Miscellanea 
403 
istic a. For such, the offspring of Crested Norwich x Crested Manchester Coppy* would be 
'pure-bred.'" As the characteristic a must mean the crest and coppy (which however are 
very dissimilar) and as the rule in Crests and Coppies is to mate a crested bird with a crest- 
bred plainhead, a coppy with a coppy-bred plainhead, it is difficult to comprehend the wort of 
fancier to which reference is made. 
I must state, however, that our respective interpretations of the term "pure-bred" have 
nothing to do with the point at issue, which was my criticism of a result of Dr Davenport's — on 
the ground that pure-bred birds had 7wt been used— purporting to overturn one of Russ', 
in which he (Russ) expressly stipulated that pure-bred birds were necessary. (See Davenport's 
quotation, p. 15.) 
Concerning the "crude speculation" of my evolution theory, I sincerely hope that it will 
form the subject of serious discussion by competent authorities. I have many more facts 
in its support than could be produced in a paper which had to include such an amount of 
spade work. 
With one statement in the Rejoinder, I am in complete agreement with Dr Davenport, viz. 
"I would defy anybody to classify my canaries now." He has succeeded in proving by his 
so-called scientific method, not only that a crestless bii'd is a perfect crest, but that it has 
a more perfect crest than three-fourths of the existing prize crested birds. By some ingenious 
allelomorphic scheme he might as easily prove that all bipeds were quadrupeds. From his false 
assumptions both in the matter of crest, and plumage-colour, it is quite impossible for him to 
arrive at any scientifically correct, or practically useful conclusion. 
III. Inheritance in Canaries : A Study in Mendelism. 
By DAVID HERON, iM.A. 
In 1908 there appeared a paper t on Inheritance in Canaries, by Chas. B. Davenport, 
Director of the Station for Experimental Evolution, Carnegie Institute of Washington, in which 
the author discusses the inheritance of form of crest and plumage-colour in canaries and con- 
cludes that those characters follow Mendelian rules. 
These results are accepted without criticism by Bateson|, and although Galloway^ has 
ofiered some criticism, it seems desirable that further attention should be called to the methods 
used in this paper. 
It will not however be necessary to consider Davenport's discussion of the inheritance of 
plumage-colour; his treatment of the inheritance of form of crest on pages 8 to 13 and the 
general list of matings in Table E, p. 24, so far as it deals with crest-form, will be found quite 
suSicient for the present purpose, i.e. to measure the scientific weight of this paper. 
Before considering his results a strong protest must be made against the very defective 
condition in which they have been presented. From the material we can only arrive at a lower 
limit to the number of blunders in those few pages (in all 114 lines of text and four Tables) by 
noting those cases in which the tables are self- contradictory, e.g., when -a bird is said to be 
crested in one table and non-crested in another. 
* A redundant term ! 
t Carnegie Institute of Washington Publication, No. 9-5. Papers of the Station for Experimental 
Evolution, No. 10. 
X MendeVs Principles of Heredity, pp. 37 and 43. 
§ Biometrika, Vol. vii. p. 1 et seq. 
