a
common and cheap mineral, and presents, as closely as anything can, the
milky bloom and color of a torrent wave: and if they can conquer the
opal, they may at last have some chance with the stream, as far as the
stream is in any wise possible. But, as I have just said, the bright
parts of it are _not_ possible, and ought, as much as may be, to be
avoided in choosing subjects. A great deal more may, however, be done
than any artist has done yet, in painting the gradual disappearance and
lovely coloring of stones seen through clear and calm water.

Students living in towns may make great progress in rock-drawing by
frequently and faithfully drawing broken edges of common roofing slates,
of their real size.

NOTE 6, p. 125.--"_Nature's economy of color._"

252. I heard it wisely objected to this statement, the other day, by a
young lady, that it was not through economy that Nature did not color
deep down in the flower bells, but because "she had not light enough
there to see to paint with." This may be true; but it is certainly not
for want of light that, when she is laying the dark spots on a foxglove,
she will not use any more purple than she has got already on the bell,
but takes out the color all round the spot, and concentrates it in the
middle.

NOTE 7, p. 138.--"_The law of repetition._"

253. The reader may perhaps recollect a very beautiful picture of
Vandyck's in the Manchester Exhibition, representing three children in
court dresses of rich black and red. The law in question was amusingly
illustrated, in the lower corner of that picture, by the introduction of
two crows, in a similar color of court dress, having jet black feathers
and bright red beaks.

254. Since the first edition of this work was published, I have
ascertained that there are two series of engravings from the Bible
drawings mentioned in the list at p. 50. One of these is inferior to the
other, and in many respects false to the drawing; the "Jericho," for
instance, in the false series, has common bushes instead of palm trees
in the middle distance. The original plates may be had at almost any
respectable printseller's; and ordinary impressions, whether of these or
any other plates mentioned in the list at p. 50, will be quite as useful
as proofs: but, in buying Liber Studiorum, it is always well to get the
best impressions that can be had, and if possible impressions of the
original plates, published by Turner. In case these are