ted fowls the nostrils
offer a remarkable peculiarity in being raised with a crescentic
outline. The primary wing-feathers are short in Cochins; in a male,
which must have been more than twice as heavy as _G. bankiva_, these
feathers were in both birds of the same length. I have counted, with
Mr. Tegetmeier's aid, the primary wing-feathers in thirteen cocks and
hens of various breeds; in four of them, namely in two Hamburghs, a
Cochin, and Game Bantam, there were 10, instead of the normal number 9;
but in counting these feathers I have followed the practice of
fanciers, and have _not_ included the first minute primary feather,
barely three-quarters of an inch in length. These feathers differ
considerably in relative length, the fourth, or the fifth, or the
sixth, being the longest; with the third either equal to, or
considerably shorter than the fifth. In wild gallinaceous species the
relative length and number of the main wing and tail-feathers are
extremely constant.

The tail differs much in erectness and size, being small in Malays and
very small in Cochins. In thirteen fowls of various breeds which I have
examined, five had the normal number of 14 feathers, including in this
number the two middle sickle-feathers; six others (viz. a Caffre cock,
Gold-spangled Polish cock, Cochin hen, Sultan hen, Game hen, and Malay
hen) had 16; {259} and two (an old Cochin cock and Malay hen) had 17
feathers. The rumpless fowl has no tail, and in a bird which I kept
alive the oil-gland had aborted; but this bird, though the os coccygis
was extremely imperfect, had a vestige of a tail with two rather long
feathers in the position of the outer caudals. This bird came from a
family where, as I was told, the breed had kept true for twenty years;
but rumpless fowls often produce chickens with tails.[422] An eminent
physiologist[423] has recently spoken of this breed as a distinct
species; had he examined the deformed state of the os coccyx he would
never have come to this conclusion; he was probably misled by the
statement, which may be found in some works, that tailless fowls are
wild in Ceylon; but this statement, as I have been assured by Mr.
Layard and Dr. Kellaert, who have so closely studied the birds of
Ceylon, is utterly false.

The tarsi vary considerably in length, being relatively to the f