Hiram S. Maxim

Fig. 74.—Photograph of a model of my machine, showing the fore and aft horizontal rudders and the superposed aeroplanes.

In my early “whirling table”[5] experiments, the aeroplanes used were from 6 inches to 4 feet in width. They were for the most part made of thin pine, being slightly concave on the underneath side and convex on the top, both the fore and aft edges being very sharp. I generally mounted them at an angle of 1 in 14[6]—that is, in such a position that in advancing 14 feet they pressed the air down 1 foot. With this arrangement, I found that with a screw thrust of 5 lbs. the aeroplane would lift 5 × 14, or 70 lbs., while if the same plane was mounted at an angle of 1 in 10, the lifting effect was almost 50 lbs. (5 × 10). This demonstrated that the skin friction on these very sharp, smooth and well-made aeroplanes was so small a factor as not to be considered. When, however, there was the least irregularity in the shape of the aeroplane, the lifting effect, when considered in terms of screw thrust, was greatly diminished. With a well-made wooden plane placed at an angle of 1 in 14, I was able to carry as much as 113 lbs. to the H.P., whereas with an aeroplane consisting of a wooden frame covered with a cotton fabric (Fig. 75), I was only able to carry 40 lbs. to the H.P.[7]

These facts taken into consideration with my other experiments with large aeroplanes, demonstrated to my mind that it would not be a very easy matter to make a large and efficient aeroplane. If I obtained the necessary rigidity by making it of boards, it would be vastly too heavy for the purpose, while if I obtained the necessary lightness by making the framework of steel and covering it with a silk or cotton fabric in the usual way, the distortion would be so great that it would require altogether too much power to propel it through the air. I therefore decided on making a completely new form of aeroplane. I constructed a large steel framework arranged in such a manner that the fore and aft edges consisted of tightly drawn steel wires. This framework was provided with a number of light wooden longitudinal trusses, similar to those shown in Fig. 76. The bottom side was then covered with balloon fabric secured at the edges, and also by two longitudinal lines of lacing through the centre. It was stretched very tightly and slightly varnished, but not sufficiently to make it absolutely air-tight. The top of this framework was covered with the same kind of material, but varnished so as to make it absolutely airtight. The top and bottom were then laced together forming very sharp fore and aft edges, and the top side was firmly secured to the light wooden trusses before referred to. Upon running this aeroplane, I found that a certain quantity of air passed through the lower side and set up a pressure between the upper and lower coverings. The imprisoned air pressed the top covering upward, forming longitudinal corrugations which did not offer any perceptible resistance to the air, whereas the bottom fabric, having practically the same pressure on both sides, was not distorted in the least. This aeroplane was found to be nearly as efficient as it would have been had it been carved out of a solid piece of wood. It will be seen by the illustration that this large or main aeroplane is practically octagonal in shape, its greatest width being 50 feet, and the total area 1,500 square feet.

EXPERIMENTS WITH A LARGE MACHINE.

Upon running my large machine over the track (Fig. 77) with only the main aeroplane in position, I found that a lifting effect of 3,000 to 4,000 lbs. could be obtained with a speed of 37 to 42 miles an hour. It was not always an easy matter to ascertain exactly what the lifting effect was at a given speed on account of the wind that was generally blowing. Early in my experiments, I found if I ran my machine fast enough to produce a lifting effect within 1,000 lbs. of the total weight of the machine, that it was almost sure to leave the rails if the least wind was blowing. It was, therefore, necessary for me to devise some means of keeping the machine on the track. The first plan tried was to attach some very heavy cast-iron wheels weighing with their axle-trees and connections about 1¹/₂ tons. These were constructed in such a manner that the light flanged wheels supporting the machine on the steel rails could be lifted 6 inches above the track, leaving the heavy wheels still on the rails for guiding the machine. This arrangement was tried on several occasions, the machine being run fast enough to lift the forward end off the track. However, I found considerable difficulty in starting and stopping quickly on account of the great weight, and the amount of energy necessary to set such heavy wheels spinning at a high velocity. The last experiment with these wheels was made when a head wind was blowing at the rate of about 10 miles an hour. It was rather unsteady, and when the machine was running at its greatest velocity, a sudden gust lifted not only the front end, but also the heavy front wheels completely off the track, and the machine falling on soft ground was soon blown over by the wind.

