82 Cincinnati Society of JSfatnral History. 



appendages of the larva are homologous with thase of the imago." 

 In Lithocolletis, I have seen the head and its appendages of the 

 forthcoming larva retracted from the corresponding parts of the 

 skin that is about to be cast. I have seen a larva cease to feed, and 

 when no trace of the contained larva of the next stage could be 

 detected in it, then the skin of the contained larva was gradually 

 loosened from the old skin, and the antennae, tropin, and legs of the 

 forthcoming larva were gradually retracted from their corresponding 

 parts of the old skin; and at the fifth moult the antennae of the pupa 

 (and imago), as long as the body of the insect, and composed of a 

 multitude of joints, either of which is as large as the entire antennae 

 of the larva, are nevertheless withdrawn from the antennae of the 

 larva. Part b}^ part, and organ by organ, I have under the microscope, 

 in these small larvae, seen the external parts and organs of each stage 

 withdrawn from the corresponding parts and organs of the preceding 

 stage. Each organ, when it is so withdrawn, differs at first but little 

 in form or size from the part from which it is withdrawn, and within 

 which it was formed, but the}^ grow rapidly, and in a few hours I have 

 seen the long and many-jointed antennte of the pupa grow from the 

 mere rudiment, as it were, which was withdrawn from the larval an- 

 tennae. On one occasion, I saw a rather surprising demonstration of 

 the fact, that ring by ring each segment of the larva corresponds with 

 and is formed within the corresponding segment of the larva of the 

 preceding stage. A larva had ceased to feed, and retired to the middle 

 of its mine as if to moult. Wishing to stain the neural ventral ganglia, 

 as I had frequently done with other insects, I extracted the larva from 

 the mine, and cut off its head, as previous observations had shown that 

 the staining fluid would not act through the larval integument. I let 

 the body remain a short time in the staining fluid, and then observed 

 it in glycerine under the microscope. The staining process was a fail- 

 ure ; and as no trace of the new larva could be detected contained 

 within the old, I concluded that the larva was not preparing for its 

 moult, but had ceased to feed for some other reason; when happening 

 to press on the cover glass, to my astonishment the contained or new 

 larva " shot out" from the old larval skin, leaving the perfect old skin, 

 containing an almost perfect tracheal system, looking in fact almost as 

 if the larva had suddenly duplicated itself. 



To return, however, to the life histories ofLithocolIetis, as illustrated 

 b}' a species from each of the three groups. 



I have traced the life histories of many species of the fiat group, and 

 they are essentially the same. Taking Lithocolletis guttifinitella, 



