________________
182
EPIGRAPHIA INDICA.
[Vol. V.
is published herewith, I need not attempt a minute description of all individual characters; but one or two inore general points may be drawn attention to. In deciphering the text, as was stated already by Dr. Hultzsch, a difficulty is occasionally caused by the great similarity of two or aven three different letters. Thus, it is not always easy to distinguish between p and y, between n and I (compare nalini-dalasya, 1. 6), between ch and compare kari chandra-, 1:9), ch and v (compare rachobhira, 1. 32), v and dh (compare vadhú-vaidhavya-, l. 11), or between the subscript # and compare induḥ, 1. 4, and oragendrau, 1.7); and where letters like these happen to occur in proper names such as we find in lines 36 and 43-49, it is impossible to vouch for the absolute correctness of the transcribed text. Another matter which may be mentioned is, that for some letters we have two or more different forms. This is particularly the case with the subscript u, but also, e.g., with l and dh ; (for the forms of a compare Vasudev aya, 1. 1, dyrmanina, 1. 4, pånadyugé, 1. 8, induh, 1. 4, and punitu, 1.5; for those of 1, khalr, 1. 2, Lamuôdarah, 1. 3, and kili-kula, 1. 25; and for those of dh, dhrita, 1.6, and khad9-ayudha, 1, 34). I may also state that the letter r, where it immediately precedes another consonant, is written by the ordinary superscript sign, except in the coujuncts rgg, rnn, and rth, the forms of which may be seen from svargga, 1. 38, Udayakarnnah, 1. 17, and 'tyartham, 1. 15. In the word varandvalt in line 2, the superscript r has been wrongly engraved on the top of an akshara which would be rnná, already without it. The sign of aragraha is not used in the inscription; nor are there any special signs for final consonants. The sign of anu zvára is always written above the line and is nowhere employed in the interior of a simple word, instead of the nasal of one of the five classes and the sign of visarga, differing from the sign which is used in the Deopara inscriptions, is much like an English 8, except that.often, at the bottom, it has a short tail. The language
sign for r is generally used when the sign of the consonant with which is combined has a triangular top, as is the case in conjuncts like rkk, rchchh, rij, rtt, rdd, rddh, ril, roo, etc. Neither of these two peculiarities is found in the Deopara inscription or in the inscription here edited. It is true that in these inscriptions the signs of certain aksharas, such as ku, tu, tra, tri, trai, etc., more or less frequently, have an angular top, but we now here see the triangle; and never is denoted in them by the side-line, described above. [Io lines 1-46 of Vaidyadeva's inscription, recording to Mr. Venis's edition, the letter, as the first part of a conjunct, is omitted by the engraver 36 times,-twice (according to the impressions only once) before y, once before , and no less than 33 times wbere the r would ordinarily be denoted by the side-line. According to my experience, this sideline generally is very thin and shallow in the original inscriptions, so that often it does not sbew At all clearly in the impressions; and, in the case of Vaidyadéra's plates an examination of four impressions, of which I owe one to Mr. Venis himself and three to Dr. Führer, enables me to state with confidence that the engraver is not guilty of so many omissions as would seem to occur at first sight.] - As regards the letter jh, it will suffice to compare the sign for j (which is almost exactly like the jh of the modern Bengali) in jhata in line 41 of the present inscription, and that for jjh in the akshara jhi (not jhi) of wjjhitd in line 21 of the Denpara inscription, with the quite different signs for the same letters in the words jhafiti and w hila in lines 28 and 7 of Vaidyadeva's plates.
The initial i, in the Govindpur inscription, is denoted by two ciroles, placed side by side, with a kind of circumflex above them; and in Vaidyadeva's plates we have two signs for i, one with two circles below (as in iti, 1. 3), and the other with the two eireles at the top (as in ina, 1. 45, end imdi, 1. 66), both quite different from the i of the inscription here edited. [1 may mention that Vaidyadêra's plates furnish two corresponding forms of the rare initial. One of them occurs at the end of line 40, in Pai, where the photolithograpo omits the rertical line between the tico circles, by which is distinguished from i, and wbich is perfectly clear in the impressions, and the other form we have in the word lidna, in line 54, the f of which has been erroneously taken to be ai.]-If I bad to suggest special names for the two varieties of the alphabet spoken of above, I, with my present knowledge, should call that of Vaidyadeva's plates the Pala, and the otber the Sena variety.
1 The same signs, wbich of course owe their origin to the fact that the sign for W written on, not above, the top line, are used in the Deopara inscription and elsewhere.
The sime mistake was made by the engraver of the Gauhati plates of Indrapalavarman (Jour. Beng. 41. Soc. Vol. LXVI. P. I. p. 123 ff.) in the word aranapa, Plate iia, 1.6, compare the proper sign for rape, without the en nereerint rien for r, in varana, ibid. Plate lib. 1. 2. The sign transcribed by anya (corrected to rinva). ibid. Plate iia, l: 3. is really ranya in the original. Whether in the Gaubati plstes, in the conjunet reg is written on or above the line, it is difficult to decide.
The two eireles were joined, so as to enable the writer to form the sign of risarga with one stroke of the pel. To a similar process we owe the form of the initial i, bere used.
