s the
possibility, as we shall explain further on, that the part played by
chance is much greater in the variation of plants than in that of
animals, because, in the vegetable world, function does not depend so
strictly on form. Be that as it may, the neo-Darwinians are inclined to
admit that the periods of mutation are determinate. The direction of the
mutation may therefore be so as well, at least in animals, and to the
extent we shall have to indicate.

We thus arrive at a hypothesis like Eimer's, according to which the
variations of different characters continue from generation to
generation in definite directions. This hypothesis seems plausible to
us, within the limits in which Eimer himself retains it. Of course, the
evolution of the organic world cannot be predetermined as a whole. We
claim, on the contrary, that the spontaneity of life is manifested by a
continual creation of new forms succeeding others. But this
indetermination cannot be complete; it must leave a certain part to
determination. An organ like the eye, for example, must have been formed
by just a continual changing in a definite direction. Indeed, we do not
see how otherwise to explain the likeness of structure of the eye in
species that have not the same history. Where we differ from Eimer is in
his claim that combinations of physical and chemical causes are enough
to secure the result. We have tried to prove, on the contrary, by the
example of the eye, that if there is "orthogenesis" here, a
psychological cause intervenes.

Certain neo-Lamarckians do indeed resort to a cause of a psychological
nature. There, to our thinking, is one of the most solid positions of
neo-Lamarckism. But if this cause is nothing but the conscious effort of
the individual, it cannot operate in more than a restricted number of
cases--at most in the animal world, and not at all in the vegetable
kingdom. Even in animals, it will act only on points which are under the
direct or indirect control of the will. And even where it does act, it
is not clear how it could compass a change so profound as an increase of
complexity: at most this would be conceivable if the acquired characters
were regularly transmitted so as to be added together; but this
transmission seems to be the exception rather than the rule. A
hereditary change in a definite direction, which continues to accumulate
and add to itself so as to build up a more and more complex machine,
must certainly be related