I then provided a safety track of 3 × 9 Georgia pine placed about 2 feet above the steel rails, the wooden track being 30 feet gauge and the steel rails 9 feet gauge (Fig. 77). The machine was next furnished with four extra wheels placed on strong outriggers and adjusted in such a manner that when it had been lifted 1 inch clear of the steel rails, these extra wheels would engage the upper wooden track.[8]

When fully equipped, my large machine had five long and narrow aeroplanes projecting from each side. Those that are attached to the sides of the main aeroplanes are 27 feet long, thus bringing the total width of the machine up to 104 feet. The machine is also provided with a fore and an aft rudder made on the same general plan as the main aeroplane. When all the aeroplanes are in position, the total lifting surface is brought up to about 6,000 square feet. I have, however, never run the machine with all the planes in position. My late experiments were conducted with the main aeroplane, the fore and aft rudders, and the top and bottom side planes in position, the total area then being 4,000 square feet. With the machine thus equipped, with 600 lbs. of water in the tank and boiler and with the naphtha and three men on board, the total weight was a little less than 8,000 lbs. The first run under these conditions was made with a steam pressure of 150 lbs. to the square inch, in a dead calm, and all four of the lower wheels remained constantly on the rails, none of the wheels on the outriggers touching the upper track. The second run was made with 240 lbs. steam pressure to the square inch. On this occasion, the machine seemed to vibrate between the upper and lower tracks. About three of the top wheels were engaged at the same time, the weight on the lower steel rails being practically nil. Preparations were then made for a third run with nearly the full power of the engines. The machine was tied up to a dynamometer (Fig. 78), and the engines were started with a pressure of about 200 lbs. to the square inch. The gas supply was then gradually turned on with the throttle valves wide open; the pressure soon increased, and when 310 lbs. was reached, the dynamometer showed a screw thrust of 2,100 lbs.,[9] but to this must be added the incline of the track which amounts to about 64 lbs. The actual thrust was therefore 2,164 lbs. In order to keep the thrust of the screws as nearly constant as possible, I had placed a small safety valve—³/₄-inch—in the steam pipe leading to one of the engines. This valve was adjusted in such a manner that it gave a slight puff of steam at each stroke of the engine with a pressure of 310 lbs. to the square inch, and a steady blast at 320 lbs. to the square inch. As the valves and steam passages of these engines were made very large, and as the piston speed was not excessive, I believed if the steam pressure was kept constant that the screw thrust would also remain nearly constant, because as the machine advances and the screws commence to run slightly faster, an additional quantity of steam will be called for and this could be supplied by turning on more gas. When everything was ready, with careful observers stationed on each side of the track, the order was given to let go. The enormous screw thrust started the machine so quickly that it nearly threw the engineers off their feet, and the machine bounded over the track at a great rate. Upon noticing a slight diminution in the steam pressure, I turned on more gas, when almost instantly the steam commenced to blow a steady blast from the small safety valve, showing that the pressure was at least 320 lbs. in the pipes supplying the engines with steam. Before starting on this run, the wheels that were to engage the upper track were painted, and it was the duty of one of my assistants to observe these wheels during the run, while another assistant watched the pressure gauges and dynagraphs (Fig. 79). The first part of the track was up a slight incline, but the machine was lifted clear of the lower rails and all of the top wheels were fully engaged on the upper track when about 600 feet had been covered. The speed rapidly increased, and when 900 feet had been covered, one of the rear axle-trees, which were of 2-inch steel tubing, doubled up (Fig. 80), and set the rear end of the machine completely free. The pencils ran completely across the cylinders of the dynagraphs and caught on the underneath end. The rear end of the machine being set free, raised considerably above the track and swayed. At about 1,000 feet, the left forward wheel also got clear of the upper track and shortly afterwards, the right forward wheel tore up about 100 feet of the upper track. Steam was at once shut off and the machine sank directly to the earth imbedding the wheels in the soft turf (Figs. 81 and 82) without leaving any other marks, showing most conclusively that the machine was completely suspended in the air before it settled to the earth. In this accident, one of the pine timbers forming the upper track went completely through the lower framework of the machine and broke a number of the tubes, but no damage was done to the machinery except a slight injury to one of the screws (Fig. 83).

