, as being homologous structures. He says that teeth
are allied to bones, whereas horns are more nearly allied to skin
(_Hist. Anim._, iii.). This is an astonishingly happy guess,
considering that all he had to go upon was the observation that in
black animals the horns are black but the teeth white. One cannot but
admire the way in which Aristotle fixes upon apparently trivial and
commonplace facts, and draws from them far-reaching consequences. He
often goes wrong, it is true, but he always errs in the grand manner.

While Aristotle certainly recognised the existence of homologies, and
even had a feeling for them, he did not clearly distinguish homology
from analogy. He comes pretty near the distinction in the following
passage. After explaining that in animals belonging to the same class
the parts are the same, differing only in excess or defect, he says,
"But some animals agree with each other in their parts neither in form
nor in excess and defect, but have only an analogous likeness, such as
a bone bears to a spine, a nail to a hoof, a hand to a crab's claw,
the scale of a fish to the feather of a bird, for that which is a
feather in the bird is a scale in the fish" (Cresswell, _loc. cit._,
p. 2). One of these comparisons is, however, a homology not an
analogy, and the last phrase throws a little doubt upon the whole
question, for it is not made clear whether it is position or function
that determines what are equivalent organs.

In the _De Partibus Animalium_ there occurs the following
passage:--"Groups that only differ in degree, and in the more or less
of an identical element that they possess, are aggregated under a
single class; groups whose attributes are not identical but analogous
are separated. For instance, bird differs from bird by gradation, or
by excess and defect; some birds have long feathers, others short
ones, but all are feathered. Bird and Fish are more remote and only
agree in having analogous organs; for what in the bird is feather, in
the fish is scale. Such analogies can scarcely, however, serve
universally as indications for the formation of groups, for almost all
animals present analogies in their corresponding parts."[8] It is thus
similarity in form and structure which determines the formation of the
main groups. Within each group the parts differ only in degree, in
largeness or smallness, softness and hardness, smoothness or
roughness, and the like (_loc. cit._, i., 4, 644^b). These pa