290 THE INFLUENCE OF INANIMATE SURKOUNDINGS. 



larvae, absolute separation of all new varieties from the parent 

 species is thus rendered impossible. Nevertheless, these forms 

 have specific peculiarities as distinctly marked as those of 

 insects, vertebrata, and land mollusca, the only animals which 

 Wagner takes into consideration in his investigations. Accord- 

 ing to his theory, on the contrary, all such species, whose free 

 crossing with the parent form is not prevented by separation, 

 should remain very variable, and should not be distinguishable 

 into any great number of well-defined species. But, this not 

 being the case with the creatures above mentioned, it follows 

 that separation by distance cannot be, as Wagner asserts, the 

 one 'exclusive cause of the origin of new species. 



It is admitted, of course, that Wagner in his argument 

 recognises the influence of the external conditions of existence 

 and duly allows for them : but since these occasionally act as a 

 selective power as in the case of many lower marine animals 

 without the co-operation of a contemporaneous separation of the 

 varieties from the parent form, this last circumstance can never 

 be the sole cause of the process of forming a species, though it 

 may sometimes bear a principal part in it. The first question 

 is thus answered. 



The second question is : Is it indeed the fact that Wagner's 

 separation theory differs so totally from Darwin's theory of 

 selection that each completely excludes the other, as Wagner 

 seems to think 1 It does not appear so to me. Both assume 

 that different species are more or less variable ; both assert that 

 free crossing with the parent form must be prevented if a new 

 species with constant characters is to be developed ; they agree 

 in believing that a selection also must be effected among those 

 variable forms in order to induce the constancy of specific 

 characters and to increase the useful ones by accumulation. I 

 can detect only two trifling differences in their respective views. 

 Wagner appears to think that physical separation or removal, 

 which certainly is a very frequent result of migration, is 

 the means exclusively employed by nature to pi-event five 

 crossing, while Darwin says that this result may often be 

 effected by numerous other and very dissimilar causes, as for 

 instance by differences in the size of the male and female indi- 



