THE VAGABOND CRAMBUS : COCOON AND PUPATION. 137 



those in nature — the result, doubtless, of their long exclusion from the 

 light during their cocoonage. They gave no evidence of having been 

 parasitized, but were, to all appearance, in a healthy condition, dis- 

 playing, when disturbed in their retreats, a facility of locomotion fully 

 equal to that manifested during their active stage of six weeks before- 



The Cocoon. 



The general appearance of the cocoou is represented in Fig. 31. It 

 however varies considerably in size and form. The fig- 

 ure is taken from one of the largest size, while others are 

 but half so large. The average of thirty examples gives 

 a length of 0.9 inch, and a diameter at the broadest 

 part of 0.24 inch. The shape is subcylindrical, but vary- 

 ing from an almost uniform diameter, to an enlargement 

 of the lower portion to twice the diameter of the upper 

 part. The interior consists of a closely spun sheet of 

 silk, to which is connected the rather thin outer cover- 



co^J'onorcw ^"S of earth, mixed with the finer rootlets of the grass. 



BUS vDLGivAGEL- Nearly all the cocoons show curvature similar to that 



LPS, — natural , -, • , i r^ 



size. represented in the ngure. 



Delayed Pupation. 



Early in July (precise date not noted), upon opening several of the 

 cocoons for examination, the caterpillars were found within them still 

 unchanged, although they had at this time been in this inactive state 

 for a period of from one and a half to two months. The change to the 

 pupa state in the Lepidoptera usually takes place within a week after 

 the caterpillar has made its cocoon. The delayed pupation of this 

 species is an interesting feature in its life-history. A similar prolon- 

 gation of the larval stage is well known to those who are in the habit 

 of rearing Lepidoptera, to pertain to at least several species of the 

 PtilodontincB among the Bombycidce, which pass the winter in this 

 condition, pupating only a short time before the disclosure of the 

 imago in the spring.* It has also been observed in some species of the 

 NoctuidcB as in Xylina antetmata Walker (X. cinerea Kiley), in which 

 the larva lies unchanged in its earthern cocoon from June to Octo- 



*I have observed this frequently in (Edemasia coneinna (Sm.-Abb.) and in Calodasys 

 unicornis (Sm.-Abb.). Dr. H. S. Jewett informs me that Cerura borealis [? occidentalie] 

 makes cocoons late in summer and pupates the following spring ; and Ileterocampa marina 

 (Packard), after making its cocoon just below the surface of the earth or of bits of leaves, 

 twigs, etc., lies dormant through tlie winter, and pupates in the spring. 



All the Notodontians (together with Limacodes) remain a long time in their cocoons, or 

 in earth, before turning to pupae (Harris, Ent. Corr., p. 242). 



18 



