386 DARWINISM TO-DAY. 



bound to vary a little between any two individuals. And as 

 we know from the facts of experimental embryology that 

 development is, partly at least, epigenetic in character, i.e., 

 depends on and is influenced by external factors, this in- 

 evitable variation in influencing conditions is bound to pro- 

 duce variations in the individuals. . 



Is there, indeed, any need at all for assuming (i) any 

 mysterious "tendency" of the germ-plasm to vary? and (2) 

 that the individual (continuous) variation depends wholly on 

 germ-plasm structure? Why cannot the simple fluctuating 

 or Darwinian variations be chiefly the result of the inevitable 

 variation in the epigenetic factors, which, when not intruded 

 on by exceptional disturbances, w r ould themselves follow the 

 ''law of error" and hence produce "law of error" variabil- 

 ity? All normal swingings of the variation pendulum in 

 any part or character, between long and short, large and 

 small, round and angular, smooth and rough, etc., etc., 

 would result from the normal variation of the processes ; the 

 larger (extremes of range) variations being the fewer be- 

 cause the larger (extremes of range) variations in the 

 ontogenetic processes would be the fewer. Exceptionally 

 large epigenetic variations would produce exceptionally 

 large variations in the individual sports, mutations. 



Klebs, 13 as a result of his masterly experimental studies 

 on modifications of plant development, comes to the 

 conclusion that the only proved causes of variation are 

 extrinsic influences stimulating, working through, or com- 

 bined with, intrinsic conditions (not vitalistic, but physico- 

 chemical). Similarly, Tower, 12 from his protracted studies 

 on the variations in certain insects, concludes that all these 

 variations are caused by external stimuli working on the 

 germ-plasm. 



If variation is thus simply the wholly natural and un- 

 avoidable effect * of this inevitable non-identity of vital 

 process and environmental condition, why does not evolution 



