FORMATION OF THE HYMENOPTEROUS IMAGO 661 



"The position which the organs assume in the chrysalis is not due to chance, 

 everything is determined in advance, and the microscope shows us that the 

 structure of the hypodermis is specially modified in all the parts which remain 

 external. It is a fact well known to those who rear Lepidoptera that if this 

 normal arrangement is disturbed there are few chances that the perfect insect 

 will survive. A leg lifted up, or an antenna displaced, leaves a surface illy pro- 

 tected against external influences. Almost always this accident causes a drying 

 of the chrysalis. 



" Several interesting experiments may be cited as bearing on this subject. 

 If during transformation the chitinous mask of the head is separated from the 

 integument beneath, it is arrested half-way in its development, and the antennae 

 and tongue are not fully extended. When the case or skin of the caterpillar is 

 drawn, not from before backward, but in the opposite direction, all the append- 

 ages of the thorax are placed perpendicularly to the body. Dewitz and Kiinckel 

 d' Herculais, in stating that the skin of the caterpillar splits open along its whole 

 length, show that they were ignorant of the mechanism ; for it is precisely be- 

 cause the chitinous larval skin splits open only in front that it preserves suffi- 

 cient adherence to the organs beneath to draw them after it in the direction of 

 the abdomen. 



"To only read modern authors, one would suppose that the mechanism of 

 pupation had remained hitherto unknown. In reality, it did not escape the 

 notice of Swammerdam or of Re'aunmr, both of whom have described it with 

 care. The first attached too much importance to the flow of blood, the action 

 of which would be rather to push the organs out than to extend them over the 

 surface of the thorax ; the second insists on the movements of the insect. This 

 factor, very admissible in caterpillars, ' whose under side is situated on a 

 horizontal plane' (iii, 9 e Me"moire, p. 395), cannot be invoked for those which 

 suspend themselves by the tail, as in the genus Vanessa." (Gonin.) 



b. The Hymenoptera 



In the Hymenoptera, Ratzeburg was the first to figure and describe 

 the numerous intermediate stages between the larva and pupa, his 

 subjects being the ants, Cynips, and Cryptus, which pass through 

 five stadia before assuming the final pupal shape. 



In the bees, as we have observed in the larvae of Bombus (Proc. 

 Bost. Soc. Nat. Hist., 1866), after hardening a series in alcohol of 

 young in different stages of development, it will be found difficult to 

 draw the line between the different stages since they shade insensibly 

 into each other, those represented in Fig. 616 being selected stages. 

 The head of the incipient sernipupa distends the prothoracic seg- 

 ment of the larva whose head is pushed forward and the thoracic 

 segments are much elongated, while the appendages and wings are 

 well developed, and have assumed the shape of those of the pupa. 

 Development both in the head and thorax begins in the most im- 

 portant central parts, and proceeds outwards to the periphery. 

 During this period the "median segment," or 1st abdominal, has 

 begun to pass forward and to form a part of the thorax. 



