THE NEW AET OF BREEDING FISH. 35 



on the gravelly bottom of this artificial rivulet, he 

 watched with care during the six or seven weeks of 

 incubation, all the varied phases of their develop- 

 ment, with a view of discovering any hidden obstacle 

 to the success of the experiment. He found that 

 the time necessary for incubation varied with the 

 temperature : that it took a much longer time when 

 the water was cold than when it was moderate. He 

 found, too, that a sediment is deposited on the eggs, 

 which is hurtful or destructive if allowed to remain, 

 and to remedy this difficulty he cleaned them with 

 the feather of a quill. 



" Care must be taken to remove from time to 

 time the dirt which is carried by the water and de- 

 posited on these eggs ; this can be done by stirring 

 about the water with a quill feather." * 



Placed in these favorable conditions, and sub- 

 ject to his assiduous care, the eggs passed safely 

 through every stage of their development. The 

 young fish, thus hatched, seemed quite equal to those 

 hatched in the natural way. After their birth he 

 still preserved them for five weeks, and did not turn 

 them into his brooks till the umbilical bladder ap- 

 peared absorbed, that is to say, until they commenced 

 to feel hunger. 



After an experiment so successfully made, and 

 often repeated, the ingenious author of this discovery 

 had the right to say as he did, that his method qp- 



* Duhamel, op. cit, 2d part, p. 336. 



