered 4 by Armstrong, and located 26 yards(23.1m) east of his Pit 3, was also excavated by him. Armstrongnoted that it was only found 'after several baffling attempts' but doesnot elaborate on this. Once again there were no surfaceindications. Excavation revealed a shaft 12ft 6in (3.8m) deep,with the 'floorstone' encountered at 11ft 6in (3.5m). The shaftwas 10ft (3m) x 9ft (2.7m) wide at the lip, 6ft (1.8m) at about4ft (1.2m) above the bottom widening to 9ft 6in (2.9m) x 9ft(2.7m) at the floor. As with Pit 3, Armstrong referred to aseries of rough steps leading down into the shaft, though thetwo-stage working shown for Pit 3 by the British Museum may bea more appropriate explanation. Pit 4 was not described in anydetail by Armstrong, apparently because of its similarity toPit 3, although he makes no mention at all of 'chippingfloors'. Pit 4 apparently contained a greater number offinished implements, as well as a chalk 'cup' or 'lamp'.

Armstrong regarded these two pits as indicative of a more'primitive', ie earlier, phase of mining than those previouslyexcavated, in accordance with his own theories prior to theirexcavation. His evidence included the shape and form of theshafts, the lack of galleries, and the more solid and compactnature of the fill. In addition, he pointed to the absence ofantler and the discovery of bone 'hand picks', most displayingsigns of damage interpreted by Armstrong as use-wear, in thefills of both shafts. Six complete and five broken exampleswere found, as well as smaller splinters of bone. Most werefrom ox, but at least one was made from a human femur.Armstrong felt that these represented a type of diggingimplement ancestral to the antler pick.

Of the seventeen sections dug in 1923, Armstrong reported onjust one. Again he makes reference to their geologicalimportance, something which suggests that little of directarchaeological significance was recovered from the othersixteen. The sole section discusses was situated north of Pit3, at a point where the floorstone was believed to be c.5ft 4in(1.6m) below the surface. In fact, excavation revealed that thechalk and floorstone at this point had been contorted byglacial pressure and the flint seam was encountered just 6inbelow the surface. Armstrong felt that outcrops such as thismay have seen the earliest phases of mining at Grimes Graves.

Also in 1923, ET Lingwood located and excavated Floor 88.Situated 15 yards (13.7m) west of Greenwell's Shaft, an area ofsome 126 yds² (115m²) was examined (Armstrong 1926). Thetopsoil contained scattered chalk blocks and debris, flakes andlumps of flint, etc, but this was felt to represent spoil fromnearby shafts, with the 'floor' below being undisturbed. It wasinterpreted as a 'factory site' as it lacked hearths, and wasconsidered to be contemporary with or possibly earlier thanFloor 85.

Finally, and again in the summer of 1923, kendall discovered apair of carved chalk objects (Kendall 1924), both described aschalk lamps. Neither were found within the area of extantearthworks: one was recovered from 'a very small floor in the north field',and the other from 'a small floor in the west field'. It would seem that

Kendall visited Grimes Graves for the last time in September1924 (notes in Alexander Keiller Museum, Avebury). However,beyond mentioning a 'floor in the valley' where he found flakes whichfitted together, he provided no details.

July 1924-1926 - Armstrong

Three more shafts were located close to Pits 3 and 4, and twoof them were excavated. Pit 6 was located immediately to thewest of Pit 4, and was linked to it at the bottom, where theundercutting of one had broken through into the 'belling-out'of the other, an occurrence which, within his evolutionaryscheme of mining techniques, Armstrong postulated might haveled to the idea of galleries. There were no surface indicationsof Pit 6, hence its numbering - Pit 5 is to the west of Pit 6but was encountered first in trial-trenching. Another shaft,Pit 7, was also located very close to and slightly north of Pit5. Pit 7 was partially excavated to a depth of 7.5ft (2.2m)(Longworth and Varndell 1996, 65). A 'hearth' was found at 4ft6in (1.3m). The shaft was c.12ft (3.6m) wide at the surface, andthe upper fill comprised sand and chalk rubble.

