PARTULA OTAHEITANA. 147 
TasLe 84.—Partula otaheitana amabilis and P. o. otaheitana. Comparison by valleys, as to color- 
composition. (Color-classes in percentages of population, in each mode of coil.) 
Sinistral division. Dextral division. 
Plain. Plain. 
: : : -_,| Banded. : ; : : Banded. 
Light | Yellowish} Reddish | Brownish Light | Yellowish} Reddish | Brownish 
I Il. III. IV. I II. Ill. IV. 
HAutauaerteieriel ii 19.8 16.5 48.7 14.4 0.5 18.4 17.9 49.9 13.7 
Hamutayprirrciier 11.0 17.3 57.5 12.0 2.0 20.0 15.4 52.5 10.8 1.1 
ITEM oo Gg0000000 71.8 24.8 1.3 0.1 1.9 sae 
ee ee eee 
Pohaitara....... 16.8 48.6 34.6 S000 erseys Metts tyes sas 
WITH aooGg00000 14.5 3.8 80.8 0.9 Bod aN Byes [1 shell] 
It is interesting to note that the Hamuta and Pirai colonies form a couple that 
is somewhat more like the Fautaua colony than like the Pohaitara-Ururoa division; 
the latter resembles affinis of the territory further east, excepting in the matter of 
spirality. These circumstances deserve recognition, although it is not necessary to 
make a formal division of amabilis into western and eastern sections, with taxonomic 
names. 
II. The dimensional character of the four colonies varets, but the changes observed in 
passing through the range from west to east are not always consistently the same. The 
facts presented in the earlier detailed descriptions are given in summary in tables 
85 and 86, and graphically in table 87. In general, there is a decrease in absolute 
measures in passing from Fautaua to Hamuta, and on to Ururoa, and generally the 
differences are statistically significant. But the shell proportions in the Pirai- 
Pohaitara comparison show an increase, with a drop in the Pohaitara-Ururoa 
relation. The proportion of aperture length to shell length is also inconsistent. 
It is true that the shortening of the whole shell in passing from Pirai to Pohaitara 
is relatively greater than the reduction in other respects, whence the increase in 
proportionate measures; but it is difficult to see how this could be attributable to 
any influence of the environment which intensifies with further progress away from 
Fautaua. If this were so, then we should not have the decrease in the proportionate 
measure of the shell at the next geographical step. 
Ill. The dextral snails of any valley agree more closely with the reversed group of 
the same valley than with the dextral series of another valley, and the banded snails 
resemble their unbanded associates more than banded examples of a different valley. 
The figures have been given in detail in the foregoing description, and nothing 
remains to be added. ‘The data of heredity prove the real inter-relationships of the 
different color-classes and of the two types of direct and reversed shells, wherever 
those occur in the same colony. 
IV. Combining the four colonies, the variety amabilis taken as a whole may be com- 
pared with otaheitana and may be defined for future comparison with other primary 
varieties. It is unnecessary to deal with the color characters anew. ‘The statistical 
data may be summarized in two ways, with two different aims. In the first instance 
