S. A. Arendsen Hein 239 



We revert again to our two groups, the measures of which are as 

 follows : 



Males Females 



Total 1497 14-24 1503 14-00 1-017 



Second group 



Fourth thousand (June) ... 442 15-47 558 15-28 1-012 



Total of first and second group 1939 14-48 2001 14-37 1-007 



Though in each group of 1000 pupae the females are somewhat 

 smaller than the males, this difference is still so small (the average 

 sex-relation, the $ taken as unity, is only 1"007) that it may safely 

 be disregarded, when in questions of heredity as regards length the 

 average length of a brood or a strain has to be ascertained. 



The results of these measurements are also interesting in another 

 respect. When the average is calculated for each thousand (i.e. of 

 ^ and % together) of the 1st group, the figures of which are imme- 

 diately mutually comparable, the result is as follows : 



It appears therefore that the average length increases every month ; 

 the 2nd thousand measured in May is on an average + 1 mm. longer 

 than the 1st thousand of April ; the 3rd thousand of June again 

 + 1'5 mm. longer than the 2nd thousand of May. 



When each of these three variation-series is represented in a curve, 

 one observes that the curve, as a whole, shifts to the right every 

 month. This phenomenon repeated itself in the collection of the follow- 

 ing year in every culture of a series in which pupae of different lengths 

 were kept for some generations (measure-series). In many of these 

 single cultures it could be observed that the curve was gradually 

 getting two peaks. It produced the impression of two strains differing 

 geno typically (not only in their precocity and tardy maturity, but also in 

 their lengths) and separating themselves from the mixture. This is, 

 however, probably not the case, and it would seem to me that external 

 conditions alone are the principal, if not the only governing factor of 

 this phenomenon. 