Pit 6, the first to be excavated, was 8ft 2in (2.4m) x 5ft 6in(1.6m) at the mouth and 13ft (3.9m) deep. The fill was mainlyred sand, sandy chalk rubble, and fragmented chalk, apparentlytipped in mostly from the east and south. Waste flakes andcores were common throughout the fill, but no 'chipping floors'or 'occupation levels' were observed. Lumps and slabs of bothwallstone and floorstone were common. Again, rough 'steps' werepresent. In the upper part of the chalk, Armstrong referred to'numerous pipes and pockets', some of which had been clearedout and, he suggested, used as steps. Lower down a projectingbuttress had a recessed 'step' cut into it. Flint implementswere described as being fairly numerous, and fragments andsplinters of bone picks were also found. Longworth et al (1988,27) note that a single wall sherd of Deverel-Rimbury stylebucket-shaped vessel and a sherd from an Early Iron Age fineshouldered jar survive from this shaft.

The link between Pits 6 and 4 appeared not to have beendeliberate. It proved to be only 15in wide, with an irregularopening. Pit 6 was also linked to another adjoining shaft by a

larger, but again apparently accidental, opening on its northside. On the evidence of Armstrong's plan, this link would seemto be with Pit 7. Armstrong felt that the wall between the twoshafts had collapsed subsequent to their infilling.

Pit 5 had been partially excavated prior to the discovery ofPit 6, and the job of clearing it out completely was continuedafter Pit 6 had been emptied. It was 7ft (2.1m) x 6ft 3in(1.9m) at the surface and 12ft 9in (3.8m) deep. Only 1ft (0.3m)separated it from Pit 4 at the lip. A 'chipping zone...limited in botharea and thickness' was located at c. 2ft 6in depth, and contained'only massive flakes and rough cores'. Again the base of the shaft belledoutwards, but this time it took the form of a series of low,wide pockets of varying size, generally around 2ft 6in (0.75m)deep. Bone 'picks' were found, and once more a 'rude staircase'was present, this time represented by a small platform c.4ft(1.2m) from the bottom of the shaft, with three deep pocketscut into the shaft wall above it to give access to and from thesurface. Flints recovered from both shafts included whatArmstrong described as 'choppers' and 'wedges' of flint, whichhe argued had been used for breaking up the chalk and thefloorstone. Otherwise the flint assemblage consisted mainly ofthe usual flakes and cores with a small number of implements.

The West Field area was examined by digging two parallel linesof trial holes, each line 20 yards (18.2m) apart, withintervening cross trenches. It seems that fifty-two 'holes'were dug, although it is not clear whether this total includesthe cross trenches. The aim of the exercise was to see how farthe floorstone remained in situ, and to check on the presence orotherwise of 'primitive flint workings' in the valley.Armstrong reported that no workings were located in the WestField which were earlier in date than Pits 3 to 7.

'Floorstone' was found at a depth of 6ft 8in (2m) at a distanceof 222 yards (202.9m) from the eastern fence of the field. The'floorstone' was slightly contorted and the chalk above it wasdescribed as completely eroded. Armstrong continued:

Ten yards [9.1m] further to the west it was present only as a folded andcontorted mass of chalk and flint at 5ft 9in (1.7m) and upwards. Eightyards [7.3m] beyond this point the section shewed gravel and sand with

rafts of chalk but no Floor-stone down to 3ft 9in [1.1m], the line of theFloor-stone being 3ft 3in [1m]. The section in the lowest portion of thevalley revealed flinty gravel with a little sand, all much water worn. Thesurface level of the valley at this point is 3 inches below the line of theFloor-stone layer'.

He also noted that 'mine shafts were revealed in three of the upper sectionswhere the Floor-stone lies at a depth of about 18 feet', but he did not providea more accurate location for these shafts; however, the depthof the 'floorstone' suggests that they cannot have been veryfar along the line of the trial-holes. He also reported thattwo of the sections indicated an 'occupation' level, one at adepth of 1ft 6in (0.45m), the other at 4ft (1.2m). Furthertrenching showed them each to be shallow, roughly circular,hollows formed in the boulder clay. Each contained a black band4in to 6in (10cm to 15cm) thick and covered with blown sand.The only artefacts were flint flakes. Armstrong argued that 'thegeneral facies of the layer and the character of the contemporary flint chipping wassuggestive of a late date, probably Iron Age'.

