14 



muddy and the other clear; but the most interesting conclusion was a 

 notable difference in degree of specialization in the fishes inhabiting the 

 different sections of a stream system. Those from the larger river were, 

 as a rule, not only the largest, but the most primitive, or the least special- 

 ized; those preferring the bottomland lakes were, on the whole, more 

 highly differentiated; and those from the creeks were smallest and the 

 most highly specialized of all. 



It is perhaps what we ought to expect, that the creek species should 

 be more diverse and highly organized than any other fishes, for they 

 must have had longer experience of fresh-water life. As the continent 

 first began to rise from the sea, all its streams were necessarily small, 

 and its first fishes were consequently those adjusted to life in creeks. 

 As the process continued, as the surface of the land became more di- 

 versified, as the small stream systems of the coast ate their way back, 

 with many lateral branches, and united to form large rivers, and these 

 again to make rivers of the largest size, new habitats would be formed, 

 both in the uplands and along the coast, and new adaptations of fishes 

 to them would naturally lead to about the kind of classifiable diversity 

 which we actually find. 



That ecological differentiations and divisions among fresh-water 

 fishes are, as a rule, of no very compelling force is shown in a remark- 

 able manner by the fact that the whole system of such distinctions breaks 

 down almost completely at least once a year, when a great migration 

 movement up-stream and into shallow water seizes all species alike, under 

 the overpowering impulse of the breeding instinct. At this time fishes 

 of the most varied habit and habitat seem temporarily to desert or forget 

 their favorite places of resort, and throng together, indifferent to their 

 individual welfare, in search of places for the deposit of their eggs, 

 and, with many species, for the subsequent care and protection of their 

 young. Even under less extraordinary circumstances I have found, in 

 fact, that a large river like the Illinois becomes a sort of metropolis of 

 the fish population of its drainage basin, in which representatives of the 

 various groups or associations, separate and distinct in its headwaters 

 and smaller tributaries, may be found indiscriminately commingled, just 

 as in this great city we see people from scores of smaller cities and 

 hundreds of smaller towns and thousands of rural communities. Sunfish 

 species, for example, which rarely occur in each other's company in 

 collections from the smaller tributaries, were found together twice as 

 often in collections from ihe larger streams. 



