emblances to the same bone in the
Carnivorous Mammals. As has been previously noticed, Professor
Owen is of opinion that some of the Reptilian remains of the
Permian deposits will also be found to belong to this group of
the "Theriodonts."

[Illustration: Fig. 155.--Supposed footprint of a Bird, from
the Triassic Sandstones of the Connecticut River. The slab shows
also numerous "rain-prints."]

Lastly, we find in the Triassic rocks the remains of Reptiles
belonging to the great Mesozoic order of the _Deinosauria_. This
order attains its maximum at a later period, and will be spoken of
when the Jurassic and Cretaceous deposits come to be considered.
The chief interest of the Triassic Reptiles of this group arises
from the fact that they are known by their footprints as well as
by their bones; and a question has arisen whether the supposed
footprints of _birds_ which occur in the Trias have not really
been produced by Deinosaurs. This leads us, therefore, to speak
at the same time as to the evidence which we have of the existence
of the class of Birds during the Triassic period. No actual bones
of any bird have as yet been detected in any Triassic deposit;
but we have tolerably clear evidence of their existence at this
time in the form of _footprints_. The impressions in question
are found in considerable numbers in certain red sandstones of
the age of the Trias in the valley of the Connecticut River, in
the United States. They vary much in size, and have evidently been
produced by many different animals walking over long stretches of
estuarine mud and sand exposed at low water. The footprints now
under consideration form a double series of _single_ prints, and
therefore, beyond all question, are the tracks of a _biped_--that
is, of an animal which walked upon two legs. No living animals,
save Man and the Birds, walk habitually on two legs; and there
is, therefore, a _prima facie_ presumption that the authors of
these prints were Birds. Moreover, each impression consists of
the marks of three toes turned forwards (fig. 155), and therefore
are precisely such as might be produced by Wading or Cursorial
Birds. Further, the impressions of the toes show exactly the
same numerical progression in the number of the joints as is
observable in living Birds--that is to say, the innermost of the
three toes consists of three joints, the middle one of four, and
the outer one of five joints. Taking this evidence collectively,
it wo