420 



PSYCHE. 



[November — December tSgo. 



ground. The duration of the larval days. 



stages was as follows. — ist stage, 5 Food-plants: Hamavielis 



days; 2nd stage, 6 days; 3rd stage, 6 ginica., Vacciniuvi stamineicm. 

 days, 4th stage, 7 days; 5th stage, 7 Larvae from Ulster Co., N. \. 



THE NUMBER OF MOLTS OF LEPIDOPTEROUS LARVAE. 



BY HARRISON G. DYAR, RHINEBECK, N. V. 



Two articles have already appeared in 

 PsycHE on this subject* and it is evi- 

 dent from a perusal of them that con- 

 siderable confusion exists as to the 

 number of molts of certain species. 

 In this article I propose to establish a 

 criterion by which different observa- 

 tions may \)C compared and errors de- 

 tected, for it is no difficult thing to 

 overlook a molt or even to think one 

 has occurred when it has not. 



It will be seen by an examination of 

 the following figures that the widths of 

 the head of a larva in its successive 

 stages follow a regular geometrical pro- 

 gression, and if, in examining the meas- 

 urements of heads taken in following 

 out a life history, any deviation from 

 the calculated progression is shown, 

 it is evidence that an error has been 

 committed or that the larva has be- 

 haved in an abnormal manner ; but the 

 latter case can readily be distinguished 

 from the former if a moderate degree of 

 care has been exercised in taking the 

 measurements. Hence, if two sets of 

 observations show a different number 

 of stages for the same insect but each 



* W. H. Kdwards, Psyche, v. 3, p. 159. 

 A. K. Dimmock, Psyche, v. 5, p. 28. 



follows its own progression, we may 

 conclude that this variation is actual ; 

 but if either set shows a lack of regular 

 progression that one we must regard 

 with suspicion. Corroborative obsei'va- 

 tions of the kind indicated are to be de- 

 sired and until we have them we can 

 not speak with confidence about the 

 number of molts of any species. I 

 recommend that all who hei'eafter de- 

 scribe larval stages give the width of 

 the head for each stage. I have selected 

 the Iiead as the part not subject to 

 growth during the stage, and its xvidth 

 as the most convenient measurement to 

 take. 



In the following I give first the calcu- 

 lated widths of head under each species, 

 with the ratio, followed by those that 

 were actually found. All measurements 

 aie in millimeters and may be con- 

 sidered accurate to within .i mm. Cer- 

 tain irregularities are commented upon 

 in the notes to which the small num- 

 bers refer. 



FOUR STAGES. 



Callosamia promethea Drury. calc. r.i 

 1-6 2.3 3.3, r .70 

 found i.o 1.7 3.3 3.3 ^ 



