roman helmets & finds in syria

Upload: damiano-laurenzo-zamprogno

Post on 01-Jun-2018

219 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/9/2019 Roman Helmets & finds In Syria

    1/29

    Simon James

    Evidence from Dura Europos for the origins of late romanHelmetsIn: Syria. Tome 63 fascicule 1-2, 1986. pp. 107-134.

    Citer ce document / Cite this document :

    James Simon. Evidence from Dura Europos for the origins of late roman Helmets. In: Syria. Tome 63 fascicule 1-2, 1986. pp.107-134.

    doi : 10.3406/syria.1986.6923

    http://www.persee.fr/web/revues/home/prescript/article/syria_0039-7946_1986_num_63_1_6923

    http://www.persee.fr/web/revues/home/prescript/author/auteur_syria_6923http://dx.doi.org/10.3406/syria.1986.6923http://www.persee.fr/web/revues/home/prescript/article/syria_0039-7946_1986_num_63_1_6923http://www.persee.fr/web/revues/home/prescript/article/syria_0039-7946_1986_num_63_1_6923http://dx.doi.org/10.3406/syria.1986.6923http://www.persee.fr/web/revues/home/prescript/author/auteur_syria_6923
  • 8/9/2019 Roman Helmets & finds In Syria

    2/29

    EVIDENCE FROM D U R E U R O P O S

    FOR THE ORIGINS OF L TE R O M N HELMETS

    BY

    Simon JamesBritishMuseum,London

    Summary. Thedesignofhelmets usedbytheRomanarmychangedradicallyabouttheendofthethirdcenturyAD. Anhithertounpublished helmetfromDuraEuroposprovidesstrongevidencethattheridge helmetswhichmakeupthebulkoftheknownexampleareindeedofPartho-Sassanianinspiration,

    ashaslongbeensuspected.Therelationshipof ridgehelmetstotwolateRomanSpangenhelmefromEgyptisreassessed,andaDanubianoriginforthelatterisproposed.Anewmodelforthedevelopmentofthevarious lateRomanhelmettypesispresented.

    INTRODUCTION

    Of themany discoveriesmade at Dura by Gumont and,later,the Franco-Americanexpedition, the largequantities of remarkably wellpreserved militaryequipment rankhighin importance. The preservation of organic partsof shields, armour and weaponsin the dry desertconditions constituted an unparalleledtreasure trovefor the militaryarchaeologist. However, in the halfcenturysincethe great excavations, the materialhas been unaccountably almost completely neglected. The excavators' original planwas that the seriesof Final Reportson the Yale/French Academy excavationsshouldincludea volumewholly devoted to thismaterial. However,this,the projected VolumeVII, the Armsand Armour,was never written(although DonaldWright did prepare apreliminary study in the early1960s).

    *Acknowledgement.IwouldliketothankMs.Susan kind permissiontopublishtheDurahelmet,andforallMathesonandtheYaleUniversityArtGalleryfortheir theassistancetheyhavegivenmetomakethispossible.

  • 8/9/2019 Roman Helmets & finds In Syria

    3/29

    Fg 1 Smpfed

    k

    h

    o

    a

    Rmanrd

    h

    me

    o

    th

    g

    vr

    an

    p

    by

    n

    n

    yh

    me

    1

    A

    2Woms 36

    n

    r

    s

    af

    Kumb

    h1

  • 8/9/2019 Roman Helmets & finds In Syria

    4/29

    1986]EVIDENCEFROMDURAEUROPOSFORTHEORIGINSOFTHE LATEROMANHELMETS109

    The great majority of the material was sent to Yale University Art Gallery, wheremost of it still is. However, some material was exchangedwith the Royal Ontario

    Museum, and other pieces are on long-termloan to the John W. Higgins ArmoryMuseum, Worcester, Mass. The remainderof the material is in Damascus.

    Mypresent research aims to produce a catalogue andanalysis of the weaponry andits historical context. The valueof the collectionliesnot onlyin its size,diversity andstate of preservation, but also in the fact that it is todate the onlysizableassemblageofimperial Roman arms from the entire eastern empire. It is thereforepotentially ofvital importance for suchquestionsas the degree of standardisation of armsacross theempire, beside moredirectproblems suchas, for example, how Roman shields wereconstructed. Even morevaluableis the factthat most of the weapons canbe shown to

    have been deposited duringthe Persian siege whichdestroyed the city, and thus areclosely datable to the mid250sAD. This gives us areliable picture of the equipment ofat least one Roman garrisonin the midstof the great period of upheavalsof the thirdcentury.

    Most of the material is clearly Roman. However, thereis a scattering of objectswhich are clearly anomalous,and whichmustbelong to othertraditions, local or foreignto the empire. Far and awaythe mostimportant of these is a remarkable ironhelmet,which is the keystone of the presentpaper. This onewellprovenanced andsoundlydated object provides the missinglink in a chain of evidence which,in my opinion,elucidates the whole historyof one important and muchdisputed subject; namely thedevelopment of Roman helmet design and its debt to foreignprototypes.

    LATEROMANHELMETS

    A considerable number of helmets usually dated to the fourthand early fifth

    centuries ADhave been discovered in Europe sincethe nineteenth century. A corpuswas publishedby Klumbachon all the piecesknown up to the early1970s1. To thismay be addedJohnson's recent identificationof a helmet fromBurgh Castle in England2.

    All these helmets share certain common features,most notably a skull or bowl ofcompositeconstruction. They generallypossessa continuousfore-and-aftstrip or ridgepiece to which are attached two half-skulls. The latter may each consist of a single

    1.H. Klumbach(d.),SptrmischeGardehelme,2. J.S.Johnson, ALateRomanHelmetfromBurghMunich 1973. Castle , BritanniaXI,1980, pp.303-312.

  • 8/9/2019 Roman Helmets & finds In Syria

    5/29

    Fig.3.Simplified sketchesoflateRomanridgehelmetsoftheheaviertype,perhapsintentedforcavalrymen andofficers.1,BerkasovoI;2,BerkasovoII;3,Deurne;4,Budapest;5,Concesti;6,BurghCastle(afterKlumbach1973,

    andJohnson1980).

  • 8/9/2019 Roman Helmets & finds In Syria

    6/29

    1986]EVIDENCEFROMDURAEUROPOSFORTHEORIGINSOFTHE LATEROMANHELMETS111

    piece of metal3,or of three plates4. In structural terms, the helmets fall into twogroups.

    In the simpler form, the skullhalvesconsist of singlepiecesof metal rivettedto theridgestrip (fig.1and 2). They usually possessa simpleplate neckguard muchsmallerthan those seen on early Imperial helmets. The cheek guards werealso distinctlydifferentfromearlier types,being plain curved plates whichextended behind the ear.Hearing was usually facilitated by means of an aperture,in the rimof the bowland theupper edgeof the cheek piece. Neck and cheekguardswereattached by straps or laces,metal hinges not being employed.

    Fig.2.TheAugsthelmet,bestpreserved,andbestmadepecimenofthelightertypeoflateRomanridgehelmet(from

    Klumbach1973).Fig.4.HelmetI fromtheBerkasovofind,Yugoslavia.Thisalmostintacthelmetisthefinestandmostornateoftheheavier

    typeofridgehelmet.

    3.Forexample,Augst(Klumbach,op.cit,note1,pp.115-7andplates61-4),BerkasovoII(ibid., pp.15-38,andplates6-9).For theBerkasovofinds,seealsoM.Manojlovic-Marijanski,KasnorimskiSlemoviizBerkasova,NoviSad,1964.

