FROM CENTRAL INDIA TO POLYNESIA. 169 



pendix. In the following" cases there seems to he really no 

 doubt that the roots are identical (I give extracts only) : No. 6 

 Malay rckat (to which I think should he added lekat and ikat) 

 = Bahnar hit, kot, "to hind", Mon dakat, "to knot"; No. 7 

 Malay, etc. talc at = Khmer kot, Mon takilt, "to fear"; No. 182 

 Malay telut == Khmer lut, " to bend (the knee)", Bahnar lot, 

 " to enter in a bowed attitude"; No. 183 Malay tclan, etc. = 

 Bahnar ITibn, Stieng laon, "to swallow." There are a good 

 many more that could be quoted. On the other hand some of 

 the comparisons seem to me exceedingly dubious. Even 

 when the correspondence in form is to all appearance very 

 close, it often happens that the connexion in meaning strikes 

 one as being uncommonly far-fetched. There seems to be here 

 a gap in the method of such investigations. What we want 

 to guide us through the mrzes of derived and cognate words, 

 is a science of the relations of the secondary meanings of words 

 to their primary ones. I believe that the groundwork of such 

 a science has already been laid down for some of the better 

 known families of language. But it is clear that its lines must 

 be retracted for all the different cultural strata of mankind. 

 What strikes the savage in one way would probably strike the 

 more civilised man in quite another fashion. It requires a 

 really intimate acquaintance with the primitive mind to be able 

 to produce its processes with any approach to certainty. 



In many cases the etymologies suggested by Professor 

 Schmidt are nothing more or less than highly ingenious guesses. 

 Thus, for example, he sees in the Malay clakut, " grass," (which 

 does not occur in Wilkinson's Dictionary and is perhaps a 

 Javanese loanword) a root meaning "green," for which how- 

 ever the only authority given is the Stieng kiit, " green." 

 Malay akar, "root," etc., is similary traced to the Bahnar kor, 

 " to go down to the ground." Malay pandan, " pandanus," is 

 connected with the idea of sweetness : I believe the juice of 

 its fruit gives a drinkable liquor, but is it particularly sweet? 

 Malay Ixndak, "porcupine", again, is explained as the animal 

 which "rolls itself up: "one knows that hedgehogs roll them- 

 selves into a ball when attacked, but I am not enough of a 

 naturalist to be sure that porcupines do the like, though it 



R. A. Soc, No. 53, 1909- 



