birds rocked back to an even keel; but although we thought the
action was clearly automatic, and were willing to learn, our
teachers were too far off to show us just how it was done, and
we had to experiment for ourselves.'

Chanute provided his multiple glider with a seat, but, since
each glide only occupied between eight and twelve seconds, there
was little possibility of the operator seating himself. With
the multiple glider a pair of horizontal bars provided rest for
the arms, and beyond these was a pair of vertical bars which the
operator grasped with his hands; beyond this, the operator was
in no way attached to the machine. He took, at the most, four
running steps into the wind, which launched him in the air, and
thereupon he sailed into the wind on a generally descending
course. In the matter of descent Chanute observed the sparrow
and decided to imitate it. 'When the latter,' he says,
'approaches the street, he throws his body back, tilts his
outspread wings nearly square to the course, and on the cushion
of air thus encountered he stops his speed and drops lightly to
the ground. So do all birds. We tried it with misgivings, but
found it perfectly effective. The soft sand was a great
advantage, and even when the experts were racing there was not a
single sprained ankle.'

With the multiple winged glider some two to three hundred glides
were made without any accident either to the man or to the
machine, and the action was found so effective, the principle so
sound, that full plans were published for the benefit of any
experimenters who might wish to improve on this apparatus. The
American Aeronautical Annual for 1897 contains these plans;