328 



THE ENTOMOLOGIST. 



THE NUMBER OP LARVAL INSTARS OF CYANIRIS 

 (LYCMNA) SEMI ARGUS. 



By T. a. Chapman, M.D. 



I OUGHT, perhaps, to say a word in elucidation of Mr. Fro- 

 hawk's paper in the 'Entomologist' for November (p. 305). Mr. 

 Frohawk had a larger number of larvae surviving in the spring 

 than I had, so that the presumption is in favour of his being 

 able to make more satisfactory observations, as I could not 

 afford to take undue liberties with my examples. 



The discrepancy between us is not as to the autumnal his- 

 tory, where, Mr. Frohawk very truly says, observation is often 

 difficult — we agree that hybernation takes place in the third 

 stage. If I am in error, I missed a spring moult, which is not 

 impossible, owing to my scanty material, though it is not out of 

 place to say that this rendered me very unlikely to confound one 

 specimen with another — and I kept a very close watch on my 

 specimens. 



On the simple question whether Mr. Frohawk or I have 

 fallen into error, I think the probabilities are in favour of the 

 lapse having been on my part. Though variation in the number 

 of moults is by no means uncommon in many Lepidoptera, I 

 cannot remember any butterfly clearly proved so to vary, and, 

 certainly so far as my experience goes, no Lycaenid ; still, it is 

 of course possible. 



That I am perchance right has a rather weak support from 

 the fact that most Lycaenids have five stages, not a few only 

 four. I cannot at the moment remember one with six, but this 

 is no reason why L. semiargus should not have six. 



There is, however, one very important point that convinces 

 me that my observations are correct. It is one with which I 

 did not supply Mr. Tutt ; indeed, I fear I too often do not make 

 my notes for his use so complete as they ought to be. 



This point is one I have worked out in a number of species, 

 and always found it determine correctly any case of doubt like 

 the present as to number of moults. Whatever may be the case 

 in such instances as Arctia caja, where the number of moults 

 varies in different individuals to a great extent, apparently to 

 meet various exigencies arising in hybernation, I find, wherever 

 the number of moults is constant, that the size of the larval 

 head after each moult shows a constant percentage of increase 

 on that of the head in the preceding instar. 



Now, on looking up my preserved specimens of larvas and 

 larva- skins of L. semiargus, and measuring as closely as I can 

 the heads at each stage that I have preserved, I find the following 

 results. The measurements are from the base of the jaws to 

 the vertex, as I found that this was the easiest measurement to 