    4.Forexample, Deurne(Klumbach,op.cit.,note1,

    pp.51-84andplates19-20),BerkasovoI (ibid., pp.15-38andplates1-5;seealsonote 3),andBurghCastle(Johnson,op.cit.,note2). Concesti(Klumbach,op.cit.,note1,pp.91-94andplates32-37)isexceptionalinhavingacontinuoustransversestriporribpassingunderthefore-and-aftridge.

  • 8/9/2019 Roman Helmets & finds In Syria

    7/29

    112 SYRIA [LXITI

    The second type was heavier and gavemuch greaterprotection (fig.3 and 4). Thebowl halves could be either singleforgingsor composite, made from three plates.

    Rivetted around the inside of the rim of the bowlassemblywas a browband of severalcentimetres' depth. To thiswas attached a neckguard of similar formto that seenonthe lightertype, and cheek piecesof a differentform offeringbetterprotection to thethroat, but usually lacking hearing apertures. Again, hingeswere not used. Inaddition, the better preserved examples possessa featurewhichwas probably commonto all thisgroup, and whichis unknown on earlierRoman helmets. This is a nasal,fixedto the front edge of the bowl by twowingsshaped to represent eyebrows.

    Klumbachidentifiedthe simplerform as being the infantryhelmet pattern andassigned the heavier, moreelaborate and generally morelavishly embellishedgroup to

    the cavalry and officers5.With a singleexception, all the known examples consist entirelyof iron plates6.All except the Burgh Castle piece werecertainlyor probably originally covered withsilver plating, sometimes gilded, with rouletted decoration and even imitationgemstones7 (fig.4).

    No satisfactory collective namefor thesehelmets has yet appeared. Prunkhelme(display helmets)8 or Gardehelme(guard helmets)9are clearly inappropriate since theidentification of a relatively plain helmet from a limitanean fort at Burgh Castle.Perhaps Scheitelbandhelme (roughly, 'crest-band helmets')10 or Kammhelme (ridgehelmets)11are to be preffered. Here the unlovely term 'ridge helmets' is used, as itfocusses onwhat seems to me tobe the most important feature,the fore-and-aftcrestorstrip.

    Allknow examplesmay be dated to the fourthcenturyADor the beginning of thefifth on grounds of style of decoration, inscriptionsor associated artifacts12.

    5.Thosefallingintothe'infantry'classincludeAugst,Worms,andtheIntercisafinds.Tothe'cavalry'typebelongBerkasovoI andII, Deurne,Budapest,BurghCastleand Concesti(Klumbach,op.cit.,note1,andJohnson,op.cit.,note29). Thepairofhelmetsfrom

    Augsburg-Pferseeseemto be hybrids.Klumbach'sno.1 (nowinNuremburg),isof'infantry' pattern, butlackshearingcutoutsandallegedlyhasanasal(Klumbach,p.cit.,pp. 95-7andplates38-42).Helmetno.2(nowinAugsburg)alsoresemblesthe'infantry'patterninitslowskullofsimple construction,anditshearingcutouts.However,ithaseyebrowstoanchoranasal,andsurfacedecorationsimulatinga browband(ibid.,pp.99-101andplates42-44).

    6.Theridgepiece andsideplatesoftheConcestihelmet areofbronze.Theinfillplatesandbrowbandareiron(ibid.,p.92).

    7.Themost lavishlyembellishedareBerkasovoI

    (ibid.pp.15-38andplates1-5)andBudapest(ibid.pp.39-51andplates12-18).

    8. J.Werner, Zur Herkunft derFruhmittelalter-lichenSpangenhelme ,PrhistorischeZeitschriftXXXIV,1950,p.183.

    9.Klumbach,op.cit.(note1).10.A.Alfldi,EineSpatrmischeHelmformundihreschicksaleimGermanisch-RomanischenMittelalte-r ,AdaArcheologicaV,1934, pp.99-144.

    11.Klumbach,op.cit.,(note1),p.10.12.Ibid.,andJohnson,op.cit.(note2),fordetails.

    TheAugsthelmetwasfoundassociatedwith potteryofthefirsthalfofthethirdcentury,butthiswasinashallowrubbishdepositwhich onlygivesaterminus postquern.Thereis noreasonto associateit withtheoverrunningofthelimesinthemiddledecadesofthethirdcentury ratherthanwiththenearbyfourthcenturymilitarystationatKaiseraugst.

  • 8/9/2019 Roman Helmets & finds In Syria

    8/29

    1986]EVIDENCEFROM DURAEUROPOSFORTHEORIGINSOFTHE LATEROMANHELMETS113

    Perhaps the most strikingfactabout these helmets is their utterdissimilarity to thelong-establishedand highlydevelopedRoman helmets in use up to the middle of the

    thirdcentury. The archaeologicalrecordis thenblank for about halfa century,downto the timeof Constantine theGreat,by whichtime thiscompletelydifferenttradition ofhelmet construction was fullydeveloped13.The helmets of the earlyEmpire had theirown rangeof commonfeatures. The bowlwas made in onepiecewith anintegral neckguard. The cheek pieceswereusually hinged14. Additionalplates or bars wereaddedto the skull as reinforcements.Bronze was commonly used throughout the period,alongsideall-ironhelmets15.

    There is no evidence that any traceof this tradition,whichcan be seenevolvingover morethan threehundred years downto the mid thirdcentury,survived into the

    fourth. Conversely, therehas never beenany signof any true precursors to the fourthcenturytypes in the earlierRomanmilieu. The archaeologicalrecordshows a total andrelatively sudden change in helmetdesign during the blank period at the endof the thirdcentury.

    ORIGINSANDRELATIONSHIPS

    This dramatic changehas prompted several attempts to trace the originsof thenewtypes. The consensus is that the designs musthave been foreignimports,probablyfromPersia. The picture is complicatedby tworelated factors. Firstly, thereare twomorehelmets, foundin Egypt, whichareevidently akin to the Europeangroup16(fig.5,6 and 20). However, they aredistinctin onemajor respect, namely that theirskullsconsist of several vertical stripsof iron, attached to an apical roundel. There is nocontinuous fore-and-aftridge. Thisradial rather than bipartite construction meansthat theymay be described as Spangenhelme.

    The second complicatingfactoris the group of true Dark AgeSpangenhelmefromcentral and westernEurope, notably the Baldenheim type, production of which was

    13.Deurne andthetwoBerkasovohelmetscanbeshowntohavebeendepositedinConstantiniantimes(Klumbach,op.cit.,note1,pp.36-8,66-71).

    14.Excepton.someformsof auxiliarycavalryhelmet,e.g.Robinson'stypeI(H.RussellRobinson,TheArmourofImperialHome,London1975,p.104,nos.124-6).

    15.Ibid.,p.102.OneofthelatestknownImperialItalichelmets,theexamplefromNiedermrmter,isofbronze {ibid., pp. 73-4andplates179-182).

    16.Dr-el-Medineh, nowinCairo, wasfounddansunpuitsearlythiscentury(K.H.Dittmann,EinEiserneSpangenhelminKairo ,GermaniaXXIV,1940,pp.54-8). TheLeidenhelmetisofuncertainprovenance,beyondthebarefactofitsdiscoveryinan Egyptiangrave(M.Ebert, Ein Spangenhelmaus gypten ,PrhistorischeZeitschrifl,I,1909, pp.163-70).Thetwohelmets aredifferentin form,butverysimilarinconstruction.Dr-el-Medinehisparticularlyclosetothe ridgehelmets.

