________________
No. 20.)
THE INSCRIPTION ON THE WARDAK VASE.
203
widest part. Dr. Read tells me that it bears no traces of ever having been gilt, but that parts of the brass might have presented a brilliant appearance like gold, if they had been protected by remaining in contact with certain earth while buried in the tumulus.
The inscription is in the Kharoshthi script and consists of two portions. The first is written in three lines encircling the shoulder of the vase; and the second consisting of one line is written around the broadest part of the vase below the former and separated from it by three lines, and occupies not quite half the circumference. The first is the main portion and is in letters varying generally from 1 to finch in height. In the fourth line or second portion the letters are about inch longer and correspondingly broad. They are formed of minute dents pricked into the metal, and the dents are so close together that from 8 to 10 are generally contained in inch in the first three lines, and from 7 to 8 in the last line.
The two plates annexed hereto, giving a facsimile of the record, have been prepared under Dr. Fleet's supervision from the ink-impressions which also were made under his direction. The shape of the part of the vase where lines 1 and 2 lie is such that ink-impressions of them cannot be made without introducing a gap at some point or another; the place selected was naturally that where the record begins, and a supplementary ink-impression was made of that part of the record and has been reproduced below the complete lines. It shows how the lines rap right round the vase without any blank space in line 1, and with only a small space at the end of line 2, and gives in their complete forms the final ti of line 1 and the initial i of line 2, which are not perfect in the circular impression. To have reproduced line 3 in one piece would have entailed a somewhat cumbrous folding plate, hence it has been treated in three parts over. lapping each other; thus 3-B repeats the last six syllables of 3 A, and 3 O repeats the last five syllables of 3 B and also shows at its end the bha with which 3 A begins. A fourth piece 3 D has been added to show how this line also practically runs quite round the vase, and contains the last seven and the first eight letters of the line.
The style of the script is that of the Kushana inscriptions portrayed by Bühler in Table I of his Indische Palæographie, cols. I to zii. Further specimens of the script have been published lately, namely, the Taxila plate by Bühler (E. I. iv, 54), the Taxila vase by Prof. Lüders (E. I. viii, 296) and the Mathura Lion.capital by Dr. F. W. Thomas (E. I, ix, 135). Mr. E. Thomas in digonssing this inscription spoke of "the little care that has been taken to mark the nicer shades of diversity of outline which, in many cases, constitute the only essential difference between characters of but little varying form" (Prinsep's Essays, vol. I, p. 162). But it seems to me after a careful and detailed comparison of the letters, that the inscription has been carefully made, the letters are generally well-formed, only a few clerical errors occur, and the distinctions between letters which are similar in shape are generally expressed clearly except as regards two particular letters, y and 6. For instance, t and have the same general shape, and are fairly well distinguished, in that the downward stroke of t is usually of about the same length as the horizontal stroke, while in rit is generally about twice as long; and in only one or two places is the difference neglected. No distinction is made between long and short vowels, and the only special form that need be mentioned is né in muna (1.1).
The exception just referred to occurs in the letters y and s, and these are the only letters that cause real uncertainty. Ordinarily y has a pointed top and f a flat top, but the distinction is very far from being observed here and is indeed more often ignored; especially since sometimes the top has an intermediate roundish form and sometimes the letters are distorted by
1 Characters should primarily be read as what they strictly appear to be, and the resultant words accepted and acrutinized. This is essential in dealing with the endless diversities in the different kinds of Prikrit. To rond characters so as to agree with the commoner forms of the Prikrit words intended seems to be almost certain to obscure Prikrit modifications that may be important critically and linguistically. See p. 208, puto 8.
2 D2
