96 CURIOUS GENERATIVE EXPERIMENT. 



large portion of the eggs in the body of the fish were in an 

 immature state; but such is not the fact. To prove this, 

 Mr. Keiller once took the roe in a mass from the belly of a 

 salmon recently captured, divided it transversely into three 

 equal parts, and applied to each the needful quantity of milt. 

 In due time the several portions produced fry, though it 

 is true that the portion taken from the upper part of the 

 belly where the eggs were of a somewhat less size, was less 

 productive than the other two. 



So far as Mr. Keiller's observations extended, the salmon 

 never spawns on the bare rock, or amongst very large stones, 

 for the reason, that in such situations she would be unable to 

 raise the needful turbidity to conceal her progeny. 



At the tail of a spawning-ground, the work of a single 

 salmon or at all events never occupied by more than one at 

 a time there is, towards the close of the season, an immense 

 accumulation of gravel, stones, &c. occasionally, indeed, a 

 good English cart-load. What with ice and floods, however, 

 not only is this heap in great part carried away, but the very 

 cavity from whence it came, often of great extent, is so 

 filled up, that by the succeeding summer the bed of the river 

 has assumed nearly its usual appearance. 



" What may be the case in the earlier part of the season, 

 when the fish are in the pools or in deep water, I could 

 not affirm," said my informant ; " but after the female com- 

 mences spawning, I have never, but on one occasion, seen 

 the male in actual company with her. His station at that 

 time is at six or seven feet distance, directly in her wake, 

 and just beyond the heap of stones spoken of. And the 

 only apparent part he takes in the generative process, is 

 by the deposition of his milt, which of course becomes 





