t would hardly profit us to insist on holding to
Perry as the cause of the Douma: the terms have grown too remote to
have any real or practical relation to each other. In every series of
real terms, not only do the terms themselves and their associates
and environments change, but we change, and their _meaning_ for
us changes, so that new kinds of sameness and types of causation
continually come into view and appeal to our interest. Our earlier
lines, having grown irrelevant, are then dropped. The old terms can no
longer be substituted nor the relations 'transferred,' because of so
many new dimensions into which experience has opened. Instead of a
straight line, it now follows a zigzag; and to keep it straight, one
must do violence to its spontaneous development. Not that one might
not possibly, by careful seeking (tho I doubt it), _find_ some line in
nature along which terms literally the same, or causes causal in the
same way, might be serially strung without limit, if one's interest
lay in such finding. Within such lines our axioms might hold, causes
might cause their effect's effects, etc.; but such lines themselves
would, if found, only be partial members of a vast natural network,
within the other lines of which you could not say, in any sense that
a wise man or a sane man would ever think of, in any sense that would
not be concretely _silly_, that the principle of skipt intermediaries
still held good. In the _practical_ world, the world whose
significances we follow, sames of the same are certainly not sames of
one another; and things constantly cause other things without being
held responsible for everything of which those other things are
causes.

Professor Bergson, believing as he does in a heraclitean 'devenir
reel,' ought, if I rightly understand him, positively to deny that in
the actual world the logical axioms hold good without qualification.
Not only, according to him, do terms change, so that after a certain
time the very elements of things are no longer what they were, but
relations also change, so as no longer to obtain in the same identical
way between the new things that have succeeded upon the old ones. If
this were really so, then however indefinitely sames might still
be substituted for sames in the logical world of nothing but pure
sameness, in the world of real operations every line of sameness
actually started and followed up would eventually give out, and cease
to be traceable any farther. S