THE SALMON AND ITS MIGRATIONS 181 



their long up hill fight again. Some persons con- 

 tend that because salmon ascend rivers three and 

 four months before the spawning season begins, 

 they do not do so solely for the purpose of 

 breeding. For what else can it be? It is cer- 

 tainly not to escape from their enemies, for these 

 are about as numerous in the rivers as in the sea. 

 Although it is absolutely necessary that their 

 ova should be deposited at the head waters of a 

 river to hatch out, an hereditary instinct pro- 

 bably teaches them that it is safer there. Such 

 being the case, how could one expect a gravid 

 fish, weakened by the full development of its ova, 

 to be able to surmount all the obstacles found in 

 an ordinary salmon river ? It would be no more 

 fit to do this than an oyster. Hence the neces- 

 sity of ascending the stream early and reaching 

 its spawning bed before this process is too far 

 advanced. There, this development takes place 

 in a state of comparative repose. 



^molta antr (grtls*. The migration of smolts, 

 that is to say, of parr going out to sea, is, 

 even in the same river, very irregular; not so much 

 in the dates of the migration as in the numbers 

 of the fish. This variation does not appear to 

 depend upon the abundance or scarcity of the 

 parent fish in the previous two or three years. 

 I presume that there must be in some years 

 an excessive loss of ova, either by severe winter 

 floods, disease or parasites, &c. Whatever may 



