rouble about the matter, leave the green snakes swimming
upon the surface, and plunge for it. Hence reflections, if viewed
collectively, are always clear in proportion to the distance of the
water on which they are cast. And now look at Turner's Ulleswater, or
any of his distant lake expanses, and you will find every crag and line
of the hills rendered in them with absolute fidelity, while the near
surface shows nothing but a vague confusion of exquisite and lustrous
tint. The reflections even of the clouds will be given far off, while
those of near boats and figures will be confused and mixed among each
other, except just at the water-line.

Sec. 6. The error of Vandevelde.

And now we see what Vandevelde _ought_ to have done with the shadow of
his ship spoken of in the first chapter of this section. In such a calm,
we should in nature, if we had looked for the reflection, have seen it
clear from the water-line to the flag on the mainmast; but in so doing,
we should have appeared to ourselves to be looking under the water, and
should have lost all feeling of surface. When we looked at the surface
of the sea,--as we naturally should,--we should have seen the image of
the hull absolutely clear and perfect, because that image is cast on
distant water; but we should have seen the image of the masts and sails
gradually more confused as they descended, and the water close to us
would have borne only upon its surface a maze of flashing color and
indefinite hue. Had Vandevelde, therefore, given the perfect image of
his ship, he would have represented a truth dependent on a particular
effort of the eye, and destroyed his surface. But his business was to
give, not a distinct reflection, but the colors of the reflection in
mystery and disorder upon his near water, all perfectly vivid, but none
intelligible; and had he done so, the eye would not have troubled itself
to search them out; it would not have cared whence or how the colors
came, but it would have felt them to be true and right, and rested
satisfied upon the polished surface of the clear sea. Of the perfect
truth, the best examples I can give are Turner's Saltash and Castle
Upnor.

Sec. 7. Difference in arrangement of parts between the reflected object and
its image.

Be it next observed that the reflection of all near objects is, by our
fifth rule, not an exact copy of the parts of them which we see above
the water, but a totally different view and arrang