GEOGRAPHY OF THE MIGRATIONS. 161 
still further division into specific and general clears the problem to its 
final simplicity, these terms being used to distinguish between those 
vocables traceable through the Vitiaz Strait which appear on one only 
of the shores of the great southeastern prolongation of New Guinea 
or upon both. 
We find but six vocables accredited to Melanesian tongues which 
it is possible to trace through the Dampier-Vitiaz exit. When we 
direct our attention particularly upon the northern offshore islands 
(Kiriwina, Murua, Nada) we note that one vocable comes out of the 
straits as far as Kiriwina and is not found elsewhere in the region, that 
two are recorded in Kiriwina and Murua, and that one of these appears 
also upon the north shore of the promontory; that one (Chart III) is 
met with in Kiriwina, Murua, and Nada; that its identification extends 
to Dobu in the central offshore islands; that it is found at least once 
on the adjacent north shore of the promontory, and that it is found 
quite generally along the south shore from Motu to the Louisiades. In 
Charts II and III we observe two anomalies. In the former a vocable 
carrying a Vitiaz Strait element appears in the Motu-Hula-Sinaugoro 
‘region on the Gulf of Papua and is not identified at any point in the 
intermediate region until it recurs upon the north coast. It will be 
understood that the line which connects this area with the symbolic 
rectangle is not intended to express migration over sea and land, but 
is simply indicative. In Chart III we observe one vocable whose 
identity is established from the Vitiaz Strait; thence at practically 
every datum point on the north shore and in one of the southern off- 
shore islands; it occurs at the most distant point of the Gulf of Papua 
coast after a gap in which in other cases we find much of our most 
valuable identification of Polynesian material. 
We have but one clear case in which it appears that a Melanesian 
word is traceable on the north coast of New Guinea, down through the 
Vitiaz Strait and the south-extending coast to and around the south- 
eastern tip of the great island. So far as relates to any speech element 
which seems Melanesian we have but this single instance which may 
lend itself to the idea that there was a speech movement from the 
north coast to South Cape and there a reversal of direction toward the 
Gulf of Papua. 
We next examine in Charts IV and V the sixteen Polynesian vocables 
in which we can detect the suggestion of exit from the north coast 
through the Vitiaz Strait. On Chart IV there is but a single instance 
in which we find what should appear to be the characteristic picture 
of such migration, and here we have the material identified upon the 
north shore of the promontory and in Kiriwina and Murua of the off- 
shore islands. In the other four cases the identification is restricted 
to the Gulf of Papua coast, is anomalous, and the first lines out of the 
symbolic rectangle are but indicative. On Chart V we find that 
