114 



THE ENTOMOLOGIST S RECORD. 



heard at times that the melanic form of hidentata does not " breed 

 true," the explanation is that in this species blackness is dominant, 

 not recessive as in A. ;/rosfiiilariata. 



A recessive character is easy to get pure and " breed true," and a 

 dominant character is difficult. Our domestic sheep forms a good 

 example ; — whiteness is recessive, and blackness dominant. 



To put it in a practical way for the benefit of those entomologists 

 (probably still numerous) who have not had opportunity or inclination 

 to study Mendelism : — 



All Type hidentata (whether their parents were both type, one type 

 and one black, or both black) are exactly alike as regards their powers 

 of transmitting colour to their offspring. 



Melanic specimens all look exactly alike, but really consist of two 

 sorts, differing in powers of transmitting colour : — 



A. Pure melanic. (homozygous). 



B. Impure melanic. (heterozygous). 

 Therefore, 



1. Type X type always produces offspring all types. 



2. Type X melanic produces either 



i. offspring all melanic. 

 or ii. offspring 50% melanic, 50% type. 



3. Melanic x melanic produces either 



i. offspring all melanic. 

 ii. offspring 75% melanic, 25% type. 

 In 2 i. the melanic parent must have been A. 



The melanic specimens of families 2 i. and 2 ii. are all B. 



In 3 i. although all look alike, there are two possibilities : — 



If both parents were A, the offspring are all A. 



If one parent was A, and one B, then the offspring are 50% A 

 and 50% B (although they all look alike). 



In 3 ii. the parents must have been both B. Of the melanic 

 specimens (75%), i are A, and | B. 



Thus to get a pure black strain is very difficult, because A cannot 

 be distinguished by the eye from B, but attention to the above points 

 will help. 



A common experience is as follows : — A black specimen is taken 

 and deposits ova. Larvfe carefully reared, and the following year 

 imagines appear, either 50% or all black. In either case, two blacks 

 are chosen ; paired and larvaB reared. Next year imagines are 76% 

 black and 25% type. 



(The type are really pure although parents black, so time spent in 

 breeding from them is wasted). 



Two black are chosen, and here is the difficulty. There are twice 

 as many B's as A's, so the chances of picking two A's is less than the 

 chance of picking A and B, or two B's, and thus often to the chagrin 

 of the breeder aiming at a " pure," strain, types again appear. 



Still, as I have shown, definite rules govern the heredity of melanism, 

 and mongrelisation is only apparent not real. 



[I wish to acknowledge with thanks the courtesy of Messrs. 

 Mosely of Birmingham, G. T. Porritt, Parkin, and A. Home, who 

 have given me valuable advice and helped in other ways.] 