No shaft was located in the area of the contorted 'floorstone'.The lowest shaft found (ie the westernmost), designated Pit 8,was at a point where the 'floorstone' was estimated to be some15ft (4.5m) below ground level, although there was insufficienttime to excavate the shaft fill. It is described as being atleast 30ft (9.1m) in diameter, and appears to have beenexcavated by Armstrong's workmen as far as the 'wallstone'layer at least. Finds included a red deer antler crown and twoantler picks, with other antler fragments, and a small sherd ofpottery of unknown date. One of the antler picks was markedwith a series of incised lines. Armstrong argued that 'the methodof mining resembles that adopted in 1914 [Pits 1 and 2], but not precisely so and Pit 8may prove to belong to a phase intermediate between the primitive group of pits andthose examined in 1914'. However, he did not elaborate on theevidence for this intermediate placing.

A feature named 'the Black Hole' was located by Armstrong whiledigging a trial hole in 1922, although he made no mention of itin his report on that season's work (Armstrong 1923). He foundpottery, flint implements and animal bones, and decided toleave the site until more time was available for carefulinvestigation. This happened in 1924, although unfortunately

the area had been subject to some unauthorised, or at leastunrecorded, digging in the interim. Armstrong appears to haveleft his earlier trial hole open, and shortly after his returnin 1924 it seems that 'the excavation had been opened out and drasticallyenlarged by a party of unauthorised visitors to the Graves...The section had beenextensively disturbed and undercut'. This unauthorised digger wassubsequently identified (Armstrong 1926) as a Dutch professorwho obtained 'pottery and fauna,...from which evidence he assigns the whole of themines to the Iron Age'. Much of the work in 1924 thus took the formof a tidying-up operation combined with an attempt to definethe limits of the deposit. Further work occurred in 1925 and1926.

Armstrong interpreted the feature as being a large communalhearth of early Iron Age date. It was initially thought to besituated within the depression above one of the larger mineshafts on the southern margin of the earthworks. However,subsequent work in 1926 suggested to Armstrong that the hollowwas a natural one among the sand dunes. The presence of duneswas confirmed by further trenches dug in the area. Inparticular, two large circular depressions previously regardedas representing mine shafts, one to the east of the Black Holeand the other 'beyond Floor 16', were also now thought to behollows in the dunes. The Black Hole area was found to be 18ft(5.4m) in diameter and completely filled with a

stratified deposit of charcoal ashes, and black carbonaceous sand,containing a prodigious quantity of animal bones, pot-boilers of flint andstone, worked flints of all periods, pottery in considerable quantities,bone tools, shells, carved chalk, chalk balls and perforated objects ofchalk resembling loomweights..., a huge communal workshop andcooking place, situated in a hollow on the extreme south-east margin ofthe mined area. It was roughly circular, 31 feet in diameter, and in thecentre 11 feet [3.3m] deep, of which 9ft 6in [2.8m] was an accumulationof charcoal-ash and household debris...

The pottery was only briefly described: eleven differentfabrics were recognised, and reference was made to the presenceof typical Halstatt wares. Longworth et al (1988, 27) refer to1081 sherds of Deverel-Rimbury style bucket-shaped pottery assurviving, but note the possibility that Armstrong may not haveretained smaller fragments. Some Late Bronze and Iron Age

sherds are also noted (Longworth et al 1988, 105). One piece ofmetalwork, a coiled bronze ring, was found. Bone tools arereferred to but not described by Armstrong, although aselection was illustrated in a photograph. A small chalk cupwas also recovered. Armstrong expressed some surprise at thequantity of flint implements associated with a deposit which hebelieved to belong to the Iron Age.

Longworth et al (1988, 26) refer to a feature known as 'Kent'sLow Floor', the precise location of which is unknown, but whichis thought to be an extension of the Black Hole. Survivingfinds include thirteen sherds of Deverel-Rimbury bucket-shapedvessels, eleven undecorated body sherds, loomweight fragments,a bronze spiral finger ring (possibly the one referred toabove) and a Medieval potsherd. The same volume also containsreference to a 'section' south-west of the Black Hole, againnot precisely located, but which produced more sherds fromBronze Age bucket-shaped vessels and wall sherds plus sherdsfrom an Early iron Age shouldered jar.

