estone, under the name of "calx," and that he would probably
have arranged Diatoms among animals, as part of "chaos." Ehrenberg quotes
another even more pithy passage, which I have not been able to find in
any edition of the _Systema_ accessible to me: "Sic lapides ab
animalibus, nec vice versa. Sic runes saxei non primaevi, sed temporis
filiae."]

So long as the _Globigerinoe_;, actually collected at the surface, have
not been demonstrated to contain the elements of clay, the _Challenger_
hypothesis, as I may term it, must be accepted with reserve and
provisionally, but, at present, I cannot but think that it is more
probable than any other suggestion which has been made.

Accepting it provisionally, we arrive at the remarkable result that all
the chief known constituents of the crust of the earth may have formed
part of living bodies; that they may be the "ash" of protoplasm; that the
"_rupes saxei_" are not only _"temporis,"_ but "_vitae filiae_"; and,
consequently, that the time during which life has been active on the
globe may be indefinitely greater than the period, the commencement of
which is marked by the oldest known rocks, whether fossiliferous or
unfossiliferous.

And thus we are led to see where the solution of a great problem and
apparent paradox of geology may lie. Satisfactory evidence now exists
that some animals in the existing world have been derived by a process of
gradual modification from pre-existing forms. It is undeniable, for
example, that the evidence in favour of the derivation of the horse from
the later tertiary _Hipparion_, and that of the _Hipparion_ from
_Anchitherium_, is as complete and cogent as such evidence can reasonably
be expected to be; and the further investigations into the history of the
tertiary mammalia are pushed, the greater is the accumulation of evidence
having the same tendency. So far from palaeontology lending no support to
the doctrine of evolution--as one sees constantly asserted--that
doctrine, if it had no other support, would have been irresistibly forced
upon us by the palaeontological discoveries of the last twenty years.

If, however, the diverse forms of life which now exist have been produced
by the modification of previously-existing less divergent forms, the
recent and extinct species, taken as a whole, must fall into series which
must converge as we go back in time. Hence, if the period represented by
the rocks is greater than, or co-extensive with,