note of the opposition being
vermilion against green blue, both of equal tone, and at such a
height and acme of brilliancy that you cannot see the line where
their edges pass into each other.

No colors that can be fixed in earth can ever represent to you the
luster of these cloudy ones. But the actual tints may be shown you
in a lower key, and to a certain extent their power and relation to
each other.

I have painted the diagram here shown you with colors prepared for
me lately by Messrs. Newman, which I find brilliant to the height
that pigments can be; and the ready kindness of Mr. Wilson Barrett
enables me to show you their effect by a white light as pure as
that of the day. The diagram is enlarged from my careful sketch of
the sunset of 1st October, 1868, at Abbeville, which was a
beautiful example of what, in fine weather about to pass into
storm, a sunset could then be, in the districts of Kent and Picardy
unaffected by smoke. In reality, the ruby and vermilion clouds
were, by myriads, more numerous than I have had time to paint: but
the general character of their grouping is well enough expressed.
All the illumined clouds are high in the air, and nearly
motionless; beneath them, electric storm-cloud rises in a
threatening cumulus on the right, and drifts in dark flakes across
the horizon, casting from its broken masses radiating shadows on
the upper clouds. These shadows are traced, in the first place by
making the misty blue of the open sky more transparent, and
therefore darker; and secondly, by entirely intercepting the
sunbeams on the bars of cloud, which, within the shadowed spaces,
show dark on the blue instead of light.

But, mind, all that is done by reflected light--and in that light
you never get a _green_ ray from the reflecting cloud; there is no
such thing in nature as a green lighted cloud relieved from a red
sky,--the cloud is always red, and the sky green, and green,
observe, by transmitted, not reflected light.

But now note, there is another kind of cloud, pure white, and
exquisitely delicate; which acts not by reflecting, nor by
refracting, but, as it is now called, _dif_fracting, the sun's
rays. The particles of this cloud are said--with what truth I know
not[15]--to send the sunbeams round them instead of through them;
somehow or other, at any rate, they resolve them into their
prismatic elements; and then you have literally a kaleidoscope in
the sky, with every color of the prism in