Proc Soc Antiq Scot, 112 (1982) 285-294 Romae Th n coins from Trapraiw nLa M F Sekulla* ABSTRACT cataloguw ne coie A th n f eo find s beeha s n prepared. Analysi resulte th f so s suggests that coins only circulated in the local economy in the first and second century AD when the army was present in Scotland. INTRODUCTION The extensive settlement site at Traprain Law in East Lothian has always been of great importanc archaeologicae th n i e l intra-murae studth f yo l zone durin Romae gth n period. This is almost entirely the result of the major excavations carried out there, chiefly in the first quarter of this century, which were summarized on almost an annual basis in these Proceedings between 1914 and 1923. These reports detailed the occupation deposits encountered and noted the artifact collections in a summary form. The Roman coins were also published at the same time but were generally described onl theiy yb r denominations, empero occasionalld ran reversy yb e type. Over 10 % of the total were dismissed as illegible and a proportion of the remainder were misidentified at the time of publication. As such, the published record of the coins - which comprise the largest bod numismatif yo c evidence fro nativy man e sit northern ei n Britai n- presente unsatisfactorn da y source from whic coin e workho t th l s foun.Al d between 191 192d 4an 3 were recently re-examined at the National Museum of Antiquities of Scotland, together with those from Cruden's excavations in 193 othed 9an r chance finds mad t Trapraiea n betwee presene n th 189 d 8tan day. The total number of coins from Traprain is 65, excluding those from the silver hoard theid an , r overall distributio bess ni t viewed when presente mannea n di r uniform with other sites (fi . Thi bees g1) sha n achieve constructiny db ghistograa m based hundre e upoon na d coin sample usin formule gth a devise Ravety db z (1964 Cased )an y (1974) t musI . notee b t t da the outset however that a list of 65 coins is a very small number upon which to base any calculation and that as a result the visual indicators on the histogram may be less representative of the pattern of coin loss pertaining to the site than might a larger coin sample. The general factors which might cause the fluctuations in a coin histogram have been discussed elsewhere (Casey 1974), though the degree to which these will operate on a non-Roman site is a matter of conjecture. The mechanisms by which Roman coins were acquired by the native population and how they were subsequently used and lost remains almost totally unknown. The elucidation of even basic principles of coin circulation among the local inhabitants of Scotland is handicapped b yshortaga f comparativo e e material ,laca mainlf adequatelo ko t e ydu y provenanced coins from other known settlement sites (Robertson 1971). The coin histogram for Traprain naturally divides itself into three sections: up to 160 AD, 160 to 250, and finally from 250 to c 400 AD. * Department of Archaeology, University of Durham. 286 PROCEEDING SOCIETYE TH F SO , 1982 7 2 6 2 5 2 4 2 3 1 2 2 22 13-1617121 0 1 2 1 9 1 0 81 9 8 7 123 6 5 4 coin issue periods Fro 1 Traprai annua- w nLa l los10r 0spe coins (n=65) SECTION ONE, 78 TO 160 AD Before the major campaigns of Agricola in Scotland the local population in the area had little or no contact with Roman coinage. Their first chance to acquire coins will have been as a result of contact with the army. During the initial movement of the army through the Lowlands t unlikelii t y that much coin will have changed hand s- supplie more ar s e likel havo yt e been brough frop reate u mrequisitioner th ro d rather than bough perhaps cashr wa tfo t I . s only when the Lowlands were garrisoned and a more stagnant military presence established that the army, or their civilian followers, are likely to have had such contact with the natives that trade for local goods or services may have come about and coin changed hands. It is excessively difficult to demonstrate with confidence which coinnativa n so e site could direc a hav d eha t military origin. Coin numbe catalogue rth 1n 1i e provide instancn sa e whic poiny hsucma o t hlinka . Assesf o Domitian dating to 86 AD are commonly found on Flavian military sites in Scotland in unworn conditio Professod nan r Robertson (1968; 1975 argues )ha d that these