PARTULA OTAHEITANA. 239 
As in the neighboring valleys, two color-classes are represented. Banded 
snails corresponding in general with cestata constitute 33 per cent of the collection 
from the lower valley and 38.6 per cent of the more inland series. Aside from 
variations in size, well indicated by the specimens figured (plate 31, figs. 35 to 38), a 
distinct color-type is displayed, where, instead of a zonate medium band, there is a 
narrow revolving stripe of dark brown (figs. 35, 36); the narrowed band corresponds 
in position with the lower, or morphologically right, portion of a zonate marking. 
The unbanded shells (plate 31, figs. 29 to 34) also display considerable variation 
in size and form, but even more in the ground-colors; indeed, this division is hetero- 
geneous to an unusual degree. Some shells (fig. 29) are typical rich seal brown, 
like the original phea of the colonies already treated; from this extreme they grade 
through progressively lighter conditions, with or without strigations, to shells that 
approach apex of Titaviri (fig. 33), which, as we have seen, possesses a somewhat 
darkened color on the larger whorls. A novel feature is the “fleshy” cast in many 
cases (fig. 34). Decortication also produces a peculiar appearance, especially in 
the case of a strigated shell (plate 31, fig. 32). Taking the unbanded class as a 
whole, it may be regarded as phea, but it displays wider variations; its general 
average of tone is light on account of the paucity of the original dark shells, while 
the strigated subtype attains greater prominence than before. From what follows 
it seems certain that the condition observed here is transitional to that of bandless 
classes to the west, which have subordinate color-types more clearly differentiated 
from one another. 
The dextral specimens assigned to sinistrorsa are different from one another. 
The first (plate 31, fig. 40) is a typical cestata, with bands that do not agree at all 
with those of P. producta; furthermore, its single embryo was sinistral. The band- 
less specimen (plate 31, fig. 39) is more or less like unstriped producta as well as 
light-colored phea, but in form it agrees with the latter only. 
The statistics (table 185) show that the two color-divisions do not differ greatly 
in either of the two series. When these latter are compared, however, it appears 
that the shells of the lower valley (1907) are collectively longer, wider, and stouter, 
with longer, wider, and squarer apertures, as compared with those of the inward 
locality. The proportion of aperture-length to shell-length is substantially the 
same, and the tooth averages about the same in degree of development. 
Fecundity (table 186) is noticeably higher for the 1907 collection. ‘The figures 
do not indicate with reliability a positive advantage in favor of a particular color- 
class in the way of differential fertility. 
In the heredity of color-characters (table 187) the individuals with unbanded 
shells prove to be the more conservative. ‘The cestata snails produce a comparatively 
high percentage of unbanded young, so that the total numbers of the two types among 
the young agree quite closely with the figures for the gravid adults. It is worthy 
of note that the single young snail produced by the dextral banded adult was banded 
like its parent, but was sinistral! ‘The two directly coiled individuals are presum- 
ably sporadic mutants from the sinistral series, and not migrants from a distant 
colony comprising a large number of dextral individuals. 
