PARTULA FILOSA. 85 
the shells of one class depart widely from the average in one character and not in 
others, while a second class deviates in the latter more than in the former. The 
figures relating to C, D, and E are particularly significant in this connection. 
FECUNDITY AND HEREDITY. 
Many of the snails were kept alive for several months for the study of their 
habits and their responses to environmental influences. Hence the numbers of 
complete records of fecundity is reduced to 117 (55 per cent). Of this number, 
76 (64.9 per cent) were gravid. The numbers of eggs and young borne by these 
individuals are given in table 35. The series of 1909, taken during a somewhat 
drier period of the year, shows a low percentage of fecundity, while that of 1906 
gives a relatively high value for the wetter season. Apparently, then, external 
influences, such as moisture, temperature, etc., exert an influence upon the time of 
breeding in this instance which they do not have in the cases of the widely spread 
clara and hyalina. 
The general average for both series is 1.16 per cent of eggs and young for all 
adults, and 1.63 per cent for the gravid snails. These figures are far below the 
average for hyalina and clara and also somewhat below the percentage for ota- 
heitana. Either filosa is not subject to severe eliminative influences, so that the 
low birth-rate is sufficient to keep the species alive, or else it is in imminent danger 
of extinction. The former supposition seems more probable. 
TABLE 35.—Partula filosa. Fecundity. 
No. Percent| No. of | No. of | Total | Average Average 
gravid. | gravid. eggs. | young. |contents.|for Bene for all. 
Series. | Records. 
20 
1909 - 97 
TaB Le 36.—Partula filosa. Pirai Valley. Heredity of color. 
Lower valley (1906). Upper valley (1909). 
Color 
classes of Classes of young. Classes of young. 
parents. Total. Total. 
A B Cc D E A B Cc D E 
The unborn snails possess shells which are usually colored like those of the 
parent, but this is not invariably the case, for they sometimes belong to a lighter or 
to a darker class than their parents. ‘The statistics are given in table 36. It is 
interesting to find that when two well-advanced young are present, they usually 
belong to the same color-class, but sometimes they differ; in such rare cases one of 
them is like the parent, while the other belongs to a color-group only one degree 
