1901.] 
Section III. Manuscripts. 
33 
of its own which connect it with the dialects of the Western Highlands of 
Central Asia. To me it appears that it has its nearest congeners in the 
so-called Ghalchah dialects of the Pamir, the Sariq-qoli, Shighni, 
Wakhi. Munjani, Sanglichi. For an account of these may be consulted 
Dr. Grierson’s ‘‘Languages of the North-Western Frontier,” in the 
Linguistic Survey of India, where also references to other authorities 
will be found. 
In the phonetics of the language the most striking point is that it 
possesses no sonant aspirates (gh,jh, dh , dh , bh) : the guttural y, as a 
non-conjunct, 6 is also absent. With these exceptions, to judge by the 
alphabetic system, it seems to possess all the ordinary sounds of the 
Sanskrit phonetic system, including the cerebrals, the three sibilants 
s, g and s, the four nasals h, n , m, n , and the anusvara. The palatal 
nasal h (initial as well as medial) and the cerebral nasal n (only 
medial) do not occur often, and, as a rule, only in names (e.g., 
Puhadatto , Raliaja) or technical terms (e.g., ksdna) which are sugges¬ 
tive of an Indian origin. The exact force of the palatal and cerebral 
sibilants is uncertain; thus we have gumdasa ‘ sixteen ’ for Sanskrit 
sodaga , and sausa or ssausa ‘ six ’ corresponding to Shighni khhaushkh 
and Wakhi shadh or shaz. Moreover the existence of peculiar dialecting 
sounds seem to be indicated by the occurrence in the script of a special 
sign, consisting in a subscript curve or hook, which is found with 
certain words and letters, and even with these not uniformly, and the 
exact signification of which I have not been able to discover.? The 
quality of the vowels seems to be rather undefined. Thus o and i 
are often confounded; e.g., the word homi, as spelled in the documents 
Nos. 2, 9, 10, appears as himo in Nos. 1 and 4, and as himi in No. 18. 
No. 12, which throughout uses for both vowels but one sign, a kind of 
double dot (properly a modification of the ordinary sign of the vowel 
o, see Table II), spells it hdma. 
With regard to the Vocabulary, I have succeeded in determining 
a considerable number of words, either names, or terms, or numerals. 
Some obviously suggest Indian, Persian or Ghalchah affinities, others 
are peculiar. I may give a few examples. Indian names are : Puhadatto 
(Skr. 8 Punyadatta ), Suhadatto (Skr. Qubhadatta), Barmapuho (Skr. 
Bharmapunya), Pharsapuho (Skr. Spargapunya), Budasamgo (Skr. 
* 
8 The anusvara in such words as samga, may represent a conjunct guttural nasal. 
T An example of the use of the hook may be seen in the word si in Plate I, 
fig. 6, line 1. The letters with which it is found are a, d, k, d, p, ph, b, g, s, h. In the 
transcripts, given in the sequel, they are indicated by an apostrophe placed under them. 
8 Sg. = Sanglichi; M. = Munjani; W. = Wakhi; Sr. = Sariq-qoli; Sh. - Shighni; 
Prs. =*Persian; Ind. = Indian; Skr. =*Sanskrit; Pr.“Prakrit, 
J. i. 5 
