a 2 Bulletin Wisconsin Natural History Society. [Vol. 8. No. 1. 



ceptibly; perhaps it would always be within the probable error 

 for observations, either in regard to the temperature or in regard 

 to the time. Moreover, the variation is chanced. 



But it is indicated that the egg-stage shows less variation 

 than other stages. Thus, the larva's development may be greatly 

 retarded or hastened in the case of numerous individuals in a large 

 lot of similar origin because of varying supplies of food to the 

 individual, less or more favorable external environmental factors 

 liable to the individual, inherent variability, and so on, which 

 would tend to increase the limits of the period of pupation even 

 to several or more days. In such a case, there would be a per- 

 ceptible difference in the effective temperatures for the limits of 

 the period though it is doubtful, even here, if this difference in 

 temperature would really mean anything, the general tendency 

 not being affected. The pupae are probably not so variable, that is, 

 as far as actual development is concerned, but the final ecdysis 

 may be indefinitely prolonged for a number of individuals in a 

 lot of similar or simultaneous origin by the action of various 

 external factors other than temperature. In any case, it is obvious 

 that periods of time less than a minute are of no consequence 

 and it is indicated that periods of time even as long as 12 hours 

 may be ignored always, at least in most cases, without affecting 

 the realness of the results. How can we prove that this variation 

 in the duration of similar developmental periods is eliminated 

 as a factor by the determined probable error of observation ? Are 

 the two equal or so nearly so as to eliminate one of them? As 

 both are as liable to occur one way as another, in a large series of 

 observations, undoubtedly they could be ignored on the basis of 

 the law of chances? 



The following observations which bear directly upon some 

 of the stated questions are submitted for examination. They are 

 of the simplest kind, but I believe, unique in their way. They 

 are not intended to prove anything more than the fact of marked 

 variation in the duration of similar periods of embryonic 

 development. 



A female moth of Sanninoidea 1 exitiosa (Say) was allowed 

 to mate with a male, both recently emerged, and as soon as the 

 male left her she was confined in a glass vial of convenient size 



1 Or Sanvina. 



