in Animals as afforded by Observations on certain Salmonids. 283 



salvation of the various species as a whole. It may be that 

 the fish which die have, before doing so, produced more spawn 

 than those few which survive, and if that be so, it seems probable 

 that the production of a greater quantity of ova or milt is of far 

 more importance to the species than the mere survival of a few 

 individuals. Moreover, there can be little doubt that the influx 

 into the rivers of such a jostling crowd of hungry fish would act 

 most prejudicially to the young fry by consumption of all avail- 

 able food, and not only that, but by proving a stupendous cause 

 of destruction to the fry themselves. The death, on the contrary, 

 of the old fish should lead to the rapid appearance of a rich 

 invertebrate life, which finding abundant sustenance on the rotting 

 salmon carcases, must in its turn prove an ample food-supply 

 to the young. 



Next to come to the fish that do not die — it is said that one 

 species 1 alone, 0. orientalis (Pall.) the King-Fish or Chervichi, 

 survives the spawning season, and it is most significant that this 

 is one of the species which become the least distorted and dis- 

 coloured of the whole genus. Besides the King- Fish a few 

 individuals of the other species are reported to survive the 

 winter in some of the hot streams of the interior. 



Now I would suggest that we have in these phenomena a 

 possible source and origin of many of the highly developed 

 sexual characters met with in other animals. We have here 

 instances of the production in both sexes of the most gaudy 

 colours, the most conspicuous changes in form, and all as the 

 result of pathological conditions. 



Thus there are several lines of development ready for the use 

 of sexual selection. There are the highly discoloured and dis- 

 torted species which perish to a fish, and the less highly discoloured 

 and less distorted species which survive, although with difficulty. 

 I suggest then that the nuptial changes characteristic of so many 

 vertebrates may possibly be reminiscences of a former condition 

 of things through which their ancestors passed and in which the 

 processes of reproduction were accompanied, as in the Oncorhyn- 

 chus, by disorganism of the metabolism resulting in vivid changes 

 of coloration and hypertrophy or overgrowth of certain organs. 

 There might be various degrees of such a condition, in some of 

 which the animal, as in the case of our Salmon, recovered, re- 

 taining at the same time a varying and temporary extent of 

 change of colour or form of certain organs. 



It does not necessarily follow that death was in all cases the 

 result of such disorganism, and it may be that in particular cases 



1 This I take from Gnillernard. It is, however, not supported by B. W. Evermann 

 and S. E. Meek's Report of observations in the Columbia River basin and elsewhere 

 on the Pacific coast (see Bull. U.S. Fish. Comm., Vol. xvn. 1897, pp. 1-5 — 84). 



