628 Lassen on the History traced [No. 103. 



Semitic alphabets^ the latter may be omitted in them, especi- 

 ally if they be applied to a foreign language ; but this approxi- 

 mation of the vowel system is made remarkable by the pecu- 

 liarity, that a is not treated in the same manner with i, but 

 is considered inherent as in Indian languages. Whenever any 

 Semitic language expresses the short vowels by smaller 

 signs, it does so with a as well as with i and u : whenever it 

 denotes the long vowels by quiescent consonants, similar to 

 vowels, it applies for this purpose ^^ as well as •» and 1. But 

 all Indian alphabets represent, as our coins do. A, I, U, by 

 their own signs only as initials to syllables, but never A, when 

 following a consonant, and the other vowels only by abbrevia- 

 tions. 



The diphthongs, at least 6, do not follow the Indian system, 

 according to which ^ "^ te, tai, ^^ to, as well as ^ tu, are 

 written by abbreviated signs, they do not follow a Semitic 

 system ; but the diphthongs are placed in the line with the other 

 letters, and 6 has in the writing no reference to u ; e has it not 

 to ee ; while instances of the uncontracted diphthongs di and du 

 are wanting. The instance of Eukratides can decide nothing as 

 to the system of orthography peculiar to the language. 



As the diphthongs are written in this language, so were all the 

 vowels in Zend ; but that language distinguishes between long 

 and short vowels, though the former are but amplifications of 

 the latter. 



Now supposing that the characters on the coins were a Semi- 

 tic alphabet applied to an Indo-Iranian dialect, the shapes of 

 the consonants, and the initial vowels, might be considered as of 

 Semitic origin, the principle for the medial vowels would have 

 been borrowed from the Indian system of orthography, while an 

 independent principle was invented for the diphthongs ; and if 

 the orthography of the Zendic language were likewise of Semitic 

 origin, the principle adopted on the legends for only 6, (and e) 

 would have been extended in this language to all vowels. 



This conjecture embraces the postulate, that at the period 

 when the characters on the coins were introduced, the Indian 

 alphabet had already completed the system upon which its pe- 

 cuUar mode of representing the vowels is founded. 



