4 
intermediate stages in others, and in still others is hardly discernible at 
all. Moreover, the food choices of scarcely any fishes are so definite and 
unchangeable as to be unmixed and identical under all conditions, in all 
parts of their habitat, and at all times of the year. We shall find, indeed, 
the same state of affairs when we come to deal with a habitat classifica- 
tion ; and if we look the whole field of ecology over we shall see it rather 
characteristic of an ecological classification generally, especially on the 
animal side. We can nevertheless profit greatly by such groupings of 
our heterogeneous data as are still possible, if we admit the limitations 
of the scheme and understand their significance. 
One of the most peculiar of our food habitat groups contains the 
gizzard-shad and the stone-roller as its most notable representatives, to- 
gether with a few minnows less strictly limited to it, all of them char- 
acterized by unusually long, convoluted intestines, the gizzard-shad hav- 
ing also the digestive surface still further increased by the development 
of a very large number of finger-like czeca on its anterior section. These 
fishes all discard intermediary agents, and help themselves to the raw ma- 
terials of their food in the form of the mere mud and slime of the bot- 
tom, which contains, of course, a considerable quantity of organic debris, 
mostly of vegetable origin. They form the group of the mud-eaters. 
(Figs. 2, 3, and 4.) 
The gizzard-shad, although but little eaten, is one of our most val- 
uable fishes, since it is enormously abundant in our large waters, both 
rivers and lakes, competes with no other species for food, and is itself 
the principal food of our game or predaceous fishes—the most highly 
valued products of our fisheries. It affords, also, a remarkable instance 
of a transformation or development in the food habits and resources of 
fishes, coincident with increase in size. From the time this fish hatches 
from the egg until it comes to an inch or so in length it is as slender as a 
minnow, with the alimentary canal a simple straight tube. Still more 
remarkable, although the mouth of the adult is perfectly toothless, the 
young have, at this stage, a row of conical, pointed teeth upon the upper 
jaw. (Fig. 5.) Teeth would evidently be useless to it in sucking up 
mud or straining out plankton from the water; but to the larva—if such 
it may be called—they must be very useful, for instead of being a mud- 
eater the fish is predaceous in this stage, its prey being the minute animals 
of the plankton, especially the Entomostraca, which it pursues and cap- 
tures one by one as a pike might capture minnows. With its growth and 
transformation it changes its habits slowly, its food becoming more mixed 
