PARTULA OTAHEITANA. 143 
nary class. Pohaitara Valley, in the district of Arue, is sucha gully (plates 9, 10, and 
11). When it divides, at about 24 miles from the coast, the lesser division (Faaiti) 
trends toward Pirai Valley, while its larger division (Faaroa) bends eastward toward 
the upper end of Ururoa. It is unusually narrow, with steep and high sides in 
the interior, and supports little succulent vegetation. 
The snails taken in this valley were found in numbers only in a restricted strip 
of territory bordering the broken-bedded stream, at some distance back from the 
coast; they occurred on Freycinetia, Pandanus, and “oaha,”’ as well as on other 
bushes and trees in smaller numbers. The natives stated that many plantains and 
orange-trees grew on the heights above, but that no snails were to be found there. 
The barometric level of the area of collection was from 1,100 to 1,450 feet above 
the sea. 
All of the specimens of P. 0. amabilis taken here were sinistral. The distin- 
guishable color-classes do not correspond with those of previously described colonies, 
although the primary variety is certainly the same as before. ‘The chief difference 
is in the virtual absence of the reddish color-factors expressed most clearly in Class 
III shells of Fautaua, Hamuta, and Pirai. Here the shells fall into three divisions, 
namely, “light” (plate 26, figs. 18 to 22),““medium’ (plate 26, figs. 23 to 30), and 
“dark” (plate 26, figs. 31 to 36). The first of these is intermediate between Classes I 
and II of Pirai, always lacking the tinted apex; the second is about midway between 
Classes II and III of Pirai, while the last is much like Class [V without the red tinge 
of many previous examples. Decortication sometimes results in a pinkish color of 
the shell (plate 26, fig. 30), but this is essentially different from the red hue of shells 
assigned to the darker classes in other valley colonies. 
The numerical representation of the three classes is as follows: 
The statistical data relating to the three divisions 
and to the colony as a whole are given in table 78. As 
compared with the Pirai series, there is a notable 
diminution in absolute measures, both of the whole 
shell and of the aperture; this is so great in the case 
of shell length as to render the figure for shell propor- 
tions far higher than in any other colony of amabilis. Of the three color divisions 
the light class is the smallest, and the aperture length bears a noticeably higher 
proportion to the total length of the shell. The tooth (table 79) is least developed 
in the light shells. 
In the matter of fecundity (table 80), the snails of this valley show a remark- 
ably low percentage of gravid individuals (27.52 per cent) and a low degree of 
fertility among these; yet the animals were collected at a time when the colonies of 
neighboring valleys were actively breeding. 
Only 25 young were sufficiently advanced to give data on heredity (table 80), 
all of which were sinistral like their parents. The dark snails bred true, but their 
class received additions from the other two groups of adults. Hence the color 
lines are not impassable, and the statistical differences of the tables do not signify 
that rigid differentiation has taken place within the colonial limits. 
