3.2 Silver

The last major issue of siliquae in the west was struck at Milan some time between AD 397 and 402 (RIC X, nos 1227-8) and these Milan pieces have been found in great numbers in Britain, both in hoards and as site-finds (PAS record DUR-CE5622). Siliquae of Constantine III struck in Lyon in AD 407-8 are also known from the Hoxne, Coleraine and Haynes hoards but they are not numerous (Guest 2005, 146, nos 752-3; Robertson 2000, 405-6, no. 1621; Inscker and Orna-Ornstein 2009, 385, nos 97-8; AVGGGG refers to Constantine III, Arcadius, Honorius and Theodosius II).

Very few silver coins dating to the period post AD 408-411 have been found in Britain, although as already noted, the production of silver coins in the Empire had declined markedly after AD 402. Only two later silver pieces of Constantine III, struck between AD 408-11 have ever been found in Britain, one at Richborough and one in the Patching Hoard (Reece 1968, 200; Orna-Ornstein 2009, 392, no. 43). The Patching hoard also contained a siliqua of Theodosius II from Trier, struck c.AD 425-30 (Orna-Ornstein 2009, 392, no. 44). Two similar siliquae of Theodosius II were found pierced for use as jewellery alongside a later siliqua of Anthemius (AD 467-72) in an Early Medieval grave at Chatham Lines in Kent (Blackburn 1988). As with the gold issues discussed above, it is unlikely that these coins formed part of a coherent currency system, but rather were viewed as bullion or jewellery or perhaps as both.

The phenomenon of 'clipping', whereby the outer circumference of the coin is removed is a characteristic of many siliquae found in hoards and as site-finds in Britain (PAS record IOW-3818C7). While it has been argued that siliquae were clipped in order to reduce the weight of coins to match silver issues of the Visigoths and Vandals (King 1981, 9), it is more likely the silver was reserved as bullion (Abdy 2013, 109) or to make copies (Guest 2005, 146ff), allowing the remaining coin to continue in circulation (Burnett 1984). The head of the emperor was not touched by clipping, although the inscriptions could be totally removed. This respect for the imperial portrait suggests that, even if of reduced weight, these clipped siliquae still played a role as currency.

The exact chronology of the phenomenon has been the subject of vigorous debate for over 30 years, although it is now generally agreed that it is in some way connected to the cessation of coin supply to Britain. The virtually unclipped Terling hoard has been used as evidence that the practice began after AD 404 (Burnett 1984), analysis of the Stanchester hoard pushed that date forward to c.AD 406 (Abdy 2006, 84; 2013, 107-9; Abdy and Robinson 2009) while a comparison of the date of clipped siliquae and imitations in the Hoxne hoard has been used to suggest clipping occurred 'for several years, perhaps decades' after AD 409 (Guest 2005, 114). Meanwhile, Portable Antiquities Scheme data illustrate that the later the date of issue of the siliqua, the more likely it is to be clipped, providing a further indicator of an early 5th-century date for the phenomenon (Walton 2012, 111).

In addition to silver coinage, it appears that ingots were increasingly used as a store of wealth and means of payment in the late Roman world and its peripheries. For example, the Coleraine hoard from Northern Ireland contained a mixture of clipped siliquae, terminating with a coin of Constantine III (AD 407-8), hack-silver from Roman plate, and ingots of fixed weights (Robertson 2000, 405-6, no. 1621). A similar hoard with hack-silver and clipped siliquae was found at Traprain Law in East Lothian (Robertson 2000, 402-3, no. 1617; Guest 2012, 100-2). In the Coleraine hoard, there were fragments of official silver 'ox-hide' ingots, but also unmarked ingots: two flat ingots of good-quality silver that weighed a pound each and three smaller finger ingots of lesser weight that contained more trace elements, suggesting a wider range of metal sources (Abdy 2013, 110-11). An 'ox-hide' and official ingot were also found in the Canterbury Hoard, and a finger ingot has been found in excavation at Vindolanda (Robertson 2000, 1541; Wiegels 2003, pl. V, 1-2; PAS record NCL-62C367). The presence of these alternative stores of precious metal in hoarding contexts may be an indication of the gradual demise of coin use and its replacement with bullion in a variety of forms.