8 
mit; and for even so brief a discussion of the topic as is possible for 
is today, a rough classification will be useful. 
Ecological classification of animals may take either one of two princi- 
pal directions, or indeed both of them in turn. Ecology being the rela- 
tions of interaction between organisms and their environment, if we take 
it up from the side of the organism we shall naturally prefer a classifica- 
tion based on differences of animal reactions to the same environmental 
conditions—-differences of behavior, that is—and our classification will 
be a habit classification; if it is the environment, on the other hand, 
which interests us primarily, we shall assemble our animals in groups 
according to their environmental preferences, and our classification will 
be a habitat classification. The habit classification is fundamentally 
physiological and the habitat classification is primarily physical or spatial, 
and the two cut across each other, often at right angles, each habitat con- 
taining associated animals of various habit, and each habit group being 
distributed, as a rule, over various habitats. A mill pond, for instance, 
is a very definite habitat, but it may contain fishes of every sort of habit; 
strictly piscivorous fishes form a very definite habit group, one or more 
of which may be found in almost any kind of aquatic habitat. 
I do not myself favor the attempt to reduce the facts and materials 
of animal ecology to one hard and fast, all-including classification, such 
as biologists attempt to establish, in the face of almost infinite difficul- 
ties, for descriptive botany and zoology; but I believe that our ecological 
classifications should be as various as the objects we have in view, being 
made sometimes on one basis and sometimes on another, as best serves 
our purpose at the time. It will serve my present purpose to classify 
fishes first in general terms according to the principal elements of their 
food, thus forming habit groups, without present reference to their 
habitats. 
I must first acknowledge, however, that fishes can not be completely 
and clearly divided into mutually exclusive groups upon this basis, for 
their choices of food and their capacities for its appropriation are not 
sufficiently fixed and definite in the different species to make this prac- 
ticable. I can best describe the actual situation by saying that fishes 
have a common body of food resources of miscellaneous character upon 
which many of them draw almost indiscriminately according to the cir- 
cumstances at the time, but that from this common mass of resources, 
habits, and capacities there is a tendency to specialize in various direc- 
tions, which tendency goes to its limit in some species, halts at various 
