206 TRANSMISSION OF ACQUIRED CHARACTERS 



metabolism, but it does not follow that it can appreciably 

 alter the architecture of the germ-plasm. Spencer's as- 

 sumption that the change in the constitutional units of 

 the body must affect the constitutional units in the germ- 

 cells remains an assumption. 

 (3) "That the change in the offspring must, other things equal, 

 be in the same direction as the change in the parent, appears 

 implied by the fact that the change propagated throughout 

 the parental system is a change towards a new state of 

 equilibrium — a change tending to bring the actions of all 

 organs, reproductive included, into harmony with these 

 new actions." 

 Comment. — It seems to us to pass the wit of man to conceive 

 how or why an improved equilibrium in, let us say, the use 

 of the hand should involve any corresponding or represen- 

 tative change of equilibrium in the germinal material. 

 The drawback to abstract biology based on first principles 

 is that it enables its devotees to develop arguments which 

 seem plausible until they are reduced to the concrete. 



§ 9. Particular Evidences in support of the Affirmative Answer 



The question is whether modification-inheritance does or does 

 not occur, and we must no longer postpone our consideration 

 of the concrete evidence used to support the affirmative position. 

 Our reason for not placing this section in the foreground of the 

 chapter is mainly that a multitude of misunderstandings have 

 had to be cleared away before the so-called direct evidence could 

 be profitably considered. When one naturalist, Dr. W. Haacke, 

 declares that instances of modification-inheritance are as plentiful 

 as sand on the shore, and another, Prof. E. Ray Lankester, 

 declares that the Lamarckian position has its only remaining 

 defence, and that no secure one, in Brown-Sequard's experiments, 

 we have obvious justification for our preliminary discussion. 



The instances adduced as evidence of modification-inheritance 

 might be classified according to the errors involved, but we have 

 arranged them rather in reference to the general nature of the 

 modifications discussed, whether environmental or functional, 