In my experiments with the small apparatus for ascertaining the power required to perform artificial flight, I found that the most advantageous angle for my aeroplane was 1 in 14, but when I came to make my large machine, I placed my aeroplanes at an angle of 1 in 8 so as to be able to get a great lifting effect at a moderate speed with a short run. In the experiments which led to the accident above referred to, the total lifting effect upon the machine must have been at least 10,000 lbs. All the wheels which had been previously painted and which engaged the upper track were completely cleaned of their paint and had made an impression on the wood, which clearly indicated that the load which they had been lifting was considerable.[10] Moreover, the strain necessary to double up the axle-trees was fully 1,000 lbs. each, without considering the lift on the forward axle-trees which did not give way but broke the upper track.

The advantages arising from driving the aeroplanes on to new air, the inertia of which has not been disturbed, are clearly shown in these experiments. The lifting effect of the planes was 2·5 lbs. per square foot. A plane loaded at this rate will fall through the air with a velocity of 22·36 miles per hour, according to the formula v200 × P = V. But as the planes were set at an angle of 1 in 8, and as the machine travelled at the rate of 40 miles an hour, the planes only pressed the air downwards 5 miles an hour (40 ÷ 8 = 5). A fall of 5 miles an hour without advancing would only exert a pressure of ·125 lb. per square foot, according to the formula (V² × ·005 = P).[11]

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.

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 × 37533,000 = 363·63,

we find that we have 363·63 horse-power in actual effect delivered on the screws of the machine, which shows that there is rather less than 22 per cent. loss in the engines, due to cutting off before the end of the stroke, to back pressure, and to friction. The actual power applied to the machine being 363·63 horse-power, it is interesting to know what becomes of it. When the machine has advanced 40 miles (which it would do in an hour), the screws have travelled 68·1 miles (375 × 16 × 605,280) = 68·1; therefore, 150 horse-power is wasted in slip, and 213·63 horse-power consumed in driving the machine through the air. Now, as the planes are set at an angle of 1 in 8, the power actually used in lifting the machine is 133·33, and the loss in driving the body of the machine, its framework and wires through the air is 90·30 horse-power.

THE ADVANTAGES AND DISADVANTAGES OF VERY NARROW PLANES.

My experiments have demonstrated that relatively narrow aeroplanes lift more per square foot than very wide ones; but as an aeroplane, no matter how narrow it may be, must of necessity have some thickness, it is not advantageous to place them too near together. Suppose that aeroplanes should be made ¹/₄-inch thick, and be superposed 3 inches apart—that is, at a pitch of 3 inches—one-twelfth part of the whole space through which these planes would have to be driven would be occupied by the planes themselves, and eleven-twelfths would be air space (Fig. 86). If a group of planes thus mounted should be driven through the air at the rate of 36 miles an hour,[12] the air would have to be driven forward at the rate of 3 miles an hour, or else it would have to be compressed, or spun out, and pass between the spaces at a speed of 39 miles an hour. As a matter of fact, however, the difference in pressure is so very small that practically no atmospheric compression takes place. The air, therefore, is driven forward at the rate of 3 miles an hour, and this consumes a great deal of power; in fact, so much that there is a decided disadvantage in using narrow planes thus arranged.

In regard to the curvature of narrow aeroplanes, I have found that if one only desires to lift a large load in proportion to the area, the planes may be made very hollow on the underneath side; but when one considers the lift in terms of the screw thrust, I find it advisable that the planes should be as thin as possible, and the underneath side nearly flat. I have also found that it is a great advantage to arrange the planes after the manner shown in Fig. 87. In this manner the sum of all the spaces between the planes is equal to the whole area occupied by the planes; consequently, the air neither has to be compressed, spun out, nor driven forward. I am, therefore, able by this arrangement to produce a large lifting effect per square foot, and, at the same time, to keep the screw thrust within reasonable limits.

