_. He may sometimes defend
this refusal by sophistical arguments, as when he says that mechanism
would require the last stage of the universe to be simultaneous with
the first, forgetting that the unit of mechanism is not a mathematical
equation but some observed typical event. The refusal itself, however,
would be honest scepticism enough were it made with no _arriere
pensee_, but simply in view of the immense complexity of the facts and
the extreme simplicity of the mechanical hypothesis. In such a
situation, to halt at appearances might seem the mark of a true
naturalist and a true empiricist not misled by speculative haste and
the human passion for system and simplification. At the first reading,
M. Bergson's _Evolution Creatrice_ may well dazzle the professional
naturalist and seem to him an illuminating confession of the nature
and limits of his science; yet a second reading, I have good authority
for saying, may as easily reverse that impression. M. Bergson never
reviews his facts in order to understand them, but only if possible to
discredit others who may have fancied they understood. He raises
difficulties, he marks the problems that confront the naturalist, and
the inadequacy of explanations that may have been suggested. Such
criticism would be a valuable beginning if it were followed by the
suggestion of some new solution; but the suggestion only is that no
solution is possible, that the phenomena of life are simply
miraculous, and that it is in the tendency or vocation of the animal,
not in its body or its past, that we must see the ground of what goes
on before us.

With such a philosophy of science, it is evident that all progress in
the understanding of nature would cease, as it ceased after Aristotle.
The attempt would again be abandoned to reduce gross and obvious
cycles of change, such as generation, growth, and death, to minute
latent cycles, so that natural history should offer a picturesque
approach to universal physics. If for the magic power of types,
invoked by Aristotle, we substituted with M. Bergson the magic power
of the _elan vital_, that is, of evolution in general, we should be
referring events not to finer, more familiar, more pervasive
processes, but to one all-embracing process, unique and always
incomplete. Our understanding would end in something far vaguer and
looser than what our observation began with. Aristotle at least could
refer particulars to their specific types, as medicin