ON THE 74oz. COMFEESSED-AIR FLYING-MACHINE. 255 



has. The large area might be considered a defect, but when we 

 consider that it consists only of a few sticks and tissue paper, and 

 that the atmosphere is not by any means crowded with flying- 

 machines, the objection ceases to have much weight. 



On Diagram 3, the apparent trajectories of trials 5 an& 6 are 

 shown. The machine in trial 5 turned up and almost stopped, 

 but resumed its course when the preponderance of the forward 

 part brought it horizontal again. A lump of lead was put on the 

 end of the breaking stick for Trial 6. This shifted the centre of 

 gravity one inch further forward and produced the undulatory 

 flight that is shown in the drawing. Each observation adds fresh 

 weight to the assumption that the true position of the centre of 

 gravity for a continuous rectangular surface is situated between 

 •25 and *2 of the length from the forward end. 



After trial 6, the machine was attached to the chronograph to 

 see what the receiver pressure was at the 38th double vibration, 

 the number registered by the counter ; but after making three or 

 four flaps at the rate of about 200 per minute, all the paper was 

 dashed out of the wings, and the port link lugs were dragged out 

 of the cylinder cover. But again knowledge is gained from failure; 

 we learn that the chronographic test of the wing speed of the 

 stationary machine is no guide to the speed of the flying-machine, 

 as the rapidly flapping wing creates a vacuum behind it of suffici- 

 ently low pressure to allow the return stroke of the wing to pass 

 so quickly through it that the shock of the wing against the air 

 at the other side of the vacuum is strong enough to destroy the 

 paper. The efficiency of the wings during the free flight is not 

 impaired by this cause as every stroke is taken in new and solid 

 air : and the wing speed is obviously not in excess of 120 per 

 minute. 



Mr. J. A. Pollock has again ascertained for the writer the 

 theoretical work done by the engine, and he finds 2,720 foot- 

 pounds were used in reducing the pressure from 150 to 70 pounds 

 per square inch with 38 double vibrations. This is terribly wasteful 

 of power when we compare it with the india-rubber driven 

 machines. 



The two observations with the compressed-air machines confirm 

 the two cross-bow model experiments recorded on page 73 of the 

 Royal Society's Proceedings 1889, which are added to the table 

 for comparison. The cross bow takes 1\ lbs. to bend it 12 inches. 

 It appears that the same power propels different weights the same 

 distance provided the area per pound weight is the same. 



