II4 VARIATION, DISTRIBUTION, AND EVOLUTION OF THE GENUS PARTULA. 
in Garrett’s time; but such errors do not materially affect the fundamental value 
of the monograph in question. Pilsbry has followed Garrett’s description in its 
essential features and has quoted the geographical details at length. 
Mayer’s illuminating study of 1899, confirming the observations of Garrett as 
regards the inhabitants of six valleys, for the first time showed how important for 
the problems under consideration were the young snails harbored by the gravid 
adults. The additional value of Mayer’s results inheres in the fact that he describes 
an intermediate period between the time of Garrett’s original and basic discoveries 
and the present; the intermediate condition is especially significant in the case of 
the Fautaua colony, on account of the taxonomic and biological primacy of the 
shells of that association. 
In describing the abundant material of this species, it will be necessary to take 
many things into account and to enter into considerable detail. First of all, 
primary varieties and their subordinate modifications must be described and dis- 
tinguished, and referred wherever possible to earlier defined types. But a primary 
variety can not be treated solely by itself without entering into detailed specifica- 
tions regarding the geographical location of its headquarters and of its outlying 
representatives. On the other hand, it would not be justifiable to follow the species 
as a whole from valley to valley, for such a course would involve the parallel con- 
sideration often of two, and sometimes of three, principal varieties, and would thus 
lead to inevitable confusion. In the present account a combination is made of the 
two orders of description, namely, the geographical and the systematic; at the 
outset the primary taxonomic divisions of the species will be described briefly, and 
then each of these will be taken up and followed throughout its geographical range, 
regardless of other forms of the species that may accompany it. 
It is necessary to add a further preliminary word regarding the general analysis 
that follows. This species is very complex and variable, but as a whole it is 
absolutely discontinuous with any other, such as P. nodosa. Its members fall 
into groups that are relatively but not absolutely discontinuous; and if the whole 
otaheitana series be taken as a single species, then such component groups are 
properly designated primary varieties, like the corresponding major divisions of 
P. clara and P. nodosa. But in contrast with the situation as described for those 
species, this case differs in that the primary subdivisions of P. otaheitana are far 
more clearly differentiated inter se, while in addition each of them is more or less 
resolved into geographically circumscribed secondary varieties; these last often may 
be more diverse than the primary varieties of another species like P. nodosa. Hence 
instead of viewing the whole otaheitana series as one compound species, with almost 
equal propriety we might regard it as a group of species which do not differ from 
one another as much as each of them does from P. nodosa, P. filosa, etc. The essen- 
tial points, however, are the close genetic relationships of the groups here called 
primary varieties, and their collective demarcation from other species. Conse- 
quently, we shall treat this case as one involving a single species, P. otaheitana, in 
which differentiation has been carried to a far greater degree than in any other 
instance. 
