Routledge Library Editions: Archaeology [1 ed.] 9781138799714, 9781315751948, 9781138814899, 9781315747149, 1138799718

Table of contents :
Cover
Half Title
Title Page
Copyright Page
Original Title Page
Original Copyright Page
Table of Contents
Illustrations
Preface
I Fishermen and Farmers
II The Megalith Builders
III Traders and Metalworkers
IV The Coming of the Celts
V From Prehistory to History
Index

Citation preview

ROUTLEDGE LIBRARY EDITIONS: ARCHAEOLOGY

Volume 45

THE PREHISTORIC PEOPLES OF SCOTLAND

This page intentionally left blank:

THE PREHISTORIC PEOPLES OF SCOTLAND

Edited by STUART PIGGOTT

Routledge Taylor & Francis Group LONDON AND NEW YORK

First published in 1962 This edition first published in 2015 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN and by Routledge 711 Third Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 1962 Routledge & Kegan Paul Limited All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN: 978-1-138-79971-4 (Set) eISBN: 978-1-315-75194-8 (Set) ISBN: 978-1-138-81489-9 (Volume 45) eISBN: 978-1-315-74714-9 (Volume 45) Publisher’s Note The publisher has gone to great lengths to ensure the quality of this book but points out that some imperfections from the original may be apparent. Disclaimer The publisher has made every effort to trace copyright holders and would welcome correspondence from those they have been unable to trace.

THE PREHISTORIC PEOPLES OF SCOTLAND

edited by

STUART PIGGOTT

Routledgeand Kegan Paul LONDON

First published1962 by Routledge& Kegan Paul Limited BroadwayHouse, 68-74 Carter Lane London,E.C.4 Printed in Great Britain by W. & J. Mackay & Co. Ltd, Chatham © Routledge& Kegan Paul Limited 1962 No part of this book may be reproduced in anyform without permissionfrom the publisher,exceptfor the quotation of brief passagesin criticism

Contents page ix

Preface I

Fishermenand Farmers

II

III

IV V

J.

C. ATKINSON

1

The Megalith Builders

G. E. DANIEL

39

Tradersand Metalworkers

S. PIGGOTT

73

The Coming of the Celts

T. G. E. POWELL

105

From Prehistoryto History

C. A. R. RADFORD

125

R.

Index

155

v

This page intentionally left blank:

Illustrations

PLATES

betweenpages38 and 39 1. Obaniantools of antler, bone and stone.

2. PerforatedMaceheadsfrom Scotland. 3. Carvedstoneballs from Scotland. (a) StennessStoneCircle, Orkney. (b) Interior of MaesHowe, Orkney. 5. Bell-Beaker from Fingask,Perthshire,and Cord-ZonedBeaker from Bathgate,West Lothian 6. Beakersfrom Edzel (left) and Ermelo, Netherlands(right). 7. Late BronzeAge hoard, Braesof Gight. 8. Late BornzeAge bucket, Cardross.

4.

FIGURES

1. Neolithic pottery types and Scottishlong barrows

2. Lyles Hill Ware 3. 4. 5. 6.

7. 8. 9.

10. 11.

12. 13.

Leaf-shapedarrowheadsin Scotland SecondaryNeolithic pottery and flint types Carvedstoneballs in Scotland Perforatedmaceheadsin Scotland Chamberedcairnsin Scotland Stonecircles (unembanked)in Scotland Hengemonumentsin Scotland Long-necked,Bell and Cord-ZonedBeakersin Scotland Short-neckedBeakersin Scotland Food Vesselsin Scotland Cinerary Urns in Scotland Vll

page 9 15 20 24 29 31 51 70 71

78 83 87 90

ILLUSTRATIONS

14. Anglian-type Crossesin Scotland 15. Norse burials and housesin Orkney 16. Norse settlementin Scotland

page 129 131 133

ACKNOWLEDGEMENTS

Plates 1, 2, 3, 7 and 8 are from photographsby Mr. Malcolm Murray, Departmentof Archaeology,University of Edinburgh;nos. 5 and 6 (left) by the National Museum of Antiquities, Edinburgh; no. 4 (a) by the CentralOffice ofInformation,andno. 4 (b) by the Ministry of Works.

Vlll

Preface

T

chaptersconstitutingthis volume originatedas a seriesof lectures given to the ScottishSummerSchoolin Archaeologyat its Edinburgh meetingin 1955. On the deathof the School'sorganizer,Dr. F. T. Wainwright, I havetakenover the editing of this volume,andwhile of necessity the long interval betweenthe original constructionof the contributions, in print, hasled to changedviewpointsand an accesand their appearance sion of new material, so far as possiblethe necessaryadjustmentshave beenmade. HE

STUART PIGGOTT

The University of Edinburgh 1962

lX

This page intentionally left blank:

CHAPTER ONE

Fishermen and Farmers

I

is a soberingthought,andoneconduciveto a propersenseof historical perspective,thatbeforethe first savagessetfoot uponthesoil of Scotland the first urban,andin that sensecivilized, communityhadbeenflourishing at Jerichofor morethana thousandyears. The evidence for the earliest colonization of Scotland by man has beenexhaustivelyexaminedby Lacaille,l whose conclusionsaccordingly need only a brief discussionhere. In common with other workers, he distinguishesfour main gTOUpS of pre-neolithiccommunities,eachidentified by its own characteristicstoneand Hint industry, andeachreferable to some part of the later stagesof the Post-Glacialperiod. In terms of climatic and vegetationalhistory, thesemesolithiccommunitiesof hunterfishers belongto the final stagesof the Boreal and to the ensuingAtlantic phase;or in termsof the absolutechronologynow being established,even if still only in skeletal form, by radio-carbon determinations,to the periodfrom about5500B.C. to 3000B.C. The principal evidenceupon which these signs of early occupation have been dated is their relationshipto the 'raised beach'depositslaid down around the coastsof Scotlandat a time when the mean sea-level stood some twenty-five to thirty feet higher than it does today, and the presentlittoral was consequentlysubmerged.This submergenceappears to have begun about the time of the Boreal-Atlantic transition, around 5000 B.C., andto havereachedits maximumduring the fourth millennium. T

1

A. D. Lacaille, The StoneAge in Scotland, 1954. 1

R.

J.

C. ATKINSON

Thereafterthe sea-levelgraduallydropped,and it is likely that the shorelines were still recedingseawardswhen the first neolithic colonistsarrived during the third millennium B.C.1 The earliest mesolithic community hitherto identified is that representedby the Early Larnian flint assemblagerecoveredfrom the PostGlacial raisedbeachon both sidesof the North Channel.The incorporation of theseworked Bints in the beachdepositsshowsthat they must be the remainsoflittoral settlementswhich flourishedat an early stageof the coastalsubmergenceand were subsequentlyoverwhelmedby the rising sea.The original depositionof this material can thus be datednot much earlierthanthe endof the sixth millennium B.C. The Larnian flint industry is characterizedby the predominanceof flakes and blades, and by the marked absenceof microlithic elements. Genericallyat leastit resemblesthe final stagesof the Creswellianindustry of the British UpperPalaeolithic;andit hasin consequence beensupposed to representthe results of a north-westerly migration of palaeolithic hunters,in Late Glacial or Early Post-Glacialtimes, from their former habitatsin northernEngland,and their subsequentadoptionof a strandlooping existenceon the shoresof the North Channeland the Clyde and Solway Firths. By contrast,the microlithic industriesdiffer very markedly from the Larnian both in distribution and in content. They occur notably in the valleys of the Clyde and Tweed (betweenwhich the Biggar Gap provided an easy route acrossthe intervening watershed)and on Deesidein the neighbourhoodof Banchory.2All the known material consistsof surface finds, and no deposit has so far been discoveredwhich can be securely dated by pollen analysis or other means.The resemblancebetweenthe Scottish material, however, and the microlithic industries of northern England3 suggeststhat both are of much the same date. The Scottish sites,therefore,shouldrepresentsettlementof the main river basinsof the east and south of the country in Atlantic times (that is, during the fifth and fourth millennia B.C.), presumablyby wandering bands of hunters moving gradually northwards up the easterncoastal plain of Durham and Northumberlandto the Tweed, and spreadingthencewestwardsto the Clyde by way of the Biggar Gap. There remain two groupsof mesolithic sites to be considered,both of 1 3

LacailIe, op. cit., 71-75. 2 Lacaille, op. cit., 161-7, 178-93. Proc. Prehist. Soc., XXI (1955),3-20. 2

FISHERMEN AND FARMERS

which representspecializedforms of gathering and hunting economy closely tied, like that of the Larnians, to the sea-shore.In the valleys of the Forth and Tay, at a level correspondingto the high-watermark of the ancient Post-Glacial sea, traces of occupationhave been found in the form of large heapsof the shells of edible molluscs;1while in the carse clays of the former estuary of the Forth, depositedduring the marine transgressionof Atlantic times, numerousfinds have been made of the skeletonsof strandedwhales.1 In four instancesblubber-mattocksof reddeerantler, perforatedto take a woodenhandle,were found in association with the skeleton;and there is every reasonto supposethat thesetools belongedto the samefamilies of strand-looperswho accumulatedthe shellmoundson the shore. The monotony of the molluscandiet which these moundsrepresentdoubtlessensuredthat the fortuitous arrival of a supply of severaltons of red meatwas eagerlyawaited,and exploitedwith relish anddispatch. The Obanian cave-settlementsand shell-mounds on the coast of Argyll provide evidencefor a comparableadaptationto sea-shorelife, in a somewhatharsherenvironment.The remainsof the tools of thesestrandloopersare restrictedin type and poor in quality. The most characteristic stone forms are fingerlike pebbles,generallyworn or ground at one end to a blunt chisel-edge,to which the name 'limpet-scoop'has long been attached,presumablyon the suppositionthat they wereusedfor extracting uncookedlimpets from their shells (for the extractionof cookedlimpets no tool is required beyondthe finger, or at the most anotherlimpet-shell). Experimentreadily shows,however,that while the removalof live limpets from their shells with these tools is extremely difficult, they are ideally fashionedfor useas punchesfor detachingthe limpet from the rock, using anotherpebble,or a pieceof driftwood, as a hammer. The associatedHint industry is exceedingly poor and scanty, and consistsof little more than utilized Hakes.It is in boneand antler that the craftsmanshipof the Obanians shows itself best; though even these materialsare crudely worked by comparisonwith the sophisticatedproducts of other mesolithic communities. The principal types are broad bone splinters,groundto a smoothblunt edgeat one end, and somewhat resembling the limpet-punches,though doubtlessused for some other purpose;and fish-spearswith a doublerow of barbs,made fromred-deer antleraswell asfrom bone. 1

Lacaille, op. cit., 167-9, 175-8. 3

R.

J. C. ATKINSON

It is thesespearswhich have been regardedas the type-fossil of the Obanian culture, and have in the past been comparedsignificantly to similar weaponsin the epipalaeolithicAzilian culture of south-western France.1 But it has also beenremarkedthat the Obaniantool-kit contains perforatedantler mattocks precisely similar both to those found with whale-skeletonsin the Forth Valley and, further afield, to examplesfrom the Ertebelle'Kitchen-midden'culture of Denmark.Sincethe presenceof these Baltic elementsin the Obanianhas been recognized(initially by Mr. P. R. Ritchie, while working on the unpublishedObanianmaterial in the Hunterian Museum), there has been a tendencyto discount the earlier suggestionof an Atlantic origin for the culture in the French Azilian, and to regardit ratheras a peripheraloutpostof the later Forest Culturesof northernEurope.This tendencyhasalso beenencouragedby the obvious discrepancyin date betweenthe Azilian and the Obanian.In the sequenceof cultures the former follows immediately upon the final stagesof the Upper Palaeolithic,and is hardly likely much to outlastthe Late Glacial period, which endedaround 8000 B.C. The Obaniansites, on the other hand,are later than the maximumtransgressionof the PostGlacial sea,and cannot have beenoccupiedmuch before 4000 B.C. The intervening gap of some four thousandyears is obviously difficult to bridge, particularlysinceAzilian materialhasnot so far beenrecordedon ilie Atlantic coastselsewhere,eiilier in norili-westernFranceor in Wales or westernEngland. Recently, however, Clark has suggestedthat ilie objections to an Azilian origin arenot as cogentasthey might seem.2 The gapin geographical distributionis not necessarilysignificant,in view of the known mobility of communities basedon hunting, fishing, and gathering; and in this particularcasethe absenceof intermediatelinks in the chainmay be more apparentthanreal, sincethe mesolithiccoastsof Franceandsouth-western Britain are no longer availablefor study, having beensubmergedby the sea at some time since the middle of the secondmillennium B.C. The existenceof megalithicand other stonestructuresnow partially or wholly beneaththewavesin Scilly andBrittany is well known. The chronologicaldiscrepancyis also diminishedif, as Clark suggests, the Obanianis regardedas a late and peripheralvariantof the Larnian,3 Antiquiry, XXI (1947),84-104. Proc. Soc. Ant. Scot., LXXXIX (1955-56),91-106. 3 Journ. Roy. Soc. Ant. Ireland, LXXIX (1949), 170-81. 1

2

4

FISHERMEN AND FARMERS

with perhapsthe addition of Baltic Forest Culture elementswhich had been carried meanwhileacrossthe Midland Valley by the expansionof similar strand-loopingcommunitiesalreadyestablishedin the Forth and Tay estuaries.It is not impossible,however, that such Baltic influences might be discerniblein the Larnian itself, had we any sites in which the bone and antler work of this culture was adequatelyrepresented. In any case,the palaeobotanicalevidencefrom the Early Larnian site at Toomeon the shoresof Lough Neaghin NorthernIreland1 showsthat the culture was alreadyestablishedin the areabeforethe seventhmillennium B.C., a date not very far removed from the floruit of the Azilian cultureproper. Thc adoption of this view of the origin of the Larnian and Obanian culturesnecessarilyrequiresalso that we abandonthe ideaof deriving the former from the final Upper Palaeolithic communities of northern England.As Clark haspointedout, the Late Creswellianhas beenrather overworkedas a suggestedsourcefor the non-microlithic aspectsof the mesolithic cultures of Britain and, indeed, of the Continent;2and these doubts have now been significantly reinforced by a new and as yet unpublisheddiscovery.During 1959 blasting operationsnear Kilmelfort in Argyll exposedthe inner end of an inhabitedcavecontainingthe remains of an occupation-layerwith an associatedffint industry, a selectionfrom which was exhibited by the excavator,Dr. John Coles, to the Spring Conferenceof the PrehistoricSocietyin 1960. It is clear that this industry resemblesthe later stagesof the Creswellianvery much more closely than does any aspectof the Larnian. This discovery thus goes some way to confirm the suggestionof a north-westerlymigrationof Creswellianpeople to the west coastsof Scotland.But it also makesit the less probablethat the origin of the Larniancultureis to be soughtin sucha migration, andto this extent the alternativeof an Azilian origin is renderedall the more acceptable. Questionsof origin apart,however,it is clear that the territoriesof the earliestmesolithic colonists were not confined wholly to the west coasts, for from Kingsteps Quarry, near Nairn, there is evidence for human settlementas early,probably,asthe fifth millennium B.C. Herea very crude seriesof worked tools in quartz, chert, and sandstonehasbeenrecovered from a thin layer of peat resting on, and coveredby, blown sand,which 1

2

Lacaille, op. cit., 124-. Proc. Soc. Ant. Scot., LXXXIX (1955-56), 102. 5

R.

J.

C. ATKINSON

has been shown to have formed during the Boreal-Atlantic transition (circa 5000 B.C.) and afterwards.1 The preciselevel or levels in the peat from which the artefacts came is not recorded; but the crudity of the workmanshipand the absenceof any specific microlithic forms arguesfor an early date, before the penetrationof north-easternScotlandby true microlithic cultures. The presenceof oak charcoalin associationwith a microlithic industry at Birkwood on Deeside2 suggeststhat this penetration did not take place before the Atlantic period, as might indeedbe expectedfrom evidenceelsewhere. The numberof individual sitesof the mesolithicperiod now known in Scotlandnumbersat leastone hundred;but it shouldnot be assumedthat theserepresentmore than an exceedinglysmall population.Many of these sitesare markedonly by a scatterof ffints, which may be no more than the product of a few hours' activity by a single flint-knapper, or at best the debris of a temporarycamping-site,occupiedfor a few days by a small bandof roving hunters,and thenabandoned.Suchis the speedat which a skilled ffint-knapper works, and suchthe quantityof wasteflint produced in a very short time, that the wanderingsof a singlefamily could account, in the spaceof no more than a few years, for all the finds of microlithic ffints in the whole of southernScotland. Nor shouldit be supposedthat the very bulky accumulationsof foodwaste on coastal sites must representcommunitieslarger than a single family, or settlementof long duration.It has beensuggested,for instance, that a shell-moundmeasuring100 by 60 feet, andfrom 1 to 3 feet in thickness,'points to a fairly long occupation'.3 But it must not be forgotten that of all forms of food, shell-fishproducethe greatestbulk of refusefor a given quantity of edible substance.It is thus by no meansdifficult, where limpets, whelks, mussels,oysters, winkles or cockles form a substantial part of the diet, for the wasteshellsto accumulateat leastat the rate of a large bucketful, or half a cubic foot, per head per day. A group of ten people could thus amassa midden of the size quoted in less than seven years of continuous settlement,or within a single lifetime of seasonal campingon the sea-shore. All the evidence suggests,therefore, that the mesolithic population of Scotlandwas exceedinglysparse.By way of comparison,it may be noted Trans. Botanical Soc. Edinburgh, XXXVI (1954),224-9. Proc. Soc. Ant. Scot., LXX (1935-36),421. 3 Lacaille, op. cit., 176.

