12 
R. C. Temple — An Examination of the [No. 1, 
It is not difficult to see tlie relics of the old Sanskrit an in th© 
following 
1 ek-am, (?) bi-n. 
2 wi-ng, dd-n, (?) sa-nni, (?) sau-nan. 
3 wur-iin, tri-n, (?) te-g. 
4 sa-n, air-an, ahir-in, (?) chau-g, (?) chau-ga, 
5 phanti-an. 
6 sha-nk, shi-n. 
7 satli-an. 
8 ath-an. 
9 (?) wa-n, naw-an. 
10 zii-n, dah-an. 
15 pandah-an. 
20 woli-an. 
25 pansoh-aii. 
50 pantsah-in. 
j pa-n. 
1 (?) ka-n. 
2 do-n, 
2J day-an. 
4 sa-n. 
10 dah-an. 
We may perhaps see sati in li-sa, 20, Changari, and the dtt termi¬ 
nation in the Khurasani may be explained to be hath, han'd. 
But the majority of the terminations seem to be untraceable, espe¬ 
cially the remarkable ones of the Naqqashi, viz., ha-ndish, ye-ndir, tsau- 
ntar, ath-wotar, ath-watir. Ya-ndir, yi-ndir occur, too, in the Zargari 
Kashmiri. Taking wotar and watir into consideration and dropping the 
n of the others as phonetic, we get dish, dir, ter, wotar, watir for the true 
terminations. These are comparable with the idiomatic utar and otar 
of Hindi.t With handish may be compared the Tibetan nish, 2. 
In Naqqashi occurs hasht-berik for 8, with which may be compared 
the Khurasani numerals. 
Besides the above we have, and all apparently untraceable termi¬ 
nations, 
* [This is very improbable. Don 2, trin 3 correspond to the Prakrit doni, tinni 
(Skr. trini) with the nent. plnr. termination ni. —The wotar, ivatir is the Skr. uttara ; 
thus athwatir-elcam 9 would be Skr. ashtottaraikam “ one added to eight,” whence by 
a not uncommon mistake athivatir (‘added to eight/) is taken to mean ‘ eight ’ ! 
Similarly in the case of tsauntar. Ed.] 
f Kellogg § 184 (a), p. 166. Platt, 51, footnote. 
