LIFE HISTORY. 33 



Camponotinae, but not in the Myrmicinae and Dolichoderinae. 

 The fact that the pupae of the most primitive group always possess 

 cocoons, suggests that this habit is an inheritance from solitary 

 ancestors. Wheeler has recorded the presence of the naked pupa of 

 a Dolichoderine ant in Baltic amber, which shows that the absence 

 of a cocoon in this sub-family is not a recent development. In a 

 few genera of the Camponotinae there is no cocoon, but even with 

 those in which one is usually present, naked pupae may often 

 be found, even in nests which also contain pupae in cocoons. It is 

 very difficult to understand this, or to see what, under these cir- 

 cumstances, can be the object of a cocoon at all. 



Forel says he has found these naked pupae only in the late 

 autumn, a cocoon always being present with the summer larvae. 

 In England, however, I have found naked pupae of Donisthorpea 

 and Formica species in their nests in June, July, August and 

 September. 



When a larva is fully grown it becomes straight and rigid, the 

 legs, wings, etc., which are visible beneath its skin, are packed 

 closely to the body, and it is now a semipupa. The larval skin 

 then splits down the back, and is pushed off posteriorly. 



Those larvae which do not remain naked after pupation, natur- 

 ally spin their cocoon before the semipupal stage is reached. When 

 they are ready to spin, the workers cover them over with little bits 

 of earth, sand, etc., or bury them in shallow holes, to provide them 

 with starting-points on which to commence spinning. 



The workers dig up the cocoons when completed, render them 

 clean and smooth by removing all adhering substances, and pile 

 them in a heap. 



In observation nests where no earth or sand is present, I have 

 noticed that the workers will make use of bits of plaster, cotton- 

 wool, or anything handy, with which to cover a larva when ready 

 to spin. 



The colour and texture of the cocoon varies in different species, 

 being white, yellow, buff, or dark brown, and finely or coarsely 

 sculptured, etc. 



A black spot is usually to be seen at the posterior end of the 

 cocoon, which consists of the larval excreta, and the larval skin is 

 also present ; but with naked pupae these are removed during 

 pupation by the workers. 



The workers help the young imago to emerge the latter before 

 it has acquired the mature coloration has been called a " callow " 

 by biting a hole in the cocoon, and drawing out the ant. Some- 

 times the workers remove the pupae from their cocoons before they 

 have reached the imago stage. 



It was formerly supposed that an ant was unable to get out of 

 its cocoon without assistance, and if left would perish, but this 

 is not always the case. 



