246 DARWINISM TO-DAY. 



to in the next paragraph) which can soon result in complete 

 infertility, hence specific distinctness on the part of the 

 mutually infertile groups. 



Mutual infertility due to morphological variations has 

 been called "mechanical selection" by Karl Jordan, 19 and 



may rest on slight variations in germ-cells or 

 selection" 1 copulatory structures. Such morphological 



variations of the reproductive organs have been 

 believed to be shown for butterflies and spiders, while the 

 delicate tropismic responses of the active spermatozoids to 

 the attracting chemical substances in the egg-cells of echino- 

 derms, ccelenterates, molluscs, and fishes have been thought 

 to be conditions in which a slight chemical or physical varia- 

 tion might have a large influence in preventing or insuring 

 fertilisation. Indeed, when one examines comparatively the 

 curiously various complex accessory reproductive organs 

 (claspers, gonapophyses, intromittent organs, etc.) of almost 

 any group of insects, one's first thought is that this variety 

 must practically effectually exclude all possibility of hybri- 

 dising. But the interesting detailed comparative studies 

 by Snodgrass 1T of the accessory genitalia in various families 

 of Diptera make this confidence less certain. In the large 

 family Tipulidae, for example, he finds great complexity 

 and remarkable variety (among the different closely allied 

 species) in this complexity in the genitalia of the males, but 

 almost no variation at all in the corresponding (complement- 

 ary) parts of the females. "Throughout the entire family 

 the females present one type of structure, of which there is 

 but little modification, and certainly none to correspond with 

 the great variety of specific differences found in the genitalia 

 of the males." Now while the great variety of the copula- 

 tory parts would be extremely significant if shared by both 

 sexes so that only one kind of key could fit one kind of lock, 

 we have the inexplicable condition actually existing of the 

 keys being extremely various and complex, but the locks all 