  • 8/9/2019 Roman Helmets & finds In Syria

    9/29

    114 SYRIA [LXIII

    Fig.5.TheDr-el-Medinehall-ironSpangenhelm(fromDittmann1940). Comparewiththehelmets

    depictedonthearchofGalerius(fig.11).Fig.7.ThePlanighelmet, afineexampleoftheBaldenheimtypeSpangenhelm(fromPost1953).

    apparently centred on Ostrogothic Italy in the sixthcentury17(fig.7). Commentatorshave tried to put the ridge helmets, Romano-Egyptian Spangenhelme and theBaldenheim group into a unilinear sequence of development starting with theimportation of Persian models to Rome at the beginning of the fourth century.Klumbach'sschemeis that the Persian prototypes wereabsorbedby Rome andthe newfashion disseminated throughout the armyby Gonstantine, who was apparently given abejewelledhelmet of this typebeforethe battle of the MilvianBridge18. Ridgehelmets

    17.Fortheextensiveliterature on thesehelmets,seeR.Henning,DerhelmvonBaldenheimunddieverwand-tenHelmedesfrhenMiltelalters,Strasbourg,1907;M.Ebert, Die FruhmittelalterlichenSpangenhelmevonBaldenheimerTypus ,Pr.Zsch.I, 1909, pp.65ff;J.Werner, Zur Herkunft der fruhmittelalterlichenSpangenhelme ,Pr.Zsch.,XXXIV,1950, pp. 178-93;P.Post, Der KupferneSpangenhelm ,RGKBericht

    XXXIV,1953,pp.115-50.Foramore recentdiscussion,seeR.Pirling, EinSpangenhelmdesTypsBaldenheimaus LeptisMagnain Libyen ,WernerFestschriftII,Munich1974, pp.472-82.

    18.A.Alfldi,TheHelmetofConstantinewiththeChristianMonogram ,JournalofRomanStudies,XXII,1932, pp.9ff.

  • 8/9/2019 Roman Helmets & finds In Syria

    10/29

    1986]EVIDENCEFROMDURAEUROPOSFORTHEORIGINSOFTHE LATEROMANHELMETS115

    appear on his coin portraits soon afterwards19(fig.8). Gonstantine's personalresponsibility for the widespreadadoption of the newtype was an idea first developedby

    Alfldi20. Accordingto Klumbach'ssynthesis,the ridgehelmet developedduring thefourth centurytowards greatercomplexity. He implies that the appearanceof and

    Fig.6.TheDr-el-Medinehhelmet (afterDittman).

    increased emphasis on side plates in addition to the ridge was a relatively laterprogression,culminating in the Concesti helmet,deposited in the fifth century. Thelatter has twocontinuous strips runningover the apex, the outer fore-and-aft,the innertransverse. From hereit seemsa shortstep tothe four-and six-bar radial helmets fromEgypt. Theseweresomewhatarbitrarilyattributed to the fifth or even sixth centuries,putting them neatly at the endof the supposed developmentof Roman types and readyto be prototypes for the Ostrogothic Spangenhelme21.Johnson has rightlycriticisedthis simple single-line model,not least because there is very littlesign of progressivedevelopment amongthe fourthcenturyridgehelmets. Exampleswithside plates were

    presentfrom the outset22. Johnsonaccepts the late dating for the Romano-EgyptianSpangenhelme,despite the lack of solid evidence23. Infact, so little is known of theirprovenancesthat theycouldas wellbelong to, say, the late thirdor earlyfourthcenturyas the fifth24.

    19.Ibid.,p.11;Alfldi,op.cit.(note10).20.Ibid.,Klumbach,op.cit.(note1),p.10.21.Klumbach,op.cit.(note1),p.14.22.Deurne andBarkasovoI. Seenote 13.

    23.Johnson,op.cit.(note2),pp.309-11.24.H.R.Robinsondatedthemto thelatethird

    century, but failedtorecordthebasisofthisstatement(OrientalArmour,London,1967,p.73).

  • 8/9/2019 Roman Helmets & finds In Syria

    11/29

    116 SYRIA [LXIII

    Fig.8.AcoinofConstantinetheGreat,showingtheemperorwearingahelmetsimilartoBerkasovoI. Presumablythecheekguardisreducedandthenasalomittedtoavoidobscuringtheimperialportrait(fromKlumbach 1973).

    Fig.9.TheCheraghAliTepehelmet(nowinBrussels),oneoftheNorthIraniangroupwhichmaybeassignedtotheSassanianempire(fromOverlaet1982).

    In any case, unilinear modelsare inherentlyunrealistic. Influences may flowinboth directions between cultural zones. To nvestigate the originsof ridgehelmets andthe related Egyptian pieces it is necessary to look at threesuch zones, all of which wereproducing helmets in the third century.

    Firstly, there is Parthia, and its successor, Sassanid Persia, from which thereisarchaeologicalevidence including very important new material from Dura Europos,described below. Secondly, there is the Roman Empire itself, whose armaments

  • 8/9/2019 Roman Helmets & finds In Syria

    12/29

    1986]EVIDENCEFROMDURAEUROPOSFORTHEORIGINSOFTHE LATEROMANHELMETS117

    industryunderwent a revolution in the later third century. Lastly, there are thebarbarian peopleson the Northernfrontiersof the Roman Empire. The tribesof free

    Germany can be discounted from the presentdiscussion, as documentary sources andarchaeology combine to make it clear that helmets were notcommonly used by thesepeoplesduring the time in question25. However, the Sarmatians and Goths on theDanubecertainly didhave theirowntradition or traditions of helmet-making, and thesemustbe taken into account as a potential influenceon late Roman designs.

    THEPARTHO-SASSANIANEAST

    The search for convincing Persian prototypes has been hampered by the lack ofreliably dated archaeologicaldiscoveries, exacerbated by the near absence of good

    contemporary depictions. A number of helmets of allegedlySassanianmanufacture areknown,but none can definitelybe shown to belong to the crucial period, the thirdcentury26. Interestingly, all have compositebowls and all but onehave a continuousfore-and-aftstrip formingthe coreof the assembly. Unlike the Roman ridgehelmets(exceptConcesti),the skulls areall tall,witha paraboloidprofile.Cheekpieces and neckguards are lacking, although holes aroundthe rimof most examplesare thought to be forattaching a camail of ironrings27. However, these are too widely spacedfor this, andthey are more likelyto have served as anchor points for linings.

    Two apparently regional groupings maybe identified,one centred on North Iran,

    the other on Mesopotamia.The North Iranian groupconsists of four helmets with bronze or bronze-covered

    iron frames, wholly or partly covered with silver foil and embellished with acharacteristic featherpattern28(fig.9). While the stamped silver plating may seem topoint to a relationship with theRoman ridgehelmets, the use of the featherpattern mayrulethis out. While it probably is a Sassanianmotif29,Overlaet has recentlyarguedthat it is late, belonging to the fifth to seventh century30.

    Of theMesopotamiangroup,the onlysecurely provenanced exampleis from theTemple of Ishtar at Nineveh (and is now in Baghdad)31(fig.10). The context didnot

    25.The fewhelmetsfromlateRomanIronAgecontextsinFreeGermanyareRoman,eitherbootyorexports,e.g.Throsbjerg(M.Todd,TheNorthernBarbarians,London,1975,p.177).

    26.SeeS.V.Grancsay,A SassanianChieftain'sHelmet ,MetropolitanMuseumofArtBulletinXXI,1963,pp.253-262;B.J.Overlaet,ContributiontoSassanianArmamentinconnectionwitha DecoratedHelmet ,IranicaAntiqua,XVII,1982, pp. 189-206.

