Engineers and mathematicians who have written to prove that flying machines were impossible have generally computed the efficiency of aeroplanes moving through the air, on the basis that the lifting effect would be equal to a wind blowing against the plane at the rate at which the air was pressed down by the plane while being driven through the air. According to this system of reasoning, my 4,000 square feet of aeroplanes would have lifted only ·125 lb. per square foot, and in order to have lifted 10,000 lbs. they would have to have had an area twenty times as great. This corresponds exactly with the discrepancy which Professor Langley has found in the formula of Newton. With aeroplanes of one-half the width of those I employed, and with a velocity twice as great, the angle could be much less, and the advantages of continually running on to fresh air would be still more manifest. With a screw thrust of 2,000 lbs., the air pressure on each square foot of the projected area of the screw blades is 21·3 lbs., while the pressure on the entire discs of the screws is 4 lbs. per square foot, which would seem to show with screws of this size, that four blades would be more efficient than two. [Illustration: Fig. 85.--One pair of my compound engines. This engine weighed 310 lbs. and developed 180 H.P., with 320 lbs. of steam per square inch.] The engines, as before stated, are compound (Fig. 85). The area of the high-pressure piston is 20 square inches, and that of the low-pressure piston is 50·26 square inches. Both have a stroke of 12 inches. With a boiler pressure of 320 lbs., the pressure on the low-pressure piston is 125 lbs. to the square inch. This abnormally high pressure in the low-pressure cylinder is due to the fact that there is a very large amount of clearance in the high-pressure cylinder to prevent shock in case water should go over when the machine pitches; moreover, the steam in the high-pressure cylinder is cut off at three-quarters stroke, while the steam in the low-pressure cylinder is cut off at five-eighths stroke. If we should compute the power of these engines with the steam entering at full stroke, without any friction, and with no back pressure on the low-pressure cylinder, the total horse-power would foot up to 461·36 horse-power at the speed at which the engines were run--namely, 375 turns per minute. If we compute the actual power consumed by the screws, by multiplying their thrust, which is probably 2,000 lbs. while they are travelling, by their pitch, 16 feet, and this by the number of turns which they make in a minute, and then divide the product by 33,000, 2,000 × 16 × 375 ---------------- = 363·63, 33,000