264 
The changes proposed with respect to the first kind were 
these : 
1. The vowel a might be inserted after any primary letter, 
before a vowel, either as a distinct syllable, or as a guna to the 
vowel, as well as before a consonant. 
2. W after u, and y after 7, are in general, both in the 
middle and at the end of words, absolutely mute. When not 
so, they are to be sounded as a, which they implicitly con- 
tain. 
3. Secondary consonants, which are only used before par- 
ticular vowels, are to be sounded in the same manner as the 
corresponding primary ones; and if a secondary consonant 
exist proper to any vowel, and the corresponding primary 
consonant appears to precede that vowel, an a is always to be 
supplied. Ifa secondary consonant be used without its pro- 
per vowel after it, that vowel must be supplied ; 7 is here con- 
' sidered as a vowel. Thus the combination of the letters which 
Lassen calls mi would be mi; while his mi would be me, for 
mai. His fr would be pr; while his pr would be par. 
4. Besides his mistake in giving values to the secon- 
dary consonants generally, different from those of their cor- 
responding primary ones, Lassen has erroneously considered 
the secondary consonant corresponding to d before z to cor- 
respond to k’, i. e. ch; and he has given to three primary 
consonants the values d, z, and 2’, i.e. zh ; the true values of 
which Dr. Hincks maintains were z, zh, and 7, or dzh. 
The second Persepolitan alphabet, it is here maintained, 
consisted of characters representing nine elementary sounds : 
viz., four vowels, a, 7, uw, and er, and five consonants p, h, t, s, 
and n: and various combinations of these nine elements. In 
most cases, two or more characters, phonetically equivalent, 
represented the same element or combination. 
Westergaard supposes a much larger proportion of the 
characters to represent elementary sounds than Dr, Hincks ; 
and he supposes that an a might be inserted, as in the first 
