236 Rev. Epwarp Hincxs on the three Kinds of Persepolitan Writing. 
distinct argument in favour of the identity of sound of one primary letter and 
its secondary, those which Major Rawlinson calls, after Lassen, wand v. When 
a root ending in w, which, as I have already observed, when it stands alone, is 
terminated with ww, enters into composition with a word beginning with a con- 
sonant, or combines with a termination so beginning, the w is dropped, as par‘u- 
zandnam, am’utha, from the roots paruw, am’uw, in Sanscrit, puru, amu. 
According to my method, the roots are paru, amu, and the derivatives paru- 
zhandném, amu-za. Before a word or ending beginning with a vowel w should 
in all instances be inserted. If the vowel be a it is so; thus, in II. 8, we have 
uwaspd umartiyd, evimmos evavdpos, where the root vi bene takes a w after it 
when preceding aspa, equus, though in the following word, where it precedes a 
consonant, no w appears. Now it is well worthy of notice that, when the second 
part of the compound begins with 7, the secondary form of w before 7 is imva- 
riably interposed instead of the primary w. Thus, from Babir'ush, “ Babylon,” 
as Major Rawlinson writes it, he deduces Babir'uviya, “a Babylonian.” 
According to my method, the words would be Babirush, Babiruwiya, the 
secondary w, No. 31, expressing not v, but the simple semivowel w, introduced 
for euphony, or rather necessarily sounded after w, when another vowel follows 
it without a suspension of the utterance. 
It appears to me that these arguments completely establish the correctness of 
the three general principles contained in my paper. In the two first of them I 
have been partially anticipated by Major Rawlinson and by Holtzmann, who, 
however, do not apply them to one-half the cases that I do; but the great prin- 
ciple of secondary letters I believe to be altogether my own, and it affects the 
mode of reading of a very large proportion of the words of the language. With 
respect toa few of the letters of the alphabet, I have to correct the values which 
I gave in my paper; but no change which I have to make is inconsistent with 
the three general principles above mentioned. I will now briefly notice the 
new light which the Bisitun inscription throws on the values of individual 
letters. 
I have to notice, in the first place, the discovery of a new letter, «= » which 
Major Rawlinson values as n’. It is m(w), or the secondary form of No. 8, be- 
fore uw. Another new letter, which he calls fi, is, in fact, the Median character 
No. 4, as he himself remarks. It was, I conceive, used by mistake by a Median 
=. 
