to operate but keeps his muscles rigid and without disturbing the motion deliberately produces a condition in which the aeroplane has to control itself in gusts of small size. If the motion be stable, no great changes will occur as a result of the pilot's relinquishing of control. A small amount of pitching, rolling, yawing and side-slipping, etc., will occur but on the whole the speed of flight and the angle of incidence will remain at the same value as at the beginning; the wings will not change their angle of bank greatly nor the turning increase or decrease.
An instability, and in contradistinction to stability there are many instabilities possible, will magnify the effects of a gust with greater or less rapidity and the motion will depart from the initial state to some other stable state. It rarely happens that this second state is a comfortable one. An aeroplane which is unstable in normal flight will usually be stable upside down and may be so stable in that position as to be uncontrollable. The time taken to pass from one state to another is often only a matter of a few seconds, rarely as long as a few minutes.
In the very early days of flying the problem of getting into the air at all took first place in importance. The aviators of 1908-10 kept a very close watch on the weather and one of them had a standard test for satisfactory conditions. Standing with his feet apart, he dropped a feather from the level of his shoulders and if it fell outside his feet the wind was too great for flying. The record of these early years and the shortness of life of the aviators are sufficient testimony to the consequences of the extreme forms of instability. The revolutionary step which made it possible to keep the air for an hour instead of a few minutes was made by the Wright brothers when they intro- duced wing warping as a lateral control; there is little reason to doubt the statement that flying still remained an acrobatic feat. A study of the technical papers of the period 1908-14 will show how slowly the idea of banking * an aeroplane entered into the development of aviation. It is noted in March 1912 as a possible cause of accident that the pilot " is reported to have endeavoured to rise when making a turn." Not until April 1913 do we find vertical banking by Chevillard followed by upside-down flying and looping by Pegoud in Sept. of that year.
A prominent place in the technical journals was devoted to accidents and a perusal of these shows that all types were liable to fail as late as 1913. A series of accidents to monoplanes occurred in Britain and their flight was suppressed temporarily in Sept. 1912, whilst a committee was formed to investigate causes and to suggest lines of development. The findings of this committee 2 have had a marked influence on British aviation and the paragraph relating to stability is here quoted:
" The Committee desire to urge the importance of the general investigation into the stability of aeroplanes, whether monoplanes or biplanes. The experimental data at present available are not sufficient to allow a complete theory to be formulated. It is under- stood, however, that the work of the Advisory Committee has now been carried to the stage at which the problem can be attacked with hope of success, provided that the necessary facilities a large wind channel in a sufficiently big enclosed space be put at their disposal, and the Committee recommend that the Advisory Committee be asked to continue the further investigation into the stability of the aeroplane as a matter of great urgency, and more especially to exam- ine the question of inherent lateral stability, suggestions towards the solution of which have been given by the experiments of Lanchester and the calculations of Bryan."
The investigation here started led directly to the stability experiments on REi and BE2, a combination of full-scale flights at the Royal Aircraft Factory and model and theoretical preparatory work at the National Physical Laboratory. Before dealing with the results, a return to early times will be made to indicate the position of the theory of stability.
Up to the end of 1909 the chief writers on the stability of the aeroplane were Bryan, 3 Ferber, 4 ' Lanchester, 5 and Soreau. 6
1 Flight, Feb. 17 1912.
2 Report of Dptl. Comm. on Accidents to Monoplanes, 1912 (ed. 6506), p. 9.
8 Bryan and Williams, "The Longitudinal Stability of Aerial Gliders," Proc. R. S., vol. Ixxiii., 1904, p. IOO.
4 V Aviation. 6 A erodonelics, Lanchester.
- Socicte des Ingenieurs Civils de France, and in a volume: " Etat
actuel et avenir de 1'aviation."