In 1926, Armstrong published a synthesis of work undertaken atGrimes Graves in an attempt to resolve the outstanding problemof the date of the mining. He summarised his view that miningwas long-lived, beginning in the Palaeolithic and continuinginto the Neolithic, with Bronze Age and Iron Age occupationoccurring only after all mining had ceased. However, the sheerquantity of flint associated with these later episodes ofactivity and the presence of what were seen as 'palaeolithicforms' was clearly puzzling; Armstrong argued by way ofexplanation that the later occupants must have been 'assiduouscollectors of useful antiques in the form of flint implements'.

1927-1933 - Armstrong

These excavations were summarised in a single, short interimreport (Armstrong 1932). In that period, four pits werecompletely excavated and work was begun on a fifth. In additionthirty-two sections were cut, most of them in the West Field.In 1927, work focused on Pit 9, which had been discoveredduring the previous year's trial trenching. Located on thesouthern side of the valley, close to the north-east corner ofWest Field, it was found to be roughly circular with a diameter

at the top of 21ft (6.4m). The 'floorstone' was calculated tobe at a depth of c.22ft (6.7m) at this point. However, the shaftproved to be only 9ft 6in (2.8m) deep. The fill of the shaftcontained large quantities of apparently rejected 'floorstone'(Armstrong stated that the pieces covered an area of 110 sq ft(33.5m²) when laid out side by side). However, close inspectionof the base of the shaft suggested that the 'floorstone' whichhad been exploited was not in situ.

Careful excavation around the base of the shaft wallsultimately revealed the presence of numerous coves in thecompacted chalk and boulder clay, and when these were clearedout, a layer of 'floorstone' was exposed at the back. This wasfound to exist around the whole circumference of the base ofthe shaft, though in places it was very sparse and everywhereshowed contortion, many of the blocks resting upon their edges.

A trial hole dug into the base of the shaft suggested thelikelihood that the genuine 'floorstone' layer would indeed befound in situ at the expected depth. The material at the base ofthe shaft was suggested to have arrived there through glacialaction. Armstrong referred to it as 'rafted floorstone'.

As for the fill of the shaft, several 'hearths' wereencountered at various depths, while a 'well defined chipping floor'6in (14cms) deep spread over the top of the fill, yielding aseries of 'interesting artefacts'. Picks made of both red deer and roedeer antler occurred in the shaft fill. Among the flintimplements were 'heavy choppers and picks', which Armstrong suggestedhad also been used for breaking up the boulder clay and chalk.This, combined with the large diameter of the shaft, led to thesuggestion that the shaft belonged to his 'intermediate' miningphase. The presence of some axes with what Armstrong describedas 'modified tranchet blades' were felt to provide evidence of aMesolithic date for this particular mining phase. Armstrongalso felt that some 'wallstone' nodules had been used to makemining implements.

1928 saw the excavation of two further shafts (8 and 10), bothlocated on the West Field and apparently lacking any surfaceindication. Pit 8 had first been encountered in 1924 (seeabove). In both cases, excavation revealed a similar picture to

Pit 9, with 'rafted floorstone' having been exploited. In bothcases, trial holes were sunk through the bottom of the shaftsin order to demonstrate that the true 'floorstone' was indeed insitu at the expected depth.

Both shafts were comparatively shallow, 7ft (2.1m) to 8ft(2.4m) deep, and were connected in each case to two adjacentshafts by means of wide gangways. The shafts connected to Pit 8were numbered 8A and 8B (the latter connected to Pit 8 at itsbase), while those connected to Pit 10 were not numbered in thereport. Finds from Pits 8 and 10 included antler picks,although the former also included a number of bone 'picks'. Pit8 also yielded a chisel-shaped flint within an antler haft.Implements were not numerous among the flints from eithershaft.

Apparently, while these two shafts were being excavated, DrFavell dug 'numerous trial trenches...in various places'. One of theseuncovered 'an abnormally thick and extensive chipping floor' which ultimatelywas shown to occupy the hollow above another shaft, numbered12. The floor was 3ft (0.9m) maximum thickness, and at its basewere found some Neolithic sherds identified by Armstrong asbelonging to a 'bowl-shaped pot of Peterborough type'.

