166 SISSANO. 
Cape Vogel without suggesting the subdivision of the province which 
seems so plainly indicated in our Melanesian material. In Colling- 
wood Bay there is but a trace of affiliation between Kwagila and 
Kubiri, and the Raga-Oiun and Kubiri-Kiviri pairs are scarcely dis- 
tinguishable. The exterior relations are the same in amount and 
destination as on the Melanesian chart, a trace from Collingwood Bay, 
a trifle more from Goodenough Bay, and a more strongly marked line 
from the Awalama region. While the figures establishing these lines 
are the same on the two charts, the fact that the Polynesian material 
is smaller in sum gives these three notes of affiliation a higher percent- 
age value in the Polynesian group. 
Louisiades region (Charts XII and XIII).—The provisional estab- 
lishment of this province has probably been somewhat overmuch 
governed by the geography of the charts. It has seemed simpler to 
include Tubetube with the South Cape by reason of proximity, yet 
when the languages are better known it is quite likely that Tubetube 
will fall more satisfactorily with the Louisiades. Our Melanesian 
chart exhibits the close relation of Tubetube and Panaieti, amounting 
to about one-fifth of the recorded data. Panaieti affiliates with Misima 
to the extent of two-fifths. The affiliation of Misima and Tagula 
amounts to about one-seventh. ‘The exterior relations of the province 
are interesting. Panaieti and Misima show a trace of affiliation with 
Suau and Sariba independently, and Tagula to the same slight degree 
is associable with Sariba and Tubetube. Misima exhibits a higher 
degree of affiliation with the east coast province at Awalama. In the 
Dentrecasteaux region Panaieti and Misima show a trace of-:affiliation 
with Dobu; Panaieti has a somewhat higher relation with Kiriwina 
and Murua and a trace with Nada; Misima shows but a trace of 
association with Kiriwina, Murua, and Nada. ‘The Polynesian chart 
offers a markedly dissimilar picture; with the exception of the inter- 
relations of Tubetube-Panaieti and Panaieti-Misima of equal weight, 
the series is but a matter of traces too slight to lend weight to any 
suggestion of explanation. 
Dentrecasteaux region (Charts XIV and XV).—Both our charts at 
the first glance suggest arachnid industry, but when we give attention 
to the relative importance of the several lines we obtain a clear picture 
of the speech relations within the province and the extension of such 
relations to and from other provinces. In the Melanesian chart we 
note at the outset a strongly featured community of the northern 
islands Kiriwina, Murua, and Nada amounting to two-fifths of the 
material; associable with this community we observe the interrelation 
of Kiriwina and Dobu to one-third of this degree, attended by traces 
of interrelation of Dobu with Murua and Nada. Despite its distance 
in the Louisiades, Panaieti, as already mentioned, is appreciably in 
touch with at least Kiriwina and Murua of these northern islands. 