The most complete method was that by Bryan. The papers all advanced the study of the subject in some measure but the appearance in 1911 of Bryan's book Stability in Aviation laid the foundations of the subject as now known to us. About the same time other workers were entering the field, amongst whom may be mentioned Knoller, 7 Bothezat 8 and Reiszner. 9 From that time the theory of stability has been far ahead of practice. Developments have been made to cover circling flight, disturbed motion and the effects of gusts, but all are natural extensions of the theory of dynamical stability as given by Routh and applied by Bryan to the aeroplane. There is little doubt that further extensions will be made as required, but the immediate need is the devotion of existing knowledge to practice to a far greater extent than has hitherto occurred. As in other branches of research the World War has had an adverse effect in curtailing opportunities for reasoned progress.
In March 1913 a report *was issued showing the possible applica- tions of the theory of stability in numerical detail. The mathema- tical analysis cannot be useful unless a number of quantities, known as resistance derivatives, can be obtained from experiment. The re- port in question represents the first systematic attempt to apply experimental research to the evaluation of the quantities required for application of the theory. A discussion is given of the meaning and origin, from the physical side, of the resistance derivatives and rough estimates were made as to the ranges of the quantities for then existing aeroplanes. For one condition of flight more accurate data were obtained and a table of some 18 derivatives deduced cover- ing the longitudinal and lateral stabilities of an aeroplane in normal flight. There are a number of approximations which assist in the understanding of the relation between cause and effect which were of importance in the infancy of the subject.
By such a preliminary examination on the model scale, the phenomena to be looked for on the full scale were clearly defined. The now well-known " phugoid " oscillation was then unobserved and only indicated by calculations: It is indeed possible that up to that date longitudinal stability did not exist apart from the very special design of Dunne. The mathematical theory indicated quite clearly that special shapes were unnecessary and that aeroplanes of more usual form could be made stable by attention to the dis- position of weights and the arrangements of the aerofoil surfaces. In particular, the importance of a dihedral angle on the wings and an adequate fin and rudder were shown in relation to lateral stability.
In the course of the 12 months which followed great progress was made; in a series of papers " from the National Physical Laboratory, the effect of varying essential quantities, such as the centre of gravity of the aeroplane, the amount of area of the tail plane, the extent of the dihedral angle, rudder and fin area, etc., was examined in detail. It was shown that partial experiments on lateral stability would fail since there is a relation between the dihedral angle on the wings and the appropriate fin and rudder area.
Further, the exact method of inherent adjustment of an aeroplane to gusts was shown and the details of flight of a longitudinally stable aeroplane in a natural wind obtained. This was done not only for uncontrolled but for controlled flight. By the summer of 1914 the investigation of the effect of natural properties of an aeroplane on mechanical devices for controlling it were being envisaged, but the outbreak of the war broke the continuity and the subject still remains at that point of theoretical development.
In the meantime full-scale experiments were being made at the Royal Aircraft Factory. 12 A few extracts from these reports are of historical value and are here reproduced:
" Although completely controllable under all circumstances by means of the elevator, it has been found that the BE2A aeroplane, fitted with the old tail plane (TP l) was not stable with the elevator free or even locked. . . . Two methods of experimenting have been adopted: (a) Variation of the section and plane form of tail, (b) Variation of the position of the centre of gravity of the aeroplane relative to the position of the wings."
" Experiment (i) with tail (TPl). Area of tail 61 sq. feet. Centre of gravity at 0-38 of the mean chord behind the leading edge. At a height of about 2,000 ft. the elevator control lever was held in a fixed position. After a short time, a steady dive developed, which was allowed to continue so long as it was considered safe by the flier, in this instance during a flight of about 500 yards. There was no ten- dency for the path to revert to the horizontal. ... It was found that there was just as much tendency for a steady rearing to be developed as a dive. ..."
7 "Cber Langstabilitat der Drachenflugzeuge," Ztschr. fur Flug- technik und Motorluftschifffahrt, July and Aug. 1911.
8 tude de la Stabilite de I'aeroplane (Dimod & Pinet. 191 1).
9 Ztschr. fiir Flugtechnik und Motorluftschifffahrt, Feb. 10 1912.
10 R and M, No. 77, Advisory Committee Tor Aeronautics, 1912-3.
11 Reports, Advisory Committee for Aeronautics, 1913-4, pp. 216-286.
12 Reports, Advisory Committeefor Aeronautics, 1913-4, pp. 385-394.
