METAMORPHOSES OP BRAZILIAN LEPIDOPTERA. 27 



The caterpillar appareutly feeds upon a tree cnlled by 

 the Brazilians, Herva de lagarto, for I found great numbers 

 of cocoons surrounding a tree of this species. But, unfor- 

 tunately, I did not find a single caterpillar that had not 

 begun to spin, and therefore am not certain of the food- 

 plant. 



( Full-fed, Theresopolis, (about) 16th Nov., 1882. 

 Pupa. ^ Imago, on my return to Liverpool, Feb. to 

 i April, 1883. 



The cocoons were spun on twigs of the low shrubs about 

 the tree on which, I imagine, the caterpillars feed. The 

 silk was spun a good way down the twig, to strengthen it in 

 case of accident. 



Imago. 



a, $ Full-fed, 16th Nov., 1882. 



Imago, Feb., 1883. 

 h, ? Full-fed, 16th Nov., 1882. 



Imago, Feb., 1883. 



The moths began to appear in January, but the greater 

 part of them emerged in February, and five out of about 

 twenty pupae delayed their exit until the middle of April, one 

 not having yet made its appearance in the perfect state. 

 Many of the moths were deformed, owing probably to their 

 being shaken and disturbed by their journey while pupating. 

 Several of the moths had got turned end for end in the 

 cocoons, and were unable to make their exit. This was also, 

 most likely, owing to the fact of the cocoons being shaken 

 about while the insect was still in the larval condition, and, 

 perhaps, in an inverted position when pupation took place ; 

 for, of course, the pupa itself could not possibly reverse its 

 position in the cocoon. 