    27.Grancsay,op.cit.(note26), p.259.28.TwoaresaidtobefromtheAmlash region .

    OneisnowinNewYork(ibid.),theotherat Mainz

    (Overlaet,op.cit.,note26,pp.193-4).AhelmetnowinBrusselsis fromCheraghAliTepe (ibid., pp. 189-90). ThefourthisunprovenancedandisnowinLosAngeles(ibid.,p.194).

    29.Ibid.,p.200.30.Ibid.,especiallypp.205-6.31.R.C.ThompsonandR.W.Hutchinson,The

    BritishMuseumExcavationsintheTempleofIshtaratNineveh,1930-31 , LiverpoolAnnalsofArchaeologyandAnthropology,XIX,1932,p.78andplate51,3;Werner,op.cit.(note 17), p.184.

  • 8/9/2019 Roman Helmets & finds In Syria

    13/29

    118 SYRIA [LXIII

    Fig. 10.TheNinevehhelmetnowinBaghdad,oneoftheMesopotamiangroupofSassaniddate(fromOverlaet1982).

    produce a satisfactory date. The helmet consistsof an iron browband, with alongitudinal strip and two lateral bands of bronze. The infill plates are of iron. Thisis structurally identical to the NorthIranian group. Silver is lacking, but the scale or

    featherpattern on the browband and infillplates may again suggest a late date. Theotherexamplesin this groupare twohelmets,also said to be fromNineveh, now in theBritish Museum32.Both share with theBaghdadhelmet the flaringbronze ridgeandsideplates. Oneis in external appearanceparticularly closeto Baghdad,but the skullis actually of two ironhalf-shellsbeneath the bronze frame33. The second bears lesssuperficial resemblance but is structurallycloser. Here the ironplates simplyfillthegaps in the bronze frame34. Thishelmet has a second peculiarity not seenon theothers. There are twodeep cutouts in the rim over the wearer's eyes,and rivetholesthat suggest it had had an eyebrow/nasal plate likethe Roman 'cavalry' ridgehelmets.

    One furtherhelmet, also in the BritishMuseum,is thought to be Sassanian35. Itdiffersfromall the othersin being of radial construction. Allthe components seem tobe of iron covered withbronze plate. The assemblyconsists of fourtriangular plates,joined by fourvertical strips. Fragments of what seemsto be an ironcamailadhere tothe outside. The helmet may indeed be Sassanian,but there are good archaeological

    32. BritishMuseumnos.22497 and 22498{ibid.,p.188andTaf.7,1and7,2respectively;Grancsay,op.cit.,note26,p.259andfigs.7and8respectively).33.Basedonmyowninspectionofhelmet22498.

    34.Basedonmyowninspectionof22497.35.BMno.22495. Grancsay,op.cit.(note 26),

    p.256andfig.6.

  • 8/9/2019 Roman Helmets & finds In Syria

    14/29

    1986]EVIDENCEFROMDIRAEUROPOSFORTHEORIGINSOFTHE LATEROMANHELMETS119

    parallels for it in Russia,whichare dated to the eighth36to tenthcenturies37. Similarhelmets existed in Tibet untilrecenttimes38. On thisevidenceit seemsmostlikely that

    the helmet is post-Sassanian,and so willnot be considered furtherhere.Apart from the latter example, the archaeologicalrecordfor 'Sassanian'helmets isconsistent. The fore-and-aftstripis a standard feature,and may suggest a relationshipwith theRoman helmets. However, none is demonstrably earlyenough in date toprove that these types existed in the third centuryto allow themto serve asprototypesfor the Roman forms.

    Other sources of evidenceare notparticularly helpful. There are no reallycleardepictionsof helmets on Parthian monuments. It has been suggestedthat the Ninevehhelmets are related to the Parthian royal headdress seen,for example, on Arsacidcoinage39. This has been challenged by Overlaet40,but still seems a viablehypothesis. The onlydetailed depictions of Parthian helmets so faridentifiedare onearlySassanianrockcarvings at Firuzabad,depicting Ardashir I's defeat of Artabanus

    Fig.11.DetailsoftheearlySassanianroyalreliefsat Firuzabad.ASassanian'page'grappleswithaParthianonhorseback.Thepagewearsa simplecap(notethe'heraldic'symbol).TheParthian wearsanapparently simplehelmetwithanattachedscaleneck

    defence(afterGhirshman).

    Fig.12.TheDuraclibanariusgraffito.

    36. AhelmetfromGouvernmentWoronesh datedtotheeighthcentury(W.Arendt, DerNomadenhelmderFrhenMittelaltersinOsteuropa ,ZeitschriflfurHisto-rischeWaffn-undKostmkunde(Neue Folge)V, 1936,p.31andTaf.V).

    37.Ibid.,p.31andAbb.8.38.Grancsay,op.cil.(note 26),p.258,fig.9.39.Werner,op.cil.(note17),p.188.40.Overlaet,op.cit.(note26),pp.191-2.

  • 8/9/2019 Roman Helmets & finds In Syria

    15/29

    120 SYRIA [LXIII

    V41 (fig.11). They appear to have simpleone-piecebowls, but they are seenin strictprofileand so could be pain bipartite helmets. They have eithera scale camail or an

    armoured hood (coif)beneath. The Sassanianfigureson the carvings generally lackhelmets, except for a 'page' whowears a simplecap. No othersdepictions securelydatable to the mid-thirdcenturyor earlierare knownto me, except the graffiti fromDura Europos whichshow armoured cavalry42. The best known of these is the'charging clibanariusHS(fig.12). Theseare usually thought to depict Parthian orPersian cataphracts, but the fact that horsearmourwas employed by the Romandefenders of the city mustmake this uncertain. These illustrations, whichare quitecrudely executed, may suggest helmetswith composite bowls, and perhaps face-masks44. However, their crudeness, and the uncertainty of their identification as

    Persian makes their value very limited.In conclusion,up to the presentit has been impossibleto say withany degree ofassurancewhat sort of helmets wereused by thePersians in the third centuryAD, at thetime when theyare supposed to have been profoundly influencingRoman design.

    THEDURAHELMET

    Ironically, the best evidenceof allhasbeen storedaway in a museum vault for thelast half century. This is an unpublished helmet from the Franco-American

    excavationsat Dura-Europos45. The piece has been in Yale University MuseumandArt Gallery sincethe SecondWorld War46. It was foundin the countermine beneathTower 19 on the desertwall47. It was certainlydeposited during the siege whichdestroyed the cityin the mid-third century.Dura, the most important forwardbaseand garrisonof the Roman armyon thisstretchof the Euphrates, was attacked anddestroyed by the Persians underShapur I around AD 255-748. Of thewealth ofarchaeologicalremains resultingfromthe siege,the richestin termsof artifactsare thosefrom the operations around Tower 1949. The countermine was an attempt by the

    41.R.Ghirshman,Iran;ParthiansandSassanians,London,1962,plates165-7;G.Herrmann,TheIranianRevival,London,1977, pp. 87-9.

    42. Forexample, Dura,ReportontheFourthSeason,NewHaven,1933,plateXXI,3.

    43.Ibid.,plateXXII,2.44.In the fourthcenturyRomanclibanariiwore

    helmetswith facemasks(Julian,Or.inConst.LaudemI,37)andinearliercenturiesmaskedhelmetswereworn bycavalrymen,butapparentlynotinbattle(the so-called'cavalerysportshelmets').

