himself otherwise. Possibly: but whatever he experienced could not have
become art--significant form--till it had been withdrawn from the world
of experience and converted into poetry by some such exacting problem.

One problem in itself is as good as another, just as one kind of nib
is as good as another, since problems are valuable only as means. That
problem is best for any particular artist that serves that particular
artist best. The ideal problem will be the one that raises his power
most while limiting his fancy least. The incessant recourse of European
writers to dramatic form suggests that here is a problem which to them
is peculiarly favourable. Its conventions, I suppose, are sufficiently
strict to compel the artist to exert himself to the utmost, yet not so
strict as to present those appalling technical difficulties--the sort
presented by a sestina or a chant royal--that make self-expression
impossible to any but a consummate master. The novel, on the other
hand, as we are just beginning to suspect, affords for most writers an
unsatisfactory, because insufficiently rigorous, problem. Each age has
its favourites. Indeed, the history of art is very much the history of
the problem. The stuff of art is always the same, and always it must be
converted into form before it can become art; it is in their choice of
converting-machines that the ages differ conspicuously.

Two tasks that painters and writers sometimes set themselves are often
mistaken for artistic problems, but are, in fact, nothing of the sort.
One is literal representation: the other the supply of genius direct
from the cask. To match a realistic form with an æsthetic experience is
a problem that has served well many great artists: Chardin and Tolstoi
will do as examples. To make a realistic form and match it with nothing
is no problem at all. Though to say just what the camera would say is
beyond the skill and science of most of us, it is a task that will never