No work was carried out in 1929, but in 1930 Armstrong returnedand two more shafts (Pits 11 and 12), were completelyexcavated. Pit 11 was 'discovered by purposeful trenches' mid way up theslope of West Field. Some 14ft (4.2m) deep, the finds includedwhat Armstrong described as 'an important series of flint implements showingfurther stages in the development of the tranchet axe towards the neolithic celt' aswell as red and roe deer antler picks; bone 'picks' wereabsent. Pit 11 was still in the area of the 'raftedfloorstone', but the greater depth appears to have led to theuse of creep-holes rather than open gangways in order to linkup with neighbouring shafts. Longworth and Varndell (1996, 45)provide a few more details about Pit 11 from Armstrong'sunpublished notes. It seems that 2ft (0.6m) from the surface hehad encountered a 'black layer' c.1ft (0.3m) thick in the centreand 13ft (3.9m) in extent. Some Romano-British sherds representthe only artefacts known from this deposit. Pit 11 was anotherof those subsequently re-examined during the British Museumprogramme of work during the 1970s. The shaft was re-opened by

PJ Felder and colleagues, who assisted on some of the BritishMuseum excavations:

The Dutch mining team re-opened Pit 11,,,using a dragline to remove theprevious backfilling, disclosing the presence of a series of six furthershafts...with galleries suggesting a seventh...but revealing no trace ofany walls belonging to any original central shaft' (Longworth andVarndell 1996, 45)

A segment of Pit 12 25ft (7.6m) north-south and 9ft (2.7m)east-west was excavated by Armstrong to a depth of 5ft 6in(1.6m) in 1930, and in 1932 and 1933 the shaft was completelyexcavated (Armstrong 1934). Located at the extreme eastern endof the West Field, the boundary fence had to be moved to allowthe work to go ahead. Excavation showed Pit 12 to be c.36ft(10.9m) wide at the top and around 18ft (5.4m) deep, with anumber of galleries radiating from the base of the shaft inorder to exploit the in situ 'floorstone', although Armstrongnoted that 'its presence at that level provided a complete surprise becauseaccording to its presumed relation to the Floor Stone of Pit 1 and Pit 2 it was 8.85ft(2.6m) higher in Pit 12 than it had any right to be', something that appears tohave been accounted for by the presence of a major faultrunning north-south roughly between Pit 12 'and the visible shafts'.

As noted above, Neolithic pottery had already been recoveredfrom the chipping floor overlying the shaft fill duringpreliminary work following its initial discovery, and thisdiscovery prompted further investigation. By 1934 the floor hadbeen sub-divided into two layers (A and B), although thereappears to be nothing in the stratigraphy to confirm such adivision. In fact, the whole shaft was described as beingunusually rich in such 'floors' - in all, eight (A-G) weredescribed:

Floor A - mentioned above, 'was comparatively thin and scattered over thewhole surface of the pit', yielding 'no artifacts of importance'.

Floor B - again mentioned above, was 'astonishingly unproductive offinished implements', despite the thousands of flakes containedwithin it, which Armstrong described as being 'so closely compactedtogether that they resembled a tip of road metal'. Its maximum thickness was3ft 9in (1.1m), and in plan it was approximately circular,

c.25ft (7.6m) in diameter. It was estimated to contain c.13,000cubic feet of material, most of it apparently of 'floorstone'.near the base of the 'floor', on the eastern side, were three'hearths', one of them close to some fragments of human craniumand a single molar, none of which was burnt. Another of the'hearths' had around it teeth and split bones of sheep and ox.Two further small sherds of Neolithic pottery were also foundwithin the 'floor' material.

Floor C 'occupied the bank of the shaft from West to South and extended in a thinspread of flakes towards the centre of the pit where it terminated beneath Floor B,from which tilts of chalk and boulder clay separated it'. Several small'hearths' were found as were further Neolithic sherds,including more Peterborough ware. Again, despite the quantityof flint, 'no implements of importance' were recovered.

Floor D - described as a 'celt-making floor', and occurring ata depth of c.5-7ft (1.5-2m), up to 1ft (0.3m) thick, and runningfrom the south side of the shaft to the centre of the fill.Finds included 'large pieces of Floor stone, dressed cores, shapely flakes andworn out cores,...also roughed-out celts illustrating all the stages of production to thefinished article, but only broken halves of well-finished specimens'. Given theorigins of the chronology debate, it is intriguing to seeArmstrong describing the presence of rough-outs and variousstages of working.

