________________
288
EPIGRAPHIA INDICA.
[VOL. XIV.
tion, I prefer to see in tapuvae the genitive of a female name Tanuda which would then signify the founder of the Bodhisattva chapel.
The various forms of the name which is here written khushanasa have been discussed by Dr, Fleet, who did not then know the present record. The initial aspirate reminds us of the x in the Greek legend of some of the coins of Kajula-Kadphises ; 9nd, if we compare these two writings, it becomes probable that the first letter of the name was originally a gattural spirant, and, to judge from the g in gushanasa in the Panjtar inscription and gushanain the Mäņikiāla record, a voiced one. I have no doubt that Count Staël Holstein is right in combining the name kushana with Chinese yüe-chi, older güt-shi or gür-shi. I am further convinced that & short form kushi, corresponding to yüe-chi, occurs in koshano, which is used in the coin legends of Kanishka and his successors. Dr. Fleets considers this koshano as an adjective in the nom. sing. I hope, however, to have proved that the whole legend is composed in a language which is, to all practical purposes, ideutical with ancient Khotani. The only exoeption would, if Dr. Fleet were right be the word koshano. In old Khotani this form can only be the genitive plural of a koshi; it cannot by any means be explained as a nominative singular of an old a-base. In that case it would have sounded koshani. Dr. Fleet thinks that, if we explain koshano . 48 A genitive plural, we must arrange the words of the legend "in an order in which they were not intended to be taken." In two Junnar inscriptions, however, we find a genitive plural denoting nationality after a personal name exactly in the same way &* in the Kanishka coin legends. Compare No. 1154 in Professor Lüders' List, Yavanasa Irilasa Gatana, "of the Yavana Irila, of the Goths," and No. 1182, Yavanasa' Chitasa Gatāna, " of the Yuvana Chita, of the Goths," where the Yavanas (ie. Europeans) Irila and Chita are evidently described as belonging to the people of the Gatas, i.e. Goths.5
Though I agree with the Count in assuming the existence of a short base koshi in such furms 89 koshano, I am unable to understand how he came to deny the existence of another form kushana or kushāna. The new Taxila inscription proves, es remarked by Dr. Thomas, tho existence of such a word, which is in itself a very likely one. Compare the two forms a-she and a-she-na which the Chinese ased to denote the ancient royal family of the Turks. The only difficulty is whether the first a of this word is short or long. This question would be solved if the reading kusha naputro in the Mät inscription were certain. There is, however, a mark above the na in this word, and Count Staël Holstein reads kushānań putro. I therefore prefer to read kushana, leaving the question as to the length of the a open.
In the last line the photograph' favours the realing nivanae, as stated by Dr. Thomas. M. Boyer explains a-de as Sanskrit agratah, priacipally, and samaparichago as Pāli sammāparichchago. Instead of the latter it would be possible to assume sa me parichago," this my gift."
The most important qnestions connected with this record are its date and the identity of the Kushana ruler mentioned in it.
Sir John Marshall has shown that the record was found in strata which belong to the Kadphises kings and are deeper, i.e. oldər, than those of the Kanishka group. So far as I can see, his excavations have finally established the priority of the two Kadphises kings to the Kanishka group. On the other hand, Sir John has left the question open which of the two Kadphines kings is meant. He says, “The next important point is to determine which of the Kushan kings is referred to as reigning in that year. That he is identical with the nameless Kushan ruler mentioned in the Panjtar record of fourteen years earlier, is probable ; 1 JR48., 1914, pp. 869 1.
: SBAW., 1914, pp. 645 f. . loc. cit., p. 37).
• ZDYG., 68, PP 93 1. See Konow, JRAS., 1912, pp. 890 ff.
• JR48., 1914, p. 990. * JR48., 1914, pp. 977 t.
