1840.] from Bactrian and Indo- Scythian coins. 341 



cannot have occurred before *7, either upon this coin, or upon the 

 others of Philoxenos, I must maintain, that the native orthogra- 

 phy had no K at all in the name. But supposing, the line I 

 belonged N to the preceding H, and that there must have been a d 

 there, which in this form cannot be H, I can only admit, that 

 the cross line was half corroded, while both final lines were too 

 much protracted ; in a word, that we must recognize here, Mr. 

 Prinsep's figure, *. 



The assertion, that the second symbol in the name of Philoxe- 

 nos,* which I restore thus, 21^*^ is /w, is at variance, how- 

 ever, with the more common mode of representing the omicron 

 by an inherent A. We think this objection so important, 

 that we shall adopt + for lu at first only in the way of 

 conjecture. It is countenanced only by the name Agatho- 

 kles, written in the old Sanscrit character, which is rendered 

 Agathukla, and consequently the authority exists in at most only 

 a contiguous language. With more confidence we maintain, 

 that the natives did not retain the k in the name of Philoxenos, 

 and in that case an absorption of k with the sibilant must have 

 taken place. There is, however, no analogy with this absorption 

 in the Pracrit; here x (^f) is altered into kkh k\) 5 tne language of 

 the coins does not always follow the rules of absorption in Pracrit, 

 this is proved by hirmajo ; it does not abound, as the Pracrit, 

 with reduplicated consonants; this is proved by Apollodotos, which 

 is expressed Apaladato. We must therefore look out for other 

 analogies, and here the Zend supplies us with the very striking 

 fact, that it often substitutes sh for ^ (ksh) of the Sanscrit. 

 Thus dashina (the right) for daksina, ashi (eye) for akshi. I 

 therefore conclude V to be sh (sch) . 



But why should this sh occur in Lysias, which has but the 

 s in Greek? I shall be perhaps reminded, in answer, of the 

 Sanscrit rule which might have been originally observed in the 

 Zend, according to which s, preceded by any other vowel than a, 



* Mr. James Prinsep's last reading of this peculiar name from the Bac- 

 trian legend was as follows, T^Hl Phila-senasa or Philasinasa. See 

 page 652, vol. vii. July, 1838. It is this name and that of Lysias that proved 

 the p to be an S H. T. P. 