    45.Foranintroductionto thesiteanda

    hensive bibliography,seeC.Hopkins,TheDiscoveryofDuraEuropos,NewHavenandLondon1979.

    46.Yaleno.1981.62.28.47.ThehelmetisbrieflymentionedinDura,Reporton

    theSixthSeason,NewHaven,1936,p.194,andcanbeseeninfig.16 (heremyfig.6).

    48.Forthemostrecentdiscussionofthedateandhistoricalcontextof thesiege,seeS.James,DuraEuroposandtheChronologyofSyriainthe250sAD ,ChironXV,1985,pp.107-120.49.Op.cit.(note 47),pp. 188-98,204-5,439-66.

  • 8/9/2019 Roman Helmets & finds In Syria

    16/29

    Illustration non autorise la diffusion

    1986]EVIDENCEFROMDURAEUROPOSFORTHEORIGINSOFTHE LATEROMANHELMETS121

    > RUBBLE.(; WALL

    Fig.14.TheDurahelmet,lyinginsitun theTower19countermine(YaleUniversity rt GalleryDuraEuroposCollection).

    /

    MUD-RICKVM.L

    Fig.13.ThecountermineatTower19,DuraEuroposDuraRep.VI,fig.16).

    Roman defenders to stop the Persians undermining the foundations of the Towerpreparatory to an assault. The consequent battle underground resulted in thedefenders being worsted. The pileof unrecovered Romanbodies in the countermine,and the subsequent destruction of the towerare eloquent testimony to this.

    Thehelmet was foundcloseto a body lyingalone,separate fromthe mass of Romanbodies at the other end of the mine(fig.13). Thereis good reason to thinkthat thehelmet belonged to this individual, and that he was oneof the attackers. He hadapparently been facingthe city whenhe was struck down,and fellbackwards. This isof course hardly conclusive, but several other featurespoint towards the sameconclusion. Alsolyingcloseto the body was a large sword,the bladeof whichno longersurvives. However, it had a flat disk pommel of jade, whichwas certainly an importfrom Chinese Turkestan50.The shape of the pommel is unroman. The mail shirt

    50.Ibid.,p.194.

  • 8/9/2019 Roman Helmets & finds In Syria

    17/29

    I

  • 8/9/2019 Roman Helmets & finds In Syria

    18/29

    1986]EVIDENCEFROMDURAEUROPOSFORTHEORIGINSOFTHE LATEROMANHELMETS123

    foundon the body has a patternof bronze ringson the chest. These form a tridentdevice, which looks likean 'heraldic' crest. It is verysimilar to the deviceswornby

    Sassaniannobles on the contemporary rockcarvings51. Finally, the helmet itself istotally differentfrom the other helmets found at Dura52. These are all perfectlystandard Roman types53. The helmet fromthe mineis clearly related to the ostensiblySassanianhelmets from Mesopotamia andNorth Iran.

    The Durahelmet, then,veryprobably belongedto the attacking army. This neednot mean that it came from the Iranian heartland of the Persian empire. TheSassanianarmy was apparently organised on feudallines,including levies from subjectpeopleson the peripheries of the empire54. The helmet couldthereforehave come fromother areas such as lowerMesopotamiaor Afghanistan. Whatever its preciseorigins,

    hereis an exampleof the typeof helmet in use on and beyond the Eastern frontierof theRoman empire in the middle of the third centuryAD.The helmet is quite extensivelyoxidised,and was crushed whenfound(fig.14). It

    was lyingon its rightside,and the left had been crushedwhenthe roofof the minecavedin. It is estimated that 80-90%of the originalstructuresurvives.

    The object is entirelyof iron(fig.15 and 16). The skull is composite, consisting oftwosimple half-shells whichmeet on the longitudinalaxis but are not directlyjoined toeach other. Eachis independently rivettedto anexternal stripof ironrunningover thecrest from browto nape. The last few centimetresare missingat the back. Theexcavatorsrecorded the presence of a second joiningstrip,on the inside. The existenceof this cannot be confirmed,due to the liberal use of plaster in the restoration of thehelmet. It seems most likelythat the observation was mistaken and that what theexcavators saw wasthe left sideof the outer strip lyingin situ crushed down into theright skull-half.

    The dimensionsof the skullas currently restoredare as follows:height (rimto top ofskull) 250 mm; lengthc. 255mm; width160mm; circumference 660-680mm. The

    51For example,thesymbolsonthehorsetrappers

    andthepage's

    helmetat Firuzabad

    (note41).

    Theseare thoughttobeclan badgesorthelike,andare possiblyrelatedto Sarmatiandevicesof a similarpurpose(H.Nickel,Tamgasand Runes,MagicNumbers andMagicSymbols ,MetropolitanMuseumJournal,VIII,1973, pp.115-173.52.Severalunpublishedfragmentsarepreservedat

    Yale,includingthelefttempleareaofan ironhelmetskull(Yaleno.1982.28.55);a rightcheekpiece,alsoofiron(1982.28.56);andtwofragmentary bronzecrownreinforcements(1938.3300and1938.3301).Abronzebrow-guardis,accordingtoYalerecords,preservedinDamascus(no.unknown).

    53.All thefragmentslistedinnote52are identifiable

    aspiecesof Roman

    helmetsofRobinson'sauxiliary

    cavalrytypes EorF(op.cit.,note14,pp.96-9).54.R.Frye, ThepoliticalhistoryofIranundertheSassanians ,CambridgeHistoryofIran,III,1983,p.154;V.Lukonin,Political, Socialand AdministrativeInstitutions;TaxationandTrade , CHI,III,1983,p.728.Anexampleis theemploymentofa contingentofChionites,undertheirkingGrumbates,at thesiegeofAmidainAD359(Ammianus,XVIII,6,22). ThispeoplehadrecentlybeensubjugatedbyShapurII(Amm.XVI,9,4;XVII,5,1).

  • 8/9/2019 Roman Helmets & finds In Syria

    19/29

    124 SYRIA [LXIII

    200mm 200mm

    Fig.16.TheDurahelmet,frontandrightsideviews.Notethemassofrusted mailadheringtotherightforeheadarea(photos: author).

    thickness of the ironplate is hard to measure due to itscorroded state, but is estimatedto have been about 4 mm, perhaps a little moreat the rim. The crest strip was of

    roughlyconstant width,about 30 mm, and 3-4mm thick. When found,the helmetweighed 4.15 kg.

    As restored,the skull is obviously distorteddue to the lateral crushingwhichshattered the left sideand somewhatflattenedthe right. The resultis that the helmetis now longerand narrowerthan originally. However, it is clear that it was tall andthat its geometrywas dominated by thelongitudinal junction of its halves. In sideelevation, the bowlis roughlyparaboloid, but in front elevation the sides sweepup to apoint. In section, thejoining strip matches thecontourof the junction of the twohalf-bowls,i.e. nearlyflat closeto the rim, and V-sectionedover the apex. The rivetsare

  • 8/9/2019 Roman Helmets & finds In Syria

    20/29

    1986]EVIDENCEFROMDURAEUROPOSFORTHEORIGINSOFTHE LATEROMANHELMETS125

    Fig.17.Adetailoftheforeheadarea,showingthesurviving portionsofthe'eyebrows'andthebaseofthesnapped-offnasal(photo: author).

    carefullypaired along its length, at approximately 65 mm intervals. They hadpronounced heads, now distortedby corrosion, which wereprobably originallyhemispherical or perhaps flattenedspheres.

