1901.] 
Section III. Manuscripts. 
35 
i or e or ai indicate the locative (or oblique) case; e.g., bista-mye salye 
4 in the 20th year ’; dasa-mye hadai 4 on the 10th day ’; ssausa-cu salya 
4 in the 6th year ’; Kaji 4 in the (month) Kaja .’ 
An interesting fact is that ten of the complete documents are fully 
dated: also several of the fragments show mutilated dates. I have 
succeeded in reading the dates, but the key to interpreting them is still 
to be discovered. In its fullest form the date is seen in the following 
opening passage of the document No. 8 (Plate VII, fig. 2). 
17-mye hsdni ssausa-cu salya Naha mdQto 17-mye liadai , 
i.e., 4 in the 17th cycle, the sixth year, the month Naha, the 17th day.’ 
But hsdni is usually omitted, as in the opening passage of the 
document No. 13 (Plate II, fig. 6). 
bista-mye salye Kaji maqto dasa-mye hadai , 
i.e., ‘in the twentieth year, the month Kaji, the tenth day.’ 
The month’s name and the numbers are frequently post-posited, as 
in the opening clause of the document No. 12. 
sali 20 mdgto Ghvdtaja hadd 23-mye, 
i.e., 4 in the year 20, the month Chvataja, the day 23.’ 
It will be noticed that the forms salya or salye and hadai are only 
used when they follow the numeral qualified by mye; otherwise sali and 
hadd are used. This seems to point to the former being inflected forms. 
Prom its position in the series, the term hadd (or hadai) can only mean 
4 day.’ For the same reason hsdni should signify a larger period than a 
year. Hence, I have provisionally translated it by 4 cycle.’ But there 
are difficulties. Two hsdnas are named in the documents: the 17th and 
the 19th; and once the term hsani occurs without any number qualifying 
it. In the latter case, as well as in that of the 19th hsana , the 20th year 
is mentioned ; and the highest number of years mentioned in any document 
is 22. It follows that none of the well-known cycles will fit in: the 
12 years’ cycle is too short, and the 100 years’ and 60 years’ cycles are 
too long. A double 12 years’ cycle might suit: from the 6th year of the 
17th to the 20th year of the 19th cycle we should have (18 + 24 + 20 = ) 
62 years. So, after all, hsdni may signify something different: possibly 
it may refer to the number of the register, or of a local division. 
The months are always quoted by their names. I have observed 
nine of these: (1) Sharh-vdri or Sharih-vari, (2) Gvataja; (3) Buhaja 
(or Muhaja ), (4) Khahsdja or Khahsd , (5) Hamtyaji , (6) Nahaja or 
Naha , (7) Jeri , (8) Kaja , (9) Pdhiji. Two others are mutilated: 
* * hhaji , and # i * Ija. Of these names Sharih-vdri or Sharh-vdri is 
evidently identical with the old Persian hsathra-vairya and the modern 
Persian Sliahrivar , the sixth month of the year. No other name seems 
to yield to a similar identification; on the contrary Gvataja (or Gvata) 
