May 9, 1912] 
and on each of these questions I will say a few 
words. First, as to the theories of stability which 
have been given from time to time. Some of these 
I believe to be correct so far as they go, but none of 
them are anything like complete, since they are all 
based on the pressures and variations of pressure act- 
ing on the up-stream surfaces of wings and omit the 
variations due to the eddy formation which goes on 
on the down-stream side. 
Before proceeding further, it will be as well to 
define what I mean by stability in connection with 
flight. A flying body is stable if, when acted on by 
a propulsive force and the reactions of the air (but not 
steered), any small angular velocity imposed about a 
horizontal axis tends to die out, and any small dis- 
placement about a vertical axis to reach a constant 
RES! Bt PROPULSIVe FORCE 
} 
| 
\ 
OF GRAVITY 
(Q@) DIRECTION OF RESULTANT 
FORCE. IN GUSTS 
ROPULSIVE 
PROPULSIV 
FORCE JM 
RESISTANCE RESISTANCE <— >= FORCE 
DOOWNWARD 
FORCE 
A 
x =-LIFT 
rre RESISTANCES AND 
LIFT ARE INCREASED TO 
HERE RESISTANCE AND LiFT 
ARE REDUCED TO J*eOFr THEIR 
ORIGINAL VALUE, THE PROPULSIVE TIMES THEIR VALUE 
FORCE BEING DOUBLED IN| STILL AIR. 
THE PROPULSIVE FCRCE 
BEING HALVED 
(6) Gust HALF SPEED OF 
“FLIGHT OVERTAKING 
MACHINE 
DOWNWARD FORCE | 
=-LIFT 
(C) GUST HALE SPEED OF FLIGHT 
MEETING MACHINE 
Fic. 5. 
value. Or, in other words, any accidental motion of | 
the nature of pitching or rolling must tend to dis- 
NATURE 
255 
of the wings (that is, in the angle a) which they can 
produce in one period is inconsiderable, and the 
stability or instability depends chiefly on the distribu- 
tion of pressure on the up-stream surfaces, but the 
case is very different when the machine is passing 
through variable currents and the angle at which the 
air meets the wings is liable to large and rapid 
changes. The alterations in the arrangement of the 
pressures on the back surfaces are then much greater 
and take longer to go through their phases—long 
enough, in fact, to make the process of correction 
exceedingly baffling. 
That flying machines should be unstable in ordinary 
circumstances is really of very little consequence. 
The same objection applies to walking. No conscious 
effort, however, is required to keep upright on terra 
firma, but on the deck of a small vessel in seaway we 
all know that sea legs are only got by practice, often 
| involving many falls. 
The flying machine in gusty weather is much in 
the same condition, but the falls have more serious 
consequences. 
I think it very unlikely that any type of flying 
machine will be evolved which, without guidance, will 
_ be safe in bad weather, but it is quite possible that 
appear, while an arbitrary twist to the right or left 
must put the machine on a new, but straight, course. 
Technically, stability is compatible with the | 
presence of forces which produce increasing oscilla- | 
tions as the result of disturbance; but for the present 
purpose not only must the average force so called 
into play be a restituent force, but the disturbing 
motions must also tend to die out. The oscillations, 
in fact, must be damped, and not maintained. 
None of the flying machines at present in use are 
stable in the sense in which the word is here used, 
but in the ordinary conditions of flight the eddies 
formed behind the wings are small and their period 
of formation so rapid that the change in the attitude 
NO. 2219, VOL. 89] 
BO WAWAR DEGREE | the necessary corrections should be applied by an 
automatic device, and if flight is to be anything but 
a fair-weather pastime, something of the kind will 
probably be found necessary. 
What is required is an apparatus which will so 
trim the wings as to keep the machine related in a 
definite manner, firstly to the true vertical, and 
secondly to the direction of the resultant force at the 
time. 
The various ways in which this could be done might 
furnish subjects for several lectures, and I will only 
say here that the many proposals which have been 
made to use pendulums or gyroscopes to act directly 
on the correcting mechanism are certainly bound to 
fail. : 
It is essential to the success of any automatic 
control that the forces called into play to make the 
corrections of trim should not react on the director 
of those forces, whether this is a pendulum or gyro- 
scope or any other equivalent device. The only 
instance in which this condition has been fulfilled is 
the ‘‘steady platform” of the late Mr. Beauchamp 
Tower. In this Mr. Tower caused a _ gyroscope 
(which, in effect, was a pendulum with a very long 
period) to direct an axial jet of water on a group of 
openings connected by pipes to a series of rams in 
such a way that if the openings did not face the jet 
symmetrically water flowed into one or other of the 
pipes, and so altered the position of the openings until 
symmetry was restored, the restituent force having 
no tendency to alter the direction of the axial jet. 
There may be other methods of attaining the same 
object in the case of wing-trimming or control for 
flying machines, but any device in which the correct- 
ing force tends to alter the position of the corrector 
is more likely to do karm than good. 
The question of stability also becomes important 
when the flying machine is coming to the ground. 
In alighting, the machine either has to touch the 
ground at full speed and trust to retardation, sup- 
plied chiefly by the ground, for coming to rest, or it 
must alter the wing attitude with reference to the 
path so as to experience a greater resistance for a 
given lift. This latter method is adopted by birds 
when pitching on the ground, and in their case at 
the last moment is generally supplemented by flapping 
the wings when the velocity is so much reduced that 
the greatest lift the wing area is good for will not 
sustain their weight. Birds when pitching on any 