    At the apex is attached a blunt cylindrical ironprojection, apparently solid,rising51 mmabove the skull. It is apparently attached by a flat tab runningdown eachsideof the bowl, each pierced by two rivets, one goingthroughthe joining strip and onethroughthe half-skull.

    Attached to the creststrip, runningfrom just in front of the 'spike' down towards

    the brow,is a partially preservedcurved plate standing perpendicularly to the helmet.It reaches a depth of 27 mmand was originally 5-6mm thick. It was apparently notattached to the bowl on discovery55,but is reliably restored. It terminates at the upperendin a flat, teardrop-shaped tab, whichhas been pierced by a largerivetholding it tothe helmet. Howfar downthe helmet thisplate reached is uncertain, as the lowerpartis lost. However, an indentation in the edge of the right half-skullsuggests that the

    55.Frominspectionoftheobjectasrestored,infig.6.

    Itmaybethe ironplate seenlyingjusttotheleftofthehelmet

  • 8/9/2019 Roman Helmets & finds In Syria

    21/29

    126 SYRIA [LXTII

    lowerrivetwas immediately above the brow. There is no indication of a similar platedown the back of the helmet,although the possibilitycannot be ruled out.

    The forehead regionis quite badly damaged (fig.17). Due to crushing,the skull-halvesmeet almost at right-angles wheretheyshouldbutt edge to edge. Furthermore,the lowerfront cornerof the left plate is lost, and with it some important detail.Nevertheless, it is possible to reconstructthe appearanceof the browarea withsomecertainty. Overall, the lowerrim of the bowlconformsroughlyto a horizontal plane.At the front, however,there are two shallow curved cutouts over the wearer's eyes.Rivetted above these, and conformingto their curvature, is a singleplate of iron,c. 4 mmthick,forminga pair of 'eyebrows'. The innerhalfof the left eyebrow was lostwiththe damageto the skull-half. The outer partof the rightone is coveredby a mass

    of rustedmail and is hidden,but enough of each is visible to reconstruct the whole.Each eyebrow stretchedc.90 mm from the centreline. The left is 11 mm deep at thetip, flaringto 15 mm at its brokenend. The right onereaches a maximum depth of23 mm at 25 mm from the centreline. The upperedge of the plate dips steeply in themiddle, reinforcingthe impression of a pair of eyebrows.

    It is clear that the loweredges of the eyebrow plate swept down intoa nose guard,which is now snapped off at the base and.lost. While reinforcingthe forehead regionand fulfillinga certain decorative function,it is clear that the main purpose of theeyebrowswas to securely attach thisnasal. The originally T-shapedplate was heldonby fourdomeheadedrivets, one throughthe tip of each wingand another 25 mmeithersideof the centreline (fig.18).

    Adheringto the loweredge of the helmet,allalong ist back and left side,is a mass ofoxidised mail. It is certainlyan integralpart of the helmet, and can be seento bedirectlyattached to the rim. It provided the neck-and cheek defence in the formof amail screen or camail, hanging from the bowl. It is much foldedand in very poorcondition, a large partof it adhering to the right forehead regionwhere it had beentrapped beneath the helmet on deposition. In placesthe mail is well/enoughpreservedto reveal that it consists of ringsc. 9 mm in diameter, but it is not possibleto determinetheir thickness or the method of joining (butting or rivetting)or whether stamped ringswereemployed. It appearsthat the uppermost rowof ringspassedthroughholes in theloweredge of the bowl(a method used on lateroriental helmets). If so, this couldexplain the raggedness of the left edge,fromwhichthe mail has apparently been tornaway, probably whenthe helmet was crushedby the roof fall. It seems thenthat themail was attached all the way around the edge except across the eye cutouts.

    The camail is far too foldedand oxidised to reveal much about its form andlength. The fieldrecordcard with theon-sitedescription gives the lengthof the camailas 220 mm,but how this was arrivedat is obscure. It is noteven certain that it is allpreserved, as the sitedrawing shows a separate mail fragmentnext to the helmet(fig.13).

  • 8/9/2019 Roman Helmets & finds In Syria

    22/29

    1986]EVIDENCEFROMDURAEUROPOSFORTHEORIGINSOFTHE LATEROMANHELMETS127

    Spangenhelmeandother helmetsworn byarchersonTrajan'sColumn.

    / /Fig. 18.A reconstructionof theDuraHelmet.Theformofthecamailisconjectural,butisbasedon

    thenear-contemporaryFiruzabadreliefs.

    There are fourrivetsin a lineacross the back of the bowl, c.15 mmabove the rim.They do not relate to the camail. It may not be coincidental that theymatch upwiththe fourrivetsattaching the nasal to the front of the bowl. It is likely thatthese eightrivetsserved as the anchor-points for a fabricor possibly leather lining. The depth ofthe helmet dictates that theremusthave been somesuch arrangement to ensure that itsat properlyon the head and did not fall downover the eyes.

  • 8/9/2019 Roman Helmets & finds In Syria

    23/29

    128 SYRIA [LXIII

    THEDURAHELMETANDTHELATEROMANTYPES

    The closerelationship betweenthe Dura helmet and the late Romanridgehelmets isimmediately apparent. Dura is made whollyof iron, in twosimple half-shells joined to

    a ridgestrip of continuous width by prominent rivets. It had a nasal, attached byeyebrow' plates. This structureis in all essentials identical to that of the bowl of

    Berkasovo II. The morecomplex examples,like Berkasovo I and Deurne, and thesimpler ones like Augst and Intercisa, can be seenas variants, at opposite ends of arepertoireof design,all ultimately derivingfrom prototypes of which Dura is anexample. The Dura helmet is from the right area and is of precisely the rightdate toprove the validity of the claim that the late Roman ridge helmets were of Partho-Sassanian,or at least oriental origin.

    THEDANUBEANDTHEPLACEOF THEROMANO-EGYPTIANSPANGENHELME

    There remains the question of the relationship of Roman ridge helmets to theRoman Spangenhelme from Egypt. Are the former theancestors of the latter, asKlumbach believes? It was noted above that Johnson has criticised Klumbach'shypothesis of ridgehelmets increasingin complexity, culminating in the Concesti helmetwhich is constructed around twocontinuous bands over the apex which, in hisopinion, isa short stepfrom the true radial Spangenhelm. Klumbach's model is not reallytenable. There is very little evidence of progressive development towards greatercomplexity in bowl construction. Two of the earliest helmets, Berkasovo I andDeurne, are among the most complicated. In fact, thevery limited dating evidencecan be made to support the opposite view. Allthe dated complexhelmets weremadein the earlyor mid fourth century. Concesti, the possible exception, was foundin abarbarian grave and could wellhave been old when buried. On the other hand, theonlydated examplesof the simple type are late (Intercisa, belonging to the earlyfifthcentury). One can thereforeargue for a progressive degeneration of design, from thecomplexto the simple. However, I donotbelievethis to have been the case. It is farmorelikelythat a variety of types werein use simultaneously throughout the fourthcentury.

    The fact remains that the attribution of the Romano-Egyptianradial helmets to thefifth centuryor later is guesswork, and can easily be challenged. There is clearevidence for the use of trueSpangenhelmein the Roman armyas earlyas the beginningof the secondcenturyAD. Someof the archerson Trajan'sColumnaredepicted wearing

  • 8/9/2019 Roman Helmets & finds In Syria

    24/29

    1986]EVIDENCEFROMDURAEUROPOSFORTHEORIGINSOFTHE LATEROMANHELMETS129

    them (fig.19). These men are usually said to be Syrians. If so, their helmets,unusual among Roman helmets of the time (and so fararchaeologically unattested)

    could implythat therewas already a localtradition of Spangenhelmconstruction in theRoman Eastby AD100. However,thereis no independent evidencefor this,and otherarchers on the Column do not wear suchheadgear. On the other hand, the RomanSpangenhelmeare verysimilar to captured barbarian helmets shown on the base of thecolumn. Perhaps these archers wereequippedwithlocally made helmets. They mayeven have been recruitedlocally56.Clearly,the Danubemustnow be consideredas thepossible source of the Spangenhelmein the Roman army.

