ETHEOSTOMIN^ — THE DARTERS 279 



which they mainly depend for food. They swim mainly by 

 means of their pectoral fins, making quick dashes in the current 

 as a bird might make a short, rapid flight against a high wind, 

 and' resting in the intervals upon their extended ventral and anal 

 fins. Unlike most of the taxonomic groups we have hitherto 

 discussed, the darters thus form a rather definite ecological 

 assemblage, assimilated by their like adaptive characters and 

 by their similar relations to like situations. There are, never- 

 theless, well-marked degress of adaptation among the different 

 genera and species; and, likewise, in the strictness of their con- 

 finement to the class of situations characteristic of the group. 

 Three of our species, for example, are often found in still or 

 sluggish waters and over a muddy bottom; one, the sand-darter, 

 is much the commonest in streams with a sandy bottom; and 

 another, Cottogaster shumardi, is most abundant along the borders 

 of the largest rivers. The species are likewise distinguishable in 

 other features of their local distribution, as may be readily seen by 

 a comparison of the distribution maps of the darters in the atlas 

 accompanying this report. The force of competition is thus more 

 or less broken among them in various waj's, no exact analysis 

 of which has ever been attempted. The origin of these species 

 is an interesting and inviting problem, particularly open to 

 solution because of the comparatively restricted range of the 

 family and the fact that there is nothing to suggest an extensive 

 migration from the place of their original differentiation. 



The food of the subfamil}^ was studied by the senior author 

 many years ago from the contents of seventy stomachs repre- 

 senting fifteen species, collected in various parts of Illinois in 

 several months of four successive years. These indicated more 

 than their number would imply, since different darters obtained 

 from the same locality and on the same day usually agreed so 

 closely in food that the study of from two to five specimens 

 gave all the facts obtainable from several times as many. 

 Furthermore, the differences between the related species in 

 respect to food are so slight that specific peculiarities were 

 scarce^ recognizable. The data obtained, therefore, really 

 apply to the food of the whole subfamily at different seasons in 

 twentj^-nine localities within this state. This was, on the whole, 

 remarkably uniform, except that two of the species, the largest 

 and the smallest of the group, were found to differ from the 

 remainder in a way to correspond to a notable difference in 

 their local distribution. 



