PARTULA NODOSA. II! 
This is the fact that the color-varieties as they exist in Punaruu and elsewhere are 
not physiologically distinct; the evidence afforded by the embryonic shells proves 
conclusively that they do not breed exactly true, but interbreed to some extent. 
The groups called composita, leva, etc., can not interbreed, excepting in so far as 
migrants may make their way from one valley to another—something which can 
be accomplished only with the lapse of long time. In short, the varieties that are 
circumscribed geographically, and are distinguished as well by their characteristic 
details of color and form, are primary in the sense that they are more differentiated 
than color varieties living in the same valley. It may be that color differentiation 
preceded in time the division here regarded as paramount, accomplished in the 
course of emigration to new localities, but as the facts stand, that diversification 
which includes the element of geographical discontinuity is more fundamental as it 
leads more directly to true differentiation of specifically separated forms. 
PAPEHUE v. 
AQUA V. fi a 
rai A \\ SS 
WS Is 
\ 
. 
x 
OROFERE ¥, 
Text-Fic. 6.—Topographical chart of the region inhabited by Partula nodosa, showing 
the present locations of the several primary varieties and the probable lines of 
their ancestral migrations. 
III. The internal evidence as to genetic relationships indicates that var. composita 
1s directly ancestral to var. leva and var. intermedia, and indirectly through the latter to 
var. exigua (text-fig.6). This follows from the facts of resemblance and difference 
