Page 86 [86]
48 LIGHT AND WATER
surface of the sea affected in tone by a bank of cloud
lying low down on the horizon ; or again, to draw
well-defined streaks as due to very high! clouds, for
we have seen that the upper parts of the sky affect
large widths of sea. Above the limit at which the sun
streak vanishes, the lower the clouds the better de¬
fined are the streaks which they cast. The latter are
often most fascinating to watch in their rapid changes,
as when thrown by brilliant white clouds floating in
a blue sky on to water rippled by a light and shifting
breeze. As in the instance of the rock reflexions de¬
scribed above, we may again be puzzled for a moment
by the presence or absence of a streak in an unex¬
pected quarter. Two or three insignificant clouds
combine to form an unmistakable streak, whilst a
bright conspicuous mass—too high, or too low, or
possibly neutralized by some dark mass above or be¬
low it—casts none. Itis curious to see how an almost
horizontal layer of clouds only slightly thicker or
darker in places throws definite streaks across the
nearly smooth water, so determined is nature to con¬
vert horizontal into vertical. ‘‘If we see on an extent
of lightly swelling water surface the image of a bank
of white clouds, with masses of higher accumulation
at intervals, the water will not usually reflect the
whole bank in an elongated form, but it will com¬
monly take the eminent parts, and reflect them in
long straight columns of defined breadth, and miss
" By “high clouds” are meant those at a great angular elevation
above the horizon. A streak cast by such a high cloud would not
reach as far as the horizon, for a streak could hardly extend deyond
the object causing it.