PARTULA OTAHEITANA. 249 
as in the other valleys of the peninsula. Under this head, there are certain facts 
that militate against the migration hypothesis; the nearest valley to Tahiti nui 
lacks sinistrorsa, the next has two classes, the succeeding valley only one, and the 
most eastward association includes all three. If, now, the last-named colony was 
founded by snails that had reached Vaiaaia by their own efforts, we ought to find 
phea as well as apex in the intervening area. It is still possible, nevertheless, that 
the peopling of Taiarapu took place by migration some time ago, and that subse- 
quently phea disappeared from all but the Vaiaaia colony, and apex from Vavii. 
In brief, when due weight is given to all of the pertinent biological and geographical 
considerations, the question at issue remains unsettled. 
Taste 195.—Partula otahettana sinistrorsa. Collective description, sinistral shells 
Mean value. Standard deviation. 
Character. 
: : Average of : , Average of 
Combined series. Combined series. 
valley types. valley types. 
Shell, length, mm............ 19.3845+ .0161 19.4915 1.3949 .0114 0.8744 
WAG, TVING bGo000000K 10.7015 .0072 10.7571 .6239+ .0051 -4604 
proportions, p. ct...... 55.1979 .0284 55.1892 2.4633+ .0201 2.2439 
Aperture, length, mm........ 9.8033 .0079 9.8615 .6872= .0056 -4856 
WAN, HVING Go00000 7.4162 .0055 7.4464 4786+ .0039 . 3786 
proportions, p. ct...| 75.6373+ .0323 75.5094 2.8003 = .0228 2.6467 
Length aperture + length shell 
PLrOpOrcions sp sCctenmeeeeeer 50.5368 .0223 50.5676 1.9312+ .0158 1.7931 
ARDOUN, WAGES, coao00c0e0b0006 1.0964 .0041 1.0750 .3573+ .0029 .1878 
1The number of individuals is 3,422; tooth 3,445. 
IV. The collective statistical description of sinistrorsa 1s useful as a basis for its 
comparison as a whole with other primary varieties. ‘The collective characters have 
been already given (table 161) for spirality and color-composition; the quantitative 
data for the characteristic sinistral shells are presented in table 195. As in the case 
of other primary varieties, it is permissible to group all individuals into one series, 
when abundant associations will have more effect on the average figure than collec- 
tions that comprise fewer shells; the result will be a general specification for the 
whole series as a part of the otaheitana population. In the second place, each 
colony may be regarded as a genetic unit, and an average of such unit-elements 
would give a figure with a different biological significance. In the case of sinis- 
trorsa, the numbers are large in all valleys, and the two sets of figures do not differ 
as greatly as in the case of P. rubescens. The detailed comparison with other 
primary varieties may be reserved until the final summary. 
V. The banded pattern seems to be dominant to the absence of bands in a Men- 
delian order of inheritance. Although the final proof must be given by actual 
breeding experiments, yet the evidence at hand is such as to indicate the truth of 
this statement. It is in accord with the results obtained through the study of 
affinis, as well as with the experimental proof derived from the case of Helix, as 
described by Lang. 