1

2

6

FISHERMEN AND FARMERS

that for Denmark,where a single large bog can containup to a hundred separatesettlementsites, the total mesolithic population has been estimatedat no more than 125 persons.The areaof Scotlandis about twice that of Denmark; but when the mountainousregions, unsuitable for colonizationeven by hunter-fishers,are subtractedthe habitable areasof the two countriesareaboutthe same.In view of the far smallernumberof sitesof the period known in Scotland,oneis forced to the conclusionthat until the neolithic colonizationthe populationat anyonetime can hardly haveexceededtwo peoplefor eachof the moderncounties. The groupsof neolithic colonistswho arrived during the third millennium B.C. mustthereforebe regardedas the earliestsettlersto havereached Scotlandin any substantialnumbers.They would be confrontedwith a terrain hitherto virtually uninhabited,and certainly unchangedby man from its natural condition; and it is thus improbablethat the patternof neolithic settlementwas in any way influenced or controlled by the minutepopulationof nativesavages. The cultural identity of the first neolithic settlersin Scotlandis not at present entirely certain. It has long been recognizedthat among the 'Primary Neolithic' communities of Britain two principal and separate elementscan be distinguished:on the one hand the 'Windmill Hill' or 'Western Neolithic' culture, characterizedby various forms of pottery and by the collective inhumation of the dead beneath unchambered earthenlong barrows,and concentratedprincipally in the Lowland Zone of England;and on the other handthe buildersof gallery graves,covered by elongatedstonecairns, centredrespectivelyupon the Bristol Channel and the North ChannelbetweenScotlandandIreland. In the pastit hasbeenassumedalmostwithout questionthat it was the latter group, the buildersof long cairnsor gallery graves,who were alone responsiblefor the introductionof the neolithic economyto the northern kingdom.l But in the courseof time a growing body of evidencehas accumulated for primary neolithic settlementin Scotland which is neither physically associatedwith chamberedtombs, nor conforms in its geographical pattern to their predominantlywesternand northerndistribution. The whole question of the settlementof Scotlandby the earliest farmersthusrequiresexamination. The problem of origins has, moreover, been further complicatedin recentyearsby the recognitionthat theso-calledWindmill Hill culture of 1

V. G. Childe, The Prehistory of Scotland(1935), 22-79.

P.P.S.-B

7

R.

J.

C. ATKINSON

England comprisesdiverse elementswhich in all probability should be derivedfrom widely separatedareasof continentalEurope;and that so far from being merely the British representativeof the 'Western'family of Atlantic neolithic cultures,it includes also 'Eastern'elementsrelated to the Funnel-Beakerand Michelsbergculturesof the North Europeanplain andthe Low Countries.1 Somedegreeof duality and diversity within the Windmill Hill culture hasindeedlong beenimplicit in the restricteddistributionof unchambered earthenlong barrows, which are largely confined to the chalk areasof easternand southernEnglandand do not occur at all in westerndistricts, even where pottery types of specifically westernand Atlantic origin are known. This geographicalpattern can hardly be explained merely in geologicalterms, since thereis no reasonwhy unchamberedlong barrows should not have beenbuilt in any district. The differencesbetweenthe chamberedtombs of the west and the unchamberedlong barrows of the southand eastmust be accountedfor not by the merepresenceor absence of suitablebuilding stone,but as the respectiveexpressionsof two distinct anddivergenttraditionsof collectiveinhumation:in the first, of successive burial in a tomb chamberto which repeatedaccesscould be had; and in the secondof the simultaneousburial of a numberof bodies(or skeletons) collectedand storedelsewherein the period precedingthe building of the barrow. In otherwords, we are dealingwith a cultural distinction between family vaultsandmassgraves. The full implications of the internal complexity of the Windmill Hill culture, loosely so called, have still to be worked out in detail. In fig. 1, however, an attempt has been made to differentiate at least part of the basic pattern of 'eastern'and 'western' elements,by meansof selected andapparentlydiagnostictypesof pottery. The forms chosento represent the westernelementare plain lugged bowls of Piggott'sshapesA, Band C,2 whose origins in the Early Chasseyculture of France are perhaps hardly in doubt. Only lugged examplesof theseforms have beenplotted, since plain bag-shapedpots without lugs are so simple and basic a form that they are unlikely in themselvesto havemuch cultural significance. To representthe easternelementI havechosenthe gracefulshouldered bowls of Piggott's shape G, in which the neck above the shoulder is distinctly concave,and thus to be distinguishedfrom superficially similar 1 Proc. Prehist. Soc., XXI (1955), g6-IOI. 2 Arch. Joum., LXXXVIII (1931),75.

8

BOWLS OF FORMS A.B&C

•

BOWLS OF FORM G ® SCOTTISH LONG BARROWS -

25

Fig.

I.

0

50

Neolithic pottery typesand Scottishlong barrows.

100 MILES

R.

J.

C. ATKINSON

carinated or shoulderedforms in which the neck is either conical or cylindrical (Piggott'sshapesD and E). Thoughin the pasttherehas been some confusion betweenPiggott's forms D and G, an attempt is made below (p. 14) to show that they are culturally as well as morphologically distinct. The bowls of form G, which aloneare plottedin fig. 1, were long ago comparedwith the Michelsberg'tulip-beakers'of Belgium,l and there seems no reason now against returning to this view of their origin. Admittedly it was at one time abandonedon the groundthat the Michelsberg culture of the Rhinelandwas chronologicallytoo late to have stood in any prototypical relationshipto our own Windmill Hill culture;2 and indeed the latest study of the Michelsberg groups on the Continent suggeststhat in Belgium at least the analoguesof our form G bowls appearafter rather than before 2000 B.C.3 Once it is accepted,however, that the Windmill Hill culture of Britain is a cultural amalgamof diverse elements,it is no longer necessary,or evenwise, to assumethat its various componentsreachedthese islands simultaneouslyfrom different directions. There is thus no inherent difficulty in positing a North European origin for our form G bowls, providedthat it is recognizedat the sametime that their arrival was subsequentto that of the 'western'componentsof the culture. The purposeof fig. 1, however,is not to illustrate a discussion of the origins of the Windmill Hill culture, but to show that within the culture itself a differentiation into regional elementsis possible,and that both regionalelementshere distinguishedextendinto Scotland. The pattern of these two contrastedtypes shows that the 'western' materialhastwo main areasof concentration,in southernand south-western England,and aroundthe North Channel,respectively,the two areas beinglinked by a few finds in SouthWalesandAnglesey.Elsewherein the north the distribution is eastern,in Yorkshire, West Lothian, and Aberdeenshire. It is perhaps significant that in the western area outside Englandmore than half the sitesfrom which theselugged typesof pottery have been recoveredare chamberedtombs, belonging or related to the Severn-Cotswoldand Clyde-Carlingford series (Ty Isaf, Tinkinswood, Bryn yr Hen Bobl, and Pant y Saerin Wales; Audleystownin Northern Ireland; and Beacharra,Clachaig,Torlin, and Sliddery Water in western Proc. Prehist. Soc. East Anglia, VII (1934), 379. S. Piggott, Neolithic Cultures qfthe British Isles (1954), 74. S Proc. Prehist. Soc., (1959), 51-134. The new 'long' chronology for the neolithic imposedby radio-carbondatesmakesit difficult now to acceptso Iowa terminuspostquem. 1

2

xxv

10

FISHERMEN AND FARMERS

Scotland).The remainingoccurrencesin the North Channelareaare all on sand-duneor othercoastalsites. It hasalreadybeensuggestedthat somegenericrelationshipmust exist betweenthe Severn-Cotswoldand the Clyde-Carlingfordseriesof gallery graves;!and this suggestionis confirmedin a specific senseby Dr. Daniel in the succeedingchapter (p. 66), where the problems posed by the megalithic tombs of Scotlandare discussedin detail. It is enoughhere to point out that the distribution of these'western'forms of pottery reinforces the conclusionsto be drawn from the morphology of the tombs themselves,namelythat the initial colonizationof south-westernScotland is likely to have had its proximal origin in the Bristol Channelarea. It must be admitted,of course,that theseforms of lugged 'western'pottery are not particularly well representedin the Severn-Cotswoldtombsthemselves,and indeed so far have not been recordedat all from the North Wiltshire group of tombs, whose semicircular forecourts most closely resemblethoseof the Clyde-Carlingfordseries.It is thus possible,though by no meansnecessary,to suggestthat the potteryand the tombs respectively mark the courseof two parallel but distinct movementsof colonization of the samenorthernareafrom the samesouthernsource. What doesseemalmostcertain,however,is that this movementup the Irish Sea,whethersimple in characteror complex,resultedin the earliest settlementof anypartof Scotlandby groupspractisingthe neolithic farming economy.The dateof this initial colonizationis still obscure;but in view of the generallengtheningof neolithic chronologiesin Europeas a whole requiredby recentradio-carbondates(noneof which are yet availablefor Scottishsites), it is likely to have beenbeforeratherthan after QOOO B.C. This view of the identity of the earliestcolonistsis thus no more than a repetitionof the conclusionreachedby Childe a quarterof a centuryago. It is rather for the later stagesof neolithic settlement,and particularly of the easterndistricts of the country hardly penetratedby the builders of chamberedtombs, that new interpretationsof the evidence are now possible.Here, it appears,two separatemovementsof neolithic immigration can be detected:the first, at scatteredpoints up the eastcoastas far as Aberdeenshire,marking a northwards extensionof the 'eastern'elementsin the Windmill Hill culture of the south; and the second,terminating in the same region but originating ultimately in the Lyles Hill cultureof NorthernIreland. 1

Piggott, Neolithic Cultures (1954), 182. 11

R.

J. C. ATKINSON

The evidencefor the first of thesemovements,so far as potteryis concerned,is scanty.Fragmentsof bowls of form G havebeenrecoveredfrom the three sites shown in fig. 1 (Cairnpapple,West Lothian; Bantaskine, near Stirling; and Powsode Cairn on Atherb Farm in Buchan). The secondof theseis a particularlyfine andwell-madeshoulderedbowl which closelyresemblesvesselsfrom Yorkshire. Were this the only evidence,one could hardly speakwith any confidenceof a separatemovementof colonistsup the eastcoastsfrom Yorkshire, particularly as pottery of this form also occurs in Ulster (fig. 1), where it apparentlyrepresentsa secondarycolonization from Yorkshire by way of the Solway Firth. It would thus be possibleto explain these isolatedoccurrencesof bowls of form G in Scotlandas one result of the immigrationfrom Ulster discussedbelow. That this view of their origin is the lessprobable,however,is suggested by the presencein easternScotland of four long barrows (fig. 1), the characteristicburial-moundof the Windmill Hill culture in southernand easternEngland.The mostsoutherlyof theScottishexamples,at Caverton Hillhead, south of Kelso, has long since been entirely destroyed,but is reliably recordedto have consisted'of fine loosemould, intermixedwith large stones,coveredover with heath'.It measured340 feet in length, the width taperingfrom about40 feet at the easternendto 30 at the western.1 This canhardlyhavebeenanythingbut a long barrow. The secondsite, on a hill-top overlooking the sea above Gourdon, Kincardineshire,has again been shown, by probing with a steel bar, to be composedmainly of earth, with only occasionalstones.It is orientatedroughly N.E.-B.W., with a length of 155 feet and a height of 9 feet, the width taperingfrom 40 to 25 feet, the broaderend lying to the N .E.2 Thereis no apparentditch. The third site has not previously been published.It lies on a small hillock, now planted with oak trees, about half a mile S.S.E. of Pitlurg Stationin the parishof Slains, Aberdeenshire,and about 3 miles N.E. of the head of the estuary of the River Ythan. The mound is of earth, containingno stonesdetectablewith a probe, and is orientated7° W. of true north. It is about 70 feet in length and 5 feet in height, and tapers fairly sharplyto the southfrom a maximumwidth of 25 feet. The mound appearsto standon a sloping bermor platform about15 feet wide on each 1

2

Roy. Comm. Anc. Mon. Scotland,RoxburghshireInventory (1956),133,no. 218. Proc. Soc. Ant. Scot., LVIII (1923-24),23-24.

12

FISHERMEN AND FARMERS

side, at the baseof which in placesthereare suggestionsof a silted ditch; but this may be an optical illusion due to the presenceof a modernraised bank surroundingthe plantation. There are no signs of previousexcavation. The last of these Scottish long barrows stands, like the Gourdon example,in a commandingposition on a hill-top overlookingthe sea.It is locally known as the LongmanCairn, and lies nearthe coastof Banffshire about 2! miles E.S.E. of Macduff. Its overall length is 220 feet, with an orientation300 E. of N. The northern end is about 14 feet high and is divided from the remainderby a transversehollow, possiblythe site of an unrecordedexcavation.The width of the main mound tapersfrom 40 to 25 feet, with a height of about 9 feet throughoutthe greaterpart of its length. Modern quarryinginto the side of the moundshowsit to be built mainly of earth,andprobingdetectedrelatively few stones.1 While admittedlynoneof thesesiteswould be entirely at homeamong the long barrowsof southernEngland,it is nonethe lessdifficult to assume that they are merely earthenversionsof long megalithiccairns,since true long cairns, built wholly of stones,do occur in the sameregions. In the south-eastthereare the Mutiny Stonesin Berwickshire2 and the Kirshope Cairn in Roxburghshire;3while Aberdeenshirecontains at least three similar sites, at Cloghill, westof Aberdeen,4 at BalnagowannearAboyne,5 and at Knapperty Hillock, near Auchmacharin Buchan.6 It seemsfar more satisfactoryto acceptthesefour sites for what they appear to be, namelyearthenlong barrowsof Windmill Hill type, and to relatethem to the sporadicoccurrencesof pottery of form G which is appropriateto the sameculture. It is thus possibleto envisagetwo divergentmovementsin the northwards spreadof the Windmill Hill culture, both originating in Yorkshire. The first took a north-westerlyroute to the shoresof the Solway Firth, and thenceto Ulster; while the secondcontinuednorthwardsup the east coast.Findsof plain undifferentiatedWindmill Hill potteryat Old Bewick, Northumberland,on the Gullane Sandsin East Lothian, and at Roslin, near Edinburgh,help to fill out the meagreevidencefor this northward Proc. Soc. Ant. Scot., LIX (1924-25),21-28. Proc. Soc. Ant. Scot., LIX (1924-25), 198-204. 3 Roy. Comm. Anc. Mon. Scotland,RoxburghshireInventory (1956),94,no. 4 Proc. Soc. Ant. Scot., LIX (1924-25),21-22. 6 Proc. Soc. Ant. Scot., LIX (1924-25),26. 6 Proc. Soc. Ant. Scot., XXXVIII (19°3-04),273. 1

2

13

IIO.

R.

J.

C. ATKINSON

penetration; but even so the pattern is sufficiently discontinuous to suggestmovementprincipally by sea. It is noteworthy that all the sites where the relevant pottery or long barrows occur are within easy reach eitherof the seaitselfor of major river valleys. The probability of this northwards extensionof the Windmill Hill culture ha,s beenrecognizedin the past, and indeedits farthest limit has beencarried beyondthe Buchanregion to the coastof the Moray Firth, to include the well-known finds of pottery from Eastertonof Roseisle 1 It seemsmore likely, however, that this and the other near Burghead. finds of relatedforms of pottery from Scotlandshownin fig. 2 originated not in the Windmill Hill culture of the south and east,but in the Lyles Hill cultureof NorthernIreland. The diagnosticforms and decorationof Lyles Hill pottery used here are well illustrated at the type-site.2 They consistof shoulderedbowls of Piggott's forms D and E, in which the neck above the shoulder (itself often of ledge or steppedform) is respectivelyconical or approximately cylindrical. Both theseforms should be distinguishedfrom the bowls of form G already discussed,in which the neck has the form of a trumpetmouth, with a concaveflare. The rims of Lyles Hill ware are frequently bent outwards sharply, and sometimesexpandedinternally as well, in contrastto the plain or beadedrims of the form G bowls. The commonest form of ornamentis a highly characteristic'finger-tip fluting' of the surface, usually on the exterior or flattened rim only, but occasionallyinside the neckaswell. The areashownin fig. 2 includesthe majority of siteson which pottery of this type is known.3 Elsewherethere are only scatteredoccurrencesin centraland southernIreland, the Isle of Man, west Wales and south-west England.In the northernpart of the Windmill Hill provincein England these characteristic shapesof pottery are unknown; and thoughfinger-tip fluting has been recordedoccasionallyon Heslertonware in Yorkshire, it is there confined entirely to the rims of pots, and may perhapsbe accountedfor by influence from Northern Ireland, a reflex of the movement which carriedthe form G bowls from Yorkshire to Ulster. In Northern Ireland itself, Lyles Hill ware occurs fairly frequently in Piggott, Neolithic Cultures (1954), 271-2. E. Estyn Evans,Lyles Hill (1953), 32-43. 3 The map was preparedsome time before the appearanceof H. J. Case's study of Irish Neolithic pottery (Proc. Prehist. Soc.,XXVII (1961), 174-233).The term 'Lyles Hill' is usedherein a lessspecificsense. 1

2

14

LYLES HILL WARE LYLES HILL WARE WITH FLUTED ORNAMENT

10

Fig.

2.

Lyles Hill Ware.

0

Ware.

