470 PROCEEDINGS OP SECTION G. 



Etruscans, not having this letter, have rejected the d of d v a 

 and retained the next letter, the labial v, which is equivalent to 

 b h, b, m ; thus the root-word dva may become either (1) 

 l)h a, or (2) ba, or (3) ma ; and all these forms occur; for, (1) 

 S. ubha, Gr. amp ho, L. am bo mean "both;"* (2) in 

 Gothic b a i, and in German b e i-d e, mean " both ;" m Hindustani 

 ba-rah is "twelve"; and in L. also bis, bi-ni, b i-d ens, 

 b i - f a r i u s, &c., have the b only ; in the dialects of the Australian 

 aborigines, the word bula is almost universally used for "two," 

 and I consider the b u there to be a surviving proof of their 

 connection with the rest of mankind ; (3) the form m a I recog- 

 nise in the Gr. words mak-kor, mad-ibos, mak-ella (vide 

 Hesychius), " a mattock with 'two' teeth ;" and perhaps in the 

 Gr. m a c h e, " a tight, a battle," when it is compared with the L. 

 b e 1 1 u m for d u e 1 1 u m ; three of these words have the k, c h, 

 of the Etr. m a c h, and the other has d ; for, while (by Result 1) 

 ma-ad or ma-adh would be the usual form for the number 

 " two," yet the I.-K. often changes d h into g h ; besides this, 

 the guttural remains in the Gr. d i c h a, "in two," and in the L. 

 b i g-n i, which Eestus says means " twin ;" and I think that the 

 H. sh e-b a (g), "seven," (see par. XI.) also contains this numeral. 

 To account for the forin m a instead of b a, we must remember that 

 on Oriental lips the b and the m are sounded almost alike, and, 

 as the Etruscan alphabet had no b, the nearest labial m was 

 adopted. 



I believe, moreover, that the original idea hidden in the syllable 

 m a is that of repetition, as in the H. s h e n a i m, fem. s h e t a i m, 

 " two," from the verb s h a n a h, "to bend, fold, change, repeat ;" 

 ma would thus give us the G. muth (for ma-adh), "a change," 

 "an alteration," and the L. muto, " I change ;" perhaps also 

 the L. mov-eo, " I move," and the Gr. ameibo, "I change, 

 exchange;" the L. v i x, vicis, (for mhiks), from the same 

 root, also has the guttural of the Eti-. m a c h. The same idea 

 of " change," "repetition," shows itself in the L. alter, "second," 

 " the other of "two" (a lterplex = duplex, says Festus), Fr. 

 alterer, "to change," E. "alter," which are congeners to 

 G. eile, "another," L. alius, and to the T.-K. aile, a 1 a, 

 "second " (vide Zeuss, Gram. Celt., II.-3) ; and, lastly, the same 

 idea is found in the I.-K. tanise, "second" (Zeuss, ibid.), 

 which, in my view, is the Etr. t ii u n s, " second." Words of 

 the same root-origin as thuns are the modern G. tionnd- 

 adh, "to turn, come back," and tiunn-al, " a match, like- 

 ness." Festus says that paribus equis = d u o b u s equis ; it 

 is therefore possible that the pa in the L. word par, "equal, a 



•jThe word ubha, a ni p h o, is probably equivalent to gam -dva, " all two," which, 

 like the Fr. tous d e u .\, would mean " both ;" g- a m= S. sam, "all"=L. cum = 

 L. om-nis. As in Etruscan, so in Sanskrit, the d of S. dva, dvi, is rejected and the v 

 retained in the S. v i m - g a t, " twenty." 



