peared
before the scale papillae. Evidence of this was especially afforded by
the largest embryo, which had reached about the same stage in its
development as a 16-days goose embryo.

"In the largest Emperor embryo feather papillae occur all over the
hind-quarters and on the legs to within a short distance of the tarsal
joint. Beyond the tarsal joint even in the largest embryo no attempt had
been made to produce the papillae which in older penguin embryos
represent, and ultimately develop into, the scaly covering of the foot.
The absence of papillae on the foot implied either that the scale
papillae were fundamentally different from feather papillae or that for
some reason or other the development of the papillae destined to give
rise to the foot scales had been retarded. There is no evidence as far as
I can ascertain that in modern lizards the scale papillae above the
tarsal joint appear before the scale papillae beyond this joint.

"The absence of papillae below the tarsal joint in Emperor embryos,
together with the fact that in many birds each large feather papilla is
accompanied by two or more very small feather papillae, led me to study
the papillae of the limbs of other birds. The most striking results were
obtained from the embryos of Chinese geese in which the legs are
relatively longer than in penguins. In a 13-days goose embryo the whole
of the skin below and for some distance above the tarsal joint is quite
smooth, whereas the skin of the rest of the leg is studded with feather
papillae. On the other hand, in an 18-days goose embryo in which the
feather papillae of the legs have developed into filaments, each
containing a fairly well-formed feather, scale papillae occur not only on
the foot below and for some distance above the tarsal joint but also
between the roots of the feather filaments between the tarsal and the
knee joints. More important still, in a 20-days goose embryo a number of
the papillae situated between the feather filaments of the leg were
actually developing into scales each of which overlapped the root
(calamus) of a feather just as scales overlap the foot feathers in grouse
and other feather-footed birds.

"As in bird embryos there is no evidence that feather papillae ever
develop into scales or that scale papillae ever develop into feathers it
may be assumed that feather papillae are fundamentally different from
scale papillae, the difference presumably being due to the presence of
spe