R. ]. C. ATKINSON

chamberedtombs of the Clyde-Carlingford series at all stagesin their typological development,and canhardly be dissociatedfrom the spreadof the tombs themselves;though it is surely the casethat in their ultimate origin the potteryand the burial rite are distinct.1 In view of their association in Ulster, however,it is hardly surprisingthat of the sixteensitesin Scotlandat which pottery of this type has beenfound, five are certainly chamberedtombs and three more are possibly or probably of the same class. It seemslikely that the spreadof Lyles Hill ware to Scotlandtook placeonly at a late stagein the history of the Clyde-Carlingfordculture, since the tombs concernedall exhibit featuresof typological development or degenerationfrom what are generallyregardedas the earliestforms. Cairnholy I, for instance,belongsto the secondarygroupof chambered tombsin Galloway; and the pottery was found, moreover,in the blocking of the forecourt, whereit was contemporarywith Peterboroughware and marks the final stagein the useof the tomb. Glecknabaein Bute is again a cairn of 'degenerate'sub-circularform, in which severalsingle chambers were set radially; while the three sites of Nether Largie, Kilchoan, and Achnacreein Argyll all lack the diagnosticfeaturesof the 'classic'ClydeCarlingfordtombs. At three other sites, all in easternScotland (Cultoquhey,near Criefl', Perthshire; and Knapperty Hillock and PowsodeCairn in Aberdeenshire) the natureof the associatedstructureis uncertain.The former may be a chamberedcairn of Clyde-Carlingfordtype, but this is by no means certain. Knapperty Hillock appearsto have been a massive flat-topped long cairn with raisedends,but thereis no evidencefor internalchambers. The natureand even the preciseposition of the PowsodeCairn are unrecorded. Of the remaining eight sites (listed, together with the foregoing, in the Appendix, p. 34) six appearto be settlements.Three,at Rothesayin Bute and at Whitemossand Knappers,on either side of the mouth of the Clyde, lie in an areain which influencefrom Ulster is to be expected.The other three, at Eastertonof Roseisleand Urquhart in Elginshire and on Speysidein the adjoining areaof Inverness-shire,can plausibly be interpretedas the result of movementup the Great Glen by a route on which the sites in Argyll mark the proximal, and those in Aberdeenshirethe terminal,stagesof progress. The two remaining sites in Aberdeenshire,Loanheadof Daviot and 1

Piggott, Neolithic Cultures (1954), 167-70, 182. 16

FISHERMEN AND FARMERS

East Finnercy, require more detailed comment. The RecumbentStone Circle at the former site is well known; but the existenceamong the pottery from the site of a few sherdsof Lyles Hill ware has only recently been recognized.l The interpretation and dating of these enigmatic monumentsof north-easternScotlandis admittedlystill obscure;but the discoveryof Beakersherdsat Loanheadof Daviot itself and at Old Keig,2 andof a fragmentarypolishedstonebracerof Beakertype, associatedwith sherds 'of a reddish colour' at Old Rayne,S make it overwhelmingly probablethat this form of sepulchralcairn was currentduring the Beaker occupationof Aberdeenshire,with which at least someof the local Lyles Hill ware must be contemporary,as is shown by the finds from the East Finnercycairn discussedbelow. In spite of the occurrenceof Beakersherdsin three of the Recumbent Stone Circles (and very few more than this have been even partially excavatedby moderntechniques),it is difficult to regardthem primarily as Beaker monuments,since such structuresare unknown in the continental homelandof the Beakerpeoples.It is surely more satisfactoryto seethe occurrenceof theseBeakersherdsas yet anotherillustration of the ability of the Beakerculture to adaptitself to indigenousinstitutions, an ability alreadywell demonstratedby the frequentfinds of Beakerpottery in megalithic tombs, and by the developmentwithin Britain of a specifically Beakervariantof the SecondaryNeolithic form of hengemonument.4 If the origin of the RecumbentStone Circle is to be sought outside Aherdeenshire,it is perhapsnot wholly fanciful to look to the type-siteof the Lyles Hill Culture in Ulster, where the cairn on the hill-top exhibits many of the specific features of the Aherdeenshiresites, and was undoubtedlybuilt by the makersof Lyles Hill pottery. The site consistsof a low ring-cairn surroundinga central spacecontaininga crematedburial, and borderedby a boulderkerb which is brokenat one point by a miniature 'recumbentstone' set betweena pair of 'fiankers'.5 These are all featureswhich are repeated,thoughin more massiveform, in the RecumbentStoneCirclesproper.The only elementclearlylacking at Lyles Hill is 1 The recognitionof this pottery is due to Miss Audrey Henshall,to whom the writer is mostgratefulfor theinformation. 2 Proc. Soc.Ant. Scot.,LXVII (1932-33),45. 3 Proc. Soc. Ant. Scot., XXXVI (19°1-02),530. 'R. J. C. Atkinson, C. M. Piggott,and N, K. Sandars,Excavationsat Dorchester, Oxon.,1(1951),93. 5 Evans,Lyles Hill (1953), 5.

17

R.

J. C. ATKINSON

the surroundingcircle of free-standinguprights; but this may well be an accretionfrom native sourcesin Scotland,as it appearsto be in the Clava passagegravesand ring cairns.! In the absenceof any more convincing evidenceit may therefore be suggested,though with reserve,that the developmentof the RecumbentStone Circle in Aberdeenshireis to be of the Lyles relatedto the colonizationof the sameareaby representatives Hill cultureof Ireland. Supportingevidencefor this colonization,and for its late date, comes from the round cairn at EastFinnercy,Dunecht,aboutelevenmiles W. of Aberdeen. Neolithic pottery, including lugged bowls, had been found during a casualexcavationof the site in the 1920s;2and when the cairn was visited by ProfessorPiggott and the writer in 1951 a short sectionof dry stone walling was visible on the surface near the centre which, if original, suggestedthe presenceof a passagegrave. Accordingly the cairn was partially excavatedby the writer in the summerof 1952.3 The dry stone walling proved to be a recent feature, built by the previous excavatorsto retain their dump, as was subsequentlyconfirmed by one of them who visited the site. The whole of the centreof the cairn, which stood to a height of some5 feet, had beendisturbed.None of the stonesfound in the central area,however,were of a size or characterto haveservedas the sidesor capstoneof a cist, nor were stone-holesfor the baseof a cist found in the old groundsurfacebeneaththe cairn. Moreover, our informant was certainthat in the earlier openingno traceof a cist, or indeedof a burial of any kind, hadbeenfound. On thesegroundsalone it was improbablethat the site was a normal cairn of the BronzeAge; and this was confirmedby the quantityof sherds of neolithic pottery, almostall of Lyles Hill ware,and by a well-madeleafshapedarrowheadof reddishBuchanffint, which occurredprincipally on the old surfacebeneaththe cairn, and to a lesserextentin the soil filling the intersticesbeneaththe stones.This neolithic materialwas confined to the cairn itself, and was not found evena shortway outsideits edgein the surroundingfield. It is thus clearthat the cairn was built by the makersof the pottery,andthat this was deliberatelydeposited. The pottery included a single sherd of cord-zonedBeaker, sealed beneathan undisturbedpart of the cairn and only a few inches from Proc. Soc. Ant. Scot., LXXXVIII (1954-55), 197. Proc. Soc. Ant. Scot., LXIII (1928-29),62-63. a Reportto be publishedin Proceedingsofthe SocietyofAntiquariesofScotland. 1

2

18

FISHERMEN AND FARMERS

fragmentsof Lyles Hill ware. There can be no doubt that hereboth types of pottery were strictly contemporary,and this associationconfirms the apparentassociationof the sametypes at Loanheadof Daviot. The Lyles Hill occupationof Aberdeenshiremust thus overlap with at least the earlierstagesof the Beakercolonization. In view of the connexionsbetweenUlster and Aberdeenshirealready suggestedabove, it is significant that the best parallels for the East Finnercycairn are to be found in severalneolithic round cairnsexcavated in Northern Ireland, and notably that at Knockiveaghin Co. Down, in which the characteristicfeatureis the quantityof brokenneolithic pottery and other debris of occupationdepositedbeneathor in the cairn.1 The samepracticealsotypifies the cairn on Lyles Hill itself. The sherds with lug-handlesfound earlier at East Finnercy and at Loanheadof Daviot may possibly be connectedwith the movementfrom the Windmill Hill province up the east coast referred to above (p. ll). But it is perhapsbetter to regard them as an integral componentof the Lyles Hill material from the sametwo sites, particularly as the examples from East Finnercy have lugs of the 'eared'or upward-pointingvariety which seemto be particularly characteristicof Ireland. One final piece of evidence for cultural links between Ulster and Aberdeenshirerequiresbrief mention,namelythe distributionin Scotland of stone axes of porcellanitefrom the axe factories of Tievebulliagh in Antrim and Rathlin Island off the Antrim coast.2 Though a detailed picture of the occurrenceof theseaxesin Scotlandis not yet available,it is surely significant that outsidethe westerncoastalareascentredon the Clyde mouth, where their presenceis readily to be understood,the principal concentrationlies preciselyin the north-east,in Aberdeenshireand the adjacentcounties.3 The evidencediscussedabove for the neolithic settlementof eastern Scotlandis too diverse, and numerically too scanty, to give any reliable picture of the relative density of population in different areas.On this point, however, it is instructive to examinethe distribution of the commonestof all forms of neolithic artefact,the leaf-shapedflint arrowhead. The collections of the National Museum of Antiquities contain a very Ulster Journ. Arch., XX (1957), 8--20. Ulster Journ, Arch., XV (1952), 31-60. 3 A map, including unpublishedidentifications,is exhibitedin the Kelvingrove Museum,Glasgow. 1

2

19

R.

J.

C. ATKINSON

largenumberof specimensof this type, and the resultsof a countby areas are given in fig. 3. It should be noted that the horizontal scale of the diagramexpressesrelative density per unit areaof habitableland, on the assumptionthat the presentdistribution of arable cultivation is, by and

Beauty Firth Moray Banff Aberdeen Kincardine Angus ray & Forth Basins Lothians Tweed Basin

o

100

200

300

400

560

LEAF-SHAPED ARROWHEADS, per 100,000 acres of arable land, in the National Museum, Edinburgh Fig. 3. Leaf-shapedarrowheadsin Scotland.

large, a measureof the territory availablefor primary settlementby early agriculturalists.It is to be expected,of course,that arrowheadswill be found over a wider areathan that suitablefor primitive tillage, sinceit is in the marginalareasthat huntingis likely to havebeenmostintensive. It may be objected,of course,that the apparentdistribution of artifacts of this kind reHects the relative density of modern Hint-collectors ratherthanof prehistoricsettlement;andit is certainlytrue in the present 20

FISHERMEN AND FARMERS

casethat the exceptionaldensityof finds in the countyof Moray is due to the activities of collectorson a single site, the Culbin Sands,long known as a happy hunting ground. None the less, it is also true that collectors are most active and most successfulpreciselywhere Hints themselvesare most numerous;and to that extent the diagram may be put forward at least as a generalizedpicture of the relative density of primary neolithic settlement.It is at once apparentthat the main concentrationis in the north-east,and that the frequency falls progressivelysouthwardsto the Firth of Forth. In spiteof the known fertility of the soil (andof the opportunities of discovery afforded by widespread arable cultivation) the Lothians,andparticularlyEastLothian, appearto havebeenonly sparsely settled; and it is only in the Tweed basinthat the density of occupation approachesthat of the regionsnorth of the Tay. The overall picturegiven by this kind of evidencethus reinforces the indications of the pottery andothermaterialdiscussedabove. In summary,therefore,the primary neolithic colonizationof Scotland can be separatedinto three phasesof settlement,each representedby characteristicforms of pottery and of sites in the field. The first had its primary impactin the south-west,and cannotbe separatedfrom the introduction of megalithic gallery graves to the same region, probably well before2000 B.C. The resultsof this impact, and of the arrival in Scotland of the builders of passagegraves, are discussedfully in the following chapter.The secondmovement,originatingprobablyin Yorkshire,carried aspectsof the'eastern'elementin the Windmill Hill cultureof Englandas far north as Buchan; while the third saw the implantationin the northeastof an offshoot of the regionalneolithic culture of Ulster, probablyby way of the GreatGlen andthe coastof the Moray Firth. The initial datesof thesetwo latter movementsare at presentdifficult to determine.In southernEnglandat leastone unchamberedlong barrow, at Nutbane near Andover, appearsto have been completed, at latest, soon after 2500 B.C. and perhaps several centuries earlier.1 Yet the Skendleby long barrow in Lincolnshire must be contemporarywith sherdsof Cord-zonedBeakerpottery,2 which even on the Continentcan hardly be datedmuch earlier than 2200 B.C. 8 The introductionof earthen Antiquiry, XXXIII (1959), 289. Archaeologia,LXXXV (1935),53. 3 Antiquiry, XXXIV (1960), 17. The date there given requiresthe addition of approximatelytwo centuriesto correctit for the Suesseffect. 1

2

21

R.

J.

C. ATKINSON

long barrowsinto easternScotlandmay thus havetakenplaceat any time during the last three-quartersof the third millennium B.C., though on geographicalgroundsa date towards the end of this period is the more probable. The possiblechronologicallimits for the expansionof the Lyles Hill culture are if anything evenwider. The date of the Knockiveaghcairn in Ulster, alreadyreferred to (p. 19), probably lies in the range 3400-2700 B.C.;1 whereasat East Finnercy the associationwith Cord-zonedBeaker pottery again gives an initial date of around2200 B.C. at the earliest.The minimumdistancein time betweenthesetwo very similar sites must thus be of the order of half a millennium. If we are to acceptthe long chronology for the British Neolithic imposedby the radio-carbondates,we must equallyacceptthat the lives of individual cultures, andof individual types of construction,must be far longer than we have hitherto beenwilling to allow. For both of theselater movementsthe terminal areais Aberdeenshire and the adjacentcounties.Whateverthe factors that renderedthis district particularly attractive to the earliest farmers, they must have operated with equalforce upon the SecondaryNeolithic population,now to be discussed,whosedistribution showsa similar bias towardsthe north-east. The conceptof the SecondaryNeolithic, as the result of the cultural and technologicalimpact of Primary Neolithic agriculturalistsand herdsmen upon surviving mesolithic groups of huntersand fishermen,is now well establishedand has beenfruitfully applied outsideas well as within Britain, where it was first evolved.2 In the United Kingdom the basic diagnostic elements of these cultures include, first, specific forms of pottery, of the Peterborough,Rinyo-Clacton, and Sandhill types, each with a markedly regional distribution; secondly, tools and weaponsin flint and stone,which both morphologicallyand geographicallyare more uniform than the potteryand reflect, in manyinstances,the persistenceof mesolithic traditions in fashion and craftsmanship;and thirdly, ritual sites including embankedopen-airsanctuaries,or henge-monuments, and individual or collective burials. The latter occur occasionally both in caves and in round barrows, and rarely as collective cremationsassociatedwith monumentsof'henge'type. In Scotlandthe manifestationsof SecondaryNeolithic culture can be 1 2

Antiquity, XXXIV (1960), 112. Piggott, Neolithic Cultures (1954),276-3°1.

22

FISHERMEN AND FARMERS

divided into two main groups.On the one handthereare thosewhich are likely to haveoriginatedin England,wherethey arefar betterrepresented, and whose presencein the north can be explainedby the migration of cultural groupsalreadyfully formed in the south; theseare plotted in fig. 4. On the other hand, there are types which seemto be largely or wholly of native Scottish origin, of which the two leading forms, perforated maceheads andcarved stoneballs,areseparatelymapped. Of the potterytypes,Peterboroughware (fig. 4) occursonly on the few siteslisted in the Appendix (p. 35), and on noneof theseis it represented by more thana few sherds.Apart from a doubtful exampleat Scotstarvitin Fife, it is restrictedto the region south of the Clyde-Forth line, which accordingly marks the extreme northern frontier of a cultural group specifically adaptedto the more genial climate and landscapeof southern England. Rinyo-Clacton ware has a much wider, though very discontinuous, distributionin Scotland,and one which on the west coast,at any rate, can hardly be interpretedexceptby a sea-bornemovement.The origins of this form of pottery, outsideBritain, have still satisfactorilyto be determined. But in view of the generallyacceptedanaloguesfor its decorationamong the late neolithic ChasseyII waresof westernFrance,l and of the occurrencein passagegravesin Orkney and Caithnessof stoneand bonetypes which accompanyRinyo-Clacton ware at the sites of Skara Brae and Rinyo, onemay hazardthe guessthat the spreadof this potteryup the west coaststo the far north is in someway to be connectedwith the spreadof the passagegravesthemselves. The great value of the excavationsat the two principal sites of the Rinyo-Clactonculture in the north should not be allowed to obscurethe fact that Skara Brae and Rinyo are sui generis, and cannot be taken as typical of SecondaryNeolithic settlementseven on the mainland of Scotland, let alone elsewherein Britain. Structurally they representa specific adaptationto the exigenciesof a treelessenvironment,in which the only organic building materials were driftwood and the bones of whales,both of uncertainandfortuitous supply; while the finds from both sites,andpartietdarlythe boneandstoneartifacts,showmarkedinfluences from the Circumpolar Stone Age cultures of the Sub-Arctic Zone, in a degreeunparalleledin any other Rinyo-Clactonsite farther to the south. The remainingtypes which can be regardedas of ultimately southern 1

Piggott, Neolithic Cultures (1954),344-6.

P.P.S.-c

23

PETERBOROUGH VVARE RINYO-CLACTON WARE DISCOIDAL FLINT KNIVES LONG POLISHED FLINT KNIVES

4. 4.

Fig. 4. SecondaryNeolithic pottery and flint types

pottery

FISHERMEN AND FARMERS

origin are flint knives with polishededges;jet 'sliders' or belt-fasteners; and perhapsflint arrowheadsof the chisel-edgedand lop-sided forms derived from the petit-tranchet type of the mesolithic Northern Forest cultures. The distribution of two types of flint knives with polished edgesis shownin fig. 4. The first is generallyelongated,of blade-likeform, often with only one long edgetreatedby grinding or polishing, but sometimes polished all over. Such knives occur in England in well-established SecondaryNeolithic contexts,particularly with burials of the Dorchester culture.l The secondis the discoidalknife, generallylargerand sometimes more carefully finished than the former type. On groundsof distribution only (for no specimenhas yet been found in unequivocalassociation) Clark ascribedtheseto the Beakercultures.2 But they are not known in Beakercontextson the Continent,and their distribution accordsequally well with that of othervarietiesof SecondaryNeolithic materialin Britain. In Scotlandboth typesoccur predominantlyin the east,but with some differencesof local distribution. The long knives are found right up to the extremenorth, in Caithnessand Orkney, where in three instancesthey are associatedwith passagegraves (Ormiegill, Camster Round, and Unstan); in southernScotlandthey are rare. The discoidal type, on the other hand,is chiefly concentratedin the Tweed basin,and has not been found north of the CromartyFirth. Though the numberof both types is small, the differing distributions may reflect the same kind of regional specialization that is apparent in other types of SecondaryNeolithic materialin Scotland. The sizeof theseknives,andparticularlyof the discoidaltype, suggests that they were madefrom mined ratherthan from beach-pebbleflint; and since there is no marked concentrationof either type in the Buchan district of Aberdeenshire,the only sourceof massiveflint in Scotland,it is likely that many of them were imported from the south, either from Yorkshire or even from the flint mines of East Anglia, near which the greatestdensity of discoidal knives has beenrecorded. It may be suggestedin passingthat the well-known 'Pict's knives' of Shetlandmay constitutea specializedregional variant, in local material, of the samebasicform. Thoughno Pict's knife has so far beenfound in a chamberedtomb or other satisfactorily dateablecontext, the apparent 1 2

Piggott, Neolithic Cultures (1954), 359. Proc. Prehist. Soc. East Anglia, VI (1929),52.

25

R.

J.

