he very
day that the machine was completed the wind blew from 25 to 30 miles per
hour, and we took it out for a trial as a kite. We found that while it
was supported with a man on it in a wind of about 25 miles, its angle
was much nearer 20 degrees than three degrees. Even in gusts of 30 miles
the angle of incidence did not get as low as three degrees, although the
wind at this speed has more than twice the lifting power of a 21-mile
wind. As winds of 30 miles per hour are not plentiful on clear days, it
was at once evident that our plan of practicing by the hour, day after
day, would have to be postponed. Our system of twisting the surfaces
to regulate the lateral balance was tried and found to be much more
effective than shifting the operator's body. On subsequent days, when
the wind was too light to support the machine with a man on it, we
tested it as a kite, working the rudders by cords reaching to the
ground. The results were very satisfactory, yet we were well aware that
this method of testing is never wholly convincing until the results are
confirmed by actual gliding experience.

We then turned our attention to making a series of actual measurements
of the lift and drift of the machine under various loads. So far as we
were aware, this had never previously been done with any full-size
machine. The results obtained were most astonishing, for it appeared
that the total horizontal pull of the machine, while sustaining a weight
of 52 lbs., was only 8.5 lbs., which was less than had previously been
estimated for head resistance of the framing alone. Making allowance for
the weight carried, it appeared that the head resistance of the framing
was but little more than 50 per cent. of the amount which Mr. Chanute
had estimated as the head resistance of the framing of his machine. On
the other hand, it appeared sadly deficient in lifting power as compared
with the calculated lift of curved surfaces of its size. This deficiency
we supposed might be due to one or more of the following causes:--(1)
That the depth of the curvature of our surfaces was insufficient, being
only about one in 22, instead of one in 12. (2) That the cloth used in
our wings was not sufficiently air-tight. (3) That the Lilienthal tables
might themselves be somewhat in error. We decided to arrange our machine
for the following year so that the depth of the curvature of its
surfaces could be varied at will and its covering air-proofed.

Our attention