1853.] Sifdn and Sorsok Vocabularies. 127 



that the analogies and affinities indicated by the last series of voca- 

 bularies between the Himalayan and Tibetan tongues on one hand 

 and the Indo-Chinese on the other, are carried on and confirmed by 

 some of the present series, whilst others extend the links to the 

 Altaic group of languages ; the Gyarung, Takpa, and Manyak car- 

 rying the chain of connexion onwards from the South-east, and the 

 Thochu, Horpa, and Sokpa, transmitting it over the Kwanleun to 

 the North and West ; the Gyarung by its grammatical structure 

 exhibiting also marvellous correspondencies with remoter regions ; 

 with Caucasus, as has been separately shown already, and with 

 Oceanica, as will appear in the sequel of this communication. How 

 far precisely the other languages now submitted may participate 

 these express and peculiar features of grammatical affinity I am not 

 yet prepared to say. But the whole of them certainly exhibit a great 

 general resemblance in the broader traits of syntactic,* and yet a 

 greater in those of etymological construction. In a word they are 

 evidently members of that single and vast family of languages the 

 singleness and the vastness of which I conceive to be justly inferrible 

 even from its vocables : 1 st, because of the similarity of the roots, 

 2nd, because of the similarity of the serviles, 3rd, because of the 

 similar principles governing the uses and the mutations of both, and 

 the consequent composition and the character of the integral words 

 which exhibit an essential identity in numberless terms of prime 

 necessity after due allowance for synonymous changes in their roots 

 and for euphonic and differential changes in their serviles within 

 known limits and upon a demonstrably single plan. And I infer that 

 the differences characterising this vast family of languages, however 

 striking at first sight, are subordinate, because when the languages 

 are examined upon abroad enough scale these differences are seen to 

 pass away by insensible gradations. Such as they are, they arise 



* I may instance the universal substitution of a continuative participle in lieu of 

 conjunctions and of conjunctive (relative) pronouns, because this feature has been 

 supposed to be specially characteristic of the Altaic group. It is no more so than 

 the vocalic harmony of Turki, or than the inverted style and tonic system of the 

 Indo-Chinese tongues. These appear to me to be blending differences of degree 

 only, not absolute differences of kind, and to have been used to sever unduly the 

 several groups. 