C. ATKINSON

associationof hoardsof theseobjectswith polishedstoneaxesl and with a stoneadze2 suggestthat a neolithic dateis at leastprobable. Of the jet 'sliders', shuttle-shapedobjectsperforatedwith an oval or lenticular apertureand perhapsused to fasten a belt or sash,only three exampleshavebeenfound in Scotland.Two comefrom Skye and Kintyre respectively,3the latter as a secondarydepositin the Clyde-Carlingford tomb at Beacharra.The third is a stray find from a peat-bogat Balgone, near North Berwick.4 Though the westernexamplesmay be of Scottish material,the type occursmorecommonlyin England,whereit is associated with burials of the Peterboroughand Dorchestercultures.s In consideringthe flint arrowheadsof petit-trancket-derivativeform, generallyregardedas a specific type of the SecondaryNeolithic cultures, it must be bornein mind that no comprehensivestudy has beenmadeof their distribution. However, in so far as the collections of the National Museumrepresenttheir relative density in various parts of the country, it is clear that they presenta very different picture from the leaf-shaped arrowheadsdiscussedabove (p. 20). Like the latter, they occur almost entirely in easternareas,from the Beauly Firth to the Border, and only a barescoreof specimenshas beenfound in the west and north. But unlike the leafform, thereis no markedconcentrationin the north-east,apartfrom the singlesite of the Culbin Sands.Elsewherethey occurin large numbers only in the Tweedbasin,where they actually outnumberthe leaf form by about sevento two. In the countiesof Banff, Aberdeen,Kincardine, and Angus, on the other hand,the leaf shapeoccursover thirty times as often as the petit-trancket-derivative.The restricted distribution of the type thus matchesthat of the discoidal flint knives; and in view of its scarcity near the principal sourceof flint in Aberdeenshire,it may perhapsbest be regardedas exotic to Scotland, even though many of the specimens in the Tweedbasinare undoubtedlyof local flint. 6 TheSecondaryNeolithic materialdiscussedabovecanthus all plausibly be regardedas intrusive, representingthe northwards migration from England of cultures already fully formed before their expansion. In supportof this view it may be noticedthat the distribution of thesetypes Proc. Soc. Ant. Scot., XXIX (1894-95),48-54. Proc. Soc. Ant. Scot., LXXX (1945-46), 140-1. 8 Proc. Soc. Ant. Scot., L (1915-16), 221. 4 Proc. Soc. Ant. Scot., VI (1867-8), 107-8. 5 Piggott, Neolithic Cultures (1954), 3Il. 6 Proc. Soc. Ant. Scot., LXXXI (1946-47), 181-2. 1

2

26

FISHERMEN AND FARMERS

(fig. 4) conformsvery little to the recordedpatternof mesolithicsettlement in Scotland,and that they do not occur at all on the coastof Argyll, in the Forth Valley or aroundthe head-watersof the Clyde and Tweed. It is only indeedin the middle and lower areasof the Tweedbasinthat any substantialoverlap can be seen. It is possible, however, to distinguish elements in the Secondary Neolithic materialof Scotlandwhich appearto be indigenous,or at least mainly if not entirely of Scottish origin. These comprise the Sandhill type of pottery, and two forms of stone object: carved stone balls and cylindrically perforatedmaceheads.The latter are plottedin figs. 5 and 6. Sandhill pottery has been recovered in quantity only from three coastal sites, the Luce Sands in Wigtownshire and the Gullane and HedderwickSandsin EastLothian.1 The waresare very variablein shape, fabric, and decoration,and no comprehensivestudy of the material has so far beenmade.Two shapesappear,however,to be noticeablyfrequent; a hemisphericalbowl with plain rim, sometimesslightly inturned,decorated with coarsecord-impressions;anda larger and coarserbarrel-shapedjar, with Hat baseand Hattenedrim, decoratedwith irregular linear patterns scoredwith a point or with a bird bone,the latter giving a shallowdouble groove. At Glenluce trial excavationsby the writer in 1951 showedthat thesewareswere contemporarywith fragmentsof cord-zonedBeaker;and a similar associationappearsto have existedon the EastLothian sites. Genericallythis pottery may be comparedwith the Sandhill waresof Northern Ireland,2 and there are indeedresemblancesof a specific kind as well. But even the material from Glenluce,when seenin bulk, has its on own characteristicswhich distinguishit from comparableassemblages the othersideof the North Channel.The Sandlllll waresof Scotland,therefore, thoughrelatedto thoseof Ireland, may nonethe less be regardedas an indigenous phenomenon.It is on these sites, if anywhere,that we should seek for a survival of the traditions of the mesolithic strandloopersof the Atlantic period. The carvedstoneballs haveoften beendiscussedin the past,3generally 1 Proc. Soc. Ant. Scot., LXXXVI (1951-52), 43-69 (Glenluce); ibid., XLII (1907-08),3°8--19(Gullane); ibid., LIII (1928--29), 68--72 (Hedderwick). Much of the materialfrom Glenlucehasnot beenpublished. 2 Piggott, Neolithic Cultures (1954), 317-21. 3 Proc. Soc. Ant. Scot., XI (1875-76),2g--62, 313-19; ibid., XXXVI (1901-02), 11-16; ibid., XLI (1916-07), 290-3°0;ibid., XLVIII (1913-14), 407-20; Proc. Roy.Soc.Edinburgh,L (1929-30),72-73.

27

R.

J.

C. ATKINSON

without any firm conclusion,since apart from the specimensfrom Skara Brae! nonehasso far beenfound in satisfactoryassociationwith any other significant type of object. The assumptionthat theseballs belong to the SecondaryNeolithic culture of Scotland thus rests upon an admittedly slenderfoundation. But a date Within the first half of the secondmillennium B.C. is supportedby the spiralornamenton the examplesfrom Towie, Aberdeenshire,from Elgin and from Glasterlaw,Forfarshire,which has been comparedto motifs in the repertory of passage-grave 'art' of this period. Moreover, the distribution of carvedstoneballs accordswell with that of the maceheads, whoseSecondaryNeolithic contextis not in doubt. The map in fig. 5 is basedon the literature cited above and on a censusof the unpublishedexamplesin the principal Scottish museums. It includesabouttwice the numberof sitespreviouslyplotted by Childe,2 but excludesat least twenty specimenswhose preciseprovenanceis not recorded.Almost all of thesecome from somewherein Aberdeenshire,so that the concentrationof the type in that countyis evenmore markedthan the map itself suggests. It will be noticed that south of the Clyde-Forth line carved stone balls are conspicuouslyabsentfrom the areasin which 'intrusive' Secondary Neolithic materialoccurs(fig. 4), and especiallyfrom the basinof the Tweed, a deficiency which servesto underlineall the more strongly the regional characterof the concentrationin Aberdeenshire. The polished stone maceheadswith cylindrical perforation, mapped in fig. 6, belongto typeswhich havewell-establishedSecondaryNeolithic 3 The remarkabledensity of theseobjects in Orkney (where associations. curiously the greaternumber are broken specimens)comprisesover 30 per cent of the total known in Scotland,and it is difficult not to assumea northern origin for the type. Though petrological examinationhas not beencarriedout systematically,it should be notedthat two specimensof the 'cushion'type studiedby Gibson,4found in Lewis and Fife, appearto be of a rock native to Shetland.5 Maceheadsof these types are also widespread,of course,in England and Wales, and it would be foolish to suggestthat they all originate north of the Border. But it shouldbe rememberedthat almostall of them V. G. Childe, Skara Brae (1931), IOo-g. Proc. Roy. Soc. Edinburgh, L (1929-30),73. 3 Piggott, Neolithu Cultures (1954), 285, 353, 354. 'Proc. Soc. Ant. Scot., LXXVIII (1943-44), 16-25. 5 Proc. Soc. Ant. Scot., LXVIII (1933-34),428-32. 1

2

28

""

""

Fig. 5. Distribution of carvedstoneballs.

MILES

R.

J.

C. ATKINSON

are of igneousor metamorphicrocks (examplesin flint are exceedingly rare), so that their placesof manufacture,and hencethe origin of the type itself, shouldlie within the Highland Zone. Apart from the exceptionalconcentrationof maceheadsin Orkney, the only othermarkedareaof densityis in the Buchanregionof Aberdeenshire, preciselywhere the greatestnumberof carvedstoneballs has also beenrecorded.It is difficult to avoid the conclusionthat the patternof both theseclassesof object, and of the populationwhich they must represent, is related to the presencein Buchan of the only major source of flint in Scotland.It occursprincipally on the ridge of high ground which extendswestwardsinland from BuchanNess for a distanceof about ten miles. Other minor sources have been recorded a mile or two E. of Turriff and in Boyndie Bay, immediatelyW. of Banff.1 Many flint-knapping sites have been reported from Aberdeenshire,2 usually at somelittle distancefrom the actual sourcesof Hint. At one of these a hoard of rough-outsoccurred,3 which suggeststhat as with the mined flint of southernEnglandfabrication at the sourcewas confined to preliminary Haking only. The part-finished tools would then be distributed in that form, the final processesbeing left to the eventual purchaser. The easily recognizablered and orange colours of the Buchan flint provide a meansof identifying the areaover which it was distributed.No systematicsurvey of its occurrencehas so far been made; but casual 4 observationsshowthat to the north at leastit reachedas far as Sutherland 5 and the neighbourhoodof Wick in Caithness. The prime use of Buchan flint appearsto have been not for heavy woodworking tools, such as axes and adzes,but for the smaller objects like scrapers,knives, andarrowheads,for which igneousand metamorphic rocks are unsuitableas raw materials.The value of a substancecapable of being Hakedin a controlledmannerand of giving a really sharpedgeis well illustrated by the occurrenceon severalsites in the Tweed basin of 6 which must havebeentradedright across fragmentsof Arran pitchstone, southernScotland,probably by way of the Clyde and the Biggar Gap. Proc. Soc. Ant. Scot., X (1872-73),514-18. Proc. Soc. Ant. Scot., LI (191&-17), 117-27. 3 Proc. Soc. Ant. Scot., XXX (1895-96),346-51. 4 Proc. Soc. Ant. Scot., LXXX (1945-46),32. c; Proc. Soc. Ant. Scot., LXIX (1934-35), 114. 6 Proc. Soc. Ant. Scot., LVIII (1923-24), 120, 123; ibid., LXI (1926-27), 114. 1

2

30

10 '

0

Fig. 6. Distribution of stonemaceheadscylindrically perforated.

50 MILES

R.

J.

C. ATKINSON

The importing of this material into an area in which worked flints, especiallyof SecondaryNeolithic type, are abundantsuggeststhat the beach-pebbles, which there as elsewherein Scotlandoutsidethe Buchan district are the only available form of raw flint, were not consideredan ideal raw material even for the manufactureof small tools. Indeed, modernexperimentshowsthat suchpebblesare full of unexpectedflaws, and havenoneof the uniformity or toughnessof mined flint. It can thereforebe suggestedthat the Buchanflint depositsaccountin large part alike for the earlier settlementof the north-easternangle of Scotlandby groups of migrants from the Windmill Hill and Lyles Hill cultures,andfor the developmentin the sameareaof a regionalSecondary Neolithic populationwhich owed little to influencesfrom outside. In the absenceof any settlementsites of the Secondary Neolithic culturesof Scotland,apart from SkaraBrae and Rinyo, which cannotbe regardedas typical, it is difficult to say much aboutthe economicbasisof the life of these communities.It has already been suggested,however, that the exploitation and distribution of raw materialsplayed some part, and perhapsa major one, in their economy.If this is so, it is not only inorganic materialsto which they would haveturnedtheir attention.The widespreadand predominantlycoastaland riverine distribution of their stoneartifacts arguesa familiarity with the seaand with boats,and hence with fish. Thoughtoday we can only trace the economicactivity of early man throughimperishablecommodities,we may be surethat the bulk of his commercewas in things that leave no trace. Kippers may well have beenshippedout of Aberdeenwhen Troy itself was young.

32

Appendix The Scottishfinds of pottery and other objectsplotted in figs. 1,2, and 4 arelisted hereunder. FIG.

I

LuggedbowlsofPiggott'sformsA, B, andC Loanheadof Daviot, Aberdeenshire Proc. Soc. Ant. Scot., LXIX (193435), 168-214,fig. 14, no. 12. East Finnercy,Aberdeenshire Proc. Soc. Ant. Scot., LXIII (192829),62-63,fig. 50, nos. 6 and 7. Proc. Soc. Ant. Scot., LXXXII Cairnpapple,West Lothian (1947-48),102,fig. 15,no.2. Proc. Soc. Ant. Scot., LXIII (1928Beacharra,Kintyre, Argyll 29),51-53,fig. 28. Proc. Soc. Ant. Scot., LXIII (1928Torlin, Arran 29),46,fig. 15. Clachaig,Arran Proc. Soc. Ant. Scot., LXIII (192829),46,fig. 16. Proc. Soc. Ant. Scot., LXIII (1928Sliddery Water, Arran 29),47,fig. 18. Luce Sands,Wigtownshire Proc. Soc. Ant. Scot., LXIII (192829), 66, no. 28, fig. 44, no. 23; ibid., LXVII (1932-33), 240, no. 15, fig. 6, no. II.

ShoulderedbowlsofPiggott'sForm G PowsodeCairn, Atherb, Aberdeenshire Bantaskine,Stirlingshire

Unpublished. Nat. Mus., Edinburgh,nos. EO 910, 923. Proc. Soc. Ant. Scot., LXIII (192829), 56-57,fig. 38, no. 8.

ss

R.

J.

C. ATKINSON

Cairnpapple,West Lothian

Proc. Soc. Ant. Scot., LXXXII (1947-48),102,fig. IS, no. I. Proc. Soc. Ant. Scot., LXXXIII (1948-49),118-19,fig. 7, no. I.

Cairnholy I, Kirkcudbright

FIG.

2

LylesHill Ware Urquhart, Morayshire Townhead,Rothesay,Bute Cairnholy I, Kirkcudbright

Proc. Soc. Ant. Scot., LXXX (194S46), 142, PI. XXIV, no. I. Proc. Soc. Ant. Scot., LXIII (192829), S7-59, fig. 39· Proc. Soc. Ant. Scot., LXXXIII (1948-49),119,fig·7,no.2.

LylesHill Ware with finger-tip fluting Eastertonof Roseisle,Morayshire Spey Valley, betweenGrantown and Newtonmore PowsodeCairn, Atherb, Aberdeenshire KnappertyHillock, Auchmachar, Aberdeenshire Loanheadof Daviot, Aberdeenshire East Finnercy, Aberdeenshire Cultoquhey,CriefI', Perthshire Achnachree,Benderloch,Argyll Kilchoan, Poltalloch, Argyll Nether Largie, Poltalloch, Argyll

Proc. Soc. Ant. Scot., LXIII (192829),56, figs. 37, 38. Proc. Soc. Ant. Scot., LXXI (193637),367. Unpublished. Nat. Mus., Edinburgh,no. EO 9 I7. Proc. Soc. Ant. Scot., LXIII (192829),63· Proc. Soc. Ant. Scot., LXIX (193435), 207, fig. 14, type 6, nos. 2-4. Unpublished. Report forthcoming in Proc. Soc.Ant. Scot. Unpublished. Information from ProfessorStuartPiggott. Proc. Soc. Ant. Scot., LXIII (192829),38, fig. 3. Proc. Soc. Ant. Scot., LXIII (192829),38. Proc. Soc. Ant. Scot., LXIII(I92829),37,fig. I. 34

FISHERMEN AND FARMERS: APPENDIX

Proc. Soc. Ant. Scot., LXIII (192829),49-50,figs. 22, 23. Proc. Soc. Ant. Scot., LXXXII (1947-48),236,fig. I, no. 2. Unpublished. Information from ProfessorStuartPiggott. Proc. Soc. Ant. Scot., LXIII (192B29),66,nos. 13, IB, 20.

Glecknabae,Bute Knappers,Clydebank, Dumbartonshire Whitemoss,Bishopton, Renfrewshire Luce Sands,Wigtownshire

FIG. 4

PeterboroughWare Proc. Soc. Ant. Scot., LXXXII (1947-48),262. Proc. Soc. Ant. Scot., LXIX (193435),363-4. Proc. Soc. Ant. Scot., LXIII (I 92B29),92, fig. 55, nos.4-7. Unpublished. Nat. Mus., Edinburgh.

Scotstarvit,Fife Knappers,Clydebank, Dumbartonshire HedderwickSands,East Lothian ScremerstonHill, near Berwickon-Tweed,Northumberland Ford Castle, Northumberland ShewaltonSands,Irvine, Ayrshire Cairnholy I, Kirkcudbright Luce Sands,Wigtownshire

Arch. Jouro., LXXXVIII (1931), 157· Proc. Prehist. Soc., XVII (1951),53. Proc. Soc. Ant. Scot., LXXXIII (1948-49),120,no. 7. Proc. Soc. Ant. Scot., LXIII (192B29),92,fig. 55, nos. 1-3,8.

RinyoClacton Ware Rinyo, Rousay,Orkney

Evie, Orkney Dingieshow,Deerness,Orkney

Proc. Soc. Ant. Scot.,LXXIII 39),22-25;ibid., LXXXI 47),34-39· Proc. Soc. Ant. Scot., LXXX 46), 143· Proc. Soc. Ant. Scot., LXXX 46), 142. 35

(193B(1946(1945(1945-

R.

J.

C. ATKINSON

SkaraBrae, Orkney Freswick Sands,Caithness Unival, North Uist Tentsmuir,Fife Gullane Sands,East Lothian HedderwickSands,East Lothian Knappers,Clydebank, Dumbartonshire Townhead,Rothesay,Bute Tormore, Arran Luce Sands,Wigtownshire Old Yeavering,Wooler, Northumberland

Childe, SkaraBrae (1931), 127-34. Proc. Prehist. Soc., XVII (1951), 73. Proc. Soc. Ant. Scot., LXXXII (1947-48),26-28. Unpublished. St. Andrews University Museum.

Proc. Soc. Ant. Scot., XLII (19078),312- 15. Proc. Soc. Ant. Scot., LXXX (194546), 143,fig. I, no. 5. Proc. Soc. Ant. Scot., LXXXIV (1949-50),IBo-3· Proc. Soc. Ant. Scot., LXXXIV (1949-50), 183. Piggott, Neolithic Cultures (1954), 386. Proc. Soc. Ant. Scot., LXXX (194546), 143· Unpublished. Information from Dr. Brian Hope-Taylor.

PolishedFlint Knives Rinyo, Rousay,Orkney Unstan,Orkney SkaraBrae, Orkney CamsterRound Cairn, Caithness Ormiegill, Ulbster, Caithness Urquhart, Morayshire Urquhart, Morayshire

Proc. Soc. Ant. Scot.,LXXIII (193B39),27,fig. 8, no. 1I. Proc. Soc. Ant. Scot., XIX (188485),350. Childe, Skara Brae (1931), 114, fig. 10. Anderson,Scotlandin Pagan Times: Bronzeand StoneAges(1886), 252, fig. 251. Anderson,op. cit., 247, fig. 244. Proc. Soc. Ant. Scot., IX (1870-71), 238, fig. I. Unpublished. Nat. Mus., Edinburgh. 36

FISHERMEN AND FARMERS: APPENDIX

Fintray, Aberdeenshire

Unpublished. Nat. Mus., Edinburgh.