A large number of experiments with very narrow aeroplanes have been conducted by Mr. Horatio Philipps at Harrow, in England. Fig. 88 shows a cross-section of one of Mr. Philipps’ planes. Mr. Philipps is of the opinion that the air, in striking the top side of the plane, is thrown upwards in the manner shown, and a partial vacuum is thereby formed over the central part of the plane, and that the lifting effect of planes made in this form is therefore very much greater than with ordinary narrow planes. I have experimented with these “sustainers” (as Mr. Philipps calls them) myself, and I find it is quite true that they lift in some cases as much as 8 lbs. per square foot,[13] but the lifting effect is not produced in the exact manner that Mr. Philipps seems to suppose. The air does not glance off in the manner shown. As the “sustainer” strikes the air two currents are formed, one following the exact contour of the top, and the other that of the bottom. These two currents join and are thrown downwards, as relates to the “sustainer,” at an angle which is the resultant of the angles at which the two currents meet. These “sustainers” may be made to lift when the front edge is lower than the rear edge, because they encounter still air, and leave it with a downward motion.

In my experiments with narrow superposed planes, I have always found that with strips of thin metal made sharp at both edges and only slightly curved, the lifting effect, when considered in terms of screw thrust, was always greater than with any arrangement of the wooden aeroplanes used in Philipps’ experiments. It would, therefore, appear that there is no advantage in the peculiar form of “sustainer” employed by this inventor.

If an aeroplane be made perfectly flat on the bottom side and convex on the top, and be mounted in the air so that the bottom side is exactly horizontal, it produces a lifting effect no matter in which direction it is run, because, as it advances, it encounters stationary air which is divided into two streams. The top stream being unable to fly off at a tangent when turning over the top curve, flows down the incline and joins the current which is flowing over the lower horizontal surface. The angle at which the combined stream of air leaves the plane is the resultant of these two angles; consequently, as the plane finds the air in a stationary condition, and leaves it with a downward motion, the plane itself must be lifted. It is true that small and narrow aeroplanes may be made to lift considerably more per square foot of surface than very large ones, but they do not offer the same safeguard against a rapid descent to the earth in case of a stoppage or breakdown of the machinery. With a large aeroplane properly adjusted, a rapid and destructive fall to the earth is quite impossible.

THE EFFICIENCY OF SCREW PROPELLERS, STEERING, STABILITY, &c.

Before I commenced my experiments at Baldwyn’s Park, I attempted to obtain some information in regard to the action of screw propellers working in the air. I went to Paris and saw the apparatus which the French Government employed for testing the efficiency of screw propellers, but the propellers were so very badly made that the experiments were of no value. Upon consulting an English experimenter, who had made a “life-long study” of the question, he assured me that I should find the screw propeller very inefficient and very wasteful of power, and that all screw propellers had a powerful fan-blower action, drawing in air at the centre and discharging it with great force at the periphery. I found that no two men were agreed as to the action of screw propellers. All the data or formulÆ available were so confusing and contradictory as to be of no value whatsoever. Some experimenters were of the opinion that, in computing the thrust of a screw, we should only consider the projected area of the blades, and that the thrust would be equal to a wind blowing against a normal plane of equal area at a velocity equal to the slip. Others were of the opinion that the whole screw disc would have to be considered; that is, that the thrust would be equal to a wind blowing against a normal plane having an area equal to the whole disc, and at the velocity of the slip. The projected area of the two screw blades of my machine is 94 square feet, and the area of the two screw discs is 500 square feet. According to the first system of reasoning, therefore, the screw thrust of my large machine, when running at 40 miles an hour with a slip of 18 miles per hour, would have been, according to the well-known formula,

V² × ·005 = P

18² × ·005 × 94 = 152·28 lbs.

If, however, we should have considered the whole screw disc, it would have been 18² × ·005 × 500 = 810 lbs. However, when the machine was run over the track at this rate, the thrust was found to be rather more than 2,000 lbs. When the machine was secured to the track and the screws revolved until the pitch in feet, multiplied by the turns per minute, was equal to 68 miles an hour, it was found that the screw thrust was 2,164 lbs. In this case, it was of course, all slip, and when the screws had been making a few turns they had established a well-defined air-current, and the power exerted by the engine was simply to maintain this air current. It is interesting to note that, if we compute the projected area of these blades by the foregoing formula, the thrust would be—68² × ·005 × 94 = 2,173·28 lbs., which is almost exactly the observed screw thrust.

