20C AMPHIMIXIS OR ESSENTIAL MEANING OF [XII. 



maintained that the ontogeny of Metazoa and, so far as I 

 am aware, of Metaphj^ta also, primaril}^ depends on the 

 necessity for sexual reproduction, or, better still, on the exis- 

 tence of iimcellular germs. An ontogeny must then follow ; 

 for the collective hereditary tendencies of an animal being 

 concentrated in a single cell, they must therefore, during 

 development, pass through a series of stages very similar 

 to those of their phyletic history. But, besides the germs 

 destined for sexual reproduction, there are other unicellular 

 germs, spores, &c. ; and hence it is clear that the unicellular 

 condition brings other advantages than those which amphi- 

 mixis confers ; but these unicellular agamic germs never 

 exhibit any approach to the extent of range witnessed in 

 sexual cells, and the origin and universal existence of unicel- 

 lular germs are therefore to be sought in the latter. 



I have already shown that the sexual cells, upon their first 

 appearance, in some simple cell-colony such as Pandorina, 

 would be compelled to undergo a nuclear ' reducing division,' 

 after a relatively small number of sexually reproduced genera- 

 tions ; because otherwise a continued doubling of the nuclear 

 units must have occurred in consequence of the periodicall}' re- 

 peated union of the nuclear substance of different individuals. 

 This ' reducing division,' which is now securely proved for both 

 male and female sexual cells in Metazoa, has, however, 

 another meaning. 



I proceed from the assumption that nature aims at the widest 

 possible range for amphimixis. How could this be obtained 

 more effectually than by rendering the imicellular germs incapable 

 of developing alone ? 



The male germ-cells, being specially adapted for seeking and 

 entering the ovum, are, as a rule, so ill provided with nutriment 

 that their unaided development into an individual would be 

 impossible ; but with the ovum it is otherwise, and accordingly 

 the ' reducing division ' removes half the germ-plasm, and the 

 power of developing is withdrawn. 



What happens in the unicellular organisms ? Here also our 

 theory demands that periodic amphimixis should be provided b}' 

 nature. For the attainment of this object it was indispensable 

 that, as in Metazoa and Metaphyta, the organisms should, at 

 certain periods, arrange themselves in pairs, and that their 



