PARTULA OTAHEITANA. 279 
Actually, when the combination and corrections of (5) have been made, the figures 
are 51 and 39, respectively, thus disclosing a discrepancy of 4 out of 90 cases, or 
4.4 per cent. 
The foregoing analysis does not lead to a decisive conclusion regarding the 
dominance or recessiveness of the red color, although the probability of a Mendelian 
order of inheritance is indicated. It is by analogy with the case of P. 0. rubescens 
and with that of Helix as investigated experimentally by Lang, that the red color 
factor would be regarded as the dominant member of the pair. 
NORTHERN CRASSA SERIES—ATEHI TO PUNARUU VALLEYS. 
In the three valleys to the northward of the small area occupied by crassa 
occidentalis, the otaheitana colonies display the principal features of the typical 
crassa found to the south. Atehi (which is next to Papehue) and Maruapoo are 
short, deep valleys of the fourth order of size, while the third area is Punaruu, a 
great element of the first order. Here, then, we find one and the same variety 
in three valleys, whereas in the case of P. nodosa each topographical element is 
inhabited by a distinct primary variety, viz, nodosa exigua, nodosa leva, and nodosa 
composita. The lack of differentiation in the one case and its existence in the other 
case can scarcely be referred to identical environmental conditions. 
The three colonies of this northern series are made up of phea and cestata color- 
classes, in absolute and relative numbers that have been given earlier in table 233. 
In Atehi, the former class is made up of confluens in the main (plate 33, figs. 50 to 53) 
although some intergrading striata occur; two exceptional decorticated examples 
were found, one of which is illustrated in plate 33, figure 50. In Maruapoo also 
the striata shells were few, and were not sharply separated from the confluens class 
(plate 33, figs. 59 and 60). The third colony, that of Punaruu, comprised a large 
number of very characteristic striata (plate 33, figs. 61 to 64), which as before gradu- 
ally merged into confluens. Shells of the cestata class greatly preponderate in Atehi 
(plate 33, figs. 54 to 58), but form a minority in the other two valleys (plate 34, figs. 
tand2). In general the “median” band is more clearly marked than in the crassa of 
the southern series, and the sutural and basal bands are sometimes developed so as 
to render the pattern somewhat like that of sinistralis banded examples. 
In statistical respects (table 234) the associations differ from one another in 
various ways, in different measures; the details are sufficiently described by the 
numerical data. When the plain and banded classes are compared the relations 
existing in one case are reversed in another, without any consistency. 
The several subdivisions of the Punaruu colony, taken in different years from 
the lower, middle, and upper parts of the inhabited area, are not identical in their 
morphology. The most interior specimens are not much larger than the typical 
crassa, and hence the statistical results do not accord with Garrett’s statement 
that “far above the range of nodosa in Punaruu, the otaheitana shells are similar 
to, but larger than, crassa.” Possibly at the very head of the valley they may be, 
but my own explorations could not be prosecuted beyond a distance of 5 miles 
from the coast. 