    The Trajan's Column reliefseffectivelyprove that true Spangenhelmewerein useamong the native peopleson the Northbank of the Danube,especiallythe Sarmatians,

    by AD 100. Several scenes show Sarmatian armoured cavalry equipped withbullet-shaped helmets of radial construction with cheekpiecesbut no neckguard (fig.10).Similarhelmets, possibly withnasals, are to be seenon the reliefsof captured armsonthe base of the Column. Thesecould be Sarmatian or Dacian. This is the earliestevidence so far for the appearance of the type. No archaeological examplesdemonstrably of such an earlydate are known to me. It seems most likelythat thetrue Spangenhelmcame to the Roman worldacross the lowerDanubianfrontierduringthe late first centuryAD if not earlier,and suchhelmets wereused by at least a smallproportion of the auxilia thereafter.

    The Egyptian Spangenhelme closely resemblethe Trajan's Column depictionsnotjust in basic construction but also in detailedshape. This is seenmost strikinglyin thecase of the Leiden helmet, whichis almost identical to certain examples worn bySarmatianson the Column (fig.20). There are the same apicalroundel, radial strips,simple cheekpieces andunprotected neck. Similarly, the bowl of the Der-el-Medinehhelmet is remarkably close to a number of examples on the baseof the Columnand ontrophies on the helical frieze. These share the slightlypointed top, the externalbrowband (which contrasts with the internalbrowband of the ridge helmets), andprobably also the nasal, but thesedetails, so proneto damage, are lost from themonument.

    There is independent evidencethat Spangenhelmeverylikethe Der-el-Medinehtypewere used by Roman troops,apparently cavalry, around AD300. The reliefson theArch of Galerius at Thessalonikashow this quite clearly (fig.21). There is the sameradially constructed bowl tendingtowards a point,cheekpieces, neckguardand this timeunequivocallya nasal.

    56.Danubianarcherunits areknown,e.g.cohorsI Oxford1980,p.228;alsothealaIIIAug.Thracsag., andThracumsagittariorum,P.A.Holder,TheAuxiliafrom alaI Thrac.vet. sag.,ibid.p.227.Augustusto TrajanBARInternationalseriesno. 70,

  • 8/9/2019 Roman Helmets & finds In Syria

    25/29

    Fig.21.RomanSpangenhelmedepictedontheArchofGaleriusatThessalonika(afterLaubscher).Notethestrongsimilarityto theDr-el-Medi

    helmet(fig.3).

    Fig.20.TheLeiden RomanSpangenhelm,foundin Egypt(top)comparedwiththehelmetsofSarmatiansdepictedonTrajan'sColumn (centHelmetsdepictedonthetrophiesandbasalfriezeson thecolumn(bottom) resembletheDr-el-Medineh helmetinformand/or structure(fig.3).

  • 8/9/2019 Roman Helmets & finds In Syria

    26/29

    1986]EVIDENCEFROMDURAEUROPOSFORTHEORIGINSOFTHE LATEROMANHELMETS131

    This depictional evidencecan be used to tentatively reinterpretthe Romano-Egyptian Spangenhelmeon a considerablysounder footing. There is no need to see

    them as earlyByzantine. Such evidence as there is pointsto the end of the thirdcenturyas the most likelydate for theirmanufacture and use. Had they been in verywidespreaduse beforeAD 260or afterAD 320, then given the amount of other helmetmaterial foundfor these periods onewould have expected morearchaeologicalevidenceto have appeared by now. The floruitof these helmets, when theywereimportantenough to be wornby Galerius' seniorregimentsdepicted on the Thessalonikareliefs,seems to correspond with what is otherwisea blank in the archaeological record.Roman Spangenhelmethereforepreceded rather than evolved from the ridgehelmetfamily. Thisperhaps explainsthe use of hingeson the Egyptian helmets, which can be

    seenas an earlyfeature.

    ROMANHELMETDESIGN:ANEWSYNTHESIS

    These conclusionspointto a newsynthesisof the development of Roman helmets.It seems that the 'traditional' Imperial cavalryand infantrytypes weredisplaced at anunknown rate and to an unknown (but large) degree by all-ironSpangenhelmein thesecondhalfon the thirdcentury. It is suggestedthat radial helmets likethose foundinEgypt and depictedon the Archof Galeriuswerethe dominant types used by the armies

    of the Tetrarchy. However, thesein their turnwerealready at least partlydisplacedby new,bipartite ridgehelmets by the320s, as archaeology demonstrates. The latterthen seem to become the sole tradition of helmet construction used in the empirethroughout the rest of the fourthcentury. That thesebipartite helmets wereofPartho-Sassanianorigin nowseemsbeyond reasonabledoubt in the lightof the Duraevidence. However,an important question remains to be addressed. It is not at allobviouswhy averylong establishedand highlydevelopedtradition of helmet design wassuddenly abandoned to be replaced by transdanubian Spangenhelme and then byorientalising ridgehelmets.

    Klumbach, who follows Alfldiin attributing the dissemination of ridgehelmetspersonally to Constantine, seesthe employment of Persian-style helmets as part of theorientalisation of court ceremonial andthe dressof emperors,officialsand state servants,fromthe timeof the Tetrarchy57.No doubt thisis partof the explanation;one has onlyto look at BerkasovoI to believeit. However, it seems to methat it is notthe barbaricsplendour of some of the helmets that is of deepest significance,but the simplicity ofdesign which is commonto all,resultingfromthe elimination of allcomponentsrequiring

    57.Seenote20.

  • 8/9/2019 Roman Helmets & finds In Syria

    27/29

    132 SYRIA [LXIII

    high levels of skill to manufacture. Complex forgingssuch as hingesand one-piecebowls with integral neckguards wereexcised from the new designs. The simple

    components of the ridge helmets could be made by relatively unskilled andinexperiencedsmiths. The downrightcrudenessof a number of examplessuggests thattheyoftenwere. Eventhe finestpieceshardlymatch up to second centurystandardsof construction.

    Not only, then, was there a complete change in the design of Roman helmets.There was a simultaneousdecline in standards of manufacture. These changesmustbeseenagainst the historical background of contemporary developments in the armyandthe armaments industryitself.

    Little is known for certain about the organisation of the industrywhichproduced

    the fine helmets of the second and earlythird centuries AD58. It is assumed thatindividualcraftsmenor small private companies suppliedthe troops,probably through avariety of mechanisms including commissions from individual soldiers and largermultiple orders from regiments, provincial army commands or the centralgovernment. Whatever the case, presumably these were cash transactions.Armourers willhave been paid in specie for their work,and used the coin to cover rawmaterials and overheads, pay their taxes and support their families. They wereparticulary dependant on the soundnessof the currencyas laws controlling sales andpossessionof armaments restrictedtheirmarketalmost exclusivelyto the state59. Thecollapseof the coinagefrom the middle of the third centurywould have paralysed thissystemof supply. The armycouldnotaffordto buythe weapons,while the armourerscouldnotsell theirwares norbuy raw materials. The operating systemof the industry,which had endured for several centuries,collapsed.

