the better as our squares are
smaller, more numerous and more varied in tone. But an infinity of
elements infinitely small, presenting an infinity of shades, would be
necessary to obtain the exact equivalent of the figure that the artist
has conceived as a simple thing, which he has wished to transport as a
whole to the canvas, and which is the more complete the more it strikes
us as the projection of an indivisible intuition. Now, suppose our eyes
so made that they cannot help seeing in the work of the master a mosaic
effect. Or suppose our intellect so made that it cannot explain the
appearance of the figure on the canvas except as a work of mosaic. We
should then be able to speak simply of a collection of little squares,
and we should be under the mechanistic hypothesis. We might add that,
beside the materiality of the collection, there must be a plan on which
the artist worked; and then we should be expressing ourselves as
finalists. But in neither case should we have got at the real process,
for there are no squares brought together. It is the picture, _i.e._ the
simple act, projected on the canvas, which, by the mere fact of entering
into our perception, is _de_composed before our eyes into thousands and
thousands of little squares which present, as _re_composed, a wonderful
arrangement. So the eye, with its marvelous complexity of structure, may
be only the simple act of vision, divided _for us_ into a mosaic of
cells, whose order seems marvelous to us because we have conceived the
whole as an assemblage.

If I raise my hand from A to B, this movement appears to me under two
aspects at once. Felt from within, it is a simple, indivisible act.
Perceived from without, it is the course of a certain curve, AB. In this
curve I can distinguish as many positions as I please, and the line
itself might be defined as a certain mutual coordination of these
positions. But the positions, infinite in number, and the order in which
they are connected, have sprung automatically from the indivisible act
by which my hand has gone from A to B. Mechanism, here, would consist
in seeing only the positions. Finalism would take their order into
account. But both mechanism and finalism would leave on one side the
movement, which is reality itself. In one sense, the movement is _more_
than the positions and than their order; for it is sufficient to make it
in its indivisible simplicity to secure that the infinity of the
successive positio