126 



MR. L. DOXCASTER AXD REV. G. H, RAYNOR ON [Feb. 20, 



insects used. Only three of these original pairings gave larvse 

 which reached maturity. Their i-esults are given in Table I. 



Table I. 



No. of 



Exp. $ PAEBNT. $ PARENT. OffSPEING. 



03.3... 9 sordiata X <J sordiata gave 22 sordiata ^ , 25 sordiata $ . 



03.6... ? primaria X S sordiata „ 27 sordiata $ , 7 sordiata $ , 1 prtmaria $ . 



03.4... ^ prunaria X S ;prunaria „ 27 prunaria ^ ,1 sordiata ^ ,40 prunaria^ . 



These figures immediately suggested that the banded var. 

 sordiata was a simple Mendelian dominant over the unhanded 

 prunaria type. The next year's work confirmed this conclusion ; 

 and it must be supposed that the single prunaria among the 

 ofispring of 03.6 and the single sordiata in 03.4 were due to 

 accident. The larvae, when they first hatch, are exceedingly 

 minute, and when the food is changed it is difiicult to be certain 

 that no larva clings to the hands and gets tiunsferred to the 

 wrong box. 



An inspection of the moths from 03.3 showed that about half 

 of them have the bi'own bands on the wings, with plain orange or 

 yellow centres, but that the other half, in addition to the banding, 

 have the orange centres speckled as in the typical priinaria. 

 Sometimes the speckling is very faint, so that it is hard to give 

 exact numbers of each type, but approximately among the 

 offspring of 03.3 the numbers are 24 speckled and 23 plain. 

 In 03.6 all were speckled. This suggests that the speckled 

 character of prunaria is dominant over the plain of sordiata 

 at the same time that the banding of the latter dominates over 

 its absence in the former; in this way a heterozygote can be 

 distinguished from a pure sordiata. 



In 1904, 36 pairings were made, of which 24 yielded images 

 in 1905. Their results are given in Tables II.-YII. 



Table II. — Prunaria $ X prunaria S 



