COMPARATIVE LEXICOLOGY 



311^ 



whole hand", and. "fingers'', ko^aTdpa being also written lolJiai-dpa. Xow. lii-nal 

 means "his hand ", and kotjafupa or koihafdpa would soon lose its initial Ico-, from the 

 wear to which it is subjected. In hatdhiik, ki'itdpa, and satabti a new stem is to be 

 recognized; it signifies "to grasji ", or rather "grasps", and is found in auiiica 

 sataba, "fire-tongs", in which aauu-a, means "fire" and satabu "to hold, take hold". 

 The reference here is to the clasped hand as signifying the digit "five'', because in 

 counting the fingers are bent down upou the palm of the hand, the result being a 

 closed or clasped hand. Now, in aelJch-ulal and sol-vhepam, a form of the usual siil, 

 "hand", occurs, and -akai and -cliepam. liave presumptively a signification sematically 

 equivalent to liorafapa and salaba in the preceding Yuman examples, but the mcager- 

 ness of the material at hand prevents tlie sotting forth of the data necessary to jirove 

 this conjecture; yet it may be stated that if the term '■hand" is a constituent ele- 

 ment of the uamo for the digit " five", it is because of the fact that the fingers and 

 the thumb thereof are in number "five", so that "the entire hand, the whole hand, 

 the complete hand", may become the name for the digit "five"'. Hence, when the 

 word hand is an element of the name thereof, as it is in the present instance, it is 

 presumptively certain that some word like "entire, complete, whole, clasped, bent 

 down ", must form the other element of the compound. The Cochimi ( II) mutjuacogiii 

 is seemingly a combination o( mugua for the cognate knmui/a, "three", and fiogiii for 

 gogiiu, "two". And the Cochimi (I) ni/aJcirampai is a compound of gi-nyalc, "hand'' 

 [inUnijak, foot], and some element denoting the completion of the count of the digits 

 of one hand, -\-vampai or vampai. The Cochimi (III) and (IX) are self-explanatory, 

 naganna, signifying "hand", while Laymiin (IV) is not explaiuable from the acces- 

 sible data. These analyses fail to show genetic relatiousbip between the two lists, ic 

 so far as the digit "five" is concerned. 



.SIX 



Serian 



A. nahpsuk 



B. napk'schoch 

 inapsho,);' 



' limapkasho 

 D. snajikashroj 