Floor E - about 8ft (2.4m) in diameter and about 6ft (1.8m)down on the north side, with a small central 'hearth'.

Floors F and G - similar in extent and situation, each spanningthe entire width of the shaft fill in section, F at a depth of13ft (3.9m), G at 14ft 6in (4.4m), each resting on a smoothsurface of chalk debris. Small flakes were abundant in both,but tended to occur in marked concentrations.

The remainder of the shaft fill consisted of layers of chalkblocks, chalk rubble, sandy rubble and boulder clay, mostappearing to have entered the shaft from the east, south andnorth sides. Among the artefacts were half a chalk ball and twowhole ones, together with an oval-shaped 'ball', severalquartzite hammerstones, red and roe deer antler picks, flintsincluding scrapers, a knife and a rough 'celt'. A single sherd

of undecorated pottery was also found at c.12ft (3.6m) depth.

Eight galleries were observed in the area excavated. Three werecleared out and a fourth partially examined. All were packed tothe roof with chalk rubble except numbers 2 and 4, which hadonly been partially filled. Gallery 1 led directly to aneighbouring shaft. The other galleries were not followed farenough to see if they did likewise, although it seems likelythat some of them did. No flint implements came from any of theexplored galleries, but sixteen picks were found, mostly reddeer, though one was roe deer and another was made from oxbone. The shaft itself was not fully excavated because theMinistry of Works wished the shaft to be left open for publicdisplay, and it was decided to leave a full section of the fillin situ.

By the time Armstrong published his report on Pit 12 in 1934,he had already effectively lost the argument over the date offlint-mining. That doubts still lingered into the early 1930sis evident from Kendrick's discussion of the site (in Kendrickand Hawkes 1932, 74-78). However, the Palaeolithic dating ofGrimes graves was firmly rejected in print by Grahame Clark andStuart Piggott (Clark 1932; Clark and Piggott 1933), who arguedthat all phases of mining could be dated to the Neolithic(specifically to the 'Windmill Hill Culture') on the basis ofassociated finds, noting that the Palaeolithic dating restedpurely on a typological study of lithic artefacts whichinvolved comparing selected unfinished items from a mining sitewith finished objects of Palaeolithic date from elsewhere.Inevitably much of the criticism focused on Armstrong's workand interpretations, for example his claim that Floor 85Crepresented early, pre-mining activity despite the presencewithin it of pottery similar to sherds recovered from the shaftfills in 1914, those shafts being considered among the latestat Grimes Graves. Armstrong forcefully rejected Clark andPiggott's arguments in his 1934 report, which also featured aneditorial footnote appended to the paper summarising the mainpoints of the debate. However, it seems that the argument hadalready been won by the advocates of the Neolithic dating, withArmstrong left in a minority.

1934-1939 - Armstrong

Armstrong continued to excavate until the outbreak of war in1939, but did not himself publish anything on this last stageof his work, which focused on three further shafts (numbered13, 14 and 15).

Pit 13 was one of Armstrong's so-called 'primitive pits',resembling in form and content Pits 3-7 (Mercer 1981a).According to Longworth and Varndell (1996, 73), it lay to thenorth-east of Pit 14 (see below). Beneath the topsoil was a 3in(7cms) thick layer of golden sand, and below that was the chalkrubble infill of the shaft. The base varied in width from12.5ft (3.8m) to 15ft (4.5m), and had been deepened in order toextract floorstone from beneath the main nodular layer. Fourundercuttings were encountered, one of them 6ft (1.8m) long, inwhich some floorstone was still present.

