592 Scientific Proceedings, Roi/al Dublin Society. 



X' X 

 more can it make black. As yet we have no means of naming the colour 



corresponding to the other new constitution. 

 No. 28 mated with tortoise sliould give 



No. 28 mated with F 1 black should give 

 X' X xX 



y y Yy 



black. agouti. 



Thus, from the first of these matings there should result blacks, yellows, 

 and tortoises in the ratio 2:1:1, while from tlie latter there should result an 

 unknown coloui', agoutis, blacks, yellows, and tortoises in the ratio 

 1:1:4:1 : 1. These expectations are in accordance with the experimental 

 results. It is true there was only one mating of No. 28 and tortoise, from 

 which the progeny were 3 black, 1 yellow, and 2 tortoise — small in numbers, 

 though of the kinds expected — but there were 21 matings with F\ black, and 

 the progeny were agouti-black 34, agouti 36, black 113, yellow 28, and 

 tortoise 30 ; that is, these colours in the ratio 1:1:4:1: 1 approximately. 



The other colours being all in accordance with expectation, it is now a 

 fair inference that the unknown colour is agouti-black, and that its 



X' X 



constitution is „ • Let us proceed upon this hypothesis. To prove it we 



shall have to prove that the constitutions depending upon it are accompanied 



in the experiments by the colours they ought to bear; and, since only 



constitutions containing X' are involved, we may for convenience set down 



all those into which this factor can enter together with their corresponding 



X' X 

 colours. We shall have to assume meantime that y y is also agouti-black. 



This assumption is justified on the ground that yacts similarly, whether 

 homozygous or heterozygous. 



The constitutions in question are 



