One death is a tragedy; one million is a statistic. — Joseph Stalin

Twentieth Century Inventions - Charles Gibson




Gyrostatic Inventions

THE GYRO-COMPASS—PREVENTION OF SHIPS ROLLING BRENNAN MONO-RAIL CAR


The Gyro-Compass


The invention of a gyrostatic compass came as a surprise to most people, and yet the fact that a gyrostat could act in this way had been known for two generations. So long ago as 1852, the French philosopher, Foucault, demonstrated that any gyrostat, free to be turned in two directions only, will tend to set itself with its axis of rotation parallel to the axis of the earth, by reason of the relative rotations of the two bodies. He pointed out that this law would hold good for a gyrostat (with two degrees of freedom) at any place on the earth's surface other than the two poles.

The average man probably knows little more of a gyrostat than can be observed by handling one of those simple gyrostatic tops, such as are sold as toys at exhibitions. Even if one sets such a gyrostat spinning, and holds it with the axle, say, in a vertical position, one may feel that the gyrostat resists any attempt to move the axle away from its vertical position. Better still for our present purpose if the spinning gyrostat is held with its axle in a horizontal position, whereupon the same resistance is apparent if we seek to alter the direction of the axle.

A gyrostat may he given three degrees of freedom. For instance, we might suspend the toy gyrostat by a piece of string attached to its equator ring. If care is taken to balance the gyrostat in this position, with its axle horizontal, it is quite apparent that it remains in a definite position and that it resists any change in the direction of its axle. In whatever direction we place its plane of rotation, it will remain in that position in space while the earth turns round beneath it. This was the first law which Foucault laid down concerning gyrostats. He pointed out that a gyrostat, if free to turn in all three directions, would serve in the same manner as his pendulum to demonstrate the rotation of the earth; the gyrostat would take up a fixed position in space, and we should see the earth turn round beneath it.

Dr Anschutz, of Germany, began experiments in 1900 with a gyrostat having three degrees of freedom. Of course, if such a gyrostat, say on board ship, were placed with its axle pointing north and south, it would remain in that position, no matter how much the ship turned about, but it would possess no directive force. If the motor driving it had to be stopped for any reason, the direction would be lost. If while it was spinning any one happened by contact to alter its direction, it would remain pointing in some false direction.

Anschutz had no intention of using such an arrangement in place of the magnetic compass, but merely to get fixed lines in space, for obtaining bearings or maintaining a course already definitely known. However, he found it practically impossible to construct a gyrostat having its centre of gravity and centre of suspension absolutely coincident.

In 1906 Anschutz turned his attention to a gyrostat having only two degrees of freedom. At first he combined this gyrostat with his earlier one having three degrees of freedom, and he found that it directed the combined system into the meridian line. He then experimented with a single gyrostat having two degrees of freedom only, which, according to Foucault's law, should set itself with its axis of rotation parallel to the axis of the earth.

A toy or an experimental gyrostat does not exhibit any directive phenomenon. The existence of the directive force can be observed only when the speed of rotation is high, and when all precautions have been taken to eliminate friction. Although this is so, it is easy to demonstrate with an experimental gyrostat that if the spinning body has only two degrees of freedom the position of its axle and the direction of its rotation will be affected when the gyrostat as a whole is revolved.

Picture an experimental gyrostat suspended in gymbals so carefully that it has three degrees of freedom, so that it will take up a definite position in space. An experimenter may take this in his hand and turn round without affecting the gyrostat, but let him clamp the vertical spindle carrying the supporting frame, and the conditions are changed; the gyrostat has only two degrees of freedom. If the experimenter now revolves himself and carries the gyrostat with him, it will immediately set its