in such a day-dream are often congruous in character with those that
preceded, and mere prolongations of them; but this prolongation itself
modifies them, and what develops is in no way deducible or predictable
out of what exists. This situation is perfectly explicable
scientifically. The movement of consciousness will be self-congruous
and sustained when it rests on continuous processes in the same
tissues, and yet quite unpredictable from within, because the direct
sensuous report of bodily processes (in nausea, for instance, or in
hunger) contains no picture of their actual mechanism. Even wholly new
features, due to little crises in bodily life, may appear in a dream
to flow out of what already exists, yet freely develop it; because in
dreams comparison, the attempt to be consistent, is wholly in
abeyance, and also because the new feature will come imbedded in
others which are not new, but have dramatic relevance in the story. So
immediate consciousness yields the two factors of Bergsonian freedom,
continuity and indetermination.

Again, take the somewhat disconcerting assertion that movement exists
when there is nothing that moves, and no space that it moves through.
In vision, perhaps, it is not easy to imagine a consciousness of
motion without some presentation of a field, and of a distinguishable
something in it; but if we descend to somatic feelings (and the more
we descend, with M. Bergson, the closer we are to reality), in
shooting pains or the sense of intestinal movements, the feeling of a
change and of a motion is certainly given in the absence of all idea
of a _mobile_ or of distinct points (or even of a separate field)
through which it moves; consciousness begins with the sense of change,
and the terms of the felt process are only qualitative limits, bred
out of the felt process itself. Even a more paradoxical tenet of our
philosopher's finds it justification here. He says that the units of