238 Rev. Epwarp Hincks on the three Kinds of Persepolitan Writing. 
as a double letter, not as asecondary syllabic character. The class of secondary 
letters preceding 7 must, it seems, be rejected. 
I must now say a few words on the letter No. 10, as to the value of which 
I differ from Major Rawlinson, and the use of which letter has been pointed out 
to me as an objection to my theory of secondary letters. I feel quite confident 
that my value (founded on Holtzmann’s) is right; that the Major’s is wrong, 
though much less astray than Lassen’s; and that the facts connected with the use 
of this letter are quite consistent with my theory. The objection proceeds on the 
supposition that No. 10, which the Major writes ¢, is connected with No. 7, ¢ ; so 
that, if my theory were correct, it would be a secondary form of that letter ; 
whereas No. 7, ¢ itself, precedes 7 in instances where no a@ can intervene ; as e. g. 
in the third person singular of the present aséz, “he is,” and the like; and 
again, Nos. 7 and 10 appear to be interchanged. As to the first point, I stated 
in my former paper, and still state, that No. 10 is the secondary form of d, not 
of ¢ I admit that No. 7 precedes 7, both immediately and with a intervening, 
there being no secondary form of it before 7; but I deny that di, as the Major 
writes it, that is, 28, 33, ever come together, except in the inflexion of a noun 
ending in da, in the enclitic pronoun dé, or in some other similar case, where a 
guna to the @ is required by analogy. As to the alleged etymological con- 
nexion between No. 7 and No. 10, I cannot discover it. The only apparent 
instance of such a connection is in the imperative of certain verbs, where the 
2, s, the 3, s, and 2, pl, are written with fiya, fuwa, and td, according to the 
Major’s orthography ; di, tw, and ¢d, according to mine. These I compare with 
the Greek and Sanscrit forms. They are the terminations annexed to a root 
which terminates with a vowel. In Greek we have i, irw, ire. Bopp gives 
srudhi as the Veda form of the Sanscrit, corresponding to xAv@.; the other 
persons end in ¢w and ta. Now, it is admitted that, in the old Persian, da cor- 
responded to the Greek Sa or Sy, and to the Sanscrit d’ha ; daddmi was the 
equivalent of riSype and dad’himi ; but it has not yet been admitted that dz 
and du were the old Persian equivalents of 6, Ov, and d’hi, @hu. This, how- 
ever, is what I contend for, as a part of my theory of secondary letters; and it is in 
perfect accordance with this theory that I found above diird as the equivalent of 
$8pa; and that I now find dito be the termination of the second person sin- 
