lutions, which the life of the youngest among you cannot be long
enough to read through so much as once, and of which the future
infinitudes of human life, however granted or extended, never will be
fatigued in admiration.

69. I take the rude outline of sail in Fig. 3, and now considering it
as a jib of one of our own sailing vessels, slightly exaggerate the
loops at the edge, and draw curved lines from them to the opposite
point, Fig. 4; and I have a reptilian or dragon's wing, which would,
with some ramification of the supporting ribs, become a bat's or
moth's; that is to say, an extension of membrane between the ribs (as
in an umbrella), which will catch the wind, and flutter upon it, like a
leaf; but cannot strike it to any purpose. The flying squirrel drifts
like a falling leaf; the bat flits like a black rag torn at the edge.
To give power, we must have plumes that can strike, as with the flat of
a sword-blade; and to give _perfect_ power, these must be laid over
each other, so that each may support the one below it. I use the word
below advisedly: we have to strike _down_. The lowest feather is the
one that first meets the adverse force. It is the one to be supported.

[Illustration: FIG. 4.]

Now for the manner of the support. You must all know well the look of
the machicolated parapets in mediaeval castles. You know they are
carried on rows of small projecting buttresses constructed so that,
though the uppermost stone, far-projecting, would break easily under
any shock, it is supported by the next below, and so on, down to the
wall. Now in this figure I am obliged to separate the feathers by white
spaces, to show you them distinctly. In reality they are set as close
to each other as can be, but putting them as close as I can, you get
_a_ or _b_, Fig. 5, for the rough section of the wing, thick towards
the bird's head, and curved like a sickle, so that in striking down it
catches the air, like a reaping-hook, and in rising up, it throws off
the air like a pent-house.

[Illustration: FIG. 5.]

70. The stroke would therefore be vigorous, and the recovery almost
effortless, were even the direction of both actually vertical. But
they are vertical only with relation to the bird's body. In space they
follow the forward flight, in a softly curved line; the downward
stroke being as effective as the bird chooses, the recovery scarcely
encounters resistance in the softly gliding ascent. Thus, in Fig. 5,
(I can only