46 DARWINISM TO-DAY. 



species-forming without calling to its aid some effective 

 factor to control in its beginnings the variation essential 

 as the basis of the selective action. 



Pfeffer 17 and Wolff 18 have been particularly keen and 

 severe in their criticism of the selection theory on the basis 

 of this objection. And Morgan 19 in this country has also 

 made effective use of this weapon in his destructive con- 

 sideration of the Darwinian theories. 



There is an additional point about this difficulty of the 

 necessity for a certain regularity or reliability of variation 



in order to make a beginning basis for the action 



Necessity for 



coincident ap- of selection. It is this. Close scrutiny reveals 

 the necessity often of the occurrence of several 



make a certain coincident variations in order to make any one 

 characteristic positively advantageous. What 

 advantage in the way of increased speed is a slight added 

 length of leg without a simultaneously added strength of 

 mtisculation ; or an increase in size of antlers without a 

 simultaneous increase in strength of neck muscles to sup- 

 port and manipulate the heavier head? What faint prob- 

 ability of the occurrence coincidently of the necessary varia- 

 tions (if determined only by chance, that is, the law of prob- 

 ability) to produce a gradual perfecting of so complex 

 a structure as the vertebrate eye? Or, more, how incon- 

 ceivable the coincidences, if variation is purely fortuitous, 

 necessary to the simultaneous development of two exactly 

 similar eyes : two eyes so intimately associated physio- 

 logically that normal sight is a function of both these 

 separated organs working perfectly together. Is variation 

 to be assumed to be governed by some law of bilateral 

 symmetry ? But I have shown for many cases 20 that in such 

 perfectly and fundamentally bilaterally symmetrical animals 

 as insects neither the usual Darwinian fluctuating variation 

 nor the rarer discontinuous or sport variation is governed 

 at all by such a law. In fact the independence of the varia- 