Blelack, Cromar,Aberdeenshire

Proc. Soc. Ant. Soc., IX (1870-71), 239· Unpublished. Nat. Mus., Edinburgh. Proc. Soc. Ant. Scot., LXVI (193132),25,fig. 8. Proc. Soc. Ant. Scot., IX (187°-71), 239, fig. 2. Proc. Soc. Ant. Scot.,XXXVI (19012),101,fig. 25. Unpublished. Nat. Mus., Edinburgh.

Birse, Aberdeenshire Overhowden,Channelkirk, Berwickshire Strachur,Argyll Tormore, Arran Whiting Bay, Arran

DiscoidalFlint Knives Ardross, EasterRoss

Unpublished. Nat. Mus., Edinburgh. Proc. Soc. Ant. Scot., XXV (189°91 ),498-9. Unpublished. Nat. Mus., Edinburgh.

Culbin Sands,Morayshire Marnoch, Banffshire

Pitdoulzie, Turriff, Aberdeenshire Proc. Soc. Ant. Scot., XII (1876-77), 2°7· Leslie, Aberdeenshire Unpublished. Nat. Mus., Edinburgh. Kintore, Aberdeenshire Evans, Anc. Stone Implements, 2nd ed. (1897),342. Fourdoun,Kincardineshire Proc. Soc. Ant. Scot., XI (1874-75), 576. Fordoun,Kincardineshire Unpublished. Nat. Mus., Edinburgh. Airhouse, Channelkirk, Proc. Soc. Ant. Scot., LXII (1927Berwickshire 28) 170. Ninewar, Duns, Berwickshire1 Proc. Soc. Ant. Scot., LXII (192728),172. 1

Polisheddiscoidal scraper.

37

R.

J.

C. ATKINSON

Birkenside,Lauder, Berwickshire

Unpublished. Nat. Mus., Edinburgh.

Earlston,Berwickshire

Proc. Soc.Ant. Scot.,XXVIII (189394),32 4. Proc. Soc. Ant. Scot., XXVIII ( 1893-94), 32 4. Proc. Soc. Ant. Scot., LI (1916-17), 234· Proc. Soc. Ant. Scot.,LXXIV (193940),10. Unpublished. Nat. Mus., Edinburgh. Proc. Soc. Ant. Scot., XXIII (188889), 204.

Butterlaw, Coldstream, Berwickshire Whooplaw, Stow, Roxburghshire Blackhaugh,Clovenfords, Selkirkshire Milton Loch, Kirkcudbright Luce Sands,Wigtownshire

38

I.

Obaniantools of antler, bone and stone. (Chap. I).

2. Perforated mace-heads from Scotland. ( Chap. 1).

3. Carvedstone balls from Scotland. (Chap. I).

4(a). Stennessstonecircle, Orkney. (Chap. II).

4(b). Interior of MaesHowe, Orkney. (Chap. II).

5. Bell-Beaker from Fingask, Perthshireand Cord-ZonedBeaker from Bathgate,West Lothian. (Chap. III).

6. Beakersfrom Edzel (left) and Ermelo, Netherlands(right). (Chap. III).

7. Late BronzeAge hoard. Braesof Gight. (Chap. IV).

8. Late Bronze Age bucket, Cardross.(Chap. IV).

CHAPTER TWO

The Megalith Builders

S

O much hasbeenwritten aboutthe megalithicmonumentsof Scotland, particularly in the last twenty-five years, that any fresh accountof them at the momentcan do little to supplementthe comprehensivetreatment of the subject provided by Childe in his Prehistory of Scotland (1935) and Piggott in his Neolithic Cultures of the British Isles (1954), and the readeris directedto thesetwo books and the detailedreferencesthey contain at every turn of the descriptionand discussionattemptedin this presentchapter.Yet becauseso much has been written about this important and interesting subject-and on any showing the megalith builders must be one of the main ethnic elementsin the early peoplingof Scotland-andbecausethe books of Childe and Piggott we have mentioned, themselvesseparatedby nearly twenty years, are in themselves stagesin the history of archaeologicalscholarshipin Scotland,it may be convenientand profitable here first to survey someof the major attempts madeat synthesizingour knowledgeof the Scottishmegalithicmonuments. We can then see presentorthodoxiesand heresiesin the perspectiveof our growing knowledge of Scottish prehistory and of the megalithic monumentsof Europe in general. It should be said here that in this chapterour main concernis with megalithic tombs, but brief reference is madeat the end to other typesof megalithicmonumentin Scotland. There is no need, for our presenthistorical purpose, to go farther back than Daniel Wilson, one of the pioneersof Scottish, and for that matter Europeanprehistory: indeed, it was Wilson who introducedthe P.P.S.-D

39

G. E. DANIEL

word prehistory into the English language-andnot, surprisingly, his predecessorslike the Wessexfieldworkers Colt Hoare and Cunnington or the Danishhistoriansand archaeologistsVedel-Simonsenand Christian Thomsen,who propoundedthe ideasof successiveagesof stone,bronze, and iron in the most ancienthistory of man. Daniel Wilson publishedin 1851 his Archaeologyand Prehistoric Annalsof Scotland,and there refers to 'the applicationof the term prehistoric-introducedif I mistake not, for the first time in this work'. This term was not readily receivedwith enthusiasmby all workers in the field of early Scottisharchaeology:how scornful, for example,was JamesFergussonof the term prehistory. He quotedthe title of Wilson's book in his RudeStoneMonumentsand added 'whateverthat may mean',1and RobertMunro, as late as 1899 was referring to 'that obscureperiod in the history of Scotlandvaguely defined as the prehistoric'.2 Daniel Wilson very naturally had to discussthe chamberedcairns and cromlechsof Scotland, and he set these against his picture of Scottish prehistory which he divided into the primeval or stone period, the archaic or bronze period, the iron period and the Christian period. He figured somecromlechsas he called them, using the commonWelsh folk name for a megalithic tomb, and also Wideford Hill in Orkney, which was, however,to him, not a tomb but a Picl's house.Wilson had no doubt as to the sepulchral nature of the chamberedtombs and cromlechs, thinking this had beenproved by F. C. Lukis andJohn Bell. They were all part of his monolithic era of the primeval or stone period.He did not think them Celtic monumentsas others did, but 'the work of an elder race of whose languagewe have little reason to believe any relic has survived to the present day'. The chamberedcairns he thought were 'catacombsof the whole tribe . . . the memorial of the victors on some bloody battle field'. One thing Daniel Wilson did not find surprising at all, and that is the absenceof contemporarysettlementsites;it was normal he thought that the deadshould be well providedfor and that the houses of the living shouldperish; death andreligion, he insisted,were probably more important to primitive peoplesthan life. The ordinary housesof the megalith builders would, of course,have perished,in Wilson's view. 1 JamesFergusson,RudeStone Monumentsin All Countries: Their Ages and Uses, London, 1872,239. 2 Robert Munro, Prehistoric Scotlandand its Place in EuropeanCivilisation, Edinburgh and London, 1899,v.

40

THE MEGALITH BUILDERS

The same sort of view was held by the two Scottish scholarswho, between Wilson himself and modern scholars, did most to advance Scottisharchaeologyand synthesizeScottish prehistory-namelyRobert Munro and JosephAnderson,two remarkablemen who both lived to be well over eighty and who spannedin their archaeologicallives the whole developmentof nineteenth-centuryScottish archaeologyand the whole gamut of its changing perspectives.JosephAnderson came of age two yearsafter Daniel Wilson's Prehistoric Annalswas first published-itwas incidentally the year that also saw the publication of Beddoe'sScottish Ethnology: while RobertMunro died in 1920, five yearsbefore the publication of Childe'sDawn ofEuropeanCivilisation and Kendrick'sAxeAge. This long spanof three-quartersof a centuryhadseenthe first excavations in Scottishmegalithicmonuments.GeorgePetriehad excavatedWideford Hill in 1849, and it was his plan that Wilson had published.In 1851 a Captain Thomas excavateda monumenton the Holm of PapaWestray in Orkney,which he describedas 104 feet long by 41 feet broadby 10 feet high. It consistsof a passageleadingin from the middle of the southside of a long mound, and a long chamberat right-anglesto this with no less than fourteenside-chambers openingout of it. Thomasfound no artifacts and no bones except a few of sheepand rabbit. He recordedsome engravingson the stonesand otherswere found a few yearslater by Petrie. Nearly a hundred years later the Royal Commission on the Ancient Monumentsof Scotland,in their InventoryofOrkney,suggestvery plausibly that in two places these engravingsare 'crude representationsof the human face similar to that found on the chalk drums from Folkton, Yorkshire'.1This face motif on the Holm of PapaWestrayis not only to be comparedwith that on the chalk idols from Folkton, but with the face on the roof stoneof the north side-chamberat New Grange,the faces on the undersideof the capstoneof the Dehus in Guernsey,the chalk-cut tombs of Coizard and Courjeonnetin the Marne, the statue-menhirsof southernFrance,and the pottery and portableidols of the south Iberian collective tombs. It provides a link, tenuous,slight, but none the less sure, betweenthe megalith builders of the extremenorth of the British Isles and their ultimate progenitors,as we shall see most people now believe, in the Mediterranean.But this connexion, which makes an exciting historical context out of the archaeologicaldetail of thesedead tombs,was not apparentto the Scottishantiquariesof the mid-nineteenth 1

InventoryofOrkney,Edinburgh,1946, 189. 41

G. E. DANIEL

century,who were still in the processof discoveringthe megalithic tombs of their native land. In 1861 an even more exciting tomb than the Holm of PapaWestray was discovered; this was Maes Howe in the Orkneys, which Daniel Wilson had characterizedas 'the Wiltshire of Scotland,in so far as the mere number of sepulchralmounds along with megalithic groups and other aboriginal structures can constitute this distinction'. This fine monument,which the Royal Commissionon the Ancient Monumentsof Scotlanddescribewith pardonablepride as 'the supremeexampleof its class in Great Britain',! is a corbelled passagegrave with three sidechambersset in a round mound 24 feet high by 115 feet in diameter.The mounditself is surroundedby a berm and this by a broadditch averaging 45 feet in width and6 feet in depth.JamesFarrer,M.P., whenhe excavated the mound in 1861, found no artifacts or bones,but he did find that the tomb had beenbrokeninto previouslyand that recordsof this survivedon various stonesof the chamberin the form of twenty-four runic inscriptions as well as threeengravedfigures of a walrus, a dragon,and a serpent knot. Professor Haakon Shetelig dates these engravings to the early twelfth century A.D. The runic inscriptions were apparentlymade on different occasions:two inscriptionsrefer to the breakingin of the Howe by Crusaders,and theseare usually identified with the expeditionof Earl RognvaldandEindrid the Younger,which winteredin Orkneyin 1150-51. Another visit took placein January1153, whereEarl Harold and his men landed near Stromnessand were in the orkahaugr, presumablyMaes Howe, 'while a snowstormdrove over them and there two men of their band lost their wits, and that was a great hindrance to the journey'. Severalof the runic inscriptionsrefer to treasurebeing found there: one says'treasurewas carriedoff in the courseof threenights', another'a long time ago was a great treasurehidden here . . . Haakon single-handed bore treasurefrom this howe', and a third 'treasurewas carriedoff before thoseCrusadersbroke into the howe'.2Thesereferences,while not specific, seem very circumstantialand can hardly be dismissedas part of a generalfolk feeling in Viking times that therewas treasureto be found in 1 Inventory of Orkney, 306. Farrer'sMaeshowewas privately printed in 1802. For an accountof recentexcavationsat the site seeV. G. Childe, Proc. Soc. Ant. Scot., LXXXVIII (1956), 155-72. 2 Thesetranslationsof the runes are from Bruce Dickins, Proc. Orkney Ant. Soc., VIn (1929-30), 27-30. See also J. Farrer, Notice of Runic Inscriptions . .. in the Orkneys,Edinburgh,1862,and InventoryofOrkney,308-13.

THE MEGALITH BUILDERS

all ancientburial mounds.It doesseemthat thetwelfth-centuryVikings who brokeinto MaesHowe did find somethingof value. Howeverthat may be, they certainlyleft nothing of value for Farrerto find sevencenturieslater.1 We have alreadyquotedone dictum of the Royal Commissionon Ancient Monumentsin Scotlandon Maes Howe; elsewherethey refer to it as a 'magnificentdomedtomb . . . probably without a rival in Western Europe',2and they do well, as Childe has done on severaloccasions,to draw attentionto someof the very remarkablefeaturesthat exist at Maes Howe-notonly the berm and the encircling ditch which we havealready mentioned,but the almost ashlar-like dressingof the walling stones,the brilliant masonryof the jointing, the three side-chambersset not on the groundbut high up in the walls, and most of all the anglepiersof what is in plan a squarechamber.Without any doubt, Maes Howe is a tour de force of prehistoricarchitecturein westernEurope.3 Three years after Farrer's discovery and excavationof Maes Howe, Canon Greenwell was digging in the Kilmartin and Kilmichael area of Argyllshire in tombs we would now describeas south-westScottishsegmented long cists or gallery graves. In 1865 and 1866 Anderson was excavatinga seriesof chamberedcairns in Caithness;in 1867 Farrerand Petrie dug the Quoynessmonumentin Sandayin the Orkneys; in 1871 Dr. Angus Smith excavatedthe chamberedcairn of AchnacreenearLoch Etive in Argyll, and in 1884 Clouston dug Unstan. The men who, like Munro and Anderson,attemptedto synthesizeScottishprehistoryat the end of the nineteenthcenturywere in quite a different position from Daniel Wilson in 1851. They now had a considerablenumberof excavations to turn to; they had accountsof different typesof tombsin different parts of Scotland,and what is equally important, since our knowledgeof megalithsor indeedof any classof monumentcan only advanceby comparativestudy outsidea country as well as excavationand analysiswithin it, they were beginningto learn somethingof similar monumentsoutside Scotland.The Baron Bonstetten'sEssai sur les dolmenshad appearedin 1865, James Fergusson'sRude Stone Monumentswas published seven 1 On the questionof the Maes Howe 'treasure'see Piggott, Neolithic Cultures of the British Isles (Cambridge,1954),253-4. 2 Reportand Introduction to Inventory of Orkney and Shetland(Edinburgh, 1946), 19. 3 The large stonesfor blocking the entrancesto the side-chambersare anotherextraordinaryfeature. In someways Maes Howe reminds one of Iberian tombs like Tutugi and not the collective chamberedtombs of the early secondmillennium. Alcalar in southPortugal has, of course,parallelsfor the cupboard-likeside-chambers.

43

G. E. DANIEL

years later; its frontispiece was the StandingStonesof Stennis,and he devoteda whole chapterto the Scottishmegaliths,expressinghis indebtednessfor information about them to John Stuart of Edinburghand to Sir Henry Dryden,the latter a man well travelledin westernEuropein search of megaliths. James Fergussonwas astonishingly complacentabout the state of knowledge at the time when he wrote: 'Whatever may be the case as regardsIreland, it is probable that the megalithic remains of Scotland are all known and have been describedmore or less in detail'. His own accountof Scotlandis very thin, but his explanationof the origins of the Scottishmegalithsvery full and definite; they were the work of two groups of people, a 'circle building race' who came from Scandinavia between800 and 1,000A.D. [sic] andspreadsouthwardsfrom the Orkneys, where they had built Stennis and Maes Howe, and a 'dolmen-building race' who had spreadup from Ireland and Wales and were derived from the builders of the megalithic monumentsof Brittany which Fergusson dated to 'the Arthurian age-between380 and 550 A.D.' We need not spendmuch time on Fergusson'sideasexceptto realize that for long his was the only generalbook dealingwith megalithicmonumentsas a whole, and that as a result his views had a wide currency, and that the fact he was able to put forward views on dating which to us at the presentday seemtotally without any foundation whatsoeverwas some indication of the real ignoranceof comparativemegalithic scholarshipin and out of Scotlandthree-quartersof a centuryago. Let us now see what sort of a synthesiswas made by Andersonand Munro, who knew far more about the Scottish material than Fergusson did and who were unconvincedby his chronologicalarguments.Between 1879 and 1882 Andersonwas askedto give the Rhind Lectures by the Societyof Antiquariesof Scotland.Theselectureshad beenfoundedfrom the estateof Henry Rhind of Sibster, who had himself excavatedfour chamberedbarrows at Yarhousein Caithness.Anderson'slectureswere publishedfour years later as Scotlandin Pagan Times; the Bronze and StoneAges.Looking back at the two chaptersof this book which deal with Scottishchamberedcairns, the presentwriter finds it most remarkableto seewhat a graspAndersonhad of the problem.1 He first describesmonu1 But thenJosephAndersonwas a most remarkableman. It is worth reading the articles he wrote on archaeologicalsubjectsfor the 1888-92 edition of Cham-

bers's Encyclopaedia.

