PARTULA RADIOLATA. 53 
different places may be identical in morphological respects; thus the bicestata classes 
prove to be virtually the same in the length and in the proportions of their shells, 
in western Cabras and Orote, which localities exhibit minor differences in botanical 
features and a more marked difference in the matter of altitude 
While it is interesting to speculate as to the antecedent episodes in the history 
of radiolata leading to the present situation, it must be admitted that the facts of 
record do not throw much light upon this history. One fundamental point is firmly 
established, namely, that the six color-classes are unquestionably members of one 
and the same species, because they agree in the general diagnostic features and also 
because they interbreed mutually, as the evidences to be presented in the next 
section will demonstrate. It is permissible to regard some at least of the several 
color-types as mutants, showing various degrees of distinctness, but the material 
itself gives no answer to the question as to which was the original form and which 
are the products by independent or derived mutation. It is true that the flavea 
class is the most abundant at the present time, but it is conceivable that another 
class was really ancestral to flavea and that this parent class has diminished in 
numbers while its product has gained ascendancy. To make the problem more 
specific, we may consider the case of bicestata, which now occurs in the widely 
separated localities of Saucio, Agafia, the Apra Region, and on Mount Salifan. 
The hypothesis is tenable that this form was at one time abundant and wide- 
spread and that the genetic factors responsible for its manifestation have dropped 
out, save in a few associations. It is also conceivable that the mutation by which 
it may be assumed to have arisen at one place has occurred in an identical manner 
and independently in another area. A third possibility is that the type arose at a 
single center, from which it subsequently spread to other territories; such an 
episode might well have preceded the first-mentioned process of local decimation 
and extinction, to which it is by no means an exclusive alternative. If the bicestata 
pattern arose by mutation at a single place, it must have been displayed at the 
outset by only a few snails whose structural characters would unquestionably be 
different from the general colonial averages unless the parental individual were 
itself average in all respects, which never seems to be the case. Because a certain 
amount of assortative mating takes place owing to the disinclination of the animals 
to wander far from their nurse-plants, the number of individuals with the dis- 
tinctive pattern and similar measures would increase. As time passed, the larger 
numbers of the new kind would migrate over wider areas and necessarily interbreed 
with differently colored and differently formed members of the species; inevitably 
the original structural distinctness of the new type would become less marked as 
the intermixture continued, while the manifestation of the peculiar pattern would 
depend upon the requisite combination of the genetic factors responsible for it. 
Possibly there are indications of such a history in the larger size of the bicestata 
shells in the Cabras, Orote, and Salifan areas, although the differences between such 
individuals and their associates are not always statistically significant. 
However, the statistical differences between and among the several color- 
classes of radiolata, as they stand to-day, are not competent to solve the problems 
