124 VARIATION, DISTRIBUTION, AND EVOLUTION OF THE GENUS PARTULA. 
show only 50.7 per cent sinistral, although their young exhibit a slight increase over 
this figure. 
To sum up, it would seem that (a) the relative numbers of dextral and sinistral 
shells vary from time to time; (b) that a marked increase in the sinistral type 
occurred during the last quarter of the nineteenth century; and (c) that subse- 
quently a slight relative decrease of the same type has supervened. Obviously a 
study of the facts after an interval of 15 to 25 years would be of the greatest interest 
and value. 
Cotor AND CoLoraTION. 
While the comparison of the several color-classes is to be made at a later junc- 
ture, it is important at this point to note whether the shells of one direction of coil 
comprise a larger proportion of the individuals of any one class. The figures of 
table 60 give the following results: the only banded shells—3 out of 955—were 
sinistral; shells with the reddish coloration constitute nearly 50 per cent of both 
the direct and the reversed form-classes; the Class I shells are next in abundance, 
almost uniformly in the several series, and they are followed by Class II and then 
by Class IV. The above facts show that no wide differences in color composition 
distinguish the sinistral and dextral groups from one another—that is to say, no real 
differentiation on the basis of color accompanies the difference in the direction of 
the spiral. 
The hereditary relations between the two form-classes are sufficiently impor- 
tant to demand separate treatment at a later point. Here we may restate the 
essential fact that the young of either kind are sometimes of the contrasted direction 
of coil. 
DIMENSIONS. 
The next question of importance deals with the existence of real differences 
between the sinistral and dextral shells in dimensional respects that would be 
brought to light only be employing statistical methods; the two classes can not be 
absolutely distinct, like two entirely different species, such as the sinistral P. 
mooreana and the dextral P. suturalis of Vaianae Valley in Moorea, for the reason 
stated in the previous paragraph. Yet the immediate question is one of prospective 
interest, if nothing more, because other valleys of Tahiti contain varieties of P. 
otaheitana which exist in both the direct and the reversed states. 
The most striking results of comparing the two classes on the basis of dimen- 
sional statistics (table 58) is the fact that the sinistral shells are certainly shorter 
in total length, and certainly stouter in shell proportions; the differences are respec- 
tively 5 and 6 times their probable errors. The shells of the former group are 
shorter also in aperture length with probability only, wider with certainty, and 
the aperture is certainly longer with reference to the total length of the shell. In 
brief, the sinistral shells are more tightly coiled than those of the contrasted group, 
which on the whole are substantially the same in color and markings, and to which 
they sometimes contribute representatives from among their offspring. 
The curves of variation of the two groups of shells are given in table 59; they 
show graphically the degrees of resemblance and difference in all seven of the 
standard characters. 
