XII.] CONJUGATION AND SEXUAL REPRODUCTION 151 



As I have already stated, that part of my former view of the 

 significance of the polar divisions, which interprets the first as 

 an extrusion of a specific ovogenetic nucleoplasm, must be 

 abandoned. The facts of spermatogenesis, as we have recently 

 learnt them from the researches of O. Hertwig, have overthrown 

 these views, inasmuch as they prove that the nuclear idioplasm 

 of all polar bodies, as well as that which is retained in the egg, 

 must be germ-plasm. The polar divisions of the egg corre- 

 spond exactly with the two divisions of the sperm-mother-cell, 

 as will be seen at once by comparing Figs. I and II. By this 

 means, four sperm-cells arise from the sperm-mother-cell, and 

 of these four each contains half the number of idants character- 

 istic of the species (see Fig. I, F). By means of the two polar 

 divisions the egg-mother-cell similarly gives rise to the egg 

 (Fig. II. D E F, i), and the three polar cells (Fig. II. D E F, 

 2, 3, and 4), each of which contains the same number of idants, 

 viz. two. As it cannot be doubted that the idioplasm of the 

 four sperm-cells is germ-plasm, it must also follow that the 

 same is true of the three polar bodies as well as of the ovum. 



If then one polar body is always formed in regular partheno- 

 genetic eggs, it might seem that an explanation is to be found 

 by regarding it as a mere phyletic reminiscence. The question 

 arises whether such a view is a just one, and in order to gain 

 as clear a solution as is possible at the present time, I have 

 added this chapter on parthenogenesis to the essay. 



Spermatogenesis undoubtedly teaches us that the two ' re- 

 ducing divisions ' of the female germ-cell originally performed 

 the primary duty of producing four distinct germ-cells from 

 each mother-germ-cell. But spermatogenesis at the same time 

 shows us that a very remarkable reduction of the idants accom- 

 panies these two divisions. The normal number of idants 

 present in the mature spermatozoon is by this means reduced 

 to half that in the primitive sperm-cell, and the result is reached 

 by a most circuitous route, for the original number is first 

 increased to double, and then, by two successive divisions, finally 

 diminished to half. 



When, however, we recognize that in normal parthenogenesis 

 one of the two ' reducing divisions' is absent, while the other 

 persists, we can hardly regard the latter as the meaningie:;s 

 reminiscence of a process which was full of significance in an 



