OTHER THEORIES OF SPECIES-FORMING 277 



In this chapter the writer has purposely avoided entering into the 

 more elaborate intricacies of the Weismannian theories of develop- 

 ment and heredity. The theories have been so generally discredited 

 and play so small a part in modem biological thought that it seems 

 useless to burden the reader's mind with needless complexities. 



Certain other phases of Weismann's work, especially his ideas of 

 the germ plasm, its separateness and its continuity, are more appro- 

 priately studied in connection with genetics than at the present time. 



ROUX'S THEORY OF INTRASELECTION OR THE BATTLE OF THE PARTS 



In point of time this theory antedates Weismann's theories, since 

 it was proposed in i88i. In some respects it is a more acceptable 

 theory than germinal selection, but in others quite unacceptable. 

 The theory is designed primarily to explain the origin of the "fine and 

 delicate inner adaptations" of organisms, which do not come in con- 

 tact with the external environment and therefore could not be directly 

 affected by it. The idea is that there is a sort of struggle among the 

 tissues for a chance to develop in the direction of functional perfection. 

 Certain contacts, stresses, and pressures of part on part aid or hinder 

 the development of parts. Thus, where muscular pressure on bone 

 is greatest or weight borne by bone is greatest there will the most bony 

 tissue be laid down in the form of lamellae. The result is that any 

 given bone improves its structure by resistance to strain and pressure, 

 which is a case of improvement with use. We may then inquire how 

 such a change in the individual could affect the evolution of the race. 

 The only reply involves the adoption of a distinctly Lamarckian con- 

 cept, and this at present is quite unacceptable. 



COINCIDENT SELECTION OR ORGANIC SELECTION 



This theory has been masked under various guises. In addition 

 to the two titles given above, it has appeared under the names "onto- 

 genetic selection" and "orthoplasy." The main idea, according to 

 Herbert, is that "the individually acquired characters, though not 

 transmitted to the offspring, serve to tide the successive generations 

 over the critical period until germinal (inborn) variations of the 

 same kind appear which are inheritable. Ontogenetic (individually 

 acquired) adaptations and natural selection work together towards the 

 same end. 



"This hypothesis would help to account for two related difficult 

 points in the theory of natural selection. Firstly, it would explain 

 the possibUity of the slow accumulation of germinal variations in theii 



