70 BTllNOLuaY OF THB IKDO-P.iCIFIC ISL.4NBS. 



the tlcfiiiitives ihoiii8elT(?s, and Uie African afBiiiiies are [>rol)Hbly 

 owing trt deiivaiion from a like eouice. The Ai'i ican lerma jirest ni 

 more afEtiities in root? an J cumbirtatiuns than the Scytliic, hui 

 Scytliic has roinaants of similar forms. 



The African affinities connect the Draviro-Anelralian qtjinni'y or 

 basis (system with the most archaic form of the Semi tico- African 

 system more- closely than wiili any other. But the JbrmtM* is sim- 

 pler and more primitive than the latter. In whi^h the aihilant 

 fieries of term;*, fonnd in all the Asialit; systemSi blends wilh the 

 labivil and prepojjdiirates over it. The AfiicrUi systfms appear 

 to have htfen more inflneitccd by the Semitic in its later ;rvat]uii<jn^^ 

 and the Sifmiiic by the Scythico-lranian, ihan the Draviro-Aus- 

 lmlian by any foreign systems. The Draviro-Australinn would 

 appear to he the most faithful rcpresenttjttve no»' e.<«istrng of itn 

 archaic S. W. Asian system of definitives and numei-als. This 

 ays^tem is still homogeneous, the labial being the principal 

 current definitive as well as unit. In the Seniitico-African sys- 

 lema there ai-e remnauta of the labial pronouji, but the sibilaai U 

 uow the principal one* The labial unit of these sysiems is hence 

 more closLdy connected with tlie exititing Draviro-Australiunj than 

 with the existing Semiltco-Lybtan^ pronouns. 



It is worthy of remark that the Euskarian, which has close af- 

 finities with the oldest form of the Semiiico- African syHtemsi, 

 jireservea a labial 1 ha-t, bo-t, 2 bi, mid 5 bo-r-t ho-st. It is found 

 also in U, be-dera-tzi, and 10 ha-mar {ante, see. 5). The Caucasian, 

 like the Scythic, Semitic and Indo-European, is mainly sibihint, 

 hat there are some labial remnants, 2 wi-^a Abkhsisian (Eniik. bi, 

 African hi-li, bi &c.) ; Sah-a/Lps^., (hut this is probably a contrac- 

 tion of chab in which the initial is sibilant as in the Georgian 

 pa-tni &c) ; 4 wor*ts-tcho Georg., nin.uk-//rt, hoo-^w, ohw-ai 

 Leeg., p-shi-iff, p-lle Circ. ; 5, w o-chn-si Georg., p-ehi Mis. j 6, 

 f'ba Ahkb. ; 7 s-wi-di, kc. Georg,, wer-al Lesg. huor, nor-/, uosh 

 Miiaj., b-le, Circ, Awar. j 8 rwn, mo, &c. Geor^., mhl-go^ 



m^k-ffo, betel-na, heeli-ffu, mvi-ba kc. Lesg., bar, bar-/ M-sj. ; 9, 

 h-ffu, horo Circ.; 10 wii, wiili Georg., wez-al lesfr. p-she Circ, 

 In 8oni(! of these term?, however, the labial is probably prefixual. 



The labial gystem would appear to have predominated 

 in S. W. Asia and spread thence to Imliaand Africa before the 