44

THE MEGALITH BUILDERS

ments in Caithnessand notes that there are two types-long, horned cairns and round cairns; he describesmonumentsin Argyll and says of them'we recognizethe sameessentialfeaturesof constructionandcontents that give typical characterto the Caithnesscairns'; he discussessomeof the Orkney monumentsand found MaesHowe as having 'sometitle to be classedas a local variety'. Anderson'sgeneralpicturewas that all the Scottishchamberedbarrows should be consideredas variants of one structural type 'however much they may differ from eachother in the minor details of externalform or internal arrangement';and that this basic type belongedto the Stone Age. He inferred that the long barrow did not continueinto the Bronze Age, but that the round-chamberedbarrow did. He was quite obviously enormouslyimpressedby the long barrows of Wiltshire and the Cotswolds, comparing the Caithnesshorned cairns with Ablington, Belas Knap, Avening, and Eyford. All, he declared,'exhibit more or less of the same general character',and, 'in view of the excessivelypronounced peculiaritiesof that character,there is no escapefrom the inferencethat they are all the work of one race of men'. And he quoted with great approval the words of ProfessorRolleston that 'the peculiarities of a horn cairn are suchthat it is impossibleto imaginethat they do not indicateto us that one race of men, and only one, must have combinedthem as they are combined.And their geographicaldistribution shows with equal conclusivenessthat of whatever stock that race may have been, they were a homogeneouspeople spreadover the whole areaof Britain'. We shall seelater how much the presentwriter is in sympathywith this Anderson-Rollestonpoint of view which has had little supportin the last quarter century. RobertMunro'sPrehistoric Scotlandwas publishedin 1899 as a general introductionto a seriesof county historiesof Scotland,and we find here the samematerial being gone over as Andersondealt with in his Rhind lectures-Caithness, Orkney, Achnacree,and the Clava tombs, but without any fresh conclusionsemerging,except that the cairns were Bronze Age in date. Munro seemedspeciallyexercisedby the presenceof cremation in Scottish chamberedtombs; in anotherplace he says that 'in the countiesof Argyll, Inverness,Sutherland,Caithnessand the Orkneysit hasbeenconclusivelyprovedthat cremationandinhumationwere carried on simultaneously',and that this showed'that the customof constructing chamberedcairns travelled slowly northwardsand was overtakenby that 45

G. E. DANIEL

of cremation'.Thesewords were first written by Munro in 1913; I quote them from the fifth edition of his book Prehistoric Britain, publishedin the Home University Library in 1928, which I read as a schoolboyinterestedin archaeology.This fifth edition appearedthree years after a book which virtually revolutionized European prehistory-Gordon Childe's Dawn of European Civilisation. In this remarkablebook Childe set the picture of Europeanprehistoryin termswhich no one working in the British Isles had hadreadily availableto them before.He summarized what was really known about megalithic monumentsin westernEurope; no one writing after this book about Scottish chamberedtombs could write in the contextsof Andersonand Munro.1 This is not to say that no substantialadvanceswere madein the study of Scottishmegalithictombs betweenAnderson'sRhind Lecturesof 1882 and 1927, when Childe was appointed the first holder of the Abercromby Chair of Prehistoric Archaeology in Edinburgh. One personin particular had done a very great deal in the way of accuratefield survey and description; this was ProfessorT. H. Bryce, who in a seriesof articlesin the Proceedingsof the Society of Antiquaries of Scotland and in a fine chapter on Sepulchral Remainsin The Book of Arran, provided us with the first full accountof the twenty-odd sites in Arran and Bute of which some, like Carn Ban, EastBennan,Whiting Bay, andSlidderiein Arran, havebecomeas classic in the literature as were the Orkney and Caithnesssites we have already mentioned.2 Bryce saw the spreadof chamberedtombs as a unitary affair from the Mediterranean;indeed, he thought that what he called 'the Iberian origin of the Long Barrow people and chamberbuilders' was a fact which went back to Tacitus's description of the western tribes of Britain. Writing in the style and with the perspectivewith which Rice Holmeswas writing in southernBritain at the time, he saw the chambered tomb builders spreadingfrom Iberia along the west coastsof France,and then, by two routes 'first along the English channel and thenceto the 1 The gap between the Anderson-Munroregime and the ChiIde regime in Scottish archaeologyis the gap which yawnedin Europeanarchaeologyas a whole. In Englandit is the gapbetweenPitt RiversandWheeler,Fox, Chi Ide, and Crawford. In Franceit is the gap betweenG. de Mortillet and Decheletteand the presentday. 2 Bryce's paperson Arran are in Proc. Soc. Ant. Scot., XXXVI (1901-02), 74181; XXXVII (19°2-03),36-37;and XLIII (1908-09),337-70;and on Bute in the same,XXXVIII (19°3-04),17-81. The Book of Arran (Glasgow, 1910) was editedby J. A. Balfour; Bryce'schapteris pp. 33-155'

46

THE MEGALITH BUILDERS

Baltic, and secondover to Ireland and up St George'sChannel.The first streamhardly touchedEnglandbut spreadon to the Baltic; the second spentitself chiefly on Ireland . . . the chamberbuilders camefrom the south-westand spreadover the Hebridesto the PentlandFirth and the Orkneys,and possiblyalso up the GreatGlen to the Moray Firth.'1 This was being written in 1910. Twenty-one years later the new Professorof PrehistoricArchaeologyreada paperto the GlasgowArchaeological Society;it was the first time ProfessorChilde had set out his views about Scottish megalithic tombs in public. It was published two years later and in the view of the presentwriter was one of the most important contributionsevermadeto the theoryof megalithsas well as to our understandingof Scottishmegalithic architecture.2 In this paperChilde set out two principles and the first was a guide to the typology of megalithic tombs in which he said: 'The oldest types will be most accuratelyreproduced in the greatest number of distinct regions; types localized in specific areaswill be later inasmuchas they representregional variants on the original type or types. Moreover, in any given area, it might be expectedthat the older typeswould be concentratedaroundone or more foci while later variants might have a wider distribution'. The second general observationwas that there were in north-westernEurope two principal sorts of plan, 'not necessarilyunrelatedin the long run': these were the passagegrave and what Childe called 'a long cist or covered gallery'-this secondclassis now usually referredto as a gallery grave. The secondgeneralobservationof Childe's was a revolutionary one as far as the study of megalithic tombsin westernEuropewas concerned. Hitherto thesetombs had beenclassified,if at all, accordingto the Montelian threefold systemusually renderedinto English as dolmen, passage grave, and gallery grave (or long cist), and this had beenthought of, not merelyas a taxonomicdevice,but also as a typologicaldevice;the dolmenpassagegrave-gallerygrave series was conceivedof as a sequence.This threefold sequencewas certainly usedto interpretthe Iberian megalithic tombs; it was also applied to the Frenchmegalithic material and is still so applied.3 It was perhapsthe good fortune of the Scottishmaterial that Bryce in (ed. Balfour) TheBookofArran, 98-99. V. G. ChiIde, 'Scottish Megalithic Tombs and their Affinities', Trans. Glas. Arch.Soc.,newser. VIII (1933),121-37. 3 As for example by Arnal, Bull.Soc.Prehist.Franc.,1956, 518; and Giot, Menhirs et Dolmens,Chateaulin,1957. 1

2

47

G. E. DANIEL

muchof it is aberrantand local and doesnot readily fall into the passagegrave and gallery-graveclassic categories,and that there are hardly any 'dolmens'in the Monteliansensein Scotland,sothatthe threefoldsequence was hardly ever applied to Scotland.! Childe suggestedin his 1931 Glasgowlecturethat while the passagegrave and gallery gravemight not be unrelatedin the long run, as far as westernEuropewas concernedthere weretwo perhapscontemporarytraditions,the passagegravefound mainly in Iberia and westernFrance,and the gallery grave in the west Mediterraneanislandsand in southernand non-coastalFrance.2 Applying theseprinciples to Scotland,Childe found passagegravesin Orkney, Caithness,Sutherland,and the Hebrides;and gallery gravesin the south-west-inGalloway, Arran, Argyll, and also the Hebrides.He thought then that the Caithnesstombs were early in the passage-grave series and compared them with tombs in Normandy, Brittany, and southernSpain; he saw their builders spreadingup the Atlantic seaways to the Orkneysand from there acrossto Denmark,definitely rejecting, as did most writers on this subjectfor many years,Bryce's notion of a dual stream from north-westernFrance. Childe then turned his attention to the gallery gravesof south-westScotlandand comparedthem with the gallery gravesof the Pyreneesand Sardinia.He dismissedthe idea that the Clyde-Solwaygallery gravesas he describedthem could be derived from the Irish or Hebrideanpassagegravesandarguedfor a westMediterraneanderivation along the Rhone-Garonneroute. These new ideas of Childe's were developed in his Prehistory of Scotland published four years after the Glasgow lecture, and here he distinguishesfive groups: (i) the long cists as he was then calling them of the south-west,again emphasizingtheir nearestparallelsin the Pyrenees, Catalonia,and Sardinia; (ii) the chamberedcairns of Caithnessemphasizing the parallelsbetweenthe corbelledtombs of Caithnessand sites like Alcalar, Los Millares, and Palmella; (iii) the collective tombsof the coasts and islands of westernScotlandbetweenLoch Etive and Cape Wrath; (iv) the Orkney tombs,most of which he regardedas 'late and specialized variantsof the Caithnessplan'; (v) the stonecircles of Strath Nairn and Strath Spey-the Clava tombs-which he regarded as 'colonists who settledon the Nairn and Spey' by crossingthe Moray Firth from eastern 1 A half-heartedattempt was made to do so for southernBritain by M. C. Burkitt in his Our Early Ancestors,Cambridge,1926, 151. 2 This suggestionwas developedby the presentwriter in Proc.Prehist.Soc., VII (1941),1-49.

48

THE MEGALITH BUILDERS

Sutherland.His summaryof all this was, then, that 'it must be concluded, despiteall seemingimprobabilities,that adventurousvoyagersfrom South Franceand Portugal did land and settle on the Westernand Northern coastsof Scotlandrespectively'.! Fifteen years after the publication of The Prehistory of Scotland, Childe gave anothertalk to the GlasgowArchaeologicalSocietyand surveyedten yearsof researchon megalithictombs.He was now preparedto abandonsome of his earlier ideas, no longer maintaining, for example, the priority of Scottishpassagegravesover English and Irish ones, but on the whole he merely strengthenedhis generalposition, and it is this that is also set out in books which he wrote in 1940 and 1950.2 The south-westScottishgallery graves,togetherwith thosefrom north-eastern Ireland, come from the west Mediterraneanand south France-the Corracloonaportholein County Leitrim and the Beacharraware are now additional arguments,and the passagegraves come from Iberia. Childe is now, however,in theselater statementsmuch exercisedby the fact that so many of the north Scottishpassagegravesare in long, horned cairns and argues for a hybridization between the long cairn-gallery grave tradition and the passagegrave-roundmoundtradition, eitherin Scotland itself, or perhaps,he suggeststentatively,in CountySligo. The most recentand at the sametime the most comprehensivestatement describing and interpreting the Scottish megalithic tombs is that given by Piggott in his Neolithic Cultures of the British Isles, to which referencehas alreadybeenmade. Piggott acceptsthe dual natureof the megalithic colonization of Scotland,and distinguishesin detail between the Scottish passagegraves and gallery graves, but he carries much further than Childe, in ingenious,detailed, and convincing typologies, the idea of hybridizationbetweenthesetwo groups.He seesthe passagegrave builders settling in the Orkneysand in the Clava area,and also in the Hebrides,and in one or two placesin south-westScotland,and he brings them from Spain and Ireland. He seesthe gallery-gravebuilders settlingin south-westScotlandand comingdirect from south-westFrance. By complicatedfusionsof architecturaltraditions,he explainsthe aberrant Th Prehistoryof Scotland,59. The second Glasgow lecture is published in Trans.Glas.Arch.Soc.;Childe's Prehistoric Communitiesof the British Isles was published in 1940; his Prehistoric Migrations in Europe in 1950. For other statementsof his views see ScottishGeog. Mag., 1934, 18-25; Scotland Bifore the Scots, London, 1946; Prehistori6 Scotland (HistoricalAssociationpamphlet115), London, 1940. 1

2

49

G. E. DANIEL

monumentsin the Hebrides,Caithness,and the Orkneys.l Thereis little disagreementbetweenChilde and Piggott and it may be said that the Childe-Piggottaccountof the natureandorigins of the Scottishmegalithic tombsrepresentsthe current,widely accepted,and orthodoxview of these splendid monuments.Can we usefully add anything to the extensive literature of descriptionand synthesiswhich we have surveyedbriefly? Is there anything to say but to repeat in slightly different words the doctrine of the dual nature of the Scottish megalithic tombs, the extent of the developmentin Scotlandof secondarymegalithic types, and the derivation of the two primary types from the main areasof characterization of theseprimary types in south-westernEurope, the passagegrave from Iberia, and the gallery grave from southernFrance.In one respect at least it seemsto the presentwriter that current Scottish megalithic orthodoxy is questionable,and that is the relation of the Clyde-Solway gallery graves to the gallery graves of southernFrance. But before we discussthis point of interpretation,let us summarizebriefly our existing knowledgeof the Scottishmegalithictombs,as that knowledgeis presented to us by Childe in The Prehistory of Scotland(1935) and Scotlandbefore the Scots (1946), by Piggott in his Neolithic Cultures of the British Isles, and by the Inventoriesof the Royal Commissionof Ancient Monuments in Scotlandand excavationreports publishedin the Proceedingsof the Societyof Antiquariesof Scotlandin the last quartercentury. There are upwardsof 360 chambertombs in Scotland,and this figure is to be comparedwith 250 or thereaboutsin EnglandandWales,probably 1,500in Irelandandsome6,000in France.The distributionof thesetombs is shown on the map (fig. 7), and it will be seenat once that distributionally these tombs can be divided into five main groups as follows: (1) south-westScotland, (2) the WesternIsles, (3) north-westScotland, (4) the Orkneys,and (5) the Shetlands.Let us discussthe types of tombs in eachof thesefive groups. 1.

South-westScotland

This group extendsfrom the south end of the Great Glen to the Solway Firth and mainly comprisesArgyll, Kintyre, Bute, Arran, Ayr, Wigtown1 Special attentionshould be paid in assessing Piggott'sviews to Piggott and Powell, Proc. Soc. Ant. Scot., LXXXIII (I948-49), 103-6I, and Piggott, 'Excavations in PassageGravesand Ring Cairns of the Clava Group, 1952-3', Proc. Soc. Ant. Scot.,LXXXVIII (I954-56), I73-207.

50

F'Ig. 7. ChamberedC' alrns In . Scot 1 and.

G. E. DANIEL

shire, and Kirkcudbrightshire. It consists of about sixty tombs. The majority of these are parallel-sided or truncated wedge-shapedlong barrows like Cairnderry barrows, but there are also amorphous-shaped and High Gillespie in Galloway. Just as Thurnamdivided the Wiltshire chamberedlong barrows into two groups, the terminally chambered tombs and the laterally chamberedtombs, so can these Clyde-Solway tombs be so divided. Characteristicterminally chamberedbarrows are East Bennan,Carn Ban, and Whiting Bay in Arran, and Cairnholy I in Galloway. Thesemonumentsfor the greaterpart have their entranceto the east, north-eastor south-east,and the entranceto the chamberis recessedin the barrow by a forecourt defined by a semicircularor arcshapedsetting of stones. The chambersthemselvesare gallery graves broken up into betweenone and five segments;often the orthostatsare imbricated. The laterally chamberedbarrows vary from some in which there still seemsto be someformal elementat the eastend (perhapsone single chamber)and one or more pairs of laterally placedchambers(like Drannandowin Galloway) to tombs in which single rectangularchambers are scatteredaroundshapelessmoundsin an apparentlyhaphazardway. While this is the main part of the story of the south-westScottishmegalithic tombs, there are one or two passagegravesin this area.The White Cairn at Bargrennan,excavatedby Piggott and Powell, is one such, and the chamberson the Water of Deughprobablyanother. 2. The WesternIsles

This group comprises the megalithic tombs of Skye and the Outer Hebrides-theislands of Lewis, Harris, North Vist, Benbecula,South Vist, and Barra. In this group should be included the monument of Achnacreeat the mouth of Loch Etive in the Firth of Lorne. This group consistsof about forty tombs of which eight are in long cairns; four of these, though unexcavated,may well have forecourts like some of the south-west.The normal cairn in this group, however, is circular, but Vnival is square, and Rudh'an Dunain kidney-shapedwith a funnelshapedforecourt. The main type of tomb plan is the passagegrave as at Barpa Langassand Rudh'an Dunain, and these two monumentsmost nearly correspondto the classiccontinentaldefinition of a passagegrave. Elsewhere there are sites which look rather like south-west Scottish segmentedgalleriesand a monumentlike Clettraval in north Vist almost 52

THE MEGALITH BUILDERS

defies classification.Lindsay Scottwould have called it a passagegrave; the presentwriter would haveat one stageclassifiedit as a gallery grave; Piggott says it 'can only be interpretedas a structural hybrid between the passagegraveand Clyde-Carlingfordtraditions'.1This single example of Clettraval shows the difficulty if not the futility of trying to apply a too-rigid classification of megalithic tombs, particularly to what are secondarytomb typesin the sensedefinedby Childe. 3. North-westScotland

This group comprisesSutherland,Caithness,EasterRoss,Inverness,and Nairn, and includes about a hundredtombs. From the point of view of tomb types they vary very considerably.At the north-eastend of the Great Glen, in the valleys of the Beauly, Ness, Nairn, and Spey, to the south side of the Moray Firth are thirty tombs now usually referredto as the Clava group; they are all in round mounds,sevenof them contain passagegraveslike Avielochan and Clava North-East.The remainder,at leasttwenty sites, consistof a circular enclosurewithout any passage.To thesetwo typesamongthe ClavatombsPiggotthasgiven the usefulnames Balnuaran type and Gash type.2 North of these Clava tombs from the Great Glen to the Pentlandfirth are what can usefully be referredto as the Caithness-Cromartychamberedtombs. Sometwenty-six of theseare in long cairns and Piggott would refer to theseas the r arrows type; the remainder,somesixty tombs, are in round moundswhich he would describe as Camsterand derivative types. The long moundsof Yarrows type vary in length from 55 to 240 feet; they are wedge-shapedand have at one end (sometimesat both) semicircularor cuspateforecourtswith the endsprolongedinto horns-hencethe nameof 'hornedcairns' which has been given to these monuments since the nineteenth century. One unusual monument-Ormiegillin Caithness-isin what can only be described,clumsily, as a short-longor squarebarrow. Theselong, horned cairnscan be divided into two typesin the sameway as Thurnamdivided the Wiltshire and Cotswold long barrows, and we have divided the south-west Scottish tombs, namely the horned cairns with terminal 1 Neolithic Cultures, 225. For Lindsay Scott'sviews on Clettravaland on Scottish megalithic tombs in generalsee Proc. Soc. Ant. Scot., LXIX (1934-45), 480-535; Antiquity, 1942,3°1-6;Proc. Prehist.Soc., 195I, 16-82. Z Neolithic Cultures, 258-9, and Proc. Soc. Ant. Scot., LXXXVIII (1954-56), 173-20 7.

53

G. E. DANIEL

chamberslike Yarrows itself, and thoselike CamsterI which are laterally chambered.The roundmoundsof the Camstertype vary in diameterfrom 28 to 75 feet-this latter monumentis Camsterround itself. All these round mounds cover passagegraves. The passagegraves in both the Camsterand the Yarrows group are unusual; they are corbel-roofedCamsterround is 10 feet high-but are broken up by projecting slabs. These slabs break up the passageas well as the chamberinto bays or stalls, and I would like to call the characteristictype of megalithic tomb in the north-eastof Scotlandthe Yarrows-Camster stalled passagegrave. Childe would like to call thesemonumentstripartite passagegraves,but it seemsto me that the essentialpeculiarity of them is not merely that the chamberitself is divided along its length into three sections,but that the whole monumentis stalled.

