OTHER THEORIES OF SPECIES-FORMING. 207 



and thus a definite longitudinal and transverse arrangement 

 of fibres in the walls of the vessel result. Without doubt, 

 holds Plate, much advance is won in this way, but this 

 specialisation of structure is not a result of intra-struggle 

 but rests on the elementary attribute of trophic irritability. 

 Not the best-qualified but the best-situated fibres have van- 

 quished the others by robbing them of food and thus finally 

 destroying them. In the second place, "many inner struc- 

 tures belong to the great category of passive adaptations; 

 they function only through their presence and cannot thus 

 be further developed by use or disuse, that is, by functional 

 stimuli, but only by natural selection. Here belong, for 

 example, the stratification of the lens in the human eye, the 

 apodemes (inner projections of the chitinised cuticula) 

 which protect the ventral nerve-cord of the crabs, the chitin 

 hooks which hold together the fore and hind wings of many 

 insects, and the similar structures which bind together the 

 secondary branches of the feather vanes of birds. These 

 inner adaptations cannot have resulted through the influence 

 of light or of nervous function or flight. There is but one 

 explanation possible; namely, that natural selection has 

 seized on and developed fortuitously appearing germinal 

 variations. But if natural selection can produce such inner 

 adaptations why can it not then produce all the others?" 



Fifth, Plate points out that Roux's theory is based on the 



inheritance of those special body characters which are 



acquired through the battle of the parts more 



Roux's theory rightly, Plate holds, through functional adapta- 



is to accept the t j on so t } iat to acce pt the theory, one has to 



inheritance of J ' 



acqnired char- declare, to that degree, a belief in the inherit- 

 acters, ance Q f aC q U { re( j characters. Thus from the 



start, the neo-Darwinians cannot accept the theory. 



After all what is this theory of Roux's but a refinement, a 

 special case, of the broader and more general long-known 

 Lamarckian theory of the modifying and formative influ- 



