640 Revision of the Baetrian Alphabet. [July, 



diacritical marks ; of these the i is the best determined, being found 

 applied to almost all the consonants in the form of a small stroke cross- 

 ing the letter. The a is uncertain ; it may be a prolongation below in 

 the r, — a foot stroke or mdtra. The e> I judge from the Manihyala 

 inscription, to be a detached stroke behind and above ; in a few cases 

 only joined. The u may be the loop so often seen at the foot of the 

 written letters. Thus we have >i ha, -n kd 9 +H hi, "K he 9 ~h ho ? ^ ku 9 

 &c. I feel it to be a little premature thus to assign sounds without any 

 positive authority : but it was from a similar assumption of the value of 

 its vowel marks, that I was led to the discovery of the Indian pillar 

 alphabet. 



"With regard to the consonants, I ought perhaps to follow the order 

 of the Hebrew alphabet, but as the language to be expressed is allied 

 to the Sanskrit, it may be more convenient to analyze them in the order 

 of the latter. 



w n, ha. This letter on further scrutiny I find invariably to represent 

 k ; and its place is never taken on the coins by 1 as I formerly sup- 

 posed. It occurs also with the vowel affix i as *h hi ; also, but seldom^ 

 with the u 9 as 3"> hu ; and with the subjoined r as "Hi hra. In the 

 compounds, hla 9 hli, a form is adopted more like the Hebrew q p* 

 (quere • ) "T?, H : there are two or three examples in support of it. 



S, hh 9 is limited as such to the name of Autimachou — but I find it 

 also representing the g in Abngasou. In the written tablets we have 9 

 and 9 and P seemingly identical with it, yet the latter with the vowel i 9 f, 

 is used in some places for dhi (intended for the inflected t. ?) — There 

 is no small affinity between P, 5, and fj, Q, the kh of the old Sanskrit 

 written invertedly. 



"~^> ~^» In » & G*gh ? — I place these forms here because they occur se- 

 veral times in the tablets and they bear some resemblance to the g of 

 the Pehlevi. 



Of the Sanskrit palatials neither the Greek nor the Cbaldaic alphabets 

 contain any proper examples— the ch and j are modified to z and te— . 

 which letters we must expect to find substituted for the Sanskrit clasa 



T, Y 9 cha; l> 9 chha. The first of these forms is found at the close 

 of a series of words terminating each in the same vowel inflection, ', e ; 

 which makes me suppose it to be the Sanskrit conjunction cha 9 uniting a 

 string of epithets in the locative case. As yet I have no stronger argu- 

 ment for its adoption. 



J., or N,/a (tsa?J. The form of the Chaldaic ts 2?, agrees well with 

 the first ; indeed in many coins of Azes the Baetrian form is identical with 



