146 On the Inheritance of Colour in Pigeons 



They may be roughly divided into (a) those with very few white 

 feathers and (h) those with several white feathers. The first class, 

 consisting of three birds, liad a few white feathers on the head, rump, vent 

 and thighs. One of them had in addition one white flight feather and 

 one white tertiary. The second class (five birds) had white in the same 

 situations but of greater extent, the white on the head often being 

 continued on to the neck. They also had several white flight feathers 

 and tertiaries, and one or both bastard wings were white. The occurrence 

 of these two classes may possibly be connected with the heterozygosis 

 of the (f parent respecting rump character. 



Exp. 53. Blue chequer, several w.f $ 64 x Typical G. livia ^ 44. 



The $ was raised in the last experiment, and was mated with its 

 own father. Seven young resulted. Of these only one was a chequered 

 bird with a very few white feathers on the nimp and one on tlie head. 

 The remaining six were unchequered blues, of which one had a blue 

 rump and showed no white feathers. Another bird had only a very 

 few white feathers on tiie rump and none elsewhere, there being much 

 less white than is found in a typical 6'. lima. Tlie four other birds had 

 more white than in G. livia. There were white feathers on the head 

 rump, vent and thighs in all four, and in three of these also on the neck, 

 abdomen, flight feathers, tertiaries and bastard wings. These birds 

 presented a mottled appearance. The Mendelian expectation for this 

 mating is chequered and non-chequered birds in equal numbers. The 

 result obtained, 1 : 6, however falls far short of this. The only other 

 mating of heterozygous chequer and homozygous non-chequer (Exp. 59) 

 gave 8 chequered : 4 non-chequered. In this case the discrepancy is in 

 the other direction, and the total of the two matings 9 : 10 gives 

 practically equality. From the behaviour of the chequered birds in 

 the subsequent matings there is no reason to believe that there is any 

 disturbing or selective factor at work. It is to be regretted that no 

 matings were made to test the gametic composition of the blue with 

 very few white feathers. The number of birds of this type produced 

 during the series of experiments was small, and the sexes of those 

 living at the same time did not admit of their being mated together. 

 It is probable however that they would have given blues with no white, 

 but we cannot predict whether or not whites could have been produced 

 from them. On the other hand the blue birds with much white would 

 certainly have given whites, as the mating of two similar birds in 

 Exp. 55 shows. As regards the colour of the rump, both parents of 



