ina, from the bay of Naples. This hydroid consists of
a long straight main stem which grows vertically upwards and which has
at regular intervals very fine and short bristle-like lateral branches,
on the upper side of which the polyps grow. The main stem is negatively
geotropic, i.e. its apex continues to grow vertically upwards when we
put it obliquely into the aquarium, while the roots grow vertically
downwards. The writer observed that when the stem is put horizontally
into the water the short lateral branches on the lower side give rise to
an altogether different kind of organ, namely, to roots, and these roots
grow indefinitely in length and attach themselves to solid bodies; while
if the stem had remained in its normal position no further growth
would have occurred in the lateral branches. From the upper side of the
horizontal stem new stems grow out, mostly directly from the original
stem, occasionally also from the short lateral branches. It is thus
possible to force upon this hydroid an arrangement of organs which is
altogether different from the hereditary arrangement. The writer
had called the change in the hereditary arrangement of organs or the
transformation of organs by external forces HETEROMORPHOSIS. We cannot
now go any further into this subject, which should, however, prove of
interest in relation to the problem of heredity.

If it is correct to apply inferences drawn from the observation on the
frog's egg to the behaviour of Antennularia, one might conclude that the
cells of Antennularia also contain non-miscible substances of different
specific gravity, and that wherever the specifically lighter substance
comes in contact with the sea-water (or gets near the surface of the
cell) the growth of a stem is favoured; while contact with the sea-water
of the specifically heavier of the substances, will favour the formation
of roots.

VI. THE EXPERIMENTAL CONTROL OF ANIMAL INSTINCTS.

(a) EXPERIMENTS ON THE MECHANISM OF HELIOTROPIC REACTIONS IN ANIMALS.

Since the instinctive reactions of animals are as hereditary as
their morphological character, a discussion of experiments on the
physico-chemical character of the instinctive reactions of animals
should not be entirely omitted from this sketch. It is obvious that such
experiments must begin with the simplest type of instincts, if they are
expected to lead to any results; and it is also obvious that only such
animals must be selected for this purpose, th