LEPIDOPTERA. 



161 



which the pupa has just quitted the caterpillar's skin. It is then still 

 soft and tender. Its body is moistened with a liquid, which, drying 

 rapidly, becomes opaque, coloured, and of a membranous consistency. 

 The result is that the parts which did not cohere 

 in the least when the chrysalis made its first 

 appearance, are fastened together, so that al- 

 though they could at first be seen, through a 

 layer of transparent fluid, they are now hidden 

 under a sort of veil or cloak. It is necessary 

 to seize then the moment of the birth of the 

 chrysalis, to observe it accurately. 



On examining the pupa before the liquid 

 which pervades these parts has had time to dry, 

 it resembles the perfect insect. All the exterior 

 parts which belong to the imago can be dis- 

 tinguished. One recognises the head, which is 

 then resting on the thorax; the two eyes and 

 the antennae (Fig. 128), which are brought for- 

 ward like two ribbons ; the wings also brought 

 over the thorax, but these are separated arti- 

 ficially in the drawing we have given after 

 Reaumur ; * and lastly in the space left between 

 the wings, the six legs, and the body of the 

 insect. 



To sum up : the pupa, when it approaches 

 the period for being hatched is only a swaddled 

 butterfly. Directly it has strength enough to rid 

 itself of its wrapping, the insect frees itself from 

 its fetters ; it flies away, brilliant and free, and 

 its many-coloured wings glitter in the sun. 



The duration of the pupa state is variable, according to the species 

 and the temperature. Reaumur placed in a hot-house, in the month 

 of January, some pupae which, in the ordinary course of things, 

 would not have been hatched till the month of May, and a fortnight 

 afterwards the images had appeared. On the other hand, he shut 

 up some pupae in an ice-house during the whole of a summer, and 

 thus retarded their being hatched by a whole year. The influence 

 of the temperature on the period of emerging, and, consequently, the 

 influence of the seasons on the length of this period, are completely 

 brought to light by these experiments.! 



* Tome i., p. 382, planche 26, Fig. 7. 



t They hardly seem from later experiments to be so fully explained. It is a 



L 



Fig. 127. 



Chrysalis of the large 



Tortoise-shell Butterfly 



(Vanessa polychloroi), 



magnified, seen from the 



lower side. 



