XII.] CONJUGATION AND SEXUAL REPRODUCTION 199 



reptans: this is not an unending series, but we do not know 

 of any reproductive cycle which, after forty agamic generations, 

 returns to a sexual one. So far as the argument is concerned, 

 it does not signify at all whether such cases are rare, as Maupas 

 thinks, or common : even their entire failure would afford no 

 proof of the theory of rejuvenescence. For the theory of 

 minghng,— if I may so designate my hypothesis, — is founded 

 on the species-preserving influence of amphimixis, and 

 leads us to expect that, wherever it is possible, nature will 

 always introduce amphimixis into the reproductive history of 

 a species and will render its employment obligatory. We 

 should have no ground for wonder if purely agamic repro- 

 duction had no real existence. The vitalizing influence of 

 amphimixis would not be proved even if this were the case. 



On the other hand, I think a single example of continuous 

 agamic reproduction proves that amphimixis is not absolutely 

 necessary for the unlimited duration of life. 



But if amphimixis is not absolutely necessary, the rarity of 

 purely parthenogenetic reproduction shows that it must have 

 a wide-spread and deep significance. Its benefits are not to be 

 sought in the single individual ; for organisms can arise by 

 agamic methods, without thereby suffering any loss of vital 

 energy : amphimixis must rather be advantageous for the 

 maintenance and modification of species. As soon as we 

 admit that amphimixis confers some such benefits, it is clear 

 that the latter must be augmented as the method appears 

 more frequently in the course of generations ; hence we are led 

 to enquire, how nature can best have undertaken to give this amphi- 

 mixis the widest possible range in the organic world. 



The following is an attempt to supply an answer to the 

 question. The increase by means of budding and fission 

 would be retained in multicellular plants and animals, on 

 account of its great advantages, but it would only endure 

 for a shorter or longer series of generations. Moreover, 

 the expected advent of amphimixis would only take place 

 when the collective hereditary tendencies of the individual 

 are concentrated in the nucleus of a single cell ; hence the 

 mechanism of reproduction must have been associated with 

 unicellular germs, and amphimixis became bound up with 

 reproduction. I cannot remember that it has ever been 



