112 VARIATION, DISTRIBUTION, AND EVOLUTION OF THE GENUS PARTULA. 
as given in the detailed analysis. It has been shown earlier that the original 
headquarters of nodosa centered in Punaruu Valley, on the basis of evidence entirely 
independent of Garrett’s statements. Here and in Taapuna the species has retained 
what is no doubt an original composite character, though the Taapuna colony dif- 
fers from the other in its frequent production of sinistral sports. Maruapoo seems 
to have been populated directly from Punaruu; it may be that the first snails to enter 
this ravine by traversing the interposed ridge were sinistral, but as only one in 200 
from Punaruu was sinistral in Garrett’s time it is more probable that the original 
colonists were dextral, or at least mixed, and that their offspring gradually lost 
the direct twist and also changed somewhat in details of color and color-pattern. 
The variety called intermedia, in Papehue, Aoua, and Orofere, is obviously more nearly 
related to composita than to /eva, for it would have arisen from the latter only by 
completely reverting to the original conditions in color and coil. It is more probable 
that this variety is descended from migrants from the upper part of Punaruu, 
from which the line of advance, if followed directly, would pass the heads of both 
Maruapoo and Atehi. In each of the three valleys of the range, the colony of 
intermedia has assumed distinctive characters, both in structural respects and in 
the proportionate numbers of the constituent color-classes; rarely a sinistral sport 
is produced. Finally, var. exigwa is so like the small-sized and plain shells in 
Papehue and Aoua as to indicate its origin from the latter by a short northward 
migration. Certainly it has nothing in common with /eva of near-by Maruapoo 
and it is unlikely that it has arisen by direct migration into Atehi from Punaruu, 
for none of the members of var. composita deviate in the direction of exigua. 
IV. The role of the environment in the differentiation of P. nodosa seems to be 
negligible. On this point, no extended argument is necessary. The production 
of differing color-varieties (pallidior, concrescens, etc.) in one and the same valley 
could scarcely be attributed to environmental factors as causal agents, for these 
are certainly the same in one locality. We can not say that var. /eva owes its 
sinistrality to peculiar external conditions of Maruapoo when dextral shells occur 
in the same place. The small size and dextral form of exigua can not be referred 
to an adverse or otherwise factorial environment in Atehi, because P. hyalina and 
P. clara flourish here in no depauperate condition, while the variety of P. otaheitana 
that also occurs is numerous, well developed, and invariably sinistral; one and the 
same environment is not likely to produce such diverse results by ztse/f. 
It would seem, then, that the constitutional characters of the first migrants 
into a previously unpopulated valley would be the real determining causes for the 
basic differential characters of their descendants, which would subsequently vary, 
but not on account of a qualitative effect of the environment. 
