106 VARIATION, DISTRIBUTION, AND EVOLUTION OF THE GENUS PARTULA. 
P. filosa, existing in a single valley only, and P. clara on the other hand, a some- 
what differentiated species now found in 52 valleys. Regarding the present geo- 
graphical limits of P. nodosa’s range as derived, and not as unchanged for an indefi- 
nite period of time, we may ask whether the prior area of occupation was greater 
than now, or less than it is at present; that 1s, whether expansion or contraction 
has been brought about recently. 
In favor of the view that nodosa’s range has recently been expanded, we may 
cite certain intrinsic aspects of the facts as observed and also the collateral evidence 
provided by other species. Under the first head the most positive detail is the 
discovery of nodosa in limited numbers in Orofere Valley in 1907, when search in 
the same place was futile only a year earlier. It is not contended that absolutely 
no individuals of this species inhabited Orofere in 1907, but it is held that the failure 
to find it in that year and the observation that it formed nearly 7 per cent of the 
representatives of the valley population in 1908 mean its recent introduction into 
this habitable area. As for the rest, the facts are not significant. It is unfortunate 
that Garrett did not state specifically whether he explored Taapuna, and the 
smaller ravines between Punaruu and Orofere, for in that case a precise comparison 
could be made with the present state of affairs; we can not determine by the most 
direct method whether Taapuna, Papehue, and Aoua, etc., have been populated 
like Orofere since Garrett’s time. Indirectly, however, the results of analyzing in 
detail the several colonies of nodosa, as discussed below, materially support the view 
herein adopted, which for the present must rest upon the single definite fact adduced. 
As collateral evidence we may add the positive demonstration already given 
that P. clara has enlarged its territory to a great degree in the past decades, after a 
period of severe contraction (p. 74 e¢ seq.). Furthermore, it will be shown in a sub- 
sequent section that certain varieties of P. otaheitana have also migrated recently to 
valleys which they did not inhabit previously. 
We may therefore adopt as most probable the view that P. nodosa has spread 
recently from a much smaller area, where it developed its peculiar and distinctive 
specific characters. 
The location of the original area is indicated by the facts to have been toward 
the northern end of the present range, and that it is probably to be identified with 
Punaruu Valley. Had it been in the southern part of the present range, where var. 
intermedia now lives, we would expect to find colonies of nodosa south of Orofere, if 
the ability to migrate northward as far as Taapuna were exercised in journeying 
southward. If, however, Punaruu be assumed to be the original center of dispersal, 
then the present range is just what would be expected, taking the topographical 
features into account. Migrants could soon reach Taapuna, over the hills, but the 
drier and loftier ridges between that valley and Faaa and Tipaerui would be barriers 
that could be traversed only with great difficulty, if at all; it would seem that they 
have not yet been crossed, although the Taapuna colony is a flourishing one, both 
in numbers and vigor. On the same assumption, the greater spread to the south- 
ward of Punaruu is natural, for the ridges are not so abrupt and dry and there would 
be less hindrance offered to the tide of migration setting toward Orofere. In short, 
