VT DEFENCES OF INSECTS 231 



ants the several worker-castes are generally quite 

 distinct, in some cases the castes are connected 

 together by finely graduated varieties. Probably the 

 variations simply originate in the many slight differ- 

 ences which appear in the offspring from the same 

 parents. No one will suppose that all the individuals 

 of the same species are cast in the same actual mould. 

 These individual differences are often inherited, and 

 thus afford materials for natural selection to act 

 on and accumulate. It doubtless appears a very re- 

 markable fact that a female butterfly (for instance) 

 should be able to produce at the same time three 

 distinct female forms and a male. But these cases 

 " are only exaggerations of the common fact that the 

 female produces offspring of two sexes which some- 

 times differ from each other in a wonderful manner." 1 

 I must pass over other phases of di- and trimor- 

 phism,* and polymorphism, a mere extension probably 

 of the same principle. Enough has been said to show 

 that this subject is a very wide one, and in many 

 cases extremely perplexing and very obscure. 



A series of phenomena remain to be considered 

 which cannot be better described than as by far the 

 highest and most important class of di- or polymor- 

 phism. As in dimorphism the species have two or 

 more sets of appearances, but these seem, as it were, 

 under the control of each insect, which is susceptible 

 to the influence of the surroundings, and can at 

 pleasure assume the appropriate form, adjusted to 

 correspond with its own peculiar environment. In 

 dimorphism proper it is the species that tends to 

 1 Darwin. Origin of Species. 



