1 82 AMPHIMIXIS OR ESSENTIAL MEANING OF [XII. 



minute invisible masses of germ-plasm, which arise from the 

 ^gg and are afterwards carried on, mingled with the contents of 

 certain somatic cells. In these cases the origin of such masses 

 in the ^gg can only be conjectured, but in conjugation observa- 

 tion shows that a part of the idioplasm is, as a matter of fact, set 

 apart in the form of micronuclei for the use of the next genera- 

 tion. The nuclear substance of the micronucleus alone is 

 undying, and continues the vital processes without limit, while 

 the macronucleus behaves, in this respect, in an entirely dif- 

 ferent manner. 



In the Metazoa the whole cellular structure of the body — the 

 soma— is worn out by the processes of life, and suffers natural 

 death : in just the same way the Infusorian macronucleus can- 

 not continue its functions for unlimited generations, but must be 

 renewed from time to time ; and indeed, as we have already 

 seen, it is formed anew from the combination nucleus which 

 originates in the amphimixis of the two reproductive nuclei. 

 During the formation of the new macronucleus the old one is 

 destro3^ed and disappears. These are processes which have no 

 analogy with fertilization : I shall return to their deeper signifi- 

 cance later on. 



A further difference between fertiHzation and conjugation lies 

 in the fact that the reproductive nuclei of Infusoria arise from the 

 thrice-repeated nuclear division of the micronucleus, while the 

 nuclei of the ^gg- and sperm-cells of Metazoa are derived from 

 the twice-repeated division of the mother-cell. 



Meaning of the Phenomena. 



It ma}^ appear decidedly premature to attempt an explana- 

 tion of the above-mentioned differences and resemblances 

 between the two forms of amphimixis ; but I am willing 

 to undertake this responsibihty, if only to give a fixed 

 direction to further investigation. If I abandon all the theo- 

 retical conceptions of fertilization and heredity developed in 

 my earlier writings, I do not believe that we need, on this 

 account, give up all views upon the processes of conjugation 

 as they are known to-day, but rather that future research 

 will be more profitable if we endeavour to test some settled 

 theory, instead of making observations with no object in view. 



The preparatory divisions of the micronucleus have been 



