use of a table, as the frequent errors in published works prove. Thus the second year of the Seleucid era in Babylonia, if obtained by subtraction, is 311/10 minus 2, that is, 309/8. But actually, as one can see in this case by inspection, the second year is 310/9. The error is not apparent with larger figures. It automatically adjusts itself after the beginning of the Christian era, since 311 s.e. is not year 0 but 1 b.c., while 312 s.e. is a.d. 1.
Documents in Arsacid Pahlavi are so scarce that those which exist present many difficulties to translators. Besides a few inscriptions on coins, the single Pahlavi parchment from Avroman in Kurdistan remains the most important item of this character among the sources.
Dated Parthian coinage began shortly after the Parthians entered Mesopotamia, and it continued to increase in volume for some years. This has been and must remain the primary basis for Parthian chronology. Many of the coins, in addition to the year date in the Babylonian Seleucid era, bear also the month of issue, which in times of disturbance is very useful to the historian. Unfortunately relatively few of these coins bear the king's name, and when they do we may assume that two or more men were contending for the throne. The remainder of the coins must, by means of the portraits on the obverses, or on stylistic grounds, be assigned to rulers known from written sources. The task is difficult; but the work of Gard-
