XII.] CONJUGATION AND SEXUAL REPRODUCTION. 215 



eggs, and is therefore under the direct influence of amphimixis. 

 It is well known that some Medusae are budded off from a 

 polype-stock, and constitute the sexual generation of the latter, 

 marking the end of a series of asexual generations ; while 

 other Medusae invariably arise from fertilized ova, and always 

 produce eggs requiring fertilization, or, in other words, adapted 

 for amphimixis. 



The degree of organisation is, in yet another way, associated 

 with the alternation of asexual with sexual generations, and thus 

 with the periodicity of amphimixis. This new relationship 

 between organisation and the recurrence of amphimixis, depends 

 upon the fact that the asexual methods of reproduction by fission 

 or budding are not possible in the highest and most complex 

 Metazoa. They are only found in the lower groups of Metazoa, 

 — the Coelenterates, Worms, and Echinoderms ; disappearing 

 in the Arthropods, Molluscs, and Vertebrates. 



In these latter, we might well suppose that every act of increase 

 would be connected with amphimixis ; for, — since the structural 

 complexity of the animals in question has rendered fission 

 and budding impracticable and has therefore compelled a re- 

 version to the unicellular germ and the occurrence of a detailed 

 ontogeny in every generation, — it might seem probable that 

 nature w^ould not lose the advantage of connecting amphimixis 

 with such a method of reproduction. We might therefore expect 

 to find no exception to the occurrence of sexual reproduction in 

 these groups. In this anticipation we should be deceived, inas- 

 much as it only appears in the great majority of cases. In the 

 minority, amphimixis is very far from universal, in spite of a 

 development from unicellular germs which would so easily have 

 permitted it : furthermore, in this minority it was formerly 

 connected with reproduction, and has been abandoned in 

 different degrees. These cases of development from partheno- 

 genetic eggs are, above all others, fitted to prove the importance 

 of the principle of utility. The transformation of female sexual 

 cells, at first directly adapted for amphimixis, into germs no 

 longer requiring fertilization, is an artifice by which nature has 

 contrived to avoid amphimixis when a high degree of structural 

 complexity has prevented reproduction by fission and budding. 



It may be remarked here that this suggestion supplies the 

 answer to a difficultv which I was, for a long time, unable to 



