1897] RELATION OF MODIFICATIONS TO HEREDITY 249 



He supposes first that under appropriate circumstances a small 

 amount of the original substance may be capable of governing the 

 course of the future organism, just as the mathematician may con- 

 struct from a small portion of a curve its whole extent. And, 

 secondly, — 



" If in a parental organism, by long habit or constant practice, 

 something grows to be second nature, so as to permeate, be it ever 

 so feebly, its germinal cells, and if the germinal cells commence an 

 independent life, they will aggrandise and grow till they form a 

 new being, but their single parts still remain the substance of the 

 parental being." 



The objections to this theory lie, I think, in the fact that direct 

 communication with the reproductive organs becomes with increas- 

 ing specialisation increasingly difficult, and therefore heredity and 

 reproduction would cease when a certain point in the specialisation 

 was reached. 



Nevertheless, some kind of provisional theory such as the fol- 

 lowing would, I believe, explain better than any other theory most 

 of the phenomena of inheritance : — 



(1) That in every cell there are certain reproductive units 



which are necessary to the development of that par- 

 ticular cell. 



(2) That these reproductive units having a very complicated 



structure (being composed of specialised protoplasm), are 

 capable of modification when acted on by external forces. 



(3) That the various impressions made upon the cell would of 



necessity be made upon these units also, and that this 

 impression will be proportional to the length of time and 

 intensity of the impression made. 



(4) That as specialisation of tissue occurs, each reproductive 



unit will tend to reproduce its own history, past im- 

 pressions becoming with each successive addition more 

 and more blurred. 



(5) That the stronger and more numerous the past impressions, 



the more difficult will it become for present impressions 

 to affect them, hence progressively diminished power of 

 use-inheritance. 



(6) That the reproductive units have the power of self multi- 



plication when in the latent condition, and that this 

 multiplication will be difficult in proportion to their 

 specialisation and complexity. Hence latent germs 

 would tend to be carried on from one generation to 

 another, and increase the general stability of the 

 organism. 



