Many runs along the track were made to test the working of this great apparatus before trusting it to launch forth in free flight. Dynamometers gave independently the thrust of the screws, and the lift of the wings on the front and rear axles. The ascensional planes for controlling the fore and aft equilibrium were tested during the run, as also the practical operation of the propelling plant. During the trials of 1893 the machine frequently lifted clear of the lower track, and flew forward resting against the guard rails above the wheels. Finally, on a gusty day, the lift against the upper track caused this to give way, whereupon the machine rose into the air with Mr. Maxim and his assistant, then toppled over on the soft earth, suffering some damage to its framework. Here the experiments were discontinued for lack of funds, having indeed demonstrated that a large weight can be carried in dynamic flight, but having proved little as to the feasibility of controlling an aëroplane in launching, in free flight, and in landing.

Compared with the work of his contemporaries this achievement of Mr. Maxim was herculean, both in construction and expenditure, the cost being reported as nearly one hundred thousand dollars. It raised high hopes for aviation. It proved conclusively not only that a flying machine could be made to lift a pilot, but that it could carry hundreds of pounds additional weight. It still holds the world’s record for magnitude of machine and cargo. But it had two great defects; it was improperly balanced and it was inadequately powered; for, as Mr. Maxim says, “the quantity of water consumed was so large that the machine could not have remained in the air but a few minutes, even if I had had room to maneuver and learned the knack of balancing in the air.”[32] These defects, however, would soon be remedied by the work of others, and particularly by the costly experiments of the automobilists, who were rapidly developing a light gasoline motor suitable for aviation.

The inventors thus far noticed had developed most of the important features of the present-day flying machines, but had not provided adequate mechanism for preserving a steady lateral balance. The present writer had proposed the combination of a double rudder and torsional wings to steer and control a flyer, and had published a paper setting forth its general principle and describing a specific device; but inventors had little need for a third rudder till they encountered the dangers of dynamic flight in gusty weather. The paper referred to was presented to the Third International Conference on Aërial Navigation, in August, 1893, under the title, Stability of Aëroplanes and Flying Machines, and was published with the proceedings of the conference.[33] It discusses mainly the question of automatic stability and steadiness; but recommends personal control during the experimental period. It concludes as follows:

“We have been considering the question of automatic stability, in so far as it may be secured in the construction of the craft itself,[34] apart from a pilot, or special equilibrating devices. The application of the latter would give exercise to an infinite amount of ingenuity, and would, perhaps, best be left to the fancy of the individual inventor. One curious design, however, occurs to me, which, since I have not seen it described elsewhere, may be worth a moment’s notice.

“Suppose a Phillips’s machine (see Plate XIV) to be provided with a double tail, and to have a vertical fin extending longitudinally along its entire length, well above the center of gravity. These would steady its flight and promote stability. Suppose also that its sustaining slats were pivoted, so that a pilot could at pleasure change their inclination on the right and left side independently. He could then set the engine for a desired speed, sweep forward along the earth with the sustainer slats horizontal, and at will mount into the air, by giving the slats an upward inclination. Once in the air he could raise or lower the machine by slightly changing the angle of the slats; he could wheel to right or left by giving one set of slats a little different slope from the other; he could arrest all pitching, rocking and wheeling by a slight counter movement of the sustainers. It would be necessary, of course, to preserve a rapid forward motion, for it is a peculiarity of the compound aëroplane that, if it comes to a standstill in the air, it will drop plumb down with a frightful plunge until it acquires headway.”

The succeeding paragraph disclosed a specific contrivance embodying the principle just given. This showed two levers rotating drum shafts for actuating wires adapted to change the impact angles of the wing surfaces. Accordingly this much of the mechanism of control, together with the broad device of the torsion wings, has been the common property of inventors since the publication of that paper. Furthermore, the combination of torsional wings and a double rudder, either fixed or movable, has been public property since that date.[35]

Little was said about the manner of manipulating the double rudder and torsional wings; for the rules of manipulation would vary in different machines, depending upon structural design and external conditions. For example, if the proposed fin and vertical rudder were ample and suitably placed, the lateral balance could be controlled by merely twisting the wings, without touching the vertical rudder; but if the fin and rudder were not adequate, the lateral poise would be controlled by twisting the wings and working the vertical rudder conjunctively. A novice might prefer leaving the rudders fixed and controlling the poise in short flights by twisting the wings by means of a single lever having two independent movements, one to rotate the wings oppositely, the other to rotate them identically.

