for a few larvae in the uterus, a large
number of eggs perish and this number is the greater the longer the
period of gestation. It thus happens that when the animals retain their
eggs a long time, very few young ones are born; and these are in a
rather advanced stage of development, owing to the long time which
elapsed since they were fertilised. When the animal lays its eggs
comparatively soon after copulation, many eggs (from 12 to 72) are
produced and the larvae are of course in an early stage of development.
In the early stage the larvae possess gills and can therefore live in
water, while in later stages they have no gills and breathe through
their lungs. Kammerer showed that both forms of Salamandra can be
induced to lay their eggs early or late, according to the physical
conditions surrounding them. If they are kept in water or in proximity
to water and in a moist atmosphere they have a tendency to lay their
eggs earlier and a comparatively high temperature enhances the tendency
to shorten the period of gestation. If the salamanders are kept in
comparative dryness they show a tendency to lay their eggs rather late
and a low temperature enhances this tendency.

Since Salamandra atra is found in rather dry alpine regions with a
relatively low temperature and Salamandra maculosa in lower regions with
plenty of water and a higher temperature, the fact that S. atra bears
young which are already developed and beyond the stage of aquatic life,
while S. maculosa bears young ones in an earlier stage, has been termed
adaptation. Kammerer's experiments, however, show that we are dealing
with the direct effects of definite outside forces. While we may speak
of adaptation when all or some of the variables which determine a
reaction are unknown, it is obviously in the interest of further
scientific progress to connect cause and effect directly whenever our
knowledge allows us to do so.

VII. CONCLUDING REMARKS.

The discovery of De Vries, that new species may arise by mutation and
the wide if not universal applicability of Mendel's Law to phenomena of
heredity, as shown especially by Bateson and his pupils, must, for
the time being, if not permanently, serve as a basis for theories of
evolution. These discoveries place before the experimental biologist the
definite task of producing mutations by physico-chemical means. It
is true that certain authors claim to have succeeded in this, but
the writer wishes to apologise to th