308 



8Cie:nce. 



[Vol. VIII., No. 191 



that many domesticated vai'ieties, though differing 

 from one another to a greater extent than many 

 natural species, retain a perfect fertility among 

 themselves. The consideration that sterility 

 between natural species is not absolute, slightly 

 changes but does not solve the problem. Mr. Dar- 

 win admitted the difiSculty, and suggested the 

 itnprobable hypothesis that the sterility was the 

 incidental effect of uniform conditions of life on 

 the generative system. 2°. The swamping effects 

 of free intercrossing upon any individual variation 

 would more than outweigh the action of natural 

 selection ; and to answer as Mr. Darwin does, that 

 many individuals might simultaneously undergo 

 the same modification, is to appeal to a highly 

 improbable series of events, especially when it is 

 remembered that, 3°, these specific distinctions 

 are so often of a useless character. Mr. Darwin 

 frankly admitted that many of these meaningless 

 detailed distinctions, like the general distinction 

 of sterility, were not explained by natural selec- 

 tion. 



In view of these objections. Dr. Romanes thinks 

 that the theory of natural selectioivhas been mis- 

 named. It is at once a different and a much 

 broader theory, — different, because it explains 

 the origin, not of species, but of adaptations of all 

 kinds, morphological, physiological, and psycho- 

 logical ; broader, because it accounts for these 

 adaptations, whether they occur in species only, 

 or also in genera, families, orders, or classes. To 

 realize, on the one hand, that natural selection 

 does not primarily explain the origin of species, 

 but only the development of adaptations, and, on 

 the other hand, that the distinctions which it does 

 explain are not confined to species, is the key to 

 the right understanding of this great biological 

 principle. When natural selection did produce 

 species, it was because accidentally the differences 

 to which it gave rise were specific in character : 

 its business was to evolve adaptations. 



It is to one among these other causes which 

 have been shown to be necessary for accounting 

 for the origin of species that Dr. Romanes devotes 

 special attention : he calls it the prevention of in- 

 tercrossing with parent forms, or the evolution of 

 species by independent variation. 



The number of trifling variations, even in one 

 generation, is enormous. The fact that natural 

 selection preserves the useful ones alone, and yet 

 can furnish 'the whole adaptive morphology of 

 nature,' gives us a glimpse of the necessarily enor- 

 mous number of non-surviving, useless variations. 

 Now, if the possessors of any of these useless vari- 

 ations were prevented by any means from inter- 

 crossing with those who did not possess them, 

 these unusef ul variations would be perpetuated by 



heredity (witness our domesticated productions), 

 and those varieties of the old sjiecies would grad- 

 ually pass into a new species. On this principle, 

 the opportunities for independent breeding Tvith- 

 out crossing with tlie parent forms explains the 

 extraoi'dinary i^revalence of peculiar species in 

 isolated oceanic islands. Geographical barriers 

 and migrations can produce the same result. And 

 this hypothesis is made doubly strong by the con- 

 sideration, that, in these cases where the extinc- 

 tion of the variation has been prevented (by pre- 

 venting the swamping effects of intercrossing 

 with the parent form), the variations thus perpet- 

 uated are generally of a useless character. But 

 the existence of natural barriers will not account 

 for all cases of species-formation by independ- 

 ent variation, because some degree of sterility 

 occurs between even closely allied species, and 

 because closely allied species are not always sep- 

 arated by geographical barriers. The principle 

 of physiological selection must be called in to com- 

 plete the explanation. 



Probably the most variable part of the organism 

 is the reproductive system ; and these variations 

 are either in the direction of increased or of dimin- 

 ished fertility. These variations would be more 

 commonly observed, were it not that by their very 

 nature they lead to more or less immediate extinc- 

 tion. But if the sterility were confined largely to 

 crossing between the parent and the varietal 

 form, while the varietal form continued fertile 

 inter se, the conditions for the formation of a 

 new species would be furnished. The result of 

 this would be, that, as before, some individuals 

 living on the same area as the rest of their species 

 would be prevented from having jwogeny with this 

 rest. The only difference is, that in the former 

 case the barrier was geographical : here it is 

 physiological. It is understood, then, that wher- 

 ever such a variation in the reproductive system 

 occurs that diminishes the fertility between the 

 varietal and the parent form, though retaining it 

 among the varietal, this physiological barrier will 

 end in dividing the species into two sections, each 

 free to develop independent, distinct histories. 

 On this ])rinciple, variations in parts other than 

 the reiM'oductive system, unless such variations 

 were useful in character, would not be preserved ; 

 but, when the difference in respect of the repro- 

 ductive system had set in, other differences would 

 secondarily supervene by independent variation. 

 To prevent an unfair objection, it may be added 

 that this theory is not concerned with the kind 

 or cause of this variation any more than that of 

 natural selection : it sets out with the fact. 



It will be impossible in so brief a notice to do 

 more than outline the evidence which Dr. Romanes 



