147 



stages to pale greenish. The larvae of the flat group, yellowish 

 or whitish in their earlier stages, usually become more or less 

 tinged with a smoky or fuscous hue, sometimes almost blue- 

 black, but at their seventh moult thev ao-ain become white. 



' I/O 



The larva of L. ornatella, at first whitish, gradually becomes 

 tinged -with green, which deepens in its last and penultimate 

 stages to a peculiar bluish green not found in any other larva of 

 the genus. This hue persists in the earlier part of the pupa 

 state. 



The pupa has the head pointed in front, with a serrated edge 

 running back on each side, which is no doubt useful in cutting 

 through the cocoon, while it pushes itself out by the contortions 

 of its body, aided by the microscopic bristles which arm the 

 upper surface of the abdominal segments. The pupa does not 

 entirely free itself from the larval skin for nearly two days after 

 the cocoon is made, and the pupa state lasts about eight days 

 after that, at midsummer, and the moth emerges through a rup- 

 ture of the pupa skin across the back of the head and down the 

 sides of the wing-cases. 



In this latitude these insects may be found in all their stages 

 from May to the fall of the leaves. I have plucked from the 

 same plant of Mhus toxicodendron, at the same time, leaves 

 containing larvae of X. guttifinitella in all their stages, as well 

 as pupte, and pupa-skins fi'om which the imago had emerged. 

 It is therefore manifestly impossible to say how many broods 

 may succeed each other in a season. That depends on the tem- 

 perature and length of the season. In Kentucky I have found 

 Gracilaria rohiniella mining locust leaves from the first of July 

 to the fall of the leaves in October ; while I have found it doing 

 the same thing in New Orleans in December, when the leaves 

 were falling. Larvae which are in their last stage, having finished 

 feeding, may winter in that stage, and this is the condition in 

 which they are usually found in the winter, the temperature 

 preventing them from passing into the pupa state till the return 

 of warm weather. Larvae which have not finished feeding 

 when the leaves fall must of course perish, but no good reason 

 can be given why pupse might not winter. Nevertheless, I 

 have never met with a Lithocolletis pupa in the winter or spring. 



