s
motion, and until it reaches the earth is always seen perpendicularly
below the car."

An interesting phenomenon, relating to the formation of fog was
witnessed by M. Flammarion in one of his voyages. He was flying low with
a fast wind, and while traversing a forest he noticed here and there
patches of light clouds, which, remaining motionless in defiance of
the strong wind, continued to hang above the summits of the trees.
The explanation of this can hardly be doubtful, being analogous to
the formation of a night-cap on a mountain peak where warm moist
air-currents become chilled against the cold rock surface, forming,
momentarily, a patch of cloud which, though constantly being blown away,
is as constantly re-formed, and thus is made to appear as if stationary.

The above instructive phenomenon could hardly have been noticed save
by an aeronaut, and the same may be said of the following. Passing in
a clear sky over the spot where the Marne flows into the Seine, M.
Flammarion notes that the water of the Marne, which, as he says, is as
yellow now as it was in the time of Julius Caesar, does not mix with the
green water of the Seine, which flows to the left of the current, nor
with the blue water of the canal, which flows to the right. Thus, a
yellow river was seen flowing between two distinct brooks, green and
blue respectively.

Here was optical evidence of the way in which streams of water which
actually unite may continue to maintain independent courses. We have
seen that the same is true of streams of air, and, where these traverse
one another in a copious and complex manner, we find, as will be shown,
conditions produced that cause a great deadening of sound; thus, great
differences in the travel of sound in the silent upper air can be
noticed on different days, and, indeed, in different periods of the same
aerial voyage. M. Flammarion bears undeniable testimony to the manner in
which the equable condition of the atmosphere attending fog enhances,
to the aeronaut, the hearing of sounds from below. But when he gives
definite heights as the range limits of definite sounds it must be
understood that these ranges will be found to vary greatly according
to circumstances. Thus, where it is stated that a man's voice may make
itself heard at 3,255 feet, it might be added that sometimes it cannot
be heard at a considerably less altitude; and, again, the statement that
the whistle of a locomotive rises to near 10,000 fe