THE EVOLUTION PROCESS 289 



others, in fact the most important of floral 

 variations, the big lifts distinctive for the 

 evolution of orders, are thus seen no longer 

 as indefinite, and hence dependent on external 

 selection for their guidance; but, on the con- 

 trary, as parallel and definite, since determined 

 through the continued checking of the vegeta- 

 tive process by the reproductive, and thus 

 pressed along parallel and definite grooves of 

 progressive change. But if this be so, the 

 importance we have been taught by Darwin 

 to assign to natural selection becomes greatly 

 changed from selecting and accumulating 

 supposed indefinite variations, to that mainly 

 of retarding definite ones, after their maximum 

 utility has been independently reached ! 



The same simple conception unlocks innu- 

 merable problems of floral morphology, large 

 and small alike, from the inevitable develop- 

 ment of angiosperm from gymnosperm (by 

 the continuous subordination in vegetative 

 development of the reproductive carpellary 

 leaf) to the origin of many of the refined 

 minor " adaptations " of the dominant school. 

 Adaptation to insects, to wind also, thus falls 

 from a primary to at most a very second- 

 ary place as a factor in the evolution of 

 flowers; for the characteristics usually ascribed 

 to the selective action of wind and insects 



