EfHN'OIiOftT OP TUB INHO-FACIPIO ISLANM. 



73 



«(3mp1rte jw>fta<»jwlon of all the clinl'ectfi of the family, ami we mmt have 

 compared ftl) ihf'ir voi-aWea not only with efich other but with those of the 

 cognate foniiUeit, — that is with all other human lana:uaj,^eij Tiie history 

 of eVRry spokfin tanjf ue ai^cflnfl:* t* the very beffinnini:,' of lippeoh, or to tha 

 ori|.jiii of mankind.^ It ci>i^t;uiif ro*>t.s tlmt have come down throuj^li nu- 

 meraiw channela and with various thanifeA of fnrm md mt»ardii;L,'. 

 Each root hiiiA also, throuicli all etlmit' time, ftowwl in hundreds of cou- 

 temporanetoui* currents multiphiujif by Aelf division, diver^nsr tar ajjart, 

 •pj>r«HUthin^, touchiag- or coale-^cin;;* and ai^uin dtvArioatiiisj;. The g-eneu- 

 loffv of every lanj^uai^** is honce ''xceedin'i'ly iu>mpliciit<'dj and will reraaia 

 a »ubject of research fur centuries f/i uome, Wc must make a Iveirinningf 

 with imperfect vocMibnlarie.^, and .-tuch partial <n'otips a.'* th^y enahle us to 

 det4?rmibe. The rpj^nlr.-* wlitch the fin^t Jabourera in f he tiold may arrivft 

 at will appear insiirniHcaritsw the fcieiice advances j but they havethi.-t fin- 

 caurajjement that every well considered comparison, however naiTow,lfitidd 

 to ft positive hi.Htorifyil result. What ii* lesirned is a sukstautial and stable 

 g^iin. It will afterwards take it^^ place as oidy one amnn;;? many evid«n- 

 of the s:ime ethnic movemeul or influence, or internal lin{jia^tie 

 chan<^«', and connections that now appear isolated or parriiil will be explain- 

 <'d as the resulta of ethni<; alliance.-, and event:* that wer»i n'»t at. iirat a;i3- 

 pected, but which havn left other records in the vocahulary. Tho eirlier 

 gcneriili stations will be coirocted when they liave been too narrow or too 

 wide, but the auh^tautial affinities brou^dit to li-^ht will aJwayii remain 

 . iiiiDong" the facts on which the science, m all its successive developmente, 

 will be ba-s^d. 



With tfiQ ?mall samples which we poi^sess of moj^t of the Himalaic 

 langiiag-ea, we mu^t he satitified with the pxamiiaitign of a few e^roajisof 

 wortld, ajid each of thesw? excealinjirlv defiictive. Not t«.) conipTieatc the 

 enquiry, it will be coiitined to a,scert4Unin,::^: Is^t, the rnmilicjvtiona of each 

 root in all the vocidmhmeH; 2d, the vooahle^ bv which each object is at 

 preiJient known in the ditferent vcw-nbuUirifs, and the connectitms thereby 

 judioateil ; :id, the atbnitiea of each vwaliulury ^^injifly. The rclationj* indi- 

 eat-ed tnider the firsit head are to a great extent archaic : they mu.^t have 

 hi^n formed duriui,'- a <rreat Inpn-e of time i and nnuxy of them jnu-it belong- 

 tit the earliest pluuw of huiimii *<peech- TJie hi-'^tory iinlictitetl by theae 

 •tKnities is complex and mu^t embrsice many and great ethnic <:han;ie»* and 

 movements. The aniuitiea examined under the 'Jd head will tlnow .tome 

 l);rht on the later etbidc movement:* ; iind thoj^e hmiufht tof^'cthcr ujuler 

 tho 3d will help to lihow to whal extent each diulect wa.4 affected by those 

 jnovemenfc*, nnd what iu modern and ita latt;r prc-liijitoricul pelaciona to 

 the other dinlects have bccu. 



As the Chinese is, on the whole, more faithful to tho primary system 

 of iioinencluture thati other lan^-ua^es, and the Ilimakic family ' takei 

 the next platic in the order of g-losnarial diffor-ranid-ation and concre- 

 tion, it wdl be useful to take a few illadtr.ttlond fmm Chinejje of 

 theusie of generic names. The root n^au— Jialectit;ally varied to 

 g-u on the pittuml side mid to niu on the liqtnd — is applied, with 

 fipeeiiic quahtivps, to the Cow, won": ntrau (//tl'o/f nguu), Biiffitkc 

 shui n«:au {mater npiu), Yak man lurau, Zebu iwng nojau, and 

 jtkimcfrus sai n(f.iu. The root yeun^;- (yanjL? kc.) isapplieti to the Sheep 

 niin yennn^, Goat &hm yeung* {Tm/u/ituin yeuu},') or tsu (t^mu &c.) yeim{r^ 

 ChamoU liug yeuuij, and Autilope ^^uttuixjaa won^j yeuui^ {i^dhio yeuugj. 



