AND PLANTS 99 



once in every two throws ; so that if I know that both dice 

 in the first throw gave more than three points, I may 

 expect that both of them will do so in a dependent second 

 throw once in two trials, instead of once in four. 



Now those of you who know Weismann's theory of the 

 mechanism, if I may use that word for want of a better, 

 by which characters are transmitted from parent to 

 offspring, will recognize that the relation I have estab" 

 lished between my first and my second result with these 

 dice is roughly analogous to the relation he indicates 

 between parent and offspring. 



Weismann supposes that the characters, which a parent 

 can transmit, are represented in the young germ-cell, 

 whether male or female, by a certain number of elements, 

 roughly comparable with the dice I have used. Before a 

 germ-cell becomes ripe, so that it can fuse with another 

 germ-cell of opposite sex, half the elements it contains 

 are thrown out of it by what Weismann considers to be 

 a purely random selection ; so that if the constitution of 

 a young germ-cell be represented by the result of a first 

 throw of my two dice, the random removal of half its 

 determinant elements will be represented by the removal 

 of the white die. When a ripe germ-cell fuses with 

 another of opposite sex, a number of elements, equal to 

 the number it has lost in the process of maturation, is 

 restored to it, and the characters of the embryo are deter- 

 mined by these two sets of elements, one set chosen at 

 random from a young germ-cell of each parent. In 

 making my second result with the dice, I have taken one 

 element at random from the two which determined the 

 first result, this process being roughly analogous to the 

 loss of half the determinants during maturation, and I 

 have restored the original number by adding to the one 

 element which remains from the first result a second 



H 2 