When I first commenced my experiments with a large machine, I did not know exactly what sort of boiler, gas generator, or burner I should finally adopt; I did not know the exact size that it would be necessary to make my engines; I did not know the size, the pitch, or the diameter of the screws which would be the most advantageous; neither did I know the form of aeroplane which I should finally adopt. It was, therefore, necessary for me to make the foundation or platform of my machine of such a character that it would allow me to make the modifications necessary to arrive at the best results. The platform of the machine is, therefore, rather larger than is necessary, and I find if I were to design a completely new machine, that it would be possible to greatly reduce the weight of the framework, and, what is still more, to greatly reduce the force necessary to drive it through the air.

At the present time, the body of my machine is a large platform, about 8 feet wide and 40 feet long. Each side is formed of very long trusses of steel tubes, braced in every direction by strong steel wires. The trusses which give stiffness are all below the platform. In designing a new machine, I should make the trusses much deeper and at the same time very much lighter, and, instead of having them below the platform on which the boiler is situated, I should have them constructed in such a manner as to completely enclose the boiler and the greater part of the machinery.[14] I should make the cross-section of the framework rectangular and pointed at each end. I should cover the outside very carefully with balloon material, giving it a perfectly smooth and even surface throughout, so that it might be easily driven through the air.

In regard to the screws, I am at the present time able to mount screws 17 feet 10 inches in diameter (Fig. 89). I find, however, that my machine would be much more efficient if the screws were 24 feet in diameter and I believe with such very large screws, four blades would be much more efficient than two.

My machine may be steered to the right or to the left by running one of the propellers faster than the other. Very convenient throttle valves have been provided to facilitate this system of steering. An ordinary vertical rudder placed just after the screws may, however, prove more convenient if not more efficient.

The machine is provided with fore and aft horizontal rudders, both of which are connected with the same windlass.

In regard to the stability of the machine, the centre of weight is much below the centre of lifting effect; moreover, the upper wings are set at such an angle that whenever the machine tilts to the right or to the left the lifting effect is increased on the lower side and diminished on the higher side. This simple arrangement makes it automatic as far as rolling is concerned. I am of the opinion that whenever flying machines come into use, it will be necessary to steer in a vertical direction by means of an automatic steering gear controlled by a gyroscope. It will certainly not be more difficult to manoeuvre and steer such machines than it is to control completely submerged torpedoes.

When the machine is once perfected, it will not require a railway track to enable it to get the necessary velocity to rise. A short run over a moderately level field will suffice. As far as landing is concerned, the aerial navigator will touch the ground when moving forward, and the machine will be brought to a state of rest by sliding on the ground for a short distance. In this manner very little shock will result, whereas if the machine is stopped in the air and allowed to fall directly to the earth without advancing, the shock, although not strong enough to be dangerous to life or limb, might be sufficient to disarrange or injure the machinery.

THE COMPARATIVE VALUE OF DIFFERENT MOTORS.

So far I have only discussed the navigation of the air by the use of propellers driven by a steam engine. The engines that I employ are what is known as compound engines—that is, they have a large and a small cylinder. Steam at a very high pressure enters the high-pressure cylinder, expands and escapes at a lower pressure into a larger cylinder where it again expands and does more work. A compound engine is more economical in steam than a simple engine, and therefore requires a smaller boiler to develop the same horse-power, so that when we consider the weight of water and fuel for a given time, together with the weight of the boiler and the engine, the engine motor with a compound engine is lighter than a simple engine. However, if only the weight of the engine is to be considered then the simple engine will develop more power per unit of weight than the compound engine. For instance, if, instead of allowing the steam to enter the small cylinder, and the exhaust from this cylinder to enter the large or low-pressure cylinder—which necessitates that the high-pressure piston has to work against a back pressure equal to the full pressure on the low-pressure cylinder—I should connect both cylinders direct with the live steam, and allow both to discharge their exhaust directly into the air, I should then have a pair of simple engines, and instead of developing 363 H.P. they would develop fully 500 H.P., or nearly 1 H.P. for every pound of their weight. I mention this fact to show that the engines are exceedingly light, and that when compared with simple engines their power should be computed on the same basis. It will, therefore, be seen that if we do not take into consideration the steam supply or the amount of fuel and water necessary, the simple steam engine is an exceedingly light motor.