    The developmentof a crisisin armsprocurement is, in myopinion, the directcauseof the establishment of the state armsfactories,or fabricae,which start to appear underthe Tetrarchy60. It is suggested that from the 260s the state was forced tobypass thefinancial crisisand started to maintain thearmourersdirectly,by providing rations andsecurityin return for product, leading to the gradual absorption of the armourers intothe Imperial service. This process reachedits logicalconclusionwhen Diocletianput iton a regularbasis and built new factoriesto accommodate (and control) them atstrategic points across the Empire. It seemsthat the state wanted quantityproduction, not fancyquality, hardlysurprisingwhenfaced with the task of supplying

    58.R.McMullen,InscriptionsonArmourandthe reiteratedin thelateEmpire(e.g. C.Th.XV,xv,1).Supplyof Armsin theRomanEmpire ,AmericanArmsexportswerealsobanned(C.J.IV,xli,2).JournalofArchaeologyLXIV,1960, pp.25-9. 60.S.James,TheFabricae;StateArmsFactoriesof

    59.ThelexIuliadevi publica.The banwas theLaterRoman Empire(forthcoming).

  • 8/9/2019 Roman Helmets & finds In Syria

    28/29

    1986]EVIDENCEFROMDURAEUROPOSFORTHEORIGINSOFTHE LATEROMANHELMETS133

    an expanding armysufferinghigh ratesof attrition,as Diocletian'ssurelywas. Thearmourers, now called fabricenses,probably had monthly quotas to fulfill61.

    Against this background, the history of helmet design becomes explicable.Traditional Roman types ceased to be made withthe hypothetical rapid collapseof theold industry in the third quarter of the third century. The state, now directlysupporting the armourers, naturally wanted cheap, functional and above all quick-to-produce designs. Perhaps the initial resultwas the general adoptionof DanubiantypeSpangenhelme by the armourerssupplying the Illyriancavalry force of Gallienus,Aurelian and Probus. These troops,many of whom werebarbarians fromacross theDanube, formedthe elite of the Tetrarchic armies. Hence the Spangenhelme ofGalerius' household troops.

    Diocletian,who reorganised the armyand founded the new arms factories,is mostlikelyto have been responsible for the introduction of the newbipartite ridgehelmets.There seems little reason for attributing their dissemination to Gonstantine, as theBerkasovofinds show that they werealready established in Licinius' armyperhaps asearlyas AD314. It is verytempting to link their appearancewith the building of thenew factories,and to see the opening of the fabricaeas both the opportunity and thereason for the introduction of the new,and definitive ridgehelmet types. While theirEastern inspiration well in tune with the Tetrarchic switch from Illyrian austerity tooriental splendour, it seems to methat the main motivation was morepractical. ThePartho-Sassanianprototypes met the requirement for simplicity of manufacture, butwere substantially redesigned to meet Roman standards of protection. Hence theaddition of plate neck-and cheek guards to all versions. Similarly, the types withaseparate browband, whetherthis featurewas of Danubianor Persian originor both,were revised. The compositeskull was fitted to the outside of the browband ratherthan the inside, improving protection by increasingthe clearance between the plates andthe wearer's head. On the otherhand, elimination of hingesand other difficult forgingsmade them suitable for rapid mass production by even a semi-skilled workforce. Thenew designs betray much careful thought,as does the distribution of the factoriesinwhich theywere made62. Muchmorethan simplya whim of fashion,the appearanceof

    61.Nodirectevidencerelatingtoproductionquotasimposedonfabricensessurvives,but alawofAD374recordsthemonthlyhelmetproduction targetsforthebarbaricarii (precious-metalsmiths)at AntiochandConstantinople.Thesemenwere apparentlynotmaking hehelmets,butdecoratingcomponentstoproducefinepiecesfor,presumably,officersandguard units.Itmentionsthateighthelmets andthesamenumbersof[?pairsof]cheekguardsarecoveredwith silverandgilded[byeachworker]eachthirtydaysatAntioch...TheworkersatConstantinoplewere apparentlyfailingto

    matchthisrateandwerebeing instructedto increaseoutput(C.Th.X, xxii,1). Presumablythefabricensesthemselvesweremonitoredandcontrolledinmuchthesameway.62.Thefabricaebreakdowninto anumberofcategor

    i e s ccordingto thetypesofequipmenttheywereproducing.Thefactoriesmakingbasic equipmentsuchasshieldsandarmour(as opposedtomorespecialistcentresmakingheavycavalryarmour,artillery,etc.),were linkedintheirdistribution tothegeographyoftheempire asit was reorganisedundertheTetrarchy.

  • 8/9/2019 Roman Helmets & finds In Syria

    29/29

    134 SYRIA [LXIII

    newstylehelmets was a resultof the 'nationalisation' of the armsindustryat the endofthe third century.

    This reinterpretation of the development of Imperial helmet design suggests thattherewas no simple unilinearsequence. It was partof a much wider network involvingseveral cultural groupings, all with their owntraditions of helmet construction, all ofwhich toa greateror lesserdegree influencedthe other. Thus earlyImperial helmetsevolved from various currentsof Hellenistic, Italian and Gallic design. If radialhelmets came to Rome from the Danubeand ridgehelmets fromacross the Euphrates,then the Eastern Europeanand Iranian peopleswhotransmitted them may wellhaveinfluencedeachother via the nomadsof Central Asia. Further, the interaction betweenRome and her neighbours was not necessarily unidirectional. The reinforcingplate

    down the front of the Durahelmet appearsto owe its inspiration to Romanprototypes ofthe second and earlythird centuries Evidently, weare dealingwitha complexweb ofinfluencesoperating over prolonged periods.

    The redating of the Romano-EgyptianSpangenhelme to the later third centuryeffectivelyprevents them serving as prototypes for the early medieval BaldenheimSpangenhelme. Instead, it is likely that the latter derive from the continueddevelopment of the radial helmet among the barbarian peoplesof Eastern Europe andCentral Asia, perhaps as seenat Kertsch in Southern Russia63. The Baldenheimhelmets are probably versions made by Roman craftsmenfor Ostrogothic masters64.The descendantsof the Dura typeand the derivative Romanhelmets are not tobe seenin the Baldenheimpieces, but in the very differentNorthernEuropeanhelmets of thesixthto eighthcenturies. Here, the findsfromSutton Hoo, Valsgardeand Coppergatein York all reflectthe dominant fourthcenturyRomantype,the bipartite ridgehelmet65.

    AlongtheRhine andDanubefrontier,therewasone 64.Werner,op.cit.(note17)p.182.shieldfactoryperfrontierprovinceandapairofarmour 65.For SuttonHooseeR.Bruce-Mitford,Thefactoriesforeachfrontierdiocese. Clearlythiswas SuttonHooShipBurial, Volume2,Arms,Armourandcarefullyplannedanddoesnotreflectpiecemealdevelop-Regalia, London1978, pp.138-231;Valsgarde,ibid.ment(James,op.cit.,note60). pp.210-4;Coppergate,P.V.Addyman,N.Pearsonand

    63.Post,op.cit.(note17)pp.133-4;W.Arendt, D.Tweddle,TheCoppergateHelmet ,AntiquityLVIEinAltturkischeWaffenfundausKertsch ,Zeitschrift1982,pp.189-94;D.Tweddle,The CoppergateHelmet,furHistorischeWaffenkunde(NeueFolge),IV,1932; York1984.Arendt,op.cit.(note36)abb.2.