2o6 A STUDY IN HEREDITY 



that this last question, strictly speaking, is not a problem of 

 heredity at all. Nothing is supposed to be transmitted from 

 the parent to the child. On the contrary, an external 

 agency is supposed to affect both the parent organism 

 and its germs, the latter in a highly particular way. It is 

 obviously necessary, however, to discuss it, for if the germ were 

 so affected, then, since variations are transmissible, the conse- 

 quent variation would tend to be transmitted to future genera- 

 tions. It should be added that though it is not denied that 

 modifications of the parent, or external agencies, circulating in 

 the blood, may so alter the germs that the offspring arising 

 from them are also altered in this or that other way, yet no 

 instance of such an alteration has been traced, at any rate, in 

 the higher animals. The complexity of the high animal body, 

 the multitude of its characters, renders such tracing of cause and 

 effect impossible. 



The modern denial of the transmission of acquirements is 

 really founded on, or should be founded on, the cell theory. 

 Before its enunciation the different parts of the offspring were 

 supposed to be derived from similar parts of the parent. Man, 

 for example, was regarded as an individual. We know now that, 

 from the standpoint of heredity, he is a community — a com- 

 munity of specialised cells, of which one set, the germs, are 

 specialised for the production of similar cell-communities, just 

 as other sets are specialised for the production of motion or 

 of bile. We have therefore no more reason to suppose that 

 muscle or liver cells play parts in heredity than to suppose that 

 germ-cells play a part in movement or in the production of bile. 

 Formerly, when the transmission of acquirements was believed 

 without question, some remarkable hypotheses to account for 

 the supposed transmission were formulated, occasionally by very 

 distinguished men. Darwin's theory of pangenesis and Spencer's 

 theory of physiological units are examples. These speculations 

 were what are known as " working hypotheses " — hypotheses, 

 that is, in which the amount of theory (perhaps I should say 

 guess-work) is wholly disproportionate to the foundation of fact, 



