162 THE SUBANU. 
quinary by composition, the linguistic equivalent of the arithmetic of 
simple addition. There are several phases of such addition-composi- 
tion but here it will suffice to illustrate along the phase of the type 1, 2, 
2+1, 2+2, 5. This I recall from the Australian aboriginals of the 
Northwest Bend of the Murray River; 1, nitshe; 2, barcoola; 3, barcoola- 
nitshe; 4, barcoola-barcoola; 5, yentimarra. 
A stage lower than this is presented by the numeration which has 
not yet set five apart as a discrete concept even without precision. 
This is illustrated among others by the Miriam (apud Ray) of Torres 
Straits. It has but two numerals, netat 1, neis 2. Higher numbers are 
but sums in addition of these two primitives, nezs-netat 2+-1 =3, neis- 
nels 2+2=4, nets-nets-netat 2+2+1=5, nets-neis-nets 2+2+2=6. 
Without prosecuting further this research these examples will serve 
my purpose satisfactorily in showing that we need feel no surprise if in 
the Polynesian decimals we may be able to detect a composite of similar 
type. At the beginning of the sense of number we find the recognition 
of the distinction between the one and the more than one; the division is 
cut along the cleavage between unity and plurality. The next dichot- 
omy must be equally simple; it will apply to the plurality. The mind 
becoming more observant of detail will distinguish between few and 
many, the small plurality and the greater, or, as I have heard it defined 
in the Beach-la-Mar, “‘small-plenty’’ and ‘‘big-plenty.”’ The least of 
these must be one and one; if it were less it would be merged in the 
name already established for unity; there we find a name assigned to 
this one and one, a name and a signification of two. The maximum 
plurality varies with different peoples and diverse culture attainments, 
but even we retain our diffuse “‘big-plenty.’’ With piety probably 
quite as obscure as the mathematics, we obey the solemn formula and 
rise and join in singing ‘‘oh for a thousand tongues to sing”’ and mean 
no more than this maximum diffuse plurality. In like manner we 
speak of a myriad stars which dot the firmament without any sense of 
restriction to 10,000. The Australian of the Murray fixes his “‘big- 
plenty” at yentimarra, which is higher than barcoola-barcoola 2+2 by 
one unit or many. His scale of number, therefore, consists of 1, 2, 
2+1, 2+2, and infinitude, which begins at the very finger-tips. 
With this possibility in our minds let us return to the examination 
of the Polynesian tolu 3. In the discussion of tasi 1 I have explained 
at no little length the development of the sense of unity and the manner 
in which its designation has been made more and more precise by the 
method of determinant composition. Inthe tabulation which sums the 
result of that inquiry I have set aside the employment of a t-component, 
ta in 22 names of unity, te in 6 names, ti in 11 names, to in 5 and tu 
in 8; in sum the t-component appears in 52 out of the 117 names for 
unity which I have assembled. 
In dealing with /ua 2 I have advanced the opinion that this also 
