242 



THE EVOLUTION THEORY 



they are of distinct sexes, and the females are so constituted that 

 they lay eggs for the first time just when the males of the current 

 year are appearing. Thus the first batch of eggs liberated by the 

 females are fertilized by the minute free-swimming 'primordial 

 males,' but after that the females themselves develop testes, and then 

 fertilize themselves ; the males die very soon after copulation, and only 

 appear the following year in a new generation. They are therefore 

 far from being mere historic reminiscences, vestiges of the early history 

 of the modern species, for they are the instruments of a regular 

 cross-fertilization of the species, and therefore of a constant mingling 

 of new ids in the germ-plasm. This is not the place to discuss the 

 marvellous life-history of these parasites in detail ; I can only say 





Abd 



^'■a 



Fig. 112 (repeated"). Develojiment of the parasitic Crustacean SaccuUna 

 carcini, after Delage. A, Nauplius stage. An, eye. I, II, III, the three pairs 

 of appendages. B, Cy2n-is stage. VI-XI, the swimming appendages. C, mature 

 animal (Saw), attaclied to its host, the shore-crab {Carcinus mcenas), with a 

 feltwork of fine root-^^rocesses enveloping the crab's viscera. S, stalk. Sacc, body 

 of the parasite, oe, aperture of the brood-cavity. Ahd, abdomen of the crab 

 with the anus («). 



that when we inquire into the whole story, and appreciate the 

 difficulties associated with the persistence of these ' primordial males,' 

 we can no longer doubt that crossing is an indispensable feature of 

 amphimixis — a feature which must at least occasionally occur if 

 amphimixis is to retain its significance. This is shown, it seems to 

 me, especially by these numerous instances of what we may call 

 compulsory retention of ephemeral males in hermaphrodite, self- 

 fertilizing animals ; it follows also from the theory, for with continued 

 self-fertilization all the ids in the germ-plasm of an individual would 

 tend to become identical, and the mingling of two germ-plasms which 

 contained identical ids would, at least according to the germ-plasm 

 theory, have no meaning at all. 



