o 



^ ETHyOLOOT OF THE liTDO-PAClTlC ISIAKBS- 2l25 



ble, it is referable to an obsolete sibilant forfc of 11, ri, 2, or to a 

 fuller form of 2 eimilar to the Samoiede Sec, bH &c. The Tibetan 

 liquid li &c occurring in all the dialect?^ preserves the farm mw 

 obsolete in the Chinese 4 although preserved in 2, one of many 

 illustrations of lhd great antiquity of the first diffusion of the Chi- 

 na-Tibetan numerals. Yeniseian has the same 4, stm, siar/t, zmng, 

 ehe^a, sha^a. Tungusian has it in the dental form di^m, dcgen^ 

 dugun. Koriak has e and i formsi n-sha-/ta«?, ng-m-ha &c. 

 Samoiede has the double dental form teti &c which is but a varia- 

 tion of its uiti, side, sin &c 2. Mongolian and Turkish have the 

 same double def, in broader forms dur, der, dor, tir, dor-f tir-t &c. 



The Thochu broad form&, zha, ra, may be from the current 

 Chinese like the BlKttian, and probably also the Gynrung, But 

 the initial def. distinct from the Bhotiao h g- and tlie form) of 

 some of the other Thochu numerals are in favour of its being 

 older. The Horpa hia is a similar broad form. 



5. Chin, ngu, ngo, go, wu, u, ng (Gyami wu)j Tib. kuttg- 

 ngo Gyar., nha, nga Bhot. nga Many. ; gwe Hor., ware Thochu. 

 The Gyarung preserves the Chinese vowel. Bhotian, Manyak and 

 Horpa have a. In like manner where Bhotian has the Chinese 

 o of the 1st prondun^ Gyarung^ Horpa and Manyak have a. The 

 Horpa and Thochu forms are probably local vaiieties but they may 

 have an independent oonnection with Chinese through western 

 varieties similai* to tlic wu, u, of Kwan-hwa and go of Hok*kion, 

 The Thochu ^va may be merely a variation of gwe or of wu. It 

 is possible that in it and even in wu we have a remnant of an 

 archaic labial 5 (Sec. 8). In the eariier numeral systems 5 was 

 generally 1. In those binary systems ivhich went beyond 4, it 

 was 4, 1, that is 2, 2, L In the quinary or hand system 5 was I 

 hand or full tale. The Ghinese ngu is a disiinct form from the 

 def. used as 1 and from the varieties occun-ing in 2, 3 and 4 with 

 the exception of no, 2. It probably belonged primarily to a 

 system which used a similar forn^^as 1, as is sliU the case in Ko- 

 riak. Yukahiri has the nasal def. in 5, 



6. GuiN. lyeu, luh, lo', Ink, loh, la' (Gyami leu) ; Tifl., druk, 

 duk, tuk, ihu, Bhot., ftwtok Gyar.; trufti Many,, Mtf.tarc Tho- 

 chu ; chho Hor. 



The full form of the Chinese is presei vcd in the Kwang-lung luk. 



