INHERITANCE OF ACQUIRED CHARACTERS 185 



stimulus in the first instance, and this modification, whatever its 

 nature (and it might of course be quite unrecognizable by us), 

 might affect the germ cells in which case the germ cells could not 

 be entirely shut off from the influence of the soma. 



The chief obstacle in the way of our belief in the inheritance of 

 acquired characters lies, as we have already seen, in the difficulty 

 of imagining any mechanism adequate to bring about the con- 

 version of somatogenic into blastogenic modifications. If, as 

 Weismann insists, the germ cells are really incapable of being 

 influenced by the body, the difficulty does indeed seem insur- 

 mountable ; but there seems to be no valid reason why we should 

 follow Weismann in this respect. Nor does it seem necessary 

 that we should adopt any theory which postulates the migration 

 of material particles, such as the gemmules of Darwin's pan- 

 genesis, in order to get over the difficulty. 



Herbert Spencer long ago indicated the direction in which the 

 solution of this problem must be sought. " It is," he says, " an 

 unquestionable deduction from the persistence of force, that in 

 every individual organism each new incident force must work its 

 equivalent of change ; and that where it is a constant or recurrent 

 force, the limit of the change it works must be an adaptation of 

 structure such as opposes to the new outer force an equal inner 

 force. The only thing open to question is, whether such re-adjust- 

 ment is inheritable ; and further consideration will, I think, show, 

 that to say it is not inheritable is indirectly to say that force does 

 not persist. If all parts of an organism have their functions 

 co-ordinated into a moving equilibrium, such that every part 

 perpetually influences all other parts, and cannot be changed 

 without initiating changes in all other parts if the limit of 

 change is the establishment of a complete harmony among the 

 movements, molecular and other, of all parts ; then among other 

 parts that are modified, molecularly or otherwise, must be those 

 which cast off the germs of new organisms. The molecules of 

 their produced germs must tend ever to conform the motions of 

 their components, and therefore the arrangements of their com- 

 ponents, to the molecular forces of the organism as a whole ; and 

 if this aggregate of molecular forces is modified in its distribution 

 by a local change of structure, the molecules of the germs must 

 be gradually changed in the motions and arrangements of their 

 components, until they are re-adjusted to the aggregate of mole- 

 cular forces. For to hold that the moving equilibrium of an 



