518 THE EVOLUTION OF MIND 



The Germinal Origin of Impi^ovements in Instinctive Be- 

 haviour. We must linger over the difficulty, which many 

 biologists feel acutely, of trying to account for improvements 

 in instinctive behaviour by variations in the germ-cell. 



When an organ, such as the proboscis of an elephant, has 

 shown in successive ages a gradual increase and differentia- 

 tion, as the skulls of fossil Proboscidea seem to indicate, 

 the non-Lamarckian evolutionist supposes that this is due 

 to the selection of variants in the direction of elongation, 

 these variants being the expressions of appropriate changes 

 in germinal organisation. The change in germinal organisa- 

 tion, say a strengthening of certain primary constituents, 

 operates during the active process of proboscis-development, 

 or of proboscis-growth, for it need not begin to exert its 

 influence until long after the foundations have been laid. 

 Thus a long-billed bird need not show much or anything in 

 the way of a long bill until after it is hatched. The general 

 idea is that an improvement of structure comes about as 

 the expression of a germinal variation which asserts itself 

 during the activity of development or growth. It is not 

 necessary to think of it as asserting itself only once, for the 

 highly differentiated structure, such as a snail's horn or a 

 newt's lens, may be regrown if it be lost. The germinal 

 variation includes a residual capacity (localised at the base 

 of the horn or in the tissue near the lens) for reproducing 

 or regenerating what has been lost. The general idea, we re- 

 peat, is that a cumulative germinal variation, implying a per- 

 fecting of some part of the germinal organisation, expresses 

 itself in the course of generations in a cumulative improve- 

 ment of a certain routine of developmental or growth activity. 



If this be admitted as conceivable, then it is not a great 

 step to pass to the improvement of instinctive activities as 



