GEN. LEUCISCUS. THE ROACH. 73 



sea, not, I apprehend, from the sea itself. I have 

 never known a Roach to be taken in the sea, into 

 which the fish had entered voluntarily. Colonel 

 Montague also, when commenting, in his manu- 

 script, upon Mr. Donovan's statement as above re- 

 ported, expresses his belief that the Roach could 

 not exist in sea- water at all ; mentioning the follow- 

 ing fact which came under his own observation. 

 In a small river that runs into a large piece of 

 water of nearly two miles of extent, close to the 

 sea, on the south coast of Devon, there is no out- 

 let but by means of percolation through the shingle 

 which forms the barrier between it and the sea. 

 In this situation Roach thrive and multiply beyond 

 all example. About eight or nine years ago, the 

 eea broke its boundary, and flowed copiously into 

 the lake, at every tide, for a considerable time, by 

 which every species of fish was destroyed. So be 

 it, replies the facetious author of " The Rod ;" but 

 this fact is by no means of a conclusive nature, in 

 as far as there may be an essential difference in 

 relation to the effect upon a fish's constitution, be- 

 tween a forced and sudden, and a voluntary and 

 graduated contact with saline waters. In the latter 

 case, there is a physiological expectation or prepara- 

 tion for the change, and we doubt if even Salmon, 

 so remarkable for their long and vivacious continu- 

 ance in both conditions of the liquid element, would 

 suddenly suffer a sea change with entire impunity, 

 or enjoy the vice versa if instantaneously transported 

 from Ocean's blue profound, and plunged over head 



