60 



On the Mongolian Affinities of the Caucasians. [No. 1. 



Eight. — Yat-sh in 



Yat-sh in Limbu. 



g-Yet in Takpa. 

 Ka-ya in Kami. 

 Ba-ya in Tangus. 

 Ei-ya-t in Mrii. 

 Ee-ya. in Kiranti. 

 Y6 in Snnwar. 

 Or-yet in Gyariing. 



Eight. — Ewa in Georgian. 

 Ed-a in Mingrelian. 

 Ee-ya in Kiranti. 

 p-Ee in Murmi. 

 Eyie in Horpa. 

 Ea-nit in Mm. 



Nine. — b gd in Circassian, 

 d gd in Tibetan. 



r gd-re in Thochu, 



Gii-bi in Manyak. 

 Gdh in Sunwar. 

 Gu-n in Newari. 

 Go in Horpa. 



Kan-gii in Gyarung, 



Tshe- ^ 

 4. i, 4- i. I Final sh', servile. Another beauti- 

 tsnentsn. > e , ' , « «. ., 

 [ ml sample ol amnity. 



g, servile, = v, d, p, below. 



Ya the root througbont the whole 

 series, with the common vocalic 

 changes. 



' Or' servile, in Gyarnng. 



Ea, Ed, Ee, is the root beyond 

 doubt, though the Kiranti sample 



■ under both this and the preced- 

 ing head shows how readily roots 

 become serviles and vice versa. 



f Note again the wonderful accord of 

 \ root and servile. 



f The ra particle here appears both 

 \ as prefix and suffix. 



Bi servile, as in Circassian. 

 The pure root. 

 N, final, servile. 

 Nude root again. 



{Kan, double servile, Ka-na = Kam 

 in Kampa, voce skin. 



Ten. — p She-n in Circassian. 

 Zhe-ba in Abassian. 

 Swa-ba in Circassian == 

 Su-a. 

 b Chii in Tibetan, 

 t Sha-i in Burmese, 

 h Sd in Kami. 

 Chi in Garo. 

 ta-Chi in Gyarung. 

 Shi in Chinese. 

 Sha-i in Tangus. 

 ta-Shi in Tunghlhu. 

 Si-sii in Sak. 

 t-Sa-u in Talien, 

 p-Chi in Takpa. 

 Chu in Serpa. 

 Cha in Gurung. 

 Cha in Lhopa. 

 Sa-n-ho in Newari, 



fSi 



,, Cha, is the root with the usual 

 cycle of changes by aspiration 

 and by alteration of the vowel ; 

 and to the root, moreover, are 

 added the usual variety of servile 

 appendages in some cases, whilst 

 in others we have the nude root. 

 All this is perfectly conformable 

 to what has been seen in the 

 nouns, and it follows therefore 

 that the peculiarities commonly 

 ascribed to the numbers do not 

 really exist. The nature of the 

 error, as derived from the exami- 

 nation of a few only of these 

 tongues, may be appreciated by 

 adverting to the remarks in the 

 next paper on the differences 

 presented to all such observa- 

 tions. 



