Recent experiments have led me in two directions—both a good deal governed by considerations of construction, which, in any case, is always poor.

In one direction I improved the steadiness of flight, but was very dissatisfied with the angle of fall; in the other direction I improved, by diminishing, the angle of fall, but was extremely dissatisfied with the steadiness of flight.

In trying different wing profiles—such as could be easily made by bending a piece of thin cardboard, I found the steadiest forms were those that required the centre of gravity to be furthest, forward, but they did not give good angles, while the best angles were given by shapes which, wherever the centre of gravity was, would always pitch forward unless controlled by a large and well turned up tail.

It appeared probable that in the steady form with the bad angle all the lift came from a limited part of the wing near the forward edge, while the bulk of the after part of the surface acted more or less as a steadying rudder or tail.