188 BULLETIN OF THE BUREAU OF FISHERIES. 
correlated, with marked physiological differences sufficient to adapt the two to differing 
habitats. Thus the assertion so often made that the slight structural differences by 
which we distinguish one species from another are commonly of no conceivable utility, 
and therefore can never have arisen through the action of natural selection, loses much 
of its force. While it may be true that these slight structural differences in themselves 
can play no significant réle in the life of the organisms concerned, it is likewise evident 
that there are certain correlative physiological changes sufficient to adapt the organisms 
to somewhat different modes of existence. That natural selection has been the con- 
trolling factor in the origination and perpetuation of such specific differences, whether 
morphological or physiological, is far from certain. But that the characters con- 
cerned are in most cases too insignificant to be of selective value is also far from certain. 
Where we have to do merely with the adoption of a more restricted habitat by 
one species than by another, it is quite possible that the physiological difference in 
question relates merely to general constitutional vigor; i. e., the less hardy species 
may restrict itself to the more favorable portion of the habitat.¢ Where, however, the 
ranges of the two species are more or less complementary to one another, particularly 
if they do not coincide throughout any portion of their extent, such an explanation is 
of course out of question, and we are obliged to fall back upon the assumption that each 
is more or less specifically adapted to its respective habitat. 
In order to throw light upon the second of the above questions (i. e., Are members 
of the same genus less likely to be associated together than species which are not so” 
closely related ?), we have adopted a method employed by Herdman (1895). This author, 
after noting the relatively large number of genera represented by the species taken 
in a single dredge haul,’ writes: ‘‘ These figures are particularly interesting in their 
bearing on the Darwinian principle that an aminal’s most potent enemies are its own 
close allies. Is it then the case, as the above cited instances suggest, that the species 
of a genus rarely live together; that if in a haul you get half a dozen species of lamel- 
libranchs, amphipods, or annelids they will probably belong to as many genera, and 
if these genera contain other British species these will probably occur in some other 
locality, perhaps on a different bottom, or at another depth? It is obviously necessary 
to count the total number of genera and species of the groups in the local fauna, as 
known, and compare these with the numbers obtained in particular hauls.” In Liver- 
pool Bay, for example, ‘the known number of species of higher Crustacea is 90, and 
these fall into 60 genera. So the genera are to the species as 2 to 3,”’ whereas in certain 
dredging collections cited ‘‘the genera are to the species on the average about as 28 to 
31, or nearly 7 to 8. Again, the total number of species of Tunicata is 46, and these 
are referred to 20 genera; while in the case given above * * * the 12 species taken 
on one spot represented ro genera, or a little over a quarter of the species represented 
half the genera. These, and many other cases which we might quote, seem to show 
that a disproportionately large number of genera is represented by the assemblage of 
species at one spot, which means that closely related species are, as a rule, not found 
together’’ (p. 463). 
a fhis suggestion has been made to us by Prof. Herdman. 
b This fact was pointed out by Sir John Murray (Challenger “Summary,” p. 1435), who, however, restricts its application 
to great depths, concluding ‘‘in the deepest zone, therefore, the species stand to the genera in the ratio of 5 to 4, and in the 
shallowest zone nearly as 3 to 1.’2 
