of the air
itself, are comparatively rare as fossils, and the record of
the past existence of these is much more imperfect than is the
case with animals living in water. Moreover, the fossiliferous
deposits are not only almost exclusively aqueous formations, but
the great majority are marine, and only a comparatively small
number have been formed by lakes and rivers. It follows from the
foregoing that the palaeontological record is fullest and most
complete so far as sea-animals are concerned, though even here we
find enormous gaps, owing to the absence of hard structures in
many great groups; of animals inhabiting fresh waters our knowledge
is rendered still further incomplete by the small proportion
that fluviatile and lacustrine deposits bear to marine; whilst
we have only a fragmentary acquaintance with the air-breathing
animals which inhabited the earth during past ages.

Lastly, the imperfection of the palaeontological record, due to
the causes above enumerated, is greatly aggravated, especially
as regards the earlier portion of the earth's history, by the
fact that many rocks which contained fossils when deposited have
since been rendered barren of organic remains. The principal cause
of this common phenomenon is what is known as "metamorphism"--that
is, the subjection of the rock to a sufficient amount of heat to
cause a rearrangement of its particles. When at all of a pronounced
character, the result of metamorphic action is invariably the
obliteration of any fossils which might have been originally
present in the rock. Metamorphism may affect rocks of any age,
though naturally more prevalent in the older rocks, and to this
cause must be set down an irreparable loss of much fossil evidence.
The most striking example which is to be found of this is the
great Laurentian series, which comprises some 30,000 feet of
highly-metamorphosed sediments, but which, with one not wholly
undisputed exception, has as yet yielded no remains of living
beings, though there is strong evidence of the former existence
in it of fossils.

Upon the whole, then, we cannot doubt that the earth's crust, so
far as yet deciphered by us, presents us with but a very imperfect
record of the past. Whether the known and admitted imperfections
of the geological and palaeontological records are sufficiently
serious to account satisfactorily for the deficiency of direct
evidence recognisable in some modern hypotheses, may be a matter
of indivi