________________
46
THE INDIAN ANTIQUARY.
[MARCH, 1905.
In the first paper I overlooked the point that the name of the writing is found not only in the Lalitavistara but also in the Mahāvastu, I. 135. Senart reads there Brahmi Pushkarasari Kharost, and remarks, p. 484, that Kharostt, if the reading is correct, "can only be regarded as a geographical name, perhaps outside India, judging from the form of the word." Senart here agrees with Lévi's explanation, against which I have endeavoured to show that the circumstances rather point to our having the name of a person to deal with. Senart's reading Kharostt is only a conjecture. Among his MSS., N. A. C. M. and L. read Kharostri, B. has Kharastri. All MSS. therefore have stri, which seems to agree with Lévi's explanation of the word. Bendall has kindly looked into the Cambridge MSS. of the Mahävastu and Lalitavistara as well as into the MS. of the Mahavastu belonging to the Royal Asiatic Society, and he affirms that the MSS. of the Mahavastu read Kharostri, and those of the Lalitavistara read Brahmi-Kharoshți-Pushkarasarim (the oldest MS. salim) as in Lefmann's edition. So the readings Kharostri and Kharoshti stand opposed to each other. Bendall further draws my attention to the fact that in the Nepalese writing the only difference between sta and stra is that the curve in stra goes a little further to the left than in sta. Interchange of the two signs is therefore very possible. Indeed there are many cases of it in the Mahavastu. In I. 78, 14, C. M. read sasta for the correct sastra; in I. 100, 7, B. N. A. read trastro, L. M. read trasto, C. read tasto; in I. 182, 12, all MSS. have sastyägära, instead of sastryägaro; in I. 192, 11, N. has sästa for bästra; in III. 1, 6, both MSS. have sty° for stry; in III. 62, 16, instead of the correct sästä, B. M. have sastra. In other groups of letters also, r is often found wrongly. So, in I. 117, 13, grotrena for gotrena; in I. 119, 8, grotro for gotro; in I. 224, 2, érighram for ghram; in I. 364, 7, éushkra° for bushka; in III. 127, 15, sähasriko for sahasiko; in III. 251, 5, prāgr eva for prag eva; in III. 329, 12, prātra for pătra; in III. 880, 2, sagrotram for sagotram. On the other hand, r is by mistake wanting in some variations. Thus, besides in the case already given of stra, for example, in I. 137, 14; 188, 1, rashṭaḥ for rashtraḥ; in I. 280, 16, rashta for rashtro; in III. 400, 2, iotriya° for śrotriya", and others. As Kharosh!t (so the MSS.) stands to Kharostri, so stands ishfika to istrikä which are constantly interchanged in the Mahavastu and Lalitavistara, so that it is often difficult to choose between them (Senart, Mahavastu, I. 568 to 244, 5). The readings of the MSS. therefore cannot decide the matter, especially as the Mahavastu MSS. are very corrupt, and all go back to one manuscript. Just the names of the scripts are very much corrupted [745] in the MSS. of the Mahavastu, as the varia lectiones show. Thus the oldest Chinese tradition always remains the one standpoint for deciding the right name of the writing running from right to left. As to that, it does not matter whether Kharoshtha is a historical person or not. With Franke I believe that with regard to time it is quite impossible to see the writing of Kashgar in the Kharoshtht. At the time of Asoka, as the inscriptions of Shahbazgarhi and Mansehra show, it was well known in Kabulistan and the Upper Indus valleys. But, that Kashgar had at that time so highly developed a culture that its writing could affect the old culture land of Kabul and the Indus, is contradicted on every side.
To what I have remarked in the first article (p. 25 f.) about the formulatory combination of khara and ushtra to kharoshtra, I will here add, that Vamana, Kävyälamkaravṛitti, 5, 2, 28, has the following Sutra: 1 na kharoshträv ity ushtrakharam iti päthat 11: "One must not say kharoshirau as the Ganapatha prescribes ushtrakharam." If kharosh frau is not found in the Mahabharata, kharoshtram is, not ushirakharam, 2, 51, 18, as also in Mann and Yajnavalkya (p. 26 above). In the Lalitavistara, 306, 6, is aśvoshtrakhara, while Aévaghosha, Buddhacharita, 13, 19, has aśvakharoshtra, as most works written in verse, doubtless under the influence of the metre, which must also have dominated in cases like Rāmāyaṇa, 6, 53, 5, Bombay ed., nägair aivaiḥ kharair ushtraiḥ. Finally, as to sta for shtha, and sva for spa, (p. 26 f. above), let me refer to the rich collections by R. Otto Franke, Päli und Sanskrit (Strassburg, 1902), p. 114 and 117, where, with reason, particular consideration is given to the change of va into spa.
