end of the second volume of ' ' Variations of Animals 

 and Plants under Domestication ' ' (1868) , is very 

 different. He supposes that each cell of the animal 

 or plant gives off minute particles , called gemmules r 

 which are capable of multiplying by self -division. 

 These gemmules circulate through the whole system 

 and are collected together in the reproductive cells. 

 In this way they are transmitted by the parent to the 

 offspring, and each gemmule is then capable of de- 

 veloping into a cell similar to that from which it was- 

 detached. But it can only develop when it find& 

 itself in a position in the new organism similar to 

 the position of its parent cell in the old 

 organisation ; and until it finds this position 

 it may be passed on from generation to 

 generation in an undeveloped state. But it will 

 always be capable of development whenever it finds 

 the necessary conditions. 



This hypothesis was conceived, chiefly, to explain 

 not only the phenomena of heredity, but also that 

 of atavism, that is the reappearance of characters 

 which have remained latent through one or more 

 generations. And it was only proposed as a pro- 

 visional working hypothesis. 



In 1870 Professor Ewald Hering gave his cele- 

 brated address on Memory. 21 According to the 

 author, memory is the faculty of reproducing ideas ; 

 but sometimes these memories come involuntarily. 

 To account for this we must suppose that some 



21 English translation published in the "Religion of 

 Science Library." Open Court Publishing Company, 



Chicago. 



