70 MR. R. SIARLES-BROWNE OX [Feb. 18, 



SO in all cases *. From the matiugs of blues with white feathers, as 

 in Exps. 11, 24, 25, and 26, whites were produced, and a definite 

 proportion of homozygous blues was to be expected. With one 

 exception, however, all the blues produced from these matings 

 showed some white feathers. We are led to conclude that some 

 of these ai-e pi-obably homozygous, although the assertion cannot 

 definitely be made without testing a large quantity of such bii'ds, 

 of which space did not permit. On the other hand, the matings 

 of blues with some white feathei-s to whites in Series B has not 

 revealed a homozygous bird other than that in F. 2 already 

 mentioned. 



One distinctly abnormal I'esult occurred in Series B from the 

 mating of blues with white feathers to whites. The propoi'tion of 

 whites produced was here much higher than the expected equality. 

 This result, which is discussed later, evidently points to the 

 existence of a definite complication. 



The red birds obtained in F. 2 are only bi'iefly mentioned 

 in the present paper. A further sei'ies of experiments, dealing 

 Avith them, is now in progress, a full report of which will be 

 published in another communication. 



The majority of reds, produced in the F. 2 generation and from 

 subsequent matings. showed a bluish tail with a very distinct bar, 

 the under parts also having a. bluish tinge. Certain specimens 

 have, however, been pi'oduced in which the bar is absent, and the 

 amount of blue much reduced. It is possible, therefore, that tvro 

 kinds of reds may eventually be demonstrated. In addition, red 

 has shown itself to be recessive to both black and blue, but 

 dominant to white f. 



The exti'acted whites, which are shown to breed true, need no 

 comment, with the exception of five proiluced in the direct line 

 (Exps. 4, 5, and 8) in which some coloured feathers were present. 

 These are shown in Exp. 7 to produce whites with and Avithout 

 coloured feathei'S in equal numbei'S. It is pi'obable that this has 



* As Crampe and Doncaster have shown in rats, and Hurst in the case of rabbits, 

 the presence of some white in otherwise self-coloured types, may be an indication of 

 heterozj*gosis in respect of a pattern-factor, and the same possibilitj' is to be reniem- 

 bei'ed here. 



t The results and tig'ures so far obtained from the further experiments are briefly 

 as follows : — 



(a) F. 2 red X F. 2 red a,-ave 7 red. 1 white with few red feathers, 2 white. 



(fi) F.2 red X F.2 red " „ 3 red, 3 white. 



(e) F. 2 red X F. 3 red „ 3 red. 1 white with few red feathers, 1 white. 



Id) F.2 red X white „ 2 black, 8 white. 



(e) Black from (d) X white „ 12 white, blue, 2 black, 1 red. 



{/) F. 2 red X black „ 6 red, 4 black. 



(q) Red from ( f) X red from (/) ., 7 red. 



(A) F.2 red X 'blue „ 6 black, 5 blue. 



(i) Black from (//) X black from (Ji) (2 pairs) gave 10 black, 7 red, 3 blue. 



(j) Blue from (/;) X blue from (/;) gave 9 blue, 2 red. 



In the above no mention is made of white feathers occurring on coloured birds ; 

 they occurred, however, on a large number, and their distribution corresponded to 

 that in similar birds described in the present paper. It will be noticed that the 

 number of white birds produced was above the expected proportion. Exp. (/') was. 

 a brother and sister matins'. 



