rful impulse, and fall again as soon as the force of the first
impetus is entirely spent. When men endeavour to persuade you to such
folly, believe them not. For my own part, I have _seen_ the flying fish
fly--deliberately fly, and flutter, and rise again, and change the
direction of their flight in mid-air, exactly after the fashion of a big
dragonfly. If the other people who have watched them haven't succeeded
in seeing them fly, that is their own fault, or at least their own
misfortune; perhaps their eyes weren't quick enough to catch the rapid,
though to me perfectly recognisable, hovering and fluttering of the
gauze-like wings; but I have seen them myself, and I maintain that on
such a question one piece of positive evidence is a great deal better
than a hundred negative. The testimony of all the witnesses who didn't
see the murder committed is as nothing compared with the single
testimony of the one man who really did see it. And in this case I have
met with many other quick observers who fully agreed with me, against
the weight of scientific opinion, that they have seen the flying fish
really fly with their own eyes, and no mistake about it. The German
professors, indeed, all think otherwise; but then the German professors
all wear green spectacles, which are the outward and visible sign of
'blinded eyesight poring over miserable books.' The unsophisticated
vision of the noble British seaman is unanimously with me on the matter
of the reality of the fishes' flight.

Another group of very interesting fish out of water are the flying
gurnards, common enough in the Mediterranean and the tropical Atlantic.
They are much heavier and bigger creatures than the true flying fish of
the herring type, being often a foot and a half long, and their wings
are much larger in proportion, though not, I think, really so powerful
as those of their pretty little silvery rivals. All the flying fish fly
only of necessity, not from choice. They leave the water when pursued
by their enemies, or when frightened by the rapid approach of a big
steamer. So swiftly do they fly, however, that they can far outstrip a
ship going at the rate of ten knots an hour; and I have often watched
one keep ahead of a great Pacific liner under full steam for many
minutes together in quick successive flights of three or four hundred
feet each. Oddly enough, they can fly further against the wind than
before it--a fact acknowledged even by the spectacled Germa