G34 Continuations of Observations [Nov. 



to shew the identity of the two classes. The description of them in 

 detail belongs to the next plate, where instead of deteriorating, they 

 will be found to improve, while they become indianized. An oppo- 

 site effect is, however, observable in a second branch, derived from the 

 same stock, which it is difficult to account for, unless by supposing a 

 divided realm, one portion flourishing and patronising the arts, while 

 the other maintained naught but the shadow of its pristine glory and 

 ancestry. This declining gradation is exemplified in figs. 14 (Ven- 

 tura) ; 13 (Stacy) ; and 15 (Keramat-Ali) ; wherein at last it is barely 

 possible to trace the semblance of the sacrificing raja on the obverse, 

 or of the female on the reverse ; although from the insensible gradations 

 in a multitude of specimens, such are undoubtedly the figures. Fig. 

 15, is a very common coin in silver and copper : one was extracted from 

 the Mdnikyala tope, and was then supposed to bear the representation 

 of a crab and a dagger ! That coin, it will be remembered, bore the 

 obvious Nagari letters ^ffaJT. Many others have been since discovered 

 with the same ; and it should be remarked, that the form of Nagari in 

 these differs essentially from that of the collateral branch. 



Here then we have the Indo-Scythic paternity of the Kanouj coinage 

 proved by the best evidence : and now we will proceed to examine in 

 detail its Hindu offspring, before entering upon the natural enquiry 

 whether such a fact is borne out by the meagre remnants of history 

 and tradition that are applicable to this obscure period. 



Plate XXXIX. Hindu Coins — First Kanouj Series. 

 The Deva Nagari alphabet, published with Mr. Wathen's transla- 

 tion of the Guzerat copper-plates two months ago, will be found to 

 apply in every respect to the coins before us : it is also nearly identical 

 with the Gaya and Allahabad alphabets; the principal exceptions 

 having place in the m, which in the latter is written more like *>, r , 

 while in the former it is % • and the s, which is respectively jj in the 

 latter, and (L> in the former. To avoid the necessity of casting a new 

 fount of type to illustrate the following observations, I have availed 

 myself of the pervading similaiuty of the Tibetan alphabet ; which, 

 though several centuries later, can, with the alteration of a few 

 letters, be employed for our purpose much more readily than the 

 modern Deva Nagari. 



The Tibetan r, *-, answers for t of this alphabet : c; for d, and the 

 prosodial stroke J for r. The latter letter when inflected with the 

 a or i vowel, may be represented exactly by the Nagari short and long 

 i, f and "^. For the k, g, dh, and m, I have been obliged to cut new 

 tv P e %> H' Q' X 5 an d f° r the letter which I have supposed from some 

 likeness to the Tibetan j#, to be a, I have given the new symbol h 



