42 SISSANO. 
We shall clarify the Melanesian situation by applying in the first 
instance the stem formula just presented. The forms which follow 
the formula, for the present disregarding variety in the nua element, 
are I, 2, 5, 6, 8, II, 12, 14, 15, 18, 20, 21, 22, 28, 34) 35; 36, 37) 39, 
40, 41. In this list the bold-faced type distinguishes forms which are 
encountered in the Polynesian series. Here we see that the initial 
consonant modulant follows from the labial spirants f and v the same 
course of weakening as in Polynesian, even to extinction (41), for we 
note (12) the Duke of York wanua as in close approximation to the 
Maori whenua, and (34-37) we have an aspirated group from north- 
ern Melanesia and the eastern Bismarck Archipelago with an outlier 
in southern New Guinea. Likewise in eastern New Guinea, in the 
closely associated languages of Suau and Tubetube, we have (39-40) 
an interesting form intermediate between hanua and anua; as between 
eanua and ianua the difference is of the slightest; they stand together 
as a Slight transition form produced by the weakening of the spiritus 
asper, leaving but a ghost of its impress before extinction. In parts 
of New Guinea and the Bismarck Archipelago we encounter a move- 
ment of strengthening of the labial, to the mutes sonant and surd b and 
p; so far as relates to this element of the stem these are the forms 21- 
33, the clearest types being Mekeo panua and Galoma banua. 
We are now in position to consider the initial element of the final 
evolutionary stage of thisstem. Reverting to our stem formula stated 
as labial + vowel + nua and stating for the present specific inquiry the 
second member, the mutable vowel, as a, we reduce the formula to labial 
+ anua. We have now to consider the status of the mutant labial in 
two possibilities. It may have been an original possession of a stem 
which has undergone mutation in accordance with the habit of the 
several languages in which we discover it. ‘This is the common 
system of interpretation of similar cases. Against this interpretation 
I strongly incline to set the somewhat considerable series of instances 
in which the mutant labial has reached the strongest position in the 
mutes b and p. If the archetype of the stem had been banua or 
panua there would have been no need for the Samoan to weaken 
it to fanua, for in the Polynesian languages we have attained to 
the richest development of the labials to be found in the Oceanic area, 
and it would have been simple for the Samoan to employ panua and for 
the Viti toemploy mbanua. On the other hand, if we assume an arche- 
type anua we shall find our difficulties resolved. ‘The more intimately 
I prosecute these minute investigations into primordial stems the 
more convinced do I become that a theory of mere mutation of con- 
sonants fails to account for form variety and the greater support do 
I obtain for my hypothesis that, in the evolution under the play of 
conscious intelligence which picks up the animal cry and by the appli- 
cation of consonant modulants with coefficient value transforms it 
