( 522 ) 



geological revolutions, become so changed that the isolation is much more incomplete 

 than it former]}' was. And when the comparatively great diversity in the organs of 

 copulation has become so constant that intergradations in these organs between two 

 respective subspecies seldom occur in the area occupied by these forms, though the 

 subspecies may be connected with one another by intergraduate forms inhabiting 

 other districts, it is possible that, when the districts of the two subspecies, if 

 geographically close together, like the Bornean and Philip])ine Islands, become so 

 extended that they overlap, the two diverse subspecies may live together without 

 immediate and complete fusion. 



It will have been noted that the etfeci attributed to mechanical selection is two- 

 fold. The variatiiiu of the genital armature caused by transmuting factors of any 

 kind is so guided by mechanical selection that every incipient species will acijuire a 

 special armature of its own. As the variation of the organs of copulation is inde- 

 pendent of that of the wing-jjattern, mechanical selection does not directly affect the 

 latter ; but its indirect iufiueuce must be of importance, when tlie special genital 

 armature acquired by mechanical selection lias become so specialised that it is, like 

 local isolation, a means of prevention of intercrossing. 



In consequence of the demonstration of individual and geographical variation of 

 the organs of copulation it becomes very probable that Eimer's * interesting sugges- 

 tion is correct, that the variation of a species concerns also egg and sperm. If 

 geographical and individual variation in these cells would be deiuonstrablc as in 

 the case of the ectodermal sexual organs, we should most probably find tliat the 

 variation in the size and form of tlie sperm and in the structure of the microi)yle 

 ap]iaratus f of the egjj is independent of the cliaractcrs of the wing and sexual 

 armature, and that accordingly similar conclusions could be drawn as in the case of 

 the organs of copulation. As diversity in sperm and egg would partly or totally pre- 

 vent fecundation, infertility between certain individuals of the same race and between 

 the s|>e('in]eus of diverse races would be the result of the variation, and lience the 

 demonstration nf the occurrence of such variation would give an actual base to 

 Romanes's theory of physiological selection, and thus supply one explanation of that 

 hitherto unsolved ([uestion how the general infertility between species has come about. 



'• Tiie real dilHcnlty," says Darwin,]: " in our present subject is not, as it 

 appears to me, why domestic varieties have not become mutually infertile when 

 crossed, but why this has so generally occurred with natural varieties, as soon as 

 they have been permanently modified in a sufficient degree to take rank as a species. 

 We are far from precisely knowing the cause ; nor is this surprising, seeing how 

 profoundly ignorant we are in regard to the normal and abnormal action of the 

 reproductive system." 



We observe that the cause of infertility adduced by us would not be physio- 

 logical diversity, but mechanical impossibility of fertilisation. Mechanical selection 

 acting in a similar way as in the case of the organs of copulation in the direction of 

 adajiiing within each race sj)erm to egg by excluding extreme varietal specimens from 

 propagation, would raise a barrier between the diverse subspecies which would admit 

 an independent development ending in specific distinctness of mutually infertile forms. 

 Mechanical selection acts upon variation caused by other factors, and is, therefore, 

 like mechanical geographical isolation, a preservative, not a [iroductive factor. 



• Artb. bei Sehmett. 11. 189.5. p. 15. 



t StanUfuBS, llandb.f. deifmctterliiigxtammler IBM. 



J 0''igiii uf SpcciiK Gth ed. London, 1888. p. SC. 



