th 
ing teeth, the gill-rakers are short and thick, and relatively ineffective 
as a straining apparatus. This is so generally true that one may even 
tell whether or not a sucker is a mollusk-eater by looking at its gill- 
rakers, although these have nothing to do directly with the collection 
or mastication of molluscan food. 
Another terminus to the series of changes in food choice through 
which most fishes pass is in the piscivorous habit, characteristic of what 
we commonly call our game fishes; especially the pike, the pike-perch, 
and the Great Lakes trout; but the largest number of our fresh-water 
species linger in the intermediate, insectivorous stage. Indeed, taking 
our adult fresh-water fishes as they come, we find that insects are by 
far the most important general element of the food of the class, eaten 
more or less by nearly every kind of fish and the main dependence of a 
great many species which, by mere increase in size and the consequent 
coarser structure of their gill apparatus, have lost the original capacity 
of the young to strain out the plankton, without attaining to a size and 
strength sufficient for the capture of a prey larger and stronger than 
aquatic insect larve. : 
Some kinds of insects occur in such abundance in situations diffi- 
cult of access, that certain groups of fishes have become especially adapted 
to their search and capture there. Darters, for example (Fig. 15), 
live mainly on insect larvzee which hide under stones in swift water, and 
they are enabled to get at this food by virtue of their large pectoral and 
anal fins, by which they can support themselves on the bottom in a swift 
current or make their way among the ripples of a‘ rocky stream, and by 
their small heads and pointed noses which enable them to pry about 
under stones where worm-like Chironomus larve and larve of small May 
flies abound. A little cyprinoid fish—the sucker-mouthed minnow (Fig. 
16)—is very similarly equipped and to a like advantage. Access to 
the same kind of food under the heavier stones of larger streams is given 
to a sucker known as the hammerhead (Fig. 17). It has a similar develop- 
ment of the paired and anal fins, and a large square head with which it 
can push and roll about the stones under which day-fly and stone-fly 
larve may be found in great abundance; and these are its principal food. 
Besides the six food classes which J have already mentioned, namely, 
the mud-eaters, the plant-eaters, the plankton-eaters, the mollusk-eaters, 
the insect-eaters, and the fish-eaters, we may doubtfully distinguish two 
more—the garbage-eaters and the omnivora. There is, indeed, but one 
of our fresh-water fishes—the common eel—which seems to live by pref- 
