V.] THE OCEANIC NEGROES: NEGRITOES. l6l 



tribal groups. It has no clear affinities to any other tongue ', the 

 supposed resemblances to -Dravidian and Australian being ex- 

 tremely slight, if not visionary. Its phonetic system is astonish- 

 ingly rich (no less than 24 vowels and 17 consonants, but no 

 sibilants), while the arithmetic stops at two. Nobody ever attempts 

 to count in any way beyond ten, which is reached by a singular 

 process. First the nose is tapped with the finger-tips of either 

 hand, beginning with the little finger, and saying 

 ubatul (one), then ikpbr (two) with the next, after Co ^^ng f 

 which each successive tap makes ankd, "and 

 this." When the thumb of the second hand is reached, making 

 ten, both hands are brought together to indicate 5 + 5, and the 

 sum is clenched with the word ardtfru="a\l." But this feat is 

 exceptional, and usually after two you get only words answering 

 to several, many, numerous, countless, which flight of imagination 

 is reached at about 6 or 7. 



Yet with their infantile arithmetic these paradoxical islanders 

 have contrived to develop an astonishingly intricate form of speech 

 characterised by an absolutely bewildering superfluity of pro- 

 nominal and other elements. Thus the possessive pronouns have 

 as many as sixteen possible variants according to the class of 

 noun (human objects, parts of the body, degrees of kinship, &c.) 

 with which they are in agreement. For instance, my is dia, dot, 

 dbng, dig, dab, dar, ddkd, dbto, dai, ddr, ad, ad-en, 

 deb, with man, head, wrist, mouth, father, son, step- s ^^^ cal 

 son, wife, &c. &c. ; and so with thy, his, our, your, 

 their! This grouping of nouns in classes is analogous to the 

 Bantu system, and it is curious to note that the number of classes 

 is about the same. On the other hand there is a wealth of post- 

 fixes attached as in normal agglutinating forms of speech, so that 

 " in adding their affixes they follow the principles of the ordinary 

 agglutinative tongues ; in adding their prefixes they follow the 

 well-defined principles of the South African tongues. Hitherto, 

 as far as I know, the two principles in full play have never been 

 found together in any other language... In Andamanese both are 



1 "The Andaman languages are one group; they have no affinities by 

 which we might infer their connection with any other known group" (Lieut. 

 R. C. Temple, quoted by Mr Man, Anthrop. Jour. 1882, p. 123). 



K. II 



