Dispersion of Fresh-water Fishes 283 



called it, the "horned dace"; and in both were large schools 

 of shiners* and of suckers. f But in every deep hole, and espe- 

 cially in the millponds along the East Coy Creek, the horned 

 poutj swarmed on the mucky bottoms. In every eddy, or in 

 the deep hole worn out at the root of the elm-trees, could be 

 seen the sunfish,§ strutting in green and scarlet, with spread 

 fins keeping intruders away from its nest. But in the Oatka 

 Creek were found neither horned pout nor sunfish, nor have 

 I ever heard that either has been taken there. Then besides 

 these nobler fishes, worthy of a place on every schoolboy's string, 

 we knew by sight, if not by name, numerous sm.aller fishes, 

 darters !| and minnows,1i which crept about in the ' gravel on 

 the bottom of the East Coy, but which we never recognized in 

 the Oatka. 



There must be a reason for differences like these, in the 

 streams themselves or in the nature of the fishes. The sun- 

 fish and the horned pout are homedoving fishes to a greater 

 extent than the others which I have mentioned; still, where 

 no obstacles prevent, they are sure to move about. There 

 must be, then, in the Oatka some sort of barrier, or strainer, 

 which keeping these species back permits others more adven- 

 turous to pass ; and a wider knowledge of the geography of 

 the region showed that such is the case. Farther down in its 

 course, the Oatka falls over a ledge of rock, forming a consider- 

 able waterfall at Rock Glen. Still lower down its waters dis- 

 appear in the ground, sinking into some limestone cavern or 

 gravel-bed, from which they reappear, after some six miles, in 

 the large springs at Caledonia. Either of these barriers might 

 well discourage a quiet-loving fish; while the trout and its 

 active associates have some time passed them, else we should 

 not find them in the upper waters in which they alone form 

 the fish fauna. This problem is a simple one ; a boy could 

 work it out, and the obvious solution seems to be satisfactory. 



* Notropis cumulus Rafinesque. 



■\ CatOiiomus commersoui (Laccpedc). 



J Ameiurus nielas Rafinesque, 



§ Eupomotis gibbosus Linna:us. 



\\Eihcostoma flabellare Rafinescjue. 



T[ Rhinichihys alronasus MitchiU. 



