OTHER THEORIES OF SPECIES-FORMING. 355 



ation between males and females are due largely to the 

 necessity of the better protection of the young 



Wallace's producing and (in the case of birds and mam- 

 mals) young protecting and caring for female, 

 and hence the acquirement on her part of a dull incon- 

 spicuous protective colour-pattern. Wallace's large ac- 

 quaintanceship with birds and butterflies enables him to 

 illustrate his theory by many apparently confirmatory ex- 

 amples, but as soon as one stops to consider the matter 

 thoughtfully the impossibility of the general or even wide 

 application of this explanation of secondary sexual char- 

 acters is at once apparent. It is necessarily limited to one 

 single category of sexual differences. 



Barrett-Hamilton 33 has noted that both sexes of the sal- 

 mon (Onchorhynchus) become markedly discoloured during 

 the spawning season. The discoloration is accompanied by 

 overgrowth or hypertrophy, especially of the jaws. "I can- 

 not believe," he says, "that this is of an aesthetic nature, since 

 these phenomena terminate in the death of the fish. They 

 seem to be, in fact, merely the outward symptoms of what, 

 as I have persuaded myself from personal observation in 

 Kamschatka, is a pathological condition accompanying, and 

 perhaps resulting from, the growth of the ova and milt. 

 I regard the whole metamorphosis as a purely excretory 

 phenomenon resulting from the upsetting of the metab- 

 olism due to the concentration of the whole vital force 

 on the effort to produce the greatest possible amount of 

 spawn. 



"May not such a state of things be invoked to explain the 

 nuptial changes of our own salmon so strangely assumed 

 before and lost after the breeding-season ? Is it not possible 

 that in the phenomena displayed by the spawning Oncho- 

 rhynchus we may have a clue to the origin of the hitherto 

 inexplicable temporary and permanent sexual characters 

 of the vertebrates and even of some invertebrates, of which 



