i895. THE ROLE OF SEX. "247 



tends to revert to the original type, and on this account it is the 

 custom to perpetuate suitable varieties by means of grafts and 

 cuttings. In the case of the ciliated infusoria passing through several 

 generations of asexual reproduction, we find progressive changes 

 terminating in debihty and death. In reproduction without conjuga- 

 tion an individual is cut off completely from the rest of the species, 

 and gives rise to an isolated Hne of descendants. If the individual is 

 a variety its progeny will almost to a certainty be varieties too, 

 subject always to fresh and changing variation. In its line of 

 descendants every fresh individual starts again on a fresh line, and so 

 it goes on like a tree, the branches of which never touch. Sooner or 

 later every living individual will be found to have deviated in all 

 possible directions from the ancestral form. 



Now, in the case of reproduction with conjugation, the facts 

 differ widely from the above. The progeny of two parents tends to 

 what we may term, after Galton,the mid-parent ' — nay, it tends even 

 more to approach the mean average type of the species than the mid- 

 parent itself. In plants, as Darwin has shown, much the same thing 

 occurs, the progeny being either intermediate between their parents in 

 all points, resembling one parent most in one point, and another in 

 another point, or the influence of one parent may predominate over the 

 influence of the other. To take two concrete cases, the branch of a 

 tree which differs from the rest by having variegated leaves, may be 

 propagated again and again by cuttings, but a six-fingered man, 

 although he may pass on this variety to some of his more immediate 

 progeny, will not permanently impress the race. In the latter 

 case, each crossing will enable the influence of more typical members 

 of the species to assert itself, and so by degrees the variation 

 will, as it were, be overpowered. Conjugation previous to repro- 

 duction tends therefore to bring the offspring of varieties more 

 towards the mid-type of the species ; it works in distinct opposition 

 to variation, and the latter can never establish itself unless it is 

 useful and external circumstances are peculiarly favourable, pre- 

 serving only those of the offspring which continue to show in some 

 degree the parental variation. 



Now, it appears to me a matter for some surprise that, with the 

 facts so clearly stated by Darwin, Galton, and others, and accepted 

 by Weismann himself, biologists have not accepted these as 

 establishing very fully the role or function of conjugation. It has this 

 yole : it brings the offspring of the act of conjugation nearer the mid- 

 type than even the mid-parent itself, while the cutting from the 

 variegated branch deviates as markedly from the mid-type as the 

 branch itself. I venture, therefore, to bring forward this explanation 

 of the role of sex, not as a theory, but as a fact, resting on a solid 

 basis of experimental evidence hitherto ignored in this relationship. 



I imagine that Weismann, and the many biologists who follow 



1 Natural Inheritance. By Francis Galton, 1S89. 