4. The Orkneys There are some forty megalithic tombs in the Orkneys and they display a great variety of plan. The most magnificent tomb is, of course,Maes Howe to which we have alreadyreferred; it is a fine passagegrave with three side-chambers:other passagegraves are Wideford Hill, Cuween, Vinquoy, and probably the Ring of Bookan. Quoynesshas a chamber lengthenedat right-angles to the passageand has six side-chambers; Quanternessembodiesthe same idea, but is rectangularin plan. The Holm of Papa W estray, whose discovery by Thomas we have already referred to, has a chamberextendedeven further at right-anglesto the passage,and no less than fourteen side-chambers.All the sites so far mentionedhave unimpededpassagesand chambers,but there is also in the Orkneys a group of monumentsstalled as in the Yarrows-Camster manner. Taversoe Tuack and Unstan are good examples of Orkney stalled passagegraves;the stalling exists only in the chamberitself, and thesechambers,like the other Orkney passagegraves,are T -shapedand have the chamberextendedat right-anglesto the passage.Blackhammer and Midhowe are remarkableexamplesof very long stalled chambers. These sites, like the Holm of PapaWestray, are set in unhornedlong barrows. The Blackhammerchamberis enteredlike the Holm of Papa Westrayby a short passagefrom the long side of the mound;Midhowe is enteredby a passagefrom the short end of the long mound. Two of the Orkney monumentscall for special attention. One is 54

THE MEGALITH BUILDERS

TaversoeTuack, on Rousay, which Marwick has described as 'after Maes Howe . . . perhapsthe most remarkableof all the Orkney cairns'.1 It was excavatedby Grant in 1937 and presentsthe very unusualphenomenonof two passagegravesin one mound placed oneabovethe other, the lower chamberenteredfrom the south-eastand the upper from the north-west.2 The other monumentis the Dwarfie Staneon Hoy-a large block of sandstonewithin which a passageand two side-chambershave beencut. If this is a rock-cut collectivetomb, as is widely held, it is almost the only such tomb recognizedin north-westernEurope, though it is worth rememberingthat there are rock-cut featuresin other megalithic tombs,suchasSamson12 in the IslesofScilly, andthat thelower chambers at TaversoeandHuntersquoyarerock cut.s

5. The Shetlands This final group of Scottish tombs comprisessixteen sites in the Shetlands. With one exceptiontheseall belong to a type of monumentwhich Bryce, who first studied these tombs in detail, christened'heel-shaped' cairns.Piggott would divide theseinto two types, the first, which he calls the Punds Wafer type, consistsof a cruciform chamberwith a passage openingout a slightly concavefaftade; thesemonumentsarepassagegraves in round moundsincorporatedin a heel-shapedor oval mound. Piggott's secondtype is the Muckle Heog type; it has the same heel shape,but insteadof a passagegraveit coversclosedcists.4 One thing standsout very clearly from a closestudyof the distribution of the Scottishmegalithic tombs; their distribution is basically a coastal and riverine one and can only meancoastalsettlementsand penetration up the rivers like that from Dornoch Firth up to Loch Shin or up the Speyto Avielochan.We do not really needa wide comparativeknowledge of megalithic monumentsin Europe as a whole before we can interpret Hugh Marwick, Ancient Monumentsin Orkney, Edinburgh, 1952, 15. On TaversoeTuack,seeW. G. Grant,Proc. Soc.Ant. Scot.,LXXIII (1938-39), 155-66. Huntersquoy on Eday is another two-story chamber tomb (Orkney 1

2

Inventory,56-59).

3 On the Dwarfie Stanesee C. S. T. Calder,Proc. Soc. Ant. Scot., LXX (193536), 2I 7-22. The site known as St. Kevin's Bed, Glendalough,in Ireland hasbeen interpretedby Hemp (Antiquiry, 1937, 348-50) and O'Neil (AntiquariesJournal, 1947, 182-3) asanotherrock-cutcollectivetomb. 4 On the Shetlandtombs seeBryce, Proc. Soc. Ant. Scot., LXXIV (193g-40), 23-35,andPiggott,Neolithic Cultures,262-3. For the Stanydale'temple',seebelow.

P.P.S.-E

55

G. E. DANIEL

the Scottishmegalithic tombs: the map alone tells us that their builders are people who came by sea to the isles and coastalplains of Scotland. Often, as Childe first analysed,we can seethem settling on raisedbeach platforms or alluvial gravels adjacentto points of entry on the shore; he noted that on Rousay 'each tomb correspondsto a natural agricultural unit, generallystill or till recentlyfarmedby a communityand comprising in each casea stream,a strip of good arable land below, and a tract of pastureabove the tomb'.1 The actual settlementsites of the people who buriedtheir deadin the Scottishmegalithictombsare as rare as megalithic settlementsites are in westernEuropeas a whole. There is a settlement site on the slopes of Wideford Hill just below the passagegrave. The potters' kilns of Eilean an Tighe on North Dist excavatedby Lindsay Scott may well be thoseof the Hebrideanmegalith builders. The settlement site at Rothesayin Bute had pottery suchas is found in someof the south-westScottish collective tombs, but Piggott regardsthe dominant culture here as of Rinyo-Skara Brae type, which he would classify as SecondaryNeolithic. The notion of the SecondaryNeolithic Cultures of the British Isles as set out by Piggott is that they representthe acculturation of Neolithic colonists and Mesolithic indigenous hunter-fishers,as ProfessorAtkinson has discussed(p. 22). SkaraBrae on the mainlandof Orkney and Rinyo on Rousay,while formally classifiedas the settlement sites of theseSecondaryNeolithic people must surely owe somethingto the megalithic colonists,and it is certainly arguablethat the villagers of SkaraBrae and Rinyo buried their deadin the Orcadiancollective tombs. A vesselin the style describedby LindsayScottas Rinyo I was found with the latestburials in the Dnival chamberedtomb in the Hebrides,and the remarkablesherdof pottery from SkaraBrae decoratedwith a design of double spirals and lozengesis in the tradition of the art on the Boyne megalithic tombs. Our detailedknowledgeof the way of life and the material culture of the megalith builders must, however, in the main be drawn from the tombsthemselves,that is to sayfrom the materialfound during excavation, and this information is naturally limited by the amountof excavationthat has takenplace.About seventyof the 360 Scottishmegalithictombs have beenexcavatedbut material remainssurvive from only forty tombs. The material, apart from pottery, consists of stone axes, leaf-shapedflint arrowheads,plano-convex knives, hollow scrapers, flint knives with 1

ScotlandBifore theScots,34.

56

THE MEGALITH BUILDERS

polishededges,tranchetarrowheads,cylindrically perforatedstonemaceheads, stone discs, and beadsof jet and stone. No metal objects have been found in primary burials in any of the Scottish collective tombs, though, as Childe has pointed out, some of the stone types found in Scottishmegalithic tombs have beenfound in Yorkshire associatedwith bronzeweaponsin individual graves.1 The pottery from the Scottish collective tombs has been studied in detail by Piggott in his Neolithic Cultures. We may note the existenceof the following main varieties: (1) undecoratedround-bottomedbowls of what can most convenientlybe called WesternNeolithic pottery,2 (2) undecoratedhemisphericalbowls with shouldersand concaveand Hared-out necks which Piggott proposesto call Lyle's Hill ware, (3) decorated carinatedbowls with a rim of less diameterthan the shoulderand with ornamentin shallow channellingor incision, which Piggott would now call BeacharraB ware, (4) much the samekind of pottery, but with the decorationdone by impressedtwisted or whipped cord such as is not found in the WesternNeolithic traditions,andwhich is Piggott'sBeacharra C, (5) shallow open bowls with a vertical collar usually ornamentedby channellingor stab-and-dragtechniquewith oblique or horizonta~ there lines, usually referredto as Unstanware. In addition to thesefive waresthere also occurs some sherdsof Beaker,Peterborough,and Rinyo wares, but usually in late primary or secondarycircumstances.The amount of Lyle's Hill ware in the Scottish collective tombs is very slight indeed, althoughit occursoften in the analagoustombs of north-easternIreland. It occurredin the blocking of Cairnholy I, but only a single bowl and in the final period of the tombs funerary use;it occurredin the Eastertonof Roseislesite in Morayshire, if this is a collective tomb, and there are 3 There is somecorrelationbetweenthe various sporadicfinds elsewhere. types of collective tombs and the pottery types; the BeacharraB ware occurs only in south-westScotland and the Hebrides, and the Unstan ware in the north-eastof Scotlandand the Orkneys.4 The material evidence for the economy of the chambered-tomb PrehistoryofScotland,23. This includes Piggott's BeacharraA pottery and the undecoratedroundbottomedbowls from Orkney, such as the two Unstan pots figured by him in NeolithicCultures,249. 3 Callender, Proc. Soc. Ant. Scot., LXIII (1928-29), 29-98; Piggott, Neolithic Cultures, 170. 4 No sherdssurvivefrom the Caithness tombs.SeePiggott,NeolithicCultures,248. 1

2

57

G. E. DANIEL

builders in Scotlandhas been well analysedby Childe in his Scotland before the Scots.They were, he suggests,mixed farmers who huntedand fished. A typical saddlequernwas found in the Rothesaysettlement;and an impressionof a grain of hulled barley on a sherdfrom Unstanand of a grain of naked barley on a sherd from Eday. Analogous tombs to the south-westScottishmonumentsin Northern Ireland and the Isle of Man haveyielded small spelt, emmerwheatand other impressionsof Triticum monococcum-dicoccum type.1 The domesticanimalsinclude cattle, sheep, pig, and a fairly large numberof the cattle bonesbelongedto immature animals.ProfessorWatsonhas noticed this at SkaraBrae and said then it was probablydue to 'the difficulty of providing winter forage to allow all, or even a large proportion of, the calvesto be carriedon until the grass grew again in the spring'. Childe argues that hunting played a more importantpart in the economyof the megalithbuildersthanit did in, say, the villagers of Skara Brae; the animals huntedinclude red deer, horse and wildfowl. Fish bonesand birds' eggsare recordedfrom Midhowe and shell fish from Yarso and Lower Dunreay.2 We now return to the questionthat much exercisedthe early writers on Scottish megalithic tombs, and is beggedall the time when we talk about the colonizationof Scotlandby the megalith builders: the question of origins. Where did the megalithbuilderswho settledin Scotlandcome from? Although in recentyearswe havemovedaway from a monogenetic theory for the origin of megalithic monumentsin Europe,3 no one has seriously argued that the Scottish megalithic monumentsoriginated in northernBritain, and most peopleassumethat they representa movement of people from south-westernEurope along the Atlantic seawaysfrom Iberia and France.4 The questionreally is by today, Was thereone movement or many, and where preciselywas the home of theseearly settlers in Scotland?To attempt an answer we must recollect the catalogueof types of megalithic tomb that have beendistinguishedin Scotland.Here 1 K. Jessenand H. Helbaek,Cerealsin Great Britain and Ireland in Prehistoric and Early Historic Times,K. DanskeVidensk.Selskab,1944. 2 For a fuller accountofthe economyof the collective-tombbuildersseeChilde, ScotlandBeforethe Scots,34 ff., 'The Megalithic Society'. 3 Largely as a result of C. A. Becker's reinterpretationof the Danish dysser, which hasat lastremovedthedifficulties felt by so manyin derivingthesemegalithic tombsfrom westernEurope. 4 But G. F. Willmot arguesfor a northernEuropeanorigin for the megalithsof Britain and France, and on the easternelement in our Neolithic, see Piggott, Proc. Prehist.Soc., 1955,96-101.

58

THE MEGALITH BUILDERS

is a minimum list of types that we must distinguish if we are going to engagein any useful discussionof origins: (1) the terminally chambered long barrowsof south-westScotlandof the Cairnholy-EastBennantype; (2) the laterally chamberedbarrows of south-west Scotland; (3) the passagegravesof south-westScotlandof the Bargrenan-Waterof Deugh type; (4) the passagegravesof the WesternIsles; (5) the aberrantpassage gravesor passagegrave-gallerygravehybrids of the WesternIsles; (6) the Clava passagegravesof Piggott'sBalnuarantype; (7) the Clava tombs of the Cask type; (8) the Camsterstalled passagegravesin round mounds; (9) the Yarrows stalled passagegravesin long mounds; (10) the laterally chamberedlong moundsof north-westScotland;(11) the Orkney passage gravesin the MaesHowe tradition; (12) the Orkneystalledpassagegraves; (13) the Punds Water type of Shetlandheel-shapedcairn; and finally (14) the Muckle Heog type of Shetlandheel-shapedcairn. No discussion of the origins of Scottish megalithic tombs can be profitable unlessit beginsby recognizingthe existenceof thesevarieties, and their classification is a matter of objective taxonomy. But is this taxonomyany more than an ingeniousacademicexercisein nomenclature and classification?Does it meananything?It beginsto look as if it does when we apply, a little more rigidly than he did himself, the principles of interpretation which Childe laid down in his 1931 lecture to the GlasgowArchaeologicalSociety.Let us examinethe list of fourteentypes in the light of Childe'sprinciple that 'typeslocalizedin specific areaswill be later inasmuchas they representregionalvariants'.Applying this, and rememberingthe provenexampleof the successionof lateral-andterminalchamberedlong barrows in the Severn-Cotswoldregion, we can at once remove from the roster of primary megalithic types in Scotland types (2), (5), (8), (9), (10), (12), and (14). Few archaeologistswould object to the removal of the laterally chamberedtypes, namely (2), (10), and (14); and probably by now very few the monumentsof type (5), which are patently variants and hybrids and part of the secondarydevelopmentof megalithicarchitecturein the WesternIsles. It might be that a few would cavil still at the removalof types (8), (9), and (12), but it must be emphasized that these monumentshave no exact parallels outside northern Scotland; the whole principle of subdividing the chambersinto stalls shouldbe regardedas a Scottishdevelopment. We are left then with seven types of Scottish megalithic tombs as representingthe possible primary settlementof Scotland, at least on a 59

G. E. DANIEL

reasonablebasis of argument.These seven types may be conveniently groupedtogetherinto four classesposingfour separatequestions:

(a) (b) (c) (d)

the primary Scottishpassagegraves,namelytype (3), (6), and (11) the primary gallery graves,namelytype (1) the Cask type of Clava tomb, i.e. type (7) the Punds Water type of Shetlandheel-shapedcairn

Let us dealwith thelasttwo questionsfirst. Bryce! andCalder2haveargued for a direct derivation of the Shetlandtombs of Punds Water type from the Mediterranean.Piggott,while recognizingthat thesetombsarewithout parallel in the British Isles, regardsthe perfectly valid comparisonswith Mediterraneantypes such as Balearic nave/as and Maltese temples as comparable,but having 'probablyno more validity than a recognitionof the essentialunity of the collective tomb tradition in western Europe where certain ritual elements-chambers, passages,fas;ades,forecourtsrecurin variouscombinationin almosteveryareacolonizedby the builders of thesetombs'.3This statement,which I havedeliberatelyquotedin full, and which can be referred to convenientlyas the law of the recurring combinationof basic features in megalithic tomb architecture,sets out clearly a principle as importantand, to all presentappearances, as correct as the two principles which Childe adumbratedin his 1931 Glasgow lecture. It is a principle which hardly needsstressingto archaeologists well steepedin the theoretical battles of comparativeethnology, but it does need constantrestating to archaeologistswho have not learnt the long and repeatedlessons of the great anthropological controversies betweenthe diffusionists and the independentinventionists.It is in its essentialsthe principle that all cultures and all versionsof a particular culture have an overall assemblageof traits and that where a culture spreadsthesetraits may crop up in different placesand at different times, but that exact and direct connexionbetweentwo cultures can only be plausibly argued where there exists the exact reappearanceof several linked traits between areaswhich could be connected,and connected within possible chronological limits. The work of studying megalithic tombs has often been retardedextensivelyby archaeologists,who with the bestwill in the world haveseizedexcitedlyon only onetrait amongthe Proc. Soc.Ant. Soc.,LXXIV (1939-40),23-36. C. S. T. Calder,Proc. Soc. Ant., Scot.,LXXXIV (1939-40),185-205. 3 NeolithicCultures,2 63.

1

2

60

THE MEGALITH BUILDERS

many that may be included in the planning, construction,and use of megalithic tombs, and argued for intercontinentalconnexionsbecause portholesare found in the Cotswolds,the Paris Basin, and the Deccan, or semicircular forecourts in northern Ireland, northern Scotland, and Sardinia.1 We perhapsneedto stateclearly herein the presentmegalithic context yet another principle supplementaryto those of Childe and Piggott and growing out of them, that unless we can show the exact correspondencebetweenseveral material traits in two distinct areasof megalithic-tombbuilding we shouldbe very charyof arguingfor a genetic relationshipbetweenthesetwo areas.This principle, which is the commonplaceof sound argumentin the diffusion v. independentinvention controversy, arises here when we discuss the Punds Water type of monument.It will arise again when we discussthe other types. Here, as regardsthe Shetlandmonuments,the presentwriter is entirely with the view implicitly expressedby Piggott in the passagequoted from his Neolithic Culturesof the British Isles. Thereis nowhereoutsidethe British Isles where the Punds Water type of monumentcan be exactly paralleled, and it seemsto me that we must regard this type and therefore all the Shetlandmegalithsas local regionaldevelopmentsfrom traditions already existing in the north of the British Isles. What ofStanydale,that interestingShetlandmonumentwhich we have avoideddiscussinguntil now?It was excavatedin 1949 by C. S. T. Calder and consistsof an oval building with a shallow crescenticforecourt enclosing an area40 feet by 22 feet, approachedby a short passageleading from the crescenticforecourt; the areaitself containssix shallow recesses. Two large post holes containing the burnt remainsof two spruceposts2 10 inches in diameter were found. The structure was almost certainly roofed by a timber ridge roof and probably looked like the conjectural restorations prepared by Mr. Calder. Flat-based pots and sherds of Beaker were found during the excavations.The excavatorthought that Stanydalewas not a tomb, nor a house, and suggestedthat it was a religious building, and that the Stanydaletemple, as he calls it, was directly related to Mediterranean megalithic architecture. Stanydale, 1 This is the occupationaldiseaseof archaeologists and anthropologistswhich G. P. Murdock (Africa: Its Peoples and their Culture History, 1959, 40-41) calls 'trait-chasing'. 2 Spruceis not known to be a native tree in Scotland,and it is presumedthat the buildersofStanydaleuseddriftwood from North America. SeeScott,Antiquity, 1951,151-3.

