HEREDITY, VARIATION AND DEATH 23 



resultant cells, when they divided, showed fewer variations and 

 so were not so well constituted for the establishment of new 

 species." 



After having thus rounded off his hypothesis and maintained 

 it in the teeth of all objectors, Weismann proceeded in his book 

 The Germ Plasm, a Theory of Heredity to modify his theory of the 

 significance of amphimixis, a term which includes conjugation 

 and sexual union. It ceased suddenly to be the prime cause of 

 variation. " The cause must lie deeper than this, it must be due 

 to the direct action of external conditions on the biophors and 

 determinants " (in the reproduction cells, not in the soma). 1 



He had already recognised such direct external influences as a 

 minor cause of change. 2 But amphimixis had always ranked as 

 first in importance. Now it seems suddenly condemned to play 

 second fiddle ; it only combines variations, it does not originate 

 them. 3 



Why this sudden change ? It was due, no doubt, to a break- 

 down of the elaborate architecture of the germ-plasm the 

 biophors, determinants, ids, idants. A highly complex animal 

 must have more determinants than a less complex one, and 

 Weismann could not see how amphimixis alone could ever cause 

 a multiplication of them. 4 The whole difficulty has its origin in 

 the attempt to explain the inexplicable architecture of the germ- 

 plasm. Surely it would have been enough to say that it is some- 

 how there, whether we can picture it to ourselves or not. As 

 to variations, we know that crossing does produce them and, 

 since no two individuals are exactly alike, every union is of the 

 nature of a cross. This being so, I cannot help regarding 

 Weismann's earlier views upon the significance of amphimixis 

 as worthier of him than this later phase. 



I believe I have stated Weismann's views fairly, and, I hope, Estimate 

 clearly. They contain strange paradoxes that have excited Uf a 1^ 8 " 

 ridicule ; some of them will, I believe, always excite amuse- theory 

 inent and it must be owned that his shifts, refinements and 

 modifications, his endless manufacture and elaboration of hypo- 



1 See Germ Plasm, p. 415. 2 See Essay upon Heredity, vol. i. p. 420. 



3 Germ Plasm, p. 431. 4 Loc. cit. p. 415. 



