lude sentiment. As for its being
easy, those only think so who never tried it; composition being, in
fact, infinitely easier to a man who can compose, than imitation of this
high kind to even the most able imitator; nor would it exclude
sentiment, for, however sincerely we may try to paint all we see, this
_cannot_, as often aforesaid, be ever done: all that is possible is a
certain selection, and more or less wilful assertion, of one fact in
preference to another; which selection ought always to be made under the
influence of sentiment. Nor will such topography involve an entire
submission to ugly accidents interfering with the impressiveness of the
scene. I hope, as art is better understood, that our painters will get
into the habit of accompanying all their works with a written statement
of their own reasons for painting them, and the circumstances under
which they were done; and, if in this written document they state the
omissions they have made, they may make as many as they think proper.
For instance, it is not possible now to obtain a view of the head of the
Lake of Geneva without including the "Hotel Biron"--an establishment
looking like a large cotton factory--just above the Castle of Chillon.
This building ought always to be omitted, and the reason for the
omission stated. So the beauty of the whole town of Lucerne, as seen
from the lake, is destroyed by the large new hotel for the English,
which ought, in like manner, to be ignored, and the houses behind it
drawn as if it were transparent.

Sec. 8. But if a painter has inventive power he is to treat his subject in
a totally different way; giving not the actual facts of it, but the
impression it made on his mind.

And now, once for all, let it be clearly understood that an "impression
on the mind" does not mean a piece of manufacture. The way in which most
artists proceed to "invent," as they call it, a picture, is this: they
choose their subject, for the most part, well, with a sufficient
quantity of towers, mountains, ruined cottages, and other materials, to
be generally interesting; then they fix on some object for a principal
light; behind this they put a dark cloud, or, in front of it, a dark
piece of foreground; then they repeat this light somewhere else in a
less degree, and connect the two lights together by some intermediate
ones. If they find any part of the foreground uninteresting they put a
group of figures into it; if any part of the distance,