ETHNOLOGT OF THE IXDD-PACIFrc ISLANDS. 



ed with fi. of 2). ; Ugriani Im, m (comWned with kof5); Jnpim inf 

 (iiDcombineJ), It is "ot (oaii'l in ihe otIif3r N tinrl 1-1. Ajsiiitir sys-» 

 terns and h is rare in Anicricii (Catawbas m-uMm-da, Saltaptiit mi-fa/, 

 ShoshOui raanii-thit?, Sioux (liaJect^ mf-ruena, itin-h nah, ^ah-mmn). It 

 has also aprcati into AfVitm, where it is caiubined with oilier particlea as 

 in Caii0. &c Ugr,j, but it is not prevjileiit. 



Nnn(?w, Tthlu Trtni ; nflZ/r Mul., nonft TucL, naifuTul,, im?/ftt Kar., 

 na^M^aTel., nai« Goiid, (Root probably and f^ri, ku a second def. 

 postfix.) 



Tliia term m not Iranian or Semitir, (Imt or is «n eleififnlin fheSem. 4) 

 Onto of the Caiuto-sian tern»s ajiproachi^s to it, tinuicwyii (Los^iaji). But 

 the clearest and lu Mt inimi»rous, ulHiiide^ urf Utrriau imd Aa iciin. Vgr, 

 nila &c., Afr. im, » >, iii, nanivSs , ( MunJinijo &:r.), np-nuj (HnHorn), tjian 

 (Mfikn), mnai (Kosah) &e. f r »m the (iis(riijiitioii oi' the tern in Atrica 

 and Asifi, and the iuikJc hi which thH^tjlemcnt!* aru interwovt't^ hiVs >oine 

 of the Afrh;an sptetn*, it is iirobuble that na &c. wns the radix of a bi- 

 Tjary system b«5loQytiug to u bH-niulion thstt |nedominal(>ti in W. Asia 

 prior to thn epouhs oi"tht» tlitfosiou of the Caticasian, the Semilic and tins 

 Iranian, It d-u*B not a[j}te;»r to have made niucli itro^\iVLi^ to the East- 

 ward, as it is not found iu the existing DJ. E Asiatic an i th - n<Jjat;ent 

 JI. Anieriean lau^u^iges, [ti the httter b a fre«jut*rjt iuitiai, but it u 

 merely dtfluitive^ being found in 3 and other numbers. 



If the romt b nnd fa, hi h the poss. poitf., the Him] fju, Ar<, k roust 

 l>c a superadded definitive. Double and even treble flpfiiiitive postfixes 

 or prefixes are not infrequent ia the a^glomcrative fornmti'ins of Asia, 

 Europe, Africa^ Asouesia and America, aufl thf Dravirian vocabularies 

 fiipply instances. The combination of /« and rpi mny tiierefor be purely 

 Bravirian in its origin. Ou, hut Src h a cojnrjjou dotinitiv*- final in 

 ,aome of the Caucasian languages, and others wbitdi do not use it luive ini- 

 ]florted it as a substantive portion of numerals which tbi-y htive borrowed 

 from tUo^ farmer. The Creorgian es-i2^u, Ij, is au instance, the proper 

 Georgian unit beinn; ar, with or without tbi. The Caiieasiiin affinities of 

 the Dravirian numerals raise the suspicion that the final tfU f)f the (atter 

 had a sirailar Lesifinn origin. The parent system was doubtlpss formed 

 in S, W, Asia, the crrefttfo^ius ot alJ the Asiaiij European and African nu- 

 meral systems J and the prflsont raucusiaii numerals ure pro!>obly thfim- 

 solves derivRtivn from soitie archaic tbrmation that was not eon fined to 

 Caucasus and did not oriiriuate thorH^ for lbs plaiu '>t"the Kupbrales, and 

 not its bead basin or the vallies beyond, is likely to have been the earli- 

 est seat of civiiisatiou Ui tlua region. 



Five. 



AinrfK, an/K, an<?Att, Tam., anyar, unc//« Mnl., Tftj/, x^ijlthn Tod., ayiwf, 

 Tul., ayiffii, (iida Kar., ayi^/« 'J'ui., eaij/m/i Gond. (Root aln or tin pro- 

 tal>ly, but j»ossildy atfj. anch). 



This is a peculiar term. If the root is tfi«j or it nppears lo be a 

 mere flexion of Hrt, 4. W anj dfc. be the root, th^^ only affinities I can 

 disi'ovor are the Caucasian hisfdu, itself nn jnKi'.rtrd coiTiliination ol (lefi- 

 nUives, and the Iranian ptint^n^c. with i be connected \vor<fs sisrnily- 

 ing ^'haivT" (c. p. yash, t-A.-fy, Ugrmn ; du, hu. rhUy semt, ( hin< se, Ultraind. 

 Ind ; hti»ta, seste, hath, Iran.) If the numeral was not derived from 

 Cauc. or Iran.^ but was a dit'eot juoiiticatioa of a word tor *'hand|" tho 



