The National Geographic Magazine 



faces (or rather surfaces substantially 

 horizontal, as in ordinary kites), but the 

 framework is admirably adapted for the 

 support of such surfaces. Horizontal 

 aeroplanes have much greater supporting 

 power than similar surfaces obliquely ar- 

 ranged, and I have made many experi- 

 ments to combine horizontal surfaces 

 with winged cells with greatly improved 

 results, so far as lifting power is con- 

 cerned. But there is always an element 

 of instability in a horizontal aeroplane, 

 especially if it is of large size, whereas 

 kites composed exclusively of winged 

 cells are wonderfully steady in the air 

 under varying conditions, though defi- 

 cient in lifting power ; and the kites com- 

 posed of the largest number of winged 

 cells seem to be the most stable in the air. 



In the case of an aeroplane of any kind 

 the center of air pressure rarely coincides 

 with the geometrical center of surface, 

 but is usually nearer the front edge than 

 the middle. It is liable to shift its posi- 

 tion, at the most unexpected times, on ac- 

 count of some change in the inclination 

 of the surface or the direction of the 

 wind. The change is usually small in 

 steady winds, but in unsteady winds great 

 and sudden changes often occur. 



The extreme possible range of fluctua- 

 tion is of course, from the extreme front 

 of the aeroplane to the rear, or vice versa, 

 and the possible amount of change, there- 

 fore, depends upon the dimensions of 

 the aeroplane, especially in the fore- 

 and-aft direction. With a large aero- 

 plane the center of pressure may sud- 

 denly change to such an extent as to en- 

 danger the equilibrium of the whole ma- 

 chine, whereas with smaller aeroplanes, 

 especially those having slight extension 

 in the fore-and-aft direction, the change, 

 though proportionally as great, is small 

 in absolute amount. Where we have a 

 multitude of small surfaces well sepa- 

 rated from one another, as in the tetrahe- 

 dral construction, it is probable that the 

 resultant center of pressure for the whole 

 kite can shift to no greater extent than 

 the centers of pressure of the individual 

 surfaces themselves. It is, therefore, ex- 



tremely unlikely that the equilibrium of 

 a large kite could be endangered by the 

 shifting of the centers of pressure in 

 small surfaces within the kite. This may 

 be the cause of the automatic stability of 

 large structures built of small tetrahedral 

 cells. If so, one principle of stability 

 would be : Small surfaces, well separated, 

 and many of them. The converse propo- 

 sition would then hold true if we desired 

 to produce instability and a tendency to 

 upset in a squall, namely : Large surfaces, 

 continuous, and few of them. 



HARGRAVI^ BOX KITES AND TETRAHEDRAI, 

 KITES COMPARED 



Another source of danger with large 

 continuous surfaces is the fact that a sud- 

 den squall may strike the kite on one side, 

 lifting it up at that side and tending to 

 upset it; but the compound tetrahedral 

 structure is so porous that a squall passes 

 right through and lifts the other side as 

 well as the side first struck ; so that the 

 kite has not time to be upset before the 

 blow on one side is counterbalanced by a 

 blow on the other. I have flown a Har- 

 grave box kite simultaneously with a 

 large kite of many tetrahedral cells in 

 squally weather for the purpose of com- 

 paring them under similar conditions. 

 The tetrahedral structure often seemed to 

 shiver when struck by a sudden squall, 

 whereas the box kite seemed to be liable 

 to a swaying or tipping motion that 

 would be exceedingly dangerous in a 

 structure of large size forming part of a 

 flying-machine. 



Another element of stability in the 

 tetrahedral structure lies in the fact that 

 the winged surfaces are elevated at a 

 greater angle above the horizon than 45°. 



Supposing the wings of a cell to be 

 opened out until they are nearly flat, or 

 at least until they each make a compara- 

 tively small angle with the horizon — say 

 20° — then, if from any cause the cell 

 should tip so as to elevate one wing 

 (say to 25°) and depress the other (say 

 to 15°), the lifting power of the wind 

 will be increased upon the elevated wing 

 and diminished on the depressed wing; 



