1853.] Sifdn and Horsoh Vocabularies. 129 



using the prefixed serviles, whence results at once all the difference 

 of soft polysyllabism or harsh monosyllabism. The resulting dis- 

 parities of the vocables are certainly often very marked, as in the 

 "Watii and Ui instance of G-yariing and Circassian, (so singularly 

 confirmed by the Malay and Tagala itii, that) as well as in those 

 given at the end of the present series of vocabularies, so that it is no 

 great wonder that the Mongolidan tongues have been referred to 

 many groups so trenchantly separated as virtually to fall under 

 different families. And, if I incline so strongly to unitise the family, 

 it is only because, as far as my investigations have gone, I have been 

 able to discern nothing absolute and invariable in the distinctions — 

 which though no doubt distinctions proper to the vocables only and 

 not affecting structural diagnostics (in the usual narrow sense, for 

 composition of words is structure) are yet unusually and as I con- 

 ceive decisively important owing to the extremely inartificial charac- 

 ter which belongs to the grammar of these tongues with some appa- 

 rently borrowed exceptions, such as that of the Turkish verbs. Not 

 that the grammatical or the physical evidence of this assumed family 

 identity conflicts with that of the vocables* — much the contrary, as 

 we shall soon see — but that the latter has unusual relative value. 

 And, would we speak plainly, we should say that grammar relates 

 equally to the construction of words and to the construction of 

 sentences, and that the former sort of putting together or syntax is 

 always equally, and often more, important than the latter. Certainly 

 it is more so in the Mongolidan tongues which are as much distin- 

 guished by their immensity of nicely discriminated terms,t most of 



particles and hence his remarks on the silent letters want point and significance. 

 The language of Nepal proper is remarkable for its numerous tones and its scanty 

 serviles, whether literal or syllabic. 



* I may mention here an interesting sample of this identity derived from the 

 substantive verb. It is ' da' in Myamma, a-da in Malay, da in Horpa, gdah in 

 Tibetan, dan in Uraon, &c. So also it is mena in Sontal and mna in Tibetan ; 

 and again, it is dug in Tibetan, dong in Bodo and Garo and du in Newari. 



f See vocab. voce ' give' and ' take.' A Tartar cannot endure that confusion of 

 the precative, optative and imperative which our imperative mood exhibits. But 

 he remedies the defect not by the multiplication of grammatical forms but by the use 

 of distinct words, or distinct multiplications of the same word, thus Davo solicits 



