175 



the extrication of the tail beg-ins, to be immediately followed 

 by the lifting up of the body, so that the booklets of the tail 

 can grasp the silk. The body is bent sharply back while the 

 tail is pointed outward and freed, the abdominal segments are 

 stretched to their utmost and then contracted, one telescoping 

 into the next successively, so that the skin is drawn into one 

 joint after another. At the same time the chrysalis is seen to 

 rise, the tail is bent back, and by a curving mov^ement strikes 

 the silk. The chrysalis has been described as climbing by the 

 aid of the skin alone, as it is pinched between the segments. 

 But on lifting the flap of skin entirely clear of the struggling 

 segments, the same struggle continues, the chrysalis rises and 

 the tail strikes the silk surely as before. The looker-on sees 

 plainly a black ligament holding the chrysalis. This is 8.75 

 mm. long, flat, roundly excised at the broader end, the two 

 prongs thus made being attached to the anterior black knobs on 

 the last segment of the chrysalis, the other end to the extreme 

 part of the caterpillar skin. I watched the pupation of only 

 two caterpillars of D. arcMppus^ — one from beginning of the 

 process to the end. But in the other I lifted the flap of skin 

 till the lio;ament was exposed. In doing this the whole thino; 

 unhooked from the silk, and as it lay in my hand, I pulled back 

 the skin and was able to examine the ligament. I also lifted 

 the chrysalis by the skin, and the ligament did not part. It did 

 so afterwards only by a strenuous effort of the chrysalis, and 

 then remained distended. 



After the booklets of the tail have caught the silk, the 

 chrysalis whirls one way and then the other, the last segment 

 twistino; and screwing in order to fasten the booklets more se- 

 curely. 



I had dropped one of the chrysalids of Gr. inter rogationis in- 

 to glycerine, at the crisis of pupation, and this was sent to 

 an experienced microscopist for examination. He verified the 

 existence of the membrane, and suggested that it is the rectum, 

 or the external cover of it, drawn out and adhering to the 

 anal ridges of the crysalis But if it be tubular it does not 

 appear why it attaches at but two points only, namely, to the 

 two knobs in D. archipjms, and the hooked ends of the 



