lso
described his own observations and submitted some computations to
account for the observed facts. These computations were correct as
far as they went, but they were scanty. It was, for instance, shown
convincingly by analysis that a gull weighing 2.188 pounds, with a total
supporting surface of 2.015 square feet, a maximum body cross-section
of 0.126 square feet and a maximum cross-section of wing edges of 0.098
square feet, patrolling on rigid wings (soaring) on the weather side of
a steamer and maintaining an upward angle or attitude of 5 degrees to 7
degrees above the horizon, in a wind blowing 12.78 miles an hour, which
was deflected upward 10 degrees to 20 degrees by the side of the steamer
(these all being carefully observed facts), was perfectly sustained at
its own "relative speed" of 17.88 miles per hour and extracted from
the upward trend of the wind sufficient energy to overcome all the
resistances, this energy amounting to 6.44 foot-pounds per second.

Great Power of Gulls.

It was shown that the same bird in flapping flight in calm air, with an
attitude or incidence of 3 degrees to 5 degrees above the horizon and
a speed of 20.4 miles an hour was well sustained and expended 5.88
foot-pounds per second, this being at the rate of 204 pounds sustained
per horsepower. It was stated also that a gull in its observed
maneuvers, rising up from a pile head on unflapping wings, then plunging
forward against the wind and subsequently rising higher than his
starting point, must either time his ascents and descents exactly with
the variations in wind velocities, or must meet a wind billow rotating
on a horizontal axis and come to a poise on its crest, thus availing of
an ascending trend.

But the observations failed to demonstrate that the variations of the
wind gusts and the movements of the bird were absolutely synchronous,
and it was conjectured that the peculiar shape of the soaring wing of
certain birds, as differentiated from the flapping wing, might, when
experimented upon, hereafter account for the performance.

Mystery to be Explained.

These computations, however satisfactory they were for the speed of
winds observed, failed to account for the observed spiral soaring of
buzzards in very light winds and the writer was compelled to confess:
"Now, this spiral soaring in steady breezes of 5 to 10 miles per hour
which are apparently horizontal, and through which the bird maintains an
average speed of abou