TRANSACTIONS OF SECTION G. 267 



U) c»ver flif;ht in a uatiiial wind. The importance of considering the natural 

 motions of aircrnft lias an imjiortance in the design of automatic controls which, 

 in addition to irilierent stability, may be expected as aids to comfort in flying. 

 There is ample evidence in accident statistics to show that unstable aeroplanes 

 iiit" dangerous. 



3. 77(6 Variation of Engine Power with Height. By H. T. Tizard.''' 



In the early days of the war, when it was decided to reduce all officially 

 observed ' performances ' of aeroplanes to a standard atmosphere, it was assunwl 

 that the horse-power of an engine depended on the engine revolutions and the 

 density of the atmosphei'e, and was independent of atmospheric temperature so 

 long as the density remained constant. If the error involved in this assumption 

 is large, the standard method of reducing performances becomes unreliable, and 

 the accuracy of comparisons between model and full scale experiments in 

 aerodynamics is considerably affected. The opposing theory that engine power 

 depends only on the pressure, and not on the density, of the atmosphere has 

 recently found considerable support. In this paper the accumulated evidence 

 from actual flight tests, and from experiments on the ground imder artificial 

 altitude conditions, is examined, and it is concluded that, although the density 

 theory is not quite true, the standard method of reducing performances is satis- 

 factory when extreme accuracy is not desired. For accurate comparisons of full 

 scale and model experiments special corrections must be introduced. 



4. Sound Emission from Airscrews. By Professor G. H. Bryan, 



F.R.S. 



When airscrews were run on the large whirling arm of the Royal Aircraft 

 Establishment the sound emitted showed only faint traces of the low bass note, 

 resembling an organ tone, which is often conspicuous when aeroplanes are 

 flying overhead ; on the contrary, the principal sounds of definite pitch heard 

 were roughly of 400, and in lesser degree 200, vibrations per second, the 

 calculated pitch due to the revolution of the blades being about 70, 40, and 

 30 per second in the several cases. When, however, the screws were mounted 

 in fixed bearings on the spinning tower, and the sound observed in the neigh- 

 bourhood of the plane of rotation, the low bass notes were very conspicuous 

 and agreed closely with the number of beats per second calculated from the 

 revolution of the blades — namely, 50, 55, and 60 for a three-bladed screw and 

 33, 37, and 40 for a two-bladed screw at 1,000, 1,100, and 1,200 revolutions per 

 second. With the two-bladed screw running at the higher speed the octave 

 (say 80 per second) was also conspicuous. 



In the case of an airscrew of which the tip velocity exceeded the velocity 

 of sound (say 1,180 feet per second), an extremely unpleasant crackling sensa- 

 tion was experienced in the neighbourhood of the plane of rotation. 



The observations lead to the conclusion that when nn aeroplane is observed 

 flying sideways the low organ-pipe tones that are observed are due to the 

 direct action of the blades of the screw upon the air, but their intensity decreases 

 as the angular distance of the observer from the plane of rotation increases, 

 thus accounting for the rise and fall of the sound with the rotary oscillations 

 of the aeroplane, as well as for the unfavourable results obtainable in the 

 whirling arm tests. As, however, the pulsations of an eight-cylinder engine 

 agree in frequency with those of a four-blade screw, tones of nearly the same 

 pitch may also be produced by the engine, and these are often heard when an 

 aeroplane is receding. They differ in tone quality from those due to the screw, 

 and the motion of the machine gives rise to a slight difference in pitch. 



In the case in which the tip velocity exceeds the velocity of sound, the 

 disturbances produced in three different positions of the tip may reach the 

 observer at the same instant during a certain portion of each jieriod, and at 

 the beginning and end of this portion the disturbance theoreticolly becomes 



' See En.jinreriiifi. Oct. 17, 1019, p. 527. 



