44 BIOLOGICAL SOCIETY OF WASHINGTON. 



been prolonged across the former lake bed as upper branches of the 

 system, the former outlet becoming the main stream, and the whole 

 a finished river system, with its fauna derived from the lake in which 

 the system originated. 



It is a well-known fact that many fishes, in connection with their 

 breeding habits, pass from marine to fresh waters, and return to 

 marine, with entire indifference to the change. Again, some exist- 

 ino; so-called land-locked fresh-water fishes are believed by natu- 

 ralists to have become such by choice, or by a failure on their part 

 to continue their periodical returns to the sea where their kind 

 originated, even when there was no apparent physical obstacle to 

 their returning. It is probable that many similar cases have 

 occurred in former geological periods, and also probable that a few 

 mollusks and other invertebrates have, during those periods, in like 

 manner changed from a marine to a fresh- water habitat; but I 

 believe that, as a rule, fresh-water faunas have primarily become 

 such by compulsion, in the manner that has just been suggested. 



Admitting this proposition, we need not attempt to trace the 

 genetic lines of fresh-water fishes as such any further back than the 

 time of the compulsory land-locking of their progenitors ; but this 

 would not forbid speculation as to what kinds of marine fishes the 

 fresh-water forms originated from. 



It is possible that, in cases of sinking beneath the sea of land 

 areas, upon which fresh waters with their faunas had become estab- 

 lished in former geological periods, the sea has reclaimed and pre- 

 served alive some of its previously alienated mollusks and fishes ; 

 but this is a matter concerning which we can, at best, make only 

 vague conjectures. 



The reason why the animal life of fresh waters is so meagre, as 

 regards the number and variety of kinds which they contain, when 

 compared with the teeming and diversified life of the sea, are vari- 

 ous. First, the presence of sodium-chloride and other salts in 

 water has evidently been condudve of evolutional diff'erentiation ; 

 and there can be no doubt that common salt has played a remark- 



