XIV] 



Alternation of Generations 



257 



be taken as pointing at all events to the existence of special 

 features in them. In most of these, however, there is 

 nothing that can be construed as indicative of some 

 wholly different system of genetic physiology. There is 

 nevertheless one phenomenon, alternation of generations, 

 which stands out as at present incapable of factorial repre- 

 sentation. ' This is best exemplified by the case of seasonal 

 dimorphism so frequently met with in insects, especially 

 Lepidoptera. For example many species exist in two types 

 of which one emerges after the winter as a spring form. 

 This breeds and lays its eggs. The larvae feed up rapidly 

 and emerge in summer as the second or summer form. In 

 their turn individuals of the summer type give rise to the 

 larvae which are destined to become the spring type again. 

 A similar sequence of distinct types is a common feature of 

 many tropical species in which the wet and dry season 

 forms correspond to the winter and summer forms of 

 temperate countries. The cycle may be complicated in a 

 variety of ways, by the intercalation of additional genera- 

 tions and otherwise. Of these more complex cycles the 

 Aphidae and Coccidae afford many striking illustrations. 

 In none of these examples can we conceive that the distinc- 

 tions between the recurring types are due to addition or 

 removal of factors. The only other suggestion which can 

 be made is that these distinctions are ultimately referable 

 to the effects of external conditions. As yet very little if 

 any evidence can be adduced in support of this suggestion. 

 The experiments of Standfuss, Fischer, Merrifield, and 

 others have shown that the form which the individual 

 assumes can within rather narrow limits be determined by 

 altering the temperature to which it is exposed, but in 

 order to obtain a satisfactory elucidation of the phenomena 

 I am here considering this evidence is insufficient. In the 

 known examples the experimental influences were brought 

 to bear on the individuals themselves. It is not impos- 

 sible that something might be accomplished by changing 

 the conditions to which the parents are exposed. It is, for 

 example, a familiar experience of lepidopterists that when a 

 one-brooded type is bred in captivity a second brood is 

 very commonly produced in the autumn, and it is natural 

 to refer this variation to some influence of the abnormal 



B. H. 



17 



