diatoms), in which the whole multiplication takes place in this way, and
with great rapidity. These also have sexual reproduction; but in it two old
individuals are always destroyed to make a single new one! Here propagation
diminishes the number of individuals fifty per cent. Who can suppose that
such a costly process as this, and that all the exquisite arrangements for
cross-fertilization in hermaphrodite plants, do not subserve some most
important purpose? How and why the union of two organisms, or generally of
two very minute portions of them, should reenforce vitality, we do not
know, and can hardly conjecture. But this must be the meaning of sexual
reproduction.

The conclusion of the matter, from the scientific point of view, is, that
sexually-propagated varieties or races, although liable to disappear through
change, need not be expected to wear out, and there is no proof that they
do; but, that non-sexually propagated varieties, though not especially
liable to change, may theoretically be expected to wear out, but to be a
very long time about it.


II

Do Species wear out? and if not, why not?


The question we have just been considering was merely whether races are, or
may be, as enduring as species. As to the inherently unlimited existence of
species themselves, or the contrary, this, as we have said, is a geological
and very speculative problem. Not a few geologists and naturalists,
however, have concluded, or taken for granted, that species have a natural
term of existence--that they culminate, decline, and disappear through