But, as before stated, great improvements have recently been made in oil engines. I have thought much on this subject, and am of the opinion that if one had an unlimited supply of money, a series of experiments could be very profitably conducted with a view of adapting the oil engine for use on flying machines. If we use a steam engine, it is necessary to have a boiler, and at best a boiler is rather a large and heavy object to drive through the air. If we use an oil engine, no boiler is necessary, and the amount of heat carried over in the cooling water will only be one-seventh part of what is carried over in the exhaust from a steam engine of the same power. Therefore, the condenser only need be one-seventh part the size, and consequently should be made lighter with the tubes placed at a greater distance apart, and thus reduce the amount of power necessary to drive the machine through the air. Moreover, the supply of water necessary will be greatly reduced, and a cheaper and heavier oil may be employed, which is not so liable to take fire in case of an accident. It is then only a question as to whether an oil engine can be made so light as to keep its weight within that of a steam motor; that is, an oil engine in order to be available for the purpose must be as light, including its water supply, as a complete steam motor, which includes not only the engine, but also the boiler, the feed pumps, the water supply, the burner, the gas generator, and six-sevenths of the condenser. It requires a very perfect steam engine and boiler, not using a vacuum, to develop a horse-power with a consumption of 1¹/₂ lbs. of petroleum per hour; but there are many oil engines which develop a horse-power with rather less than 1 lb. of oil per hour. It will, therefore, be seen that, as far as fuel is concerned, the oil engine has a decided advantage over the more complicated steam motor. Moreover, with an oil engine, the cooling water is not under pressure, so that the waste of water would be much less than with a steam engine, where the pressure is so high as to cause a considerable amount of waste through joints and numerous stuffing-boxes.

The great advances that have been made of late years in electrical science and engineering have led many to believe that almost any knotty scientific question may be solved by the employment of electrical engineering, and a great deal has been written and said in regard to navigating the air by flying machines driven by electric motors.

Before I commenced my experiments, I made enquiries of all the prominent electrical engineering establishments where there was any likelihood of obtaining light and efficient electric motors, and found that it was impossible to obtain one that would develop a horse-power for any considerable time that would weigh less than 150 lbs. Since that time, notwithstanding that a great deal has appeared in the public prints about the efficiency and lightness of electric motors, I am unable to learn of any concern that is ready to furnish a complete motor, including a primary battery, which would supply the necessary current for two hours at a time, at a weight of less than 150 lbs. per horse-power, and as far as I have been able to ascertain from what I have myself seen, I cannot learn that there are any motors in practical use which do not weigh, including their storage batteries, at least 300 lbs. per horse-power. The last electric motor which I examined was in a boat; it was driven by a primary battery which weighed over 1,000 lbs. to the horse-power. From this I am of the opinion that we cannot at present look to electricity with any hope of finding a motor which is suitable for the purpose of aerial navigation.

ENGINES.

There is no question but what birds, and for that matter all animals, when considered as thermo-dynamic machines, are very perfect motors; they develop the full theoretical amount of energy of the carbon consumed. This we are quite unable to do with any artificial machine, but birds, for the most part, have to content themselves with food which is not very rich in carbon. It is quite true that a bird may develop from ten to fifteen times as much power from the carbon consumed as can be developed by the best steam engine, but, as an off-set against this, a steam engine is able to consume petroleum, which has at least twenty times as many thermal units per pound as the ordinary food of birds. The movement of a bird’s wings, from long years of development, has without doubt attained a great degree of perfection. Birds are able to scull themselves through the air with very little loss of energy. To imitate by mechanical means, the exact and delicate motion of their wings would certainly be a very difficult task, and I do not believe that we should attempt it in constructing an artificial flying machine. In Nature it is necessary that an animal should be made all in one piece. It is, therefore, quite out of the question that any part or parts should revolve. For land animals there is no question but what legs are the most perfect system possible, but in terrestrial locomotion by machinery, not necessarily in one piece, wheels are found to be much more effective and efficient. The swiftest animal can only travel for a minute of time at half the speed of a locomotive, while the locomotive is able to maintain its much greater speed for many hours at a time. The largest land animals only weigh about 5 tons, while the largest locomotives weigh from 60 to 80 tons. In the sea, the largest animal weighs about 75 tons, while the ordinary Atlantic liner weighs from 4,000 to 14,000 tons. The whale, no doubt, is able to maintain a high speed for several hours at a time, but the modern steamer is able to maintain a still higher speed for many consecutive days.