comprised parcona f o t- signmen coinf o t s importe furniso dt arme hth y wit thesf ho paye e occurOn . t Trapraisa onln i y slightla y worn conditio becausd nan e there direc o appearn e b t militaro st y activity evidenced site th e n ther o strona s ei g likelihood that this coin derives fro mmilitara y source, although precisely how it was acquired and then lost is impossible to detail. Indeed, the other Flavian or earlier coinhavy ma se resula arrivesite s th ea f trad o tn do e wit Flaviae hth n army-buo t t sugges tthef thao sucd l m origin ha tal h a n woul mose db t unwise hige .hTh degre weaf eo r visible on severa thef o l m suggests instea dsomewhaa t later tim arrivaf eo Scotlandn li . This, however, presumes that they circulated elsewhere for a long period before loss at Traprain. A high tota coinf o l nexs si t apparen reige th Antoninu f r no tfo s Pius (Perioestabe Th . -d7) lishment of the Forth-Clyde frontier and the incorporation of the lowland zone into the Roman Empire clearly provide contexsa movemen e th r fo t morf o t e coinage int areae oth . Muce th f ho SEKULLA: THE ROMAN COINS FROM TRAPRAIN LAW j 287 currency arriving during this time may have been of Antonine date but a representation of earlier reigns shoul expectee db d (Reece 1974). Indeed e patterne b thes th ,n a r yca f wearfa so o s n -i coine judge th Nervaf so n o d- , Traja Hadriad nan Flavia e somd th nf an eo n issues woult dno preclude a date of loss in the reign of Antoninus Pius or even later. The coins of Pius themselves show very little wea thid sran argue limitea r sfo d perio circulatiof do n before loss. This observa- tios importannha t implications pattere Th . f coinno s los t Trapraid a t 2n d an n t durin1s e gth centuries suggests tha militara t y presence brought abou briea t f influ coinagf xo e int aree oth a and that this coinage continued to circulate only as long as there was a military occupation. When this was removed the coins in native hands quickly ceased to be used, as both wear patterns on the coins - together with the overall pattern of loss - imply and that as a result there was no circulating Roman coinag Scotlann ei d very soon afte demis e Antonine th r th f eo e frontier. SECTION TWO, 160 TO c 250 AD The withdrawal from the Forth-Clyde frontier coincides with the start of an hiatus in the coin record which lasts for a period of nearly one hundred years and a corresponding break in the occupation of the site has been suggested based upon this limited evidence (Burley 1956). However, such a conclusion can only be regarded as valid if two concepts concerning the Roman coinage of the period can be positively demonstrated. The first of these is that the supply of coinage to Traprain was continuous - or at least as dependable as it appears to have been within province th f Britanniaeo bees ha ns suggesteA . d above thi unlikels si havo yt e bee case nth e becaus presence eth arme th f yeo profoundly affecte coie th dn supply removae hige Th . th hf o l intensity of military occupation from much of the lowland zone in the early 160s will have lowered at a stroke the opportunities for coinage later than the reign of Pius to have changed hands. Contact which allowe exchange th r dfo coinf eo seve- n supposing that such exchange still took place - will have been rendered more difficult if only because the closest source of coinage was more distant simila.A r patter havy nema pertained durin withdrawae gth Severae th f o l n forces of campaign from Scotland. The second concept which requires positive demonstration is that any supply of coinage during this period will be archaeologically detectable in the small coin sample mosn O . t site reductiosa frequence th n i coinf yo s los Aeviden0 s i tD 25 o t t 0 fro16 mc (Casey 1974; Reece 1973). Part of the reason for this is a change in the pattern of coin use with a trend toward highesa r unit valu currencf eo y- thi s centuroccurlatee d th agai2n rd n si y an n earle in th decrease e y 3rdTh . d frequenc coif yo n los siten s o thit sa s time make mort si e unlikely- even in the case of 'continuous' coin supply - that coins of this period will be included in a small coin sample .e cas Thith t Exete a es i s r wher a recentl n ei y publishe coin0 d9 lisf so t from excavations there was a gap from Faustina II to Postumus (Bidwell 1979).