I do not intend to comment in detail on all the errors of the previous publication of these pieces, but certain important misconceptions must be removed. At the time of the publication of Hoard I, I followed the dating of Webb in Mattingly and Sydenham, Roman Imperial Coinage, Vol. V, Part 1. This led to the confident assignment of a number of Antoniniani to the year 256, and that, in turn, has been used as a fixed point for the capture of the city. But Webb's dating is quite unreliable, as Alföldi has shown, and his authority can no longer support the year 256. Moreover, I was too much influenced by his attribution to mints in which he had followed Voetter's earlier arrangement without using Voetter's own rectification. Consequently, I was led to assign a number of pieces to the mint of Rome. But examination shows that there are only five such pieces, Nos. 36, 37, and 38 above, and three of these being dated in 253, the other two in 255-6, they will not affect our chronology. As will be seen, the single dated type from Dura is No. 1, struck between December 10 and 31, 253- It may be remarked that of the 17 reverses which Alföldi gives for Valerian from Antioch, Emission 1, Group 1, we have all but four represented; all but one of Group 2. We lack seven of the 16 for Gallienus of Emission 1.

The contents of Hoards I and X are closely parallel. So are their conditions of burial, for both were found in houses under the great embankment along the city wall. Their coins, therefore, must have come into the city not merely before its capture, but before its preparation for the final assault. All belong to the first emission from Antioch and our date for the embankment consequently depends on our calculation of the duration of this emission. of the first group, with the short obverse inscription, we know of 17 types; of the second group, 8 types, of which we lack one. Supposing that the emission began late in 253, this would seem to be a sufficient number of types to provide for the years 254 and 255,18 and as has been said, the likeliest occasion for the change from one obverse inscription to another would be the beginning of a new year. With all due recognition of the uncertainties, therefore, we may conclude that the coins in Hoards I and X were struck through a large part of 255. Allowing a reasonable time for them to come to Dura, the embankment can hardly be dated before the end of 255. Indeed, if the second emission followed the first directly and did not begin until the middle of 256 the date may be put well along in that year.

The coins found with the corpses belong to a different category. They certainly represent the pocket money of soldiers engaged in the actual defense of the town, killed in a fight in one of the besiegers' mines, and therefore might be expected to include pieces somewhat later than those buried under the embankment. What they do include is a number of coins from the second Syrian mint, not found with the hoards, as well as one Roman piece of 253. Among the chance finds there are only 8 from the second Syrian mint, and this scarcity is in itself indication that the mint was not located at Samo-

18. The Second Emission which lasted, according to Alföldi, from mid-256 through 257 has three types for Valerian, but supposing the distinguishing marks on the reverses (see Alföldi, p. 47) to represent     separate issues, there would be, in effect, six types. But no such argument from analogy can be considered conclusive.


Created by the Digital Documentation Center at AUB in collaboration with Al Mashriq of Høgskolen i Østfold, Norway.

990204 PN - Email: hseeden@aub.edu.lb