f the skeleton, ossification proceeds
from separate centres, foramina are formed by the fusion of separate
bones round them. In his memoir, _Lois d'Osteogenie_ (1819), Serres
established several laws of ossification based upon this principle of
separate formation.[132]

How is the fact of multiple formation to be reconciled with the
principle of repetition, according to which organs are simplest in the
early embryo and in the lower animals? But observation shows that, as a
rule, the further down the scale you go the more divided organs
become--the more numerous the bones of the skull, for example. There is
thus a parallel between multiple formation of organs in the embryos of
the higher Vertebrates and their subdivided state in the lower. Take,
for example, the kidney. In the genus _Felis_, and in birds, each kidney
has two lobes, in the elephant four, in the otter ten, in the ox twelve
to fourteen. The human kidney in its development starts with about a
dozen lobes, and the number diminishes as the kidney grows. Thus the
permanent state of the kidney in the animals mentioned is reproduced by
the stages of its development in man (xii., p. 126).

So, too, at the second or third month the uterus of the human embryo is
bicornuate, and afterwards passes through stages comparable to the adult
and permanent uterus of rodents, ruminants, and carnivores. There is
indeed a time in the development of the human embryo when it resembles
in many of its organs the adult stage of various lower animals. It is
about this time that it possesses a tail.

We note that Serres' theory of parallelism applies, strictly speaking,
only to organs, not to organisms, although he, too, readily fell into
the error of supposing that the organisation of an embryo could be
compared as a whole with the adult organisation of an animal lower in
the scale. Thus he wrote in one of his later papers[133]--"As our
researches have made clear, an animal high in the organic scale only
reaches this rank by passing through all the intermediate states which
separate it from the animals placed below it. Man only becomes man after
traversing transitional organisatory states which assimilate him first
to fish, then to reptiles, then to birds and mammals." Serres was not
altogether free from the besetting sin of the transcendentalists--hasty
generalisation.

The law of parallelism applied not only to Vertebrates but also to
Invertebrates. In a short paper[134] of 182