61

G. E. DANIEL

Calder argues,is 'a temple of Mediterraneanlineage'; so pronouncedis the resemblancebetween Mediterraneanstructures and this Shetland monumentthat, in his view 'it is almostimpossiblenot to assumethat the Maltese templesare the prototypesfrom which Stanydaleis derived and which solve the question of its purpose'.lNow, in discussingCalder's interestingviews let us be clear on two points: first, there is no inherent reasonwhy thereshouldnot be direct contactbetweenthe Mediterranean and northernScotlandwithout leaving any direct trace in between,and secondly,we have probably for very long beentoo rigidly bound by our terminologyof megalithic monumentsin calling all the roofed chambered structurestombs.Justas our Christianchurchesare often full of tombsso it may well be that what we call megalithic tombs were also temples,or that sometimes structures which were initially primarily funerary in purposelater gave rise to monumentsthat were non-funeraryin purpose andfunction. This is how ProfessorJohnEvanshasexplainedthe development of the Maltese temples from collective tombs,2 and it has always seemedto the presentwriter that someof the great north Frenchgallery graves like Esse and Bournand, and Bagneux, were never functioning tombs,3 and the sameidea is bound to occur when we discussthe Irish monumentsthat contain central open courts-thelobster-clawtombs of Mahr and the court cairns ofR. de Valera. Surely, Piggottwas right when he describedStanydaleas'relatedarchitecturallyto the heel-shapedcairns' of Shetland,4 and would regardit, in the light of our presentcomparative knowledgeof megalithic monuments,as a local developmentin Shetland among the people who built the heel-shapedcairns, just as he would regardthe court cairns of Ireland as a local developmentfrom the ClydeCarlingford tombs. In both casesthe developmentmay have carried with it a changeof function. Our problem of origins is, then, reducedto three questions.Let us now examinethe problem of origin of the Clava tombs of Piggott'sCask type, the so-called ring cairns. Many writers have regarded these as degenerateexamplesof the Clava passagegraves of Balnuaran type; Piggott'srecentsurveyof the Clava tombs showsthat the Balnuaranand Gasktypes are roughly equalin number-therebeing ten passagegraves Proc. Soc. Ant. Scot.,LXXXIV (1949-50),203 and205. Proc. Prehist.Soc., 1953,41. 3 ArchaeologicalJournal, 1956, 11. 4 Neolithic Cultures,263.

1

2

62

THE MEGALITH BUILDERS

to nine ring cairns. He arguesthat the degenerationtheory is hard to sustain and quotes the comparison made by the present writer and T. G. E. Powel}! betweenthe Gask-typetombs and analogousstructures in Almeria publishedby the Leisners.2 It seemsto me that the Gask ring cairns are an integral part of the Clava group of tombs and cameto that area of Scotlandas part of the movementof settlerswho built the Balnuarantype passagegraves. Our problemof origins is then reducedto the two issuesof the origin of the passage-grave buildersand the origin of the gallery-gravebuilders. The primary passagegravesare thoseof the MaesHowe type and thoseof the Balnuarantype. The Bargrenanpassagegravein south-westScotland is probablyto be connected,as Piggott and Powell argue,with the undifferentiated passagegraves of the Scilly-Tramore group and with the curious tradition that producedmonumentslike Five Wells in Derbyshire.3 The Maes Howe type, despitethe originality of its architecture,is in plan in the tradition of the cruciform passagegraves of the Boyne culture of Ireland, itself ultimately apparentedto the passagegravesof southernPortugal.It is in this connexionrelevantto recollect the spirals on Eday, the oculi ornamenton the Holm of Papa Westray, and the lozengesand spirals on the sherd from Skara Brae. The origin of the Balnuaranpassagegravesof Clavais morecomplicated;the presentwriter and T. G. E. Powell suggestedthat they wereamongthe bestexamplesof primary passagegraves in the British Isles;4 they should be compared with lIe Carn, Les Sept-lIes, La Sergente,Yvias, lIe Longue, and BarnenezSouth C and D in Brittany, with Fontenay-Ie-Marmionin Normandy,1iand with the classictombs in the Almerian cemeteryof Los builders Millares and may representa separatemovementof passage-grave comparativelyearly in the generalspreadof passagegraves.Justhow this movementtook place is anothermatter; it might have beenup the Great Glen, or roundthe north of Caithness,or possiblyup the EnglishChannel from Brittany, Normandy,and so up the eastside of Englandand Scotland. Whicheverway it happenedit left no comparablecolonieson theway. Proc. Prehist.Soc.,1949, 16g-81 . G. andV. Leisner,Die Megalithgraberder IberischenHalbinsel,-DerSuden, 1943, plates1 to 7. 3 Proc. Soc. Ant. Scot.,LXXXIII (1948-49),152. 4 Proc. Prehist.Soc.,1949, 16g-87. 6 Daniel, The Prehistoric Chamber Tombs oj France, 1960, chaptersIII and IV, passim. 1

2

63

G. E. DANIEL

We come now, finally, to the problem of the south-west Scottish gallery graves like East Bennanand Cairnholy and the problem of the origins of thesetombs, like the problem of the origin of the Carlingford tombs, strict analoguesof the south-westScottishterminally chambered long barrows, is one about which much has been written in the last twenty years,and violently opposingviews held. One school soughtthe answer in the horned cairns of Sardinia, which admittedly have some points of superficial resemblanceto the Clyde-Carlingford tombs, but now that we have fuller knowledgeof theseSardiniantombs we seethat the resemblances are superficialand can easily and properly be explained accordingto the law of the recurring combinationof basic features,and that in any case none of these Sardinian tombs is early enough to be ancestralto our Scottishand Irish tombs.1 Childe and Piggott, as we have seen,see the answerto the origin of the Clyde-Carlingfordtombs in southernFrance.It is perfectly true that hereand in Cataloniawe get someof the featureswhich recur in different combinationsin Scotland;there are segmentalslabs breakingup someof the Frenchgallery graves,though La Halliade which is most often cited, is not a happy parallel-it is rather sevenmegalithic cists set together. There are a few south French collective tombs with semicircular forecourts; Henri Martin Granel discoveredthe first on the edge of the HeraultandGard, andothershavebeenfound in Heraultand Gard. They are of two types, first, monumentslike Bois Martin with a peculiar and aberrant-shaped megalithic chamberlate in the sequenceof southFrench tombs, and secondlyutilized natural cavesor rather faults in rocks made by karstic erosion, roofed with slabs and precededby a semicircular setting of stones;a good exampleof this secondtype is Ratoul between Montpellier and Ganges.2 There are, of course, plenty of examplesof gallery gravesin southernFranceand many of theseare coveredin long barrows, sometimes the short-long barrow which appears to have a functional relationshipto the long tomb it is covering and sometimesa barrow elongatedfar beyond structuralnecessities.This constructionof a long barrow for the sakeof the barrow, itself now a visible monument 1 For a modern treatment of the Sardinian material see M. Pallottino, La SardegnaNuragica, Rome, 1950, and Zervos, La Civilisation de la Sardaignedu debut de L' Eneolithiquealafin de la periodenouragique,Paris, 1954. 2 Martin and Arnal, 'Les Tombes a antennesdu Bas Languedoc',C. R. First

Congres d'Etudes Ligures (1950), Bordighera, 1952 and Pannoux and Arnal, in Atti Gong. Int. Preistoria eProtohistoricaMed. (1950), 155-78. 64

THE MEGALITH BUILDERS

of departedfriends and relatives,is not a basic feature of the collective burial ritual of the earliest megalith-tomb builders in France, but it becomesso, and both in the passage-gravetradition in north-western France (BarnenezNorth, BarnenezSouth, Motte de la Garde, Bougon, single chamber Fontenay-Ie-Marmion)andin the gallery grave-rectangular tradition we find the growth of the barrow. This importance of the barrowis essentialin understandingour British megalithictombs,because the prehistoric British megalith-builders excelled in building large barrows,from New Grangein Ireland to West Kennet and East Kennet in Wiltshire, andYarrows and CamsterLong in north Scotland.1 Then, of course,in addition to thesefeaturesof tomb morphologytheredoesoccur in southernFrancechannelledware which Bryce first comparedwith the BeacharraB ware,2 and which comparisonhas beenfruitfully elaborated by Childe,3JacquettaHawkes,'and Piggott.5 We can point to no one area, no one group of tombs in southern France where all the features of the Clyde-Solway primary tombs are found. Piggottseemswell awareof this whenhe saysthat 'it seemsimpossays rectangular(sometimes sible . . . to derive the combinationof fa~ades, segmented)chambers,and long cairns (sometimestrapezoidal)from any one Europeansource',6and so doesHawkeswhen he proposedin 1940 a fusion of Almerian cist-burialswith the passagegravesof southernSpain area and a spreadof thesehybrid types via the Catalan-Pyrenean-Basque to the Clyde-Carlingford area.7 By today we know the date of the simple rectangularchambersof the Catalanand BasquePyreneanregionS makestheir role as parentsof the Clyde-Carlingfordtombschronologically impossible;they, like the 'dolmens'of Malta and north Portugalare late, not early in the sequenceof tomb types. How can we then resolve this apparentimpassewith regard to the origin of the south-westScottishchamberedtombs, which clearly represent, as Childe argued,a primary settlementfrom outside,and yet seem 1 On long barrowsandgallery gravesin southernFranceseeG. E. Daniel, Ant. Joum., 1939, 157; Proc. Prehist. Soc., 1941, I; (ed) Dickins and Fox, The Early CulturesofNorth-WestEurope (1950), 3; Arch.Joum., 1955, I. 2 Proc. Soc.Ant. Scot.,XXXVI (1901-02),74-181. 8 Arch.Journ.,1931,37-66. 4 Arch. Joum., 1939, 126-73. 5 Neolithic Cultures, 170 ff. 6 Piggott,Neolithic Cultures, 186-7. 7 c. F. C. Hawkes,The PrehistoricFoundationsofEurope (1940), 167. 8 On thisseePericot,LosSepulcrosMegaliticosCatalanes y la Cultura Pirenaica,I 950.

65

G. E. DANIEL

to have no direct continentalanalogues?I think the answerlies with the earlier antiquaries who stressedthe comparison between the horned cairnsof Caithnessand the monumentsof south-westScotlandon the one hand and the horned long barrows of the Cotswolds on the other. We havein the Severn-Cotswoldculture of southernBritain a vigorous group of chamber-tombbuilders whose first appearancein south Glamorgan, Wiltshire, and Gloucestershiremust be surely datedwell before2500 B.C. The tombs of thesepeoplecan be paralleledin southernBrittany and in the Vendee;herewe find long barrows,gallery graves,and gallery graves with pairs of transepts-andit matters nothing in our presentcontext whether,as the presentwriter has argued,thesetranseptedgallery graves derive from the gallery gravesof southernFrance,or as many othershave argued,are forms of passagegraves,or perhapscome from types like El Pozueloin southernSpain.l Here, in westernFrance,is a context,culturally andchronologically,that could havegiven rise to the Severn-Cotswold culture. It seemsto me likely that the Severn-Cotswoldmegalith builders moved from their initial landfall on the shoresof the Bristol Channelup St. George'sChanneland into the Irish Seauntil they reachedsouth-west Scotlandand north-eastIreland and eventuallythe WesternIsles of Scotland. We have been accustomedto think of sites like Carn Turne and PentreIfan in Pembrokeshire,Trefignathin Angleseyand CapelGarmon in Denbighshire,BallynamonaLower in County Waterford, and King Orry's Grave and Cashtalyn Ard in the Isle of Man as outlying colonies of the Clyde-Carlingfordculture, insteadof, as now seemsto me likely, stageson the way to the colonizationof the north Irish Seaby the SevernCotswold builders, who ultimately came from western France. The forecourtsat PentreIfan andCarnTurneareperhapsintermediatebetween the Severn-Cotswoldforecourtsand thoseof the early Clyde-Carlingford West of West Kennetas revealedby the excavationsof a tombs and the fa~ade few years ago is in plan closer to Scotland and Ireland than are the cuspateforecourtsof someof the Severn-Cotswoldtombs. It may well be that the Grey Mare and Her Colts near Porteshamon the Dorset coast representsan early stagein what I would now like to call the Severn-Clyde culture. Is it too fanciful to see Breton merchantventurerssetting out 1 There are, admittedly no known examplesof trapezoidbarrowswith chamberedtombsin themin Brittany, but few chamberedmoundsin Brittany havebeen properly excavated.There exist trapezoidunchamberedmounds (see S. Piggott, Antiquiry, 1937,441).

66

THE MEGALITH BUILDERS

from the Morbihan and the Loire Atlantique and the lIe d'Yeu, having a first landfall in Dorset,then settling on the shoresof the Bristol Channel and on to south-westScotland?At presentthis seemsto me the only working hypothesisto explainthe origin of the Clyde-Carlingfordculture, in reality an essentialpart of the chamberedlong barrow culture of southernEngland.Of course,this doesnot meanthat the Clyde-Carlingford culture could not have been,or indeed was, in direct contact with southernFrance or Iberia at some time; how else can we account for BeacharraB ware? But there were many other contactsand their complexity should not obscurewhat seemsto me at presentthe most likely fact of their origins.l We have, then, in discussingthe megalithic settlementof Scotland, various groupsof passage-grave builders and one group of gallery-grave builders apparentedto the Severn-Cotswoldpeople. When did these colonizing movementstake placeand how many peopledid they involve? Scotland gives no direct indication of date in absoluteyears. We have evidenceof comparativedate becauseat Unival, Clettravaland Rudh-anDunain the primary (or the initial primary) use of the tombs was preBeaker,and the samesort of chronologicalconclusioncan be drawn from the evidenceof Nether Largie and Cairnholy I and II. Altogether,of the sitesexcavatedin Scotlandthat yielded archaeologicalmaterialelevenhad Beakersassociatedwith their latestuse. It is clear then, that the megalith builders were well establishedin westernand northernScotlandbefore the arrival of Beaker-usingpeoples.For absolutedates we must look to the evidencefrom France, Ireland, southernBritain, and the Mediterraneanin general.It seemslikely that someof the Scottishmegalithsdated from 2500 B.C. or before,but that the use of some of them went on to the middle of the second millennium B.C. The pumice pendant from Unival describedby Scott is in the form of a metal axe; he comparedit with the one from the gallery grave of Kerlescanin the Morbihan, but Piggottnotesthat it is muchnearerin shapeto the flat metalaxesfound in the Breton dagger graves which date from 1500-1300 B.C. An overall bracketfor the Scottishmegalithictombsof from 2500 to 1300 B.C. would seemto agreewith the presentevidence.It must not, of course,be thought that eachtomb was in use for this long period, though it is not unlikely that anyonetomb might have beenusedfor five or six generations.Nor 1 For discussionof thesemattersseeDaniel, Proc. Prehist. Soc., 1939, 143; Morgannwg,1957,3; The PrehistoricChamberTombsofFrance, 1960.

67

G. E. DANIEL

must it be thoughtthat the tombsrepresentone movement;Scott argued for successivemovementsof people extendedover perhapsthree centuries,l and this must surely be so to explain the different groups of passage-grave builders,and the gallery-gravebuilders with their separate southernFrenchtrade connexions. It is extremely difficult to try to estimatethe numbersof people involved in thesecolonizing movements.J. F. S. Stonehas estimatedthat the total population buried in the south British long barrows is about 1,600,3 but we have no idea what proportion of chamberedtombs have survivedfrom antiquity, nor what sectionof a communitywas so interred. The number of individuals found buried in Scottish megalithic tombs varies from one at Corrimony, surviving as stainson the sandsurfaceof the floors to twenty-five in Midhowe, twenty-ninein Yarso, and thirty in Ormiegill. According to Childe's calculations the average for eleven tombs is twelve,4 but it is interestingthat the largesttombs do not necessarily contain the greatestnumber of individuals. All we can be sure of as we think about the megalithic tombs of Scotlandis that the communities that built them werelarge enoughto permit the deploymentfrom time to time of the considerableamount of labour involved in constructing them and heapinglarge barrowsaroundand on top of them. We must now turn in conclusion,and briefly, to the stone circles of Scotland,which have not hitherto perhapsreceivedthe attentionwhich the megalithictombs havehad. Thereare very many of themin Scotland, and they fall fairly easily into two categories,those without encircling banks and ditches, and those with such encircling banks and ditches which would in southern Britain be classified as Henge Monuments. The distribution of thesetwo classesof monumentis shown on the two maps(figs. 8 and9). Many of the first class of monumentsurrounda central burial which may be marked by a barrow; but more often there is just a central cist such as the sites excavatedon Arran by Bryce.5 They vary in diameter from 17 feet to 45 feet, and someof the stonesareas muchas 10 feet high. Most are single circles, but one on Mauchrie Moor consistedof two, and Proc. Soc.Ant. Scot.,LXXXV (1950-51),18. WessexBeforetheCelts (London), 1958,51. S Piggott,Proc.Soc.Ant.Scot., LXXXVIII (1954-56),182-3. 4 ScotlandBeforethe Scots,37. ~ Proc. Soc. Ant. Scot.,IV, (1954),499;BookofArran, 113.

1

2

68

THE MEGALITH BUILDERS

theremay havebeena twin circle encircling a barrow at Newbridgewest of Edinburgh.1 The date of theseunembankedstonecircles rangesfrom 1700 to 1800 B.C. through to 1000 B.C.; the gravesassociatedwith them have yielded pottery from Food Vesselsto CineraryUrns. The secondclassof stonecircle, the hengemonuments,were studied by J. G. D. Clark when publishing his excavationsat Arminghall in Norfolk,2 and later by R. J. C. Atkinson in publishing the Dorchester 3 Clark listed three sitesin Scotland-the (Oxfordshire)monuments. Ring of Brodgar and the Ring of Stennis in Orkney and the Broomend of Crichie site in Aberdeenshire.To thesethree sites Atkinson was able to add anotherfour, namely Overhowden,Oxton, Berwickshire,recognised by S. and C. M. Piggott as a henge monumentin 1949; Cairnpapple, excavatedby Piggott in 1947-48; Ballymeanoch,Argyll, excavatedby Greenwellin 1864, and the BroadleeMiddlebie, Dumfrieshire,discovered by Dr. J. K. S. St.Josephby air photographyin 1947. To this list four more monumentshaverecently beenadded,all in EasterRoss.