Notes on the Customs of Mota, Banks Islands. 137 



fashion. Their mental view is different from ours, and is 

 not readily caught by us. 



Their system is natural and clear. Up to 5 their numerals 

 are the common ones of the Pacific. On the second hand 

 they are the same with a prefix up to 10, which is " san- 

 gavul." The unit above 10 is called the numei. Tens above 

 100 are called the avaviu. 



Example. — 127 is thus expressed : — 



Melnol vatuwale, avaviu sangavul rua, numei lavea- 

 rua. 



" Hundred once, the avariu is tens two, the numei is the 

 other two" — i.e., 5 -f- 2, or seven. 



Of course they use " ten hundred" and " a thousand" 

 loosely also. A traveller knowing nothing of the language, 

 and having to deal with a stupid person through an inter- 

 preter, would very likely report that the people could not 

 count, and all the anthropologists would repeat the statement. 



[Distinguishing between counting- and calculation by using the former word 

 for the mechanical process of counting visible objects — one, two, three, and so 

 on — and the latter for the mental operation necessary in the case of Mr. 

 Codrington's dried bread-fruit, I think we may say that the Fijians are good 

 counters and bad calculators. They will count up to many thousands, but 

 few of them would have been able to subtract mentally the 164 bread-fruit 

 from the whole number required. The possession of words for high numbers 

 has been taken as a mark of superior intelligence, but I think it is not 

 necessarily so. There are tribes who count to high numbers simply because 

 they have things to count, and they are in the habit of counting them. Thus 

 root-growers like the Fijian often have enormous quantities of yams to count, 

 and they count them correctly ; but they cannot be called good arithmeticians, 

 even in the smallest way, nor do our mission-students show much capacity for 

 mental calculation. They soon learn to manipulate figures with slate and 

 pencil, but they look upon them simply as things with which certain opera- 

 tions are to be performed ; and it is extremely difficult to get into their minds 

 the fact that figures are only symbols of real things, and to make them think 

 out the simplest arithmetical process. 



I have not read of any pastoral tribe who are good counters. The savage 

 does not readily count moving objects. He must lay his hand upon the things 

 he counts, and reckon them one by one. 



There are certain numbers which are landmarks, so to speak, from which 

 the counter "takes a fresh departure." Thus in Fijian, lima, hand, is 5 ; tmi, 

 finished {i.e., both hands done with), is 10; ndrau, leaf, is 100 ; undolu is 1000, 

 omba is 10,U0(J; and vatuloa, blackstone, is 100,0U0. The etymology of 

 undolu and omba is not apparent, though the former means "a company" — a 

 many, as we still say in the provinces — in the language of Florida, one of the 

 Solomon Islands, which is closely allied with the Fijian. All these words are 

 evidently " tallies." There is no doubt that they are real numerals, and that 

 the highest of tbem has frequently been brought into service in yam-counting, 

 Mr. Codrington informs me that the Mota word for 100 (melnol) is the leaf of 

 the cycas. It will be observed that the Mota people use 5 as their first tally. 

 The Fijian goes on to 10, and then starts afresh with 10 and 1, 10 and 

 2, and so on. Though his first 10 is called tini, or finished, all subsequent 

 tens are sangavulu — the Mota sangavul, a word whose etymology it is desir- 

 able to ascertain. It is common to many of the Pacific languages. 



The word for "hard" in most parts of Fiji is linga, but lima is found in 

 some of the Fijian dialects. This is simply one of those interchanges of con- 

 sonants which are so frequent in these languages. I take this opportunity of 



l2 



