ti VARIATION, DISTRIBUTION, AND EVOLUTION OF THE GENUS PARTULA. 
or it is the result of cross-breeding, and the latter conclusion certainly seems the 
more probable; in either case, however, we may be sure that the lighter and darker 
kinds of adults are not separated by physiological barriers. 
Before discussing the subject further, we may recall the characters of the 
adolescent snails. In the first place the distinction between the rubescens and 
the purpurescens phases of castanea is manifested only late in development, and 
hence we deal with the castanea class as a whole in contrast with its lighter asso- 
ciates. The latter are not distinguishable in their earlier adolescent conditions, 
as we have shown earlier, and a fortiori it is not to be expected that their embryonic 
shells would be unlike. Presumably, therefore, the lighter young recorded in 
table 61 would have developed into unicolor, bicolor, mitella, or mitella-rubra adults, 
and the brown young would have become castanea. 
TABLE 62—Partula gibba, Saipan. Proportions of the light and dark 
classes in adult and offspring generations. 




Adult generation, per cent. | Offspring generation, per cent. 
Series. 
Light classes. | Dark classes. | Light classes. | Dark classes. 
Puntan Flores.... 82.25 Leds 83.6 16.4 
Sadog Tase...... 85.37 14.63 84.4 15.6 
Puntan Muchut.. 84.67 15730 82.8 a We 
Fanaganam...... 80.98 19.02 85.4 14.6 
TORE. chee 6 ee 97.40 2.60 97.7 Died 
Astaman: .i< je. 87.86 12.14 86.6 13.4 
Gatapany skies: 73.44 26.56 Pe DT aS 
Chalankiya...... 97.18 2.82 96.6 3.4 
ALE SS shave sitet mcs 86.50 13.50 86.7 1333 


Having established the fact that light-colored and dark-colored snails inter- 
breed, we may carry the argument further, despite the real adverse circumstances. 
Each association must be taken individually, for the proportions of the two main 
color-groups vary greatly from place to place (table 62). Now, the proportions of 
light and dark young produced by the several adult series approximate so closely 
to the figures for the two groups of adults that there is not 2 per cent of dis- 
crepancy in any case (table 62); this means that each colony is relatively stable 
from generation to generation so far as the numbers of the light and dark kinds 
are concerned. ‘This color equilibrium is maintained not by the absolute restric- 
tion of offspring to their parental modes of coloration but by reciprocal contribu- 
tions of contrasted young. But in Sadog Tase the light group of adults is made 
up largely of unicolor, in Astaman bicolor is the most abundant, while at Chalan- 
kiya the mitella and mitella-rubra adults predominate; it is therefore evident that the 
castanea Class is genetically inter-related with all of the classes more lightly colored 
than itself. While it is not possible to prove directly that the lighter classes 
such as unicolor and bicolor interbreed, by analogy it is highly probable that they 
also are capable of intercrossing. 
