oic formations,
up to the Chalk, are especially characterised by the naked-seeded
Flowering plants--the Conifers and the Cycads; whilst the higher
groups of the Angiospermous Exogens and Monocotyledons characterise
the Upper Cretaceous and Tertiary rocks.

Facts of the above nature--and they could be greatly multiplied--seem
to point clearly to the existence of some law of progression,
though we certainly are not yet in a position to formulate this
law, or to indicate the precise manner in which it has operated.
Two considerations, also, must not be overlooked. In the first
place, there are various groups, some of them highly organised,
which make their appearance at an extremely ancient date, but
which continue throughout geological time almost unchanged, and
certainly unprogressive. Many of these "persistent types" are
known--such as various of the _Foraminifera_, the _Linguloe_, the
_Nautili_, &c.; and they indicate that under given conditions, at
present unknown to us, it is possible for a life-form to subsist
for an almost indefinite period without any important modification
of its structure. In the second place, whilst the facts above
mentioned point to some general law of progression of the great
zoological groups, it cannot be asserted that the primeval types
_of any given group_ are necessarily "lower," zoologically speaking,
than their modern representatives. Nor does this seem to be at
all necessary for the establishment of the law in question. It
cannot be asserted, for example, that the Ganoid and Placoid Fishes
of the Upper Silurian are in themselves less highly organised
than their existing representatives; nor can it even be asserted
that the Ganoid and Placoid orders are low _groups_ of the class
_Pisces_. On the contrary, they are high groups; but then it
must be remembered that these are probably not really the first
Fishes, and that if we meet with Fishes at some future time in
the Lower Silurian or Cambrian, these may easily prove to be
representatives of the lower orders of the class. This question
cannot be further entered into here, as its discussion could be
carried out to an almost unlimited length; but whilst there are
facts pointing both ways, it appears that at present we are not
justified in asserting that the earlier types of each group--so far
as these are known to us, or really are without predecessors--are
_necessarily_ or _invariably_ more "degraded" or "embryonic" in
their structur