128 MR. L. D02n^CASTER AXD REV. G. H. RAYNOR ON [Feb. 20, 



Table VI. — Plain sordiata x prunaria. 



From these tables I think it is sufficiently clear that the 

 banding of the sordiata is dominant over its absence in prunaria, 

 but that the speclvling of p^una.ria is at the same time dominant 

 over the plain orange of the pui-e sordiata, giving a heterozygote 

 which is both banded and speckled (PI, YIII. fig. 2). The plain 

 sordiata, however, may have some specks along the wing-rays, 

 so that an exact determination of the numbers of " plain " and 

 " speckled " is not possible. The numbers of these two classes 

 in the tables are therefore approximate. 



The numbers are not sufficiently large to show whether the 

 different types occur in the proportions demanded by Mendel's 

 La,w, with the exception of those in Table III. Here there are 

 148 prunaria to 110 sordiata, where equality is expected; but 

 the mortality is so great among the young larvae, and also during 

 hibernation, that a very small differential mortality will account 

 for this. 



The work was paitly undertaken to find out whether there 

 was any tendency for a coi'relation of either of the types with one 

 or other of the sexes, but no evidence whatever of this has 

 appeared. 



It is impoitant to notice that no intermediates occurred ; in 

 fact the darkest prunaria bred were fi-om two prtmaria parents, 

 and the lightest sordiata from sordiata parents. 



