474 PROCEEDINGS OF SECTION G. 



I thus arrange the words quoted in this paragraph : — roots — 

 (1) gab (gabh), = (2) gam (gamh) = (3) kam (kamh) = 

 (4) k a V ; derived from (1 ) gabh are G. (g a - b h - a i r) = Gr. 

 c h e i r ; derived from (2) gam h are G. 1 a m h ; Poly, lima ; 

 L. f roper w. Lav-erna; Gr. (lev-ids) le-is; derived from 

 (3) kam are G^r. pem-pe (p for k); Goth, fim-f; C. p u m-p; 

 Ki. p a n - c h a (pan for p a m) ; G'. (k a - m h - a g) k a o g , k o i g, 

 k u - i g ; Gr. p u g - m e ; Fr. poingjZ. quinque (for ku i g = 

 k u i n g), c u n c t u s ; derived from (4 ) k a v, k a v s (= k a m h s), 

 are Etr. cez, cis, ci. 



X. There is considerable uniformity in the Aiyan woi'ds for 

 "six;" e.g., S. s h a s h, old SI. s h e s t y, Goth, s a i h s, Gr. hex, L. 

 sex, I.-K. s e , s i a ; and to these I may add H. s h e s h. I have 

 never been able to satisfy myself as to the composition of this 

 word ; for that it is a compound word I do not doubt, as are also 

 all the numerals beyond five. The k in sex, hex, I consider 

 to be radical to all, and this guttural is softened into h in 

 s a i h s . We should bear in mind that numerals, from their 

 frequent use in daily life, must have been, in the early stages of 

 language, subject to great mutilation ; the numeral sex I there- 

 fore take to be a shortened form of an original word s a k a s, 

 afterwards softened into sekes, seks; but in weak languages 

 the guttui'al sounds are often dropped from words, as is so 

 commonly seen in the Polynesian dialects ; and thus s a k a s 

 would give the S. s h a s h, and its corresponding forms. The G.-Ir. 

 s i a seems to be for s a i, and is thus allied to the Gothic s a i h s. 



The supposed word s a k a s would then be compounded of s a 

 and k a s ; the k a s I form in the usual way from S. e k a, 

 " one," like the Gr. t r i a s from the root t r i ; by Result 1, e k a 

 or k a would give k a d, kat {cf. S. vim-gat), k a t h, and from 

 k a t h there is an easy transition to k a s ; the presence of the 

 formative t h is not a mere assumption, for it exist in the I.-K. 

 words s i a t h-a m h and s i a t h - a d h, "sixth;" sias, seas also 

 probably once existed in that language, like G. t r e a s, the number 

 " three," but it may have disappeared because there is another 

 word seas, which means to "maintain." The G. c h i a d, and the 

 H. a c h a t h, "one," are a near approach to k a t h, and k a s. The 

 kas would thus mean the number "one." The initial syllable 

 sal cannot explain, unless it be for the S copulative particle 

 c h a, " and ;" the whole word s a k a s would then mean " and 

 one," that is, [the first hand, five] " and a unit" [more], a mode 

 of reckoning " six," which does exist among primitive peoples. 

 I do not know that the S. c ha, "and," ever takes the form of 

 s a, although it is certain that the Sanskrit palatals, c h and j, 

 are represented in other languages by sibilants; its corresponding 

 Hebrew prefix conjunction v e, v a, " and," (v a v = "a hook, a, 

 binding or fastening)" is the same as the Gr. digamma, and that 