Pit 14, although located close to these 'primitive pits',proved to contain developed galleries (Mercer 1981a). It waspositioned to the north of the main complex, about halfwaybetween the earthworks and Pit 4. it was excavated in 1934,1935 and 1936, and proved to be 10ft (3m) deep. At the top itwas 21ft 6in (6.5m) x 26ft 6in (8m) wide. Seven 'coves' hadbeen dug around the shaft base. More details are provided byVarndell (in Longworth et al 1991, 98). It seems that four floorswere noted in the shaft fill. Floor A 'extended up the western lip of thepit and contained charcoal and ash and large flakes'; Floor B 'lay on the easternlip at 1' and contained flakes, cores, broken axes and a discoidal knife'; Floor Cwas 'at the centre of the shaft at 5-6ft (1.6m) [and] contained flakes'; Floor D was'at the north side at 7ft 6in (2.2m), a scattered floor over the area of pit shaft;contained large flakes. Two hearths were noted, one on the west side, one on theeast'. In addition, a saucer-shaped depression in the surface ofthe south-east sector, with a maximum depth of 3ft (0.9m),contained charcoal and ash, plus animal bones including ox,pigs and sheep. Shells and a piece of red deer antler crownwere also present, along with some potsherds (now lost)described by Armstrong as 'black polished pottery of Iron Age type'. A fewobjects of chalk were also present.

Pit 15 was the last excavated by Armstrong, and afterwards wasleft open for public access, along with Pit 1, which itresembled in many respects. Pit 15 is probably best known for

the discovery of various putative votive deposits, includingthe famous female figurine of chalk. The most detailed accountof the excavation of this shaft is provided in recent BritishMuseum fascicules (Longworth et al 1991, 98, 103-5; Longworth andVarndell 1996, 51-9). The shaft was excavated over threesummers between 1937 and 1939. During the 1938 season the shaftwas cleared to a depth of 16ft below site datum, by which pointthe heads of eight galleries had been revealed. The followingyear, further clearance to a depth of 19ft (5.7m) occurred, andall the galleries were exposed. Surviving notes made byArmstrong (quoted here from Varndell in Longworth et al 1991)focus almost entirely on the 'ritual groups'. It appears thatremoval of the remainder of the shaft fill uncovered what wasdescribed as an 'altar' - 'an ogive shaped platform...constructed of blocksof mined flint packed closely together...Lying upon it were seven deer antler picks...Atthe base of the altar and resting upon the floor of the pit was a skilfully made chalkvessel'. The apex of this 'altar' pointed south-east in thedirection of the so-called 'goddess', the aforementionedfigurine, which was found 'at the right hand side of Gallery 7, in an uprightposition resting upon a flat slab of chalk which eventually proved to be the top of apedestal built of similar slabs'. A hearth, consisting of wood ash andcharcoal, was found on the north-east side of the shaft, infront of the 'wall' separating the entrances of galleries 3 and4 'and immediately opposite to the Goddess, in line with the centre of the altar'.Also, in the entrance to Gallery 7 (that immediately adjacentto the Goddess and the pedestal) 'a sculptured chalk phallus was foundupon the floor close to the left hand wall of the gallery; also in the central area, agroup of three selected small natural nodules of flint, arranged in the form of aphallus, and near by, a larger nodule of ovoid shape'.

Piggott (1954, 42) noted Armstrong's suggestion that as Pit 15was poor in flint deposits, 'this obvious shrine of an Earth Goddess mayrepresent an appeal to the chthonic powers for more abundant flint'. Since theirdiscovery, doubts have been cast on the authenticity of some ofthese objects, particularly the 'Goddess', on a number ofoccasions. The problems surrounding these items of worked chalkare discussed by Varndell (in Longworth et al).

As for the fill of the shaft, Longworth and Varndell (1996,51), referring to Armstrong's notes, state that the upperlayers comprised rain-washed silt, which was partly covered bya black carbonaceous deposit which was 4-6in (9-15cms) deep on

the south-east side and up to a maximum of 3ft (0.9m). Itfeatured animal bones (ox, sheep, deer, pig), with charcoal,flint, and coarse pottery said to be like that from the 'BlackHole'. Lower down, Armstrong encountered five 'chippingfloors', the lowest three apparently featuring some evidencefor the reduction of floorstone nodules. Some 'hearths' werealso recorded, one associated with various antler tines andcrowns, all showing signs of burning, plus other fragments ofantler and a couple of picks. Of the 'chipping floors'encountered in the fill of Pit 15, two are of interestprimarily because of the light they shed on Armstrong'sattempts to come to terms with a Neolithic date for GrimesGraves. He describes the floors a producing Levalloisian andClactonian flintwork, adding:

This fact confirmed, for me, the suspicion long held that 'Clactonian' isnot a culture, but a technique employed when flakes of a strong thicktype were required, and that this technique was adopted in all periodswhen deemed most suitable for a particular purpose...these two distinctfloors, which were unquestionably of the same age, prove that Neolithicman employed alternatively both techniques with equal skill when heconsidered it desirable to do so. They also provide, if such is still requires,a conclusive answer to the problem of the presence of 'Palaeolithic'implements at Grimes Graves, which was the subject of so muchcontroversy from 1912 onwards' (quoted from Varndell, inLongworth et al 1991).

