BTHNOLOGT OP TUB nfDO-PAClPIC TSLAITDS. 



151 



Sftvara, bn-^i/, a form preserved also in7 of Yerukala vo-y/w, Kiratiti bha- 

 y-yfi and Mon bo-k (2 tor 5), anU identical with the Telugu vo-//a of 

 xo'hu'ti I and tbecotn. S. Dr^v, 10, It is probable that in bc.ra &c, 2, 

 ra waa a second an!hai(; dt'finitive or unit. 



4. If 2 be v.rsi'dn &e., 4 must be ha-Ih, i. the secondary dement 

 without the initial oii«. The K<A pen retains the root of 1 with the 

 t;oncreted consonant of the second element or primary post f. The form 

 opun-irt is probably tVnm op-pii,n, i, e. op 3, pu,n I. The Savara con- 

 tracted form 0^4^ voM-ji (Ijor 3, 1) is evidently the fuU form of the 

 Taluva OM-ji I. 



5. The (tadaba mo-lla-tfi confirms the analysis of tie Kol mu-na, mo-r 

 &c. aa 3 (S. Drav. inu-rfw &<;.) for 3, 2. 



8, e-v(u (not en-iw) aa in 2. 



Tlie Ko I irl appears be radically i.r-i, a contractinn of the S, Drav. 



The exceptional Gond, Ts.'!ii|ru and Tuhiva forma musti in conformity 

 with the amended aualysiH of 2, be read a-na-rau-r, e.ni-nii-<i*, e.na-me 

 i, a.na, e.ni, I'.na 2, and ntu.r^ mi-t^^ me 10 iu the form of the unit 

 found in tin.* Kol 5, Drav. 3 &c. 



The Telu;^u toniiniili must Ih- to-mi-<'/(, 1, e, mi-rft for 10 as la e.fti 

 ml-di 8, and to, a distinot root for I, preserved also in the Chentsu to-ta, 



and correspond inj; with the Drav. dental 3d pron. and def. (p. 56). 



The Kol a-#- of D has a for 1, as in the 8. Drav, 6. 



The idcntifieation of the S. Drav. on ! with vo.n, and e.ra 2 with be.ra 

 &c. complete the proof of the a^rreenient between the basis aystema of S, 

 Dnivirian and of Vindyan, and belwcen both and the primitive labial 

 system preserved in Australian. In it-^ first foi-m the system was simply 

 the labial definitive, or the labial and lifjuid, repejited or rompon tided. 



It is clear that the liquid in ba.rii, bija, &c. 2, is very archaifs and that 

 it early possessed a distinct n.-meral and plural torcii. In the ISemito- 

 African and Scytbic Hysteins, in which the same compound was the prin- 

 »:ipal archaic numeral name, the litjuid, chani^ing to the sibilant &c., has 

 been more stable than the labial initial^ and there are stron-f rcasoJis? for 

 regarding it as tlie epsential element ij) 2. See App. to .Sec. 6 of ch. v., 

 The S&mitic aiid AJrim/t jVfr*Hm//s, pp. 16. li), 43, See also a Note on 

 the san(e subject inserted in the Journal with Hee. 1. of chp, v. 



12H. The note belongs to p. 120, and note f of p. 129 ahould be note * 

 of p. 128. 



132. line 3 from the bottom for pishiJfe read ^o-piaa, and In the follow- 

 ing line, Jar /o-pisa, rmd mnsa, mu9, 



137. 'The substantive root !n the name for the Bttjfttloe is the labld. 

 The liquid Is the J-oot for ivat^r^ e-ru-ma = imUr-eom. 



138, In tango cvw of J Hi (not Sinirjiho) the root is nga (ta-nguY 



140. The statejneut that in Chinese the r(Mit alone signifies hufftihe 

 j«nd the inference frrwi it are incorrect. In Cbineae, ns in Dravirian, the 

 name lor the bufl'aloe is wott-r-cutv (or ox), and it is only by eoutractioa 

 that gu &c. alone is applied (o it. 



141. The Deer grwl is identical with the Bhotian god mare^ 

 145. The !*il)ilant name of Tiberkhad &e. is Tlbeto-Ultruiiidian, 

 15.5. 7th litie from i'mA^Jor moon^ Ttadsih'tT, 



157. del 5tii line &Qm loot. 



