294 Journal of the Asiatie Society of Bengal. |[June, 1911. 
There is an important departure in this plate compared with 
the preceding two. This is the use of the bipartite form of Ya 
side by side with the usual 6th century open-hooked tripartite 
form. Dr. Hoernle has made exhaustive enquiries into the 
limit of the tripartite form of Ya, but as Dr. Kielhorn has 
shown in several places, the exact limit cannot be fixed with 
certainty, but the anomaly of the use of the bipartite with 
the tripartite is evident even to the uninitiated. We have four 
clear cases of the use of each in this grant :— 
Bipartite Form. Tripartite Form. 
ashi tine ti 1 inline 17. WN lanl in line 3. 
ayabhut pee eWinss Ripert i ee 
it dk ssa 
Visthayam 2 9a 20. Menihceahiua in lines 17. ie 
The form of lingual Sa used in this plate is the usual one 
of the Eastern variety of the early Gupta alphabet, but the 
loop is more pronounced in this plate than in the preceding 
couple. The form of Ma also is not the same as in the preced- 
ing ones. It is spice the usual form of the 6th century 
(4) The Grant Samacaradeva: the year 14. 
I have already exhaustively treated the peculiarities of the 
characters of this plate in my previous article. What remains 
for me is to compare the characters of this grant with those 
of the other three dealt with above. The perusal of Mr. 
Pargiter’s able article on the three grants from Faridpur has 
obliged me to modify portions of my reading of Mr. Stapleton’s 
aie: consequently some new statements on the paleography 
_ of this grant will be necessary. It is already evident from the 
above ry page ees that the test letters of this period are 
Ha and La. The test letters of the earlier period—the palatal, 
lingual and dental sibilants—are no longer of much value; in fact 
the palatal Sa remains practically unchanged in form from the 
lst century A.D. to the 8th. In the Eastern variety of the 
characters of the 4th and 5th centuries A.D. the form of the 
lingual Sais indeed a test to prove the Eastern or Western 
origin of the alphabet; but it is on the two letters mentioned, 
of the nature and characteristics of the alphabet. Similarly 
we have another test letter in Ya, which at this time ae, 
from the tripartite form to the bipartite form when it occurs 
wnety: In a former paper I have had to deal exhaustively 
1 This should be read Vikkriyamanakani instead of Vikkriyamanani 
to read. | 
id propose to restore it in the form stated above. We 
mgt as extra ka in the same word in platei, line 11. Ind. Ant., "1910, 
