26 
Canary Breeding 
1880, vid. p. 11) are concerned; but there is certainly, unless care be exercised, 
a tendency for the bald occiput, not to become greater in area, but to be surrounded 
by aberrant feathers, which cause a dishevelled appearance at the back of the 
head — the "horned" crest in fact. 
It is possible that this condition might develop into a structure like the shell 
crest of the Trumpeter pigeon, whose rose crest exactly resembles the crest of 
a modern crested canary, covering eyes and beak in even, more or less, circular 
radiation. In this pigeon double-cresting is the rule : there are no crest-bred 
plainheads, and no b.aldness occurs (Plate V, fig. 6). 
Plumage-colour in the Crested Variety. We have already seen that my results 
with regard to " mottled x yellow," or more correctly, variegated x clear matings 
in my non-crested or muling strain did not agree with Davenport's conclusions. 
The same is true of plumage-colour in my crested strain, the results of which 
I now give, with the additional explanation that Davenport's " mottled x green " 
corresponds with my variegated x green, and his " mottled x mottled " with my 
variegated x variegated matings. 
Progeny 
green clear 
6 12 
0 17 
13 1 
The variegated x variegated proportion is roughly 51 : 3 : 6, which does not 
correspond, even approximately, to the expected 9:3:4. 
In neither of the other two sets of matings is the expected 50 °/„ of variegateds 
and of clears, of variegateds and of greens obtained. 
(ii) Inheritance of Yellow and Buff Qualities. The quality of colour being 
comparatively unimportant in the crested variety, I regret that sufficient attention 
was not given to noting the relative proportions of yellow and butf progeny. This 
point was carefully attended to in the notes of my non-crested or muling strain, 
where colour is of great importance. 
There is a wide-spread belief that yellow or jonque feather is inherited much 
more strongly from the male than from the female parent. The figures from my 
crested strain are not sufficiently full to be of great importance in definitely 
deciding this question. 
Proportion of Yellow to Buff Progeny in the Crested Variety. 
Matings Progeny 
Yellows Buffs Unknown 
18 yellow x buflf ? = 22 + 10 -f 22 
14 buff X yellow ? = 23-1-18 + 16 
Matings 
Variegated x variegated, ) _ variegated 
36 matings J ~ 103 
Variegated x clear, 
24 matings 
Variegated x green, 
24 ma.tincrs 
68 
58 
