n be more singular
than the relation between blue eyes and deafness in cats, and the
tortoise-shell colour with the female sex; the feathered feet and skin
between the outer toes in pigeons, and the presence of more or less down
on the young birds when first hatched, with the future colour of their
plumage; or, again, the relation between the hair and teeth in the naked
Turkish dog, though here probably homology comes into play? With respect
to this latter case of correlation, I think it can hardly be accidental,
that if we pick out the two orders of mammalia which are most
abnormal in their dermal coverings, viz. Cetacea (whales) and Edentata
(armadilloes, scaly ant-eaters, etc.), that these are likewise the most
abnormal in their teeth.

I know of no case better adapted to show the importance of the laws of
correlation in modifying important structures, independently of utility
and, therefore, of natural selection, than that of the difference
between the outer and inner flowers in some Compositous and
Umbelliferous plants. Every one knows the difference in the ray and
central florets of, for instance, the daisy, and this difference is
often accompanied with the abortion of parts of the flower. But, in some
Compositous plants, the seeds also differ in shape and sculpture; and
even the ovary itself, with its accessory parts, differs, as has been
described by Cassini. These differences have been attributed by some
authors to pressure, and the shape of the seeds in the ray-florets in
some Compositae countenances this idea; but, in the case of the corolla
of the Umbelliferae, it is by no means, as Dr. Hooker informs me, in
species with the densest heads that the inner and outer flowers most
frequently differ. It might have been thought that the development of
the ray-petals by drawing nourishment from certain other parts of the
flower had caused their abortion; but in some Compositae there is a
difference in the seeds of the outer and inner florets without any
difference in the corolla. Possibly, these several differences may be
connected with some difference in the flow of nutriment towards the
central and external flowers: we know, at least, that in irregular
flowers, those nearest to the axis are oftenest subject to peloria, and
become regular. I may add, as an instance of this, and of a striking
case of correlation, that I have recently observed in some garden
pelargoniums, that the central flower of the truss often lo