THE TRANSMISSION OF ACQUIREMENTS 63 



variations by supposing, in effect, that each parental acquirement 

 tends to alter the germ-plasm in such a particular and remarkable 

 manner that the 'acquirement' is reproduced by the child, in 

 whole or part, as an * inborn ' trait. As a theory of racial change, 

 it supposes that progression is due mainly to use and to 

 such injuries and nutritive changes as result in increase of 

 size (or function), whereas retrogression is due principally to 

 disuse and to such injuries or nutritive changes as result in 

 decrease of size or function, for example from mutilations or 

 starvation. 



97. Now, whether acquirements be, or be not, transmitted, we 

 know that children often vary from their parents (differ innately at 

 birth from what their parents were at birth) quite apart from such 

 ' transmission.' Thus the child of normal parents may be born 

 with an extra digit, a character which is never acquired (i.e. de- 

 veloped) by the parent in response to the stimulus of injury or use. 

 The problem which we have to solve, then, is not whether the 

 Lamarckian theory of heredity correctly accounts for the occurrence 

 of all variations, but only whether it correctly accounts for the occur- 

 rence of any variations. Similarly many characters are present in 

 every species the evolution of which cannot conceivably have been 

 due to the ' transmitted ' effects of use, injury, or any other stimuli, 

 for example nearly all the characters of plants, and in animals, 

 such characters as brilliant colour markings, the external ears, the 

 reproductive organs, and feathers and hairs. It follows, therefore, 

 that while the Lamarckian doctrine may afford a partial explana- 

 tion of the occurrence of variations and of evolution, it cannot 

 afford a complete explanation. 



98. The unicellular organism " is a part and usually a half of 

 the parent." Its structures, therefore, are derived from similar 

 structures in the parent; that is, there is real inheritance. The 

 multicellular organism, on the other hand, is derived from a single 

 cell, which, after dwelling as a parasite in the parent cell-community, 

 leaves its home to become the ancestor of a new community. The 

 parts of this new cell-community, therefore, are not derived from 

 similar parts of the parent ; that is, there is no real inheritance of 

 the characters of the first community by the second. The two 

 resemble one another for a totally different reason because their 

 development has been directed by similar germ-plasms, the 

 hereditary tendencies of which have been awakened by similar 

 stimuli. It is evident, therefore, that the problem of the alleged 

 transmission of acquirements is not the same in both cases. Indeed, 