Pit 15 was also re-examined by the Dutch mining team during theBritish Museum's excavations at the site (Longworth andVarndell 1996, 50). Their survey of the gallery system revealedseveral discrepancies with Armstrong's own plan (published inLongworth and Varndell 1996, 50-59). While some of thedifferences stem from Armstrong's surveying technique, othersstem from the fact that the pit had been left open afterexcavation, allowing visitors to explore and cause considerabledamage. It also appears that Armstrong had unwittingly exploredsome galleries belonging to adjacent shafts. In all, tenunexcavated adjacent shafts were identified during the re-survey.

Excavations 1971-2

Excavations at Grimes Graves were not resumed after the war.Indeed it was not until 1971 that archaeological investigationof the mining complex occurred once more. The impetus for thiswork arose from unsatisfactory arrangements in place at thetime for displaying the site to the public. In particular, Pit15 had been closed to the public on safety grounds in 1970. Itwas therefore decided that a new shaft should be excavated withthe principal aims of providing a shaft for public display, andobtaining detailed environmental, cultural and dating evidencefrom the shaft and its immediate surroundings as possible. Asubsidiary aim was to examine any Bronze Age occupation,similar to Armstrong's 'Black Hole', which may be encountered.A shaft on the north-east periphery of the mine earthworks waschosen. A second neighbouring shaft was partially examined in1972. The results have been published in full (Mercer 1981a,b).

British Museum Excavations 1972-76

The British Museum programme of investigations at Grimes Gravesconsisted of a programme of survey intended to provide someindication of the extent of the mined area, and the presence ofany occupation areas; and excavation focused both on the flintmining and on the later, Middle Bronze Age activity at GrimesGraves. Publication of this work is ongoing. Sieveking haspublished papers dealing with several aspects of the campaign(Sieveking et al 1973; Sieveking 1979), while the bulk of theresearch programme is being published as a series offascicules, of which 5 have appeared to date (Clutton-Brock1984; Legge 1992; Longworth et al 1988; Longworth et al 1991;Longworth and Varndell 1996).

Cable trench excavation 1982-3

A cable trench 30cm wide and 55cm deep was dug from theentrance to the Guardianship area as far as the custodian'shut, a total length of 285m. The tops of three shafts, notindicated on the ground surface, were encountered towards thesouth-east end of the trench. The published report (Healy 1985)provides an approximate guide to their location in the form ofa 1/5000 site plan. Each shaft was flanked by chalk and sanddumps similar to those encountered on excavations elsewhere

within the mine complex. The only artefact referred to is anantler pick from a layer of chalk rubble and flint nodules.

The remainder of the trench lacked archaeological features,although the slope and depth of soil meant that it did notalways reach below the topsoil. However, 1795 struck flakeswere recovered from the trench spoil, mostly from the south-east end, indicating the possible presence of further areas ofactivity.

Greenwell's Pit - 1984 excavations

Some small-scale excavation occurred around the mouthGreenwell's Pit prior to construction of a concrete capping forthe shaft. This work was also reported on by Healy (1985). Fourcuttings were made, primarily to investigate the dumps aroundthe shaft. All produced similar results - topsoil coveringspoil from Greenwell's excavations; then spoil from theoriginal Neolithic mining derived from Greenwell's Pit andsurrounding shafts; and finally an old land surface buriedbeneath the spoil. In one cutting, this surface included adarkened area containing flecks of charcoal and burnt bone, asubstantial charcoal patch, and 'an undoubted hearth...[consisting] of adepth of nearly 100mm of blackened, hardened sand containing abundant charcoalflecks and burnt flint'. Over 3000 struck flakes were recovered fromthis old land surface.