As artificial machines for terrestrial and aquatic locomotion have been made immensely stronger and larger than land or water animals, so with flying machines, it will be necessary to construct them much heavier and stronger than the largest bird. If one should attempt to propel such a machine with wings, it would be quite as difficult a problem to solve as it would be to make a locomotive that would walk on legs. What is required in a flying machine is something to which a very large amount of power can be directly and continuously applied without any intervening levers or joints, and this we find in the screw propeller.

..........

When the Brayton gas engine first made its appearance, I commenced drawings of a flying machine, using a modification of the Brayton motor which I designed expressly for the purpose; but even this was found to be too heavy, and it was not until after I had abandoned the vertical screw system that it was possible for me to design a machine which, in theory, ought to fly. The next machine which I considered was on the kite or aeroplane system. This was also to be driven by an oil engine. Oil engines at that time were not so simple as now, and, moreover, the system of ignition was very heavy, cumbersome, and uncertain. Since that time, however, gas and oil engines have been very much improved, and the ignition tube which is almost universally used has greatly simplified the ignition, so that at the present time, I am of the opinion that an oil engine might be designed which would be suitable for the purpose.

In 1889 I had my attention drawn to some very thin, strong, and comparatively cheap tubes which were being made in France, and it was only after I had seen these tubes that I seriously considered the question of making a flying machine. I obtained a large quantity of them and found that they were very light, that they would stand enormously high pressures, and generate a very large quantity of steam. Upon going into a mathematical calculation of the whole subject, I found that it would be possible to make a machine on the aeroplane system, driven by a steam engine, which would be sufficiently strong to lift itself into the air. I first made drawings of a steam engine, and a pair of these engines was afterwards made. These engines are constructed, for the most part, of a very high grade of cast steel, the cylinders being only ³/₃₂ of an inch thick, the crank shafts hollow, and every part as strong and light as possible. They are compound, each having a high-pressure piston with an area of 20 square inches, a low-pressure piston of 50·26 square inches, and a common stroke of 1 foot. When first finished, they were found to weigh 300 lbs. each; but after putting on the oil cups, felting, painting, and making some slight alterations, the weight was brought up to 320 lbs. each, or a total of 640 lbs. for the two engines, which have since developed 362 horse-power with a steam pressure of 320 lbs. per square inch. A photograph of one of these engines is shown in Fig. 85.

..........

When first designing this engine, I did not know how much power I might require from it. I thought that in some cases it might be necessary to allow the high-pressure steam to enter the low-pressure cylinder direct, but as this would involve a considerable loss, I constructed a species of an injector. This injector may be so adjusted that when the steam in the boiler rises above a certain predetermined point, say 300 lbs. to the square inch, it opens a valve and escapes past the high-pressure cylinder instead of blowing off at the safety valve. In escaping through this valve, a fall of about 200 lbs. pressure per square inch is made to do work on the surrounding steam and to drive it forward in the pipe, producing a pressure on the low-pressure piston considerably higher than the back pressure on the high-pressure piston. In this way a portion of the work which would otherwise be lost is utilised, and it is possible, with an unlimited supply of steam, to cause the engines to develop an enormous amount of power.

..........

Boiler Experiments.—The first boiler which I made was constructed something on the Herreshoff principle, but instead of having one simple pipe in one very long coil, I used a series of very small and light pipes, connected in such a manner that there was a rapid circulation through the whole—the tubes increasing in size and number as the steam was generated. I intended that there should be a pressure of about 100 lbs. more on the feed water end of the series than on the steam end, and I believed that this difference in pressure would be sufficient to ensure a direct and positive circulation through every tube in the series. This first boiler was exceedingly light, but the workmanship, as far as putting the tubes together was concerned, was very bad, and it was found impossible to so adjust the supply of water as to make dry steam without overheating and destroying the tubes.

Fig. 90.—Steam boiler employed in my experiments. With this boiler, I had no trouble in producing all the steam that I could possibly use, and at any pressure up to 400 lbs. to the square inch.