The principle of control expressed in italics had been set forth also in a preceding paragraph. Having proposed means for securing both stability and steadiness about each of the three axes of an aëroplane, the text continued:

“These ends could probably be attained very well by mounting two compound aëroplanes on a long backbone,[36] somewhat after the manner of the Hargrave cellular kites, and adding a compound rudder to the whole.” ... “If the inclination of the sustainers, front and back, could be altered independently, it might be feasible for a pilot to preserve the equilibrium of the machine even when its center of gravity was frequently shifted, as by the moving of passengers to and fro.”[37]

At that date, 1893, an inventor doubtless could have secured a broad claim on a mechanism embodying the torsion-wing-and-double-rudder mechanism of control. But in those days aviation was pursued largely as a liberal study by scientific men who wished to hasten the advent of practical flight, by presenting important physical measurements and principles which could be freely employed by all. Accordingly the three-rudder system of control seems not to have been claimed by an inventor much before the close of the nineteenth century. Since then it has been patented in one form or other by many practical aviators, some endeavoring to claim the whole broad contrivance, others claiming more restricted devices.

The static principle of the torsion wing is a familiar one in elementary mechanics. It is this: a torque of given magnitude and direction has the same effect on a rigid body whatever its point of application. The longitudinal torque, or moment, may therefore be exerted by the wings, by suitable rudders, by forward planes, by any auxiliary planes, or fins, however placed or moved for the purpose. Accordingly there seems to be an unlimited variety of concrete patentable devices available to the inventor for securing impactual torque about the longitudinal axis, or either of the other two axes. But in planning such devices it is well to remember that the moment of a couple increases with its arm, so that in a wide aëroplane the wing tips may best furnish the torque; while in a high short-winged machine, vertical planes, fins, or rudders may give the desired longitudinal moment. Obviously such vertical guiding or controlling surfaces may be so placed as to tilt the machine toward the center of curvature of its path, at the same time opposing the centrifugal force, and exerting a torque about the vertical axis tending to steer the flyer along its path.[38]

The principle of projectile stability is another consideration of some importance in aviation, or more generally in all submerged navigation, whether of air or water. A submerged body has projectile stability if its nose tends always to forerun its centroid, and follow a steady course. A dart is a good example; a fish, a torpedo. Thus if a torpedo-shaped homogeneous solid be hurled in any manner through a fluid, obliquely or even tail foremost, it promptly turns its nose to the front and proceeds steadily along an even course; but if the body has not true dynamical balance, it may oscillate or gyrate, or flit about in the most erratic manner.

Projectile stability in a flyer, as in an arrow, may be attained by playing the centroid in or near the line of forward resistance, and well ahead of the side resistance. The reasons for this are manifest. If, however, this arrangement be neglected, a special damping, or controlling, device is required to preserve headlong and steady motion. In particular, the objections to placing the centroid too low were emphasized in the above quoted paper as follows:

“I have mentioned the advantage of placing the center of mass below the center of surface; this has also its objections. While the stability against inversion is increased, the stability against rocking is sacrificed. The aëroplane so constructed may not easily overturn; but it will sway to and fro with a pendular motion. This, when lateral, is very objectionable, when fore and aft it is fatal to uniform progress, as we shall see in studying the longitudinal stability of flying machines. We shall then see that the center of mass cannot be lowered with impunity.”

Of the various flyers and models thus far studied, some manifest fairly good, others very imperfect projectile stability. Many inventors have been more alert to the gravitational stability and safety of the parachute than to the kinetic stability and keen, direct flight of the arrow. Some of the most pretentious machines imitated the thistle down more nearly than the dart or swallow. But the exigencies of actual flight would easily rectify such imperfections of design.

Tractional balance also is a property of some importance in fluid navigation. This requires that the line of propulsive thrust coincide with the line of fluid resistance. It is a property, however, that inventors readily apprehend, and usually provide for.

In general a flyer is subject to four forces: weight, thrust, air pressure and inertia. When these balance about any axis the craft has equilibrium about that axis; when they balance about the three axes the craft is completely balanced, and preserves its orientation in flight. Devices for preserving this complete balance have already been described; as also provision for propulsion and sustentation, launching and landing safely.

Thus at the close of the nineteenth century all the essential principles and contrivances of pioneer flight were worked out, except one—a suitable motor. This was the real problem of the ages. The rest was easy by comparison. A light enduring motor, if available to the old time inventors, would have brought dynamic flight centuries ago. That only could have baffled Da Vinci, Cayley, Henson, Wenham and the long line of pioneer aviators. Eventually, of course, steam engines had come, endowed with ample power; but costly to build and wasteful to operate. The light automobile engine appeared in the latter nineties; promptly thereafter followed the dynamic flyer, the snow-winged herald of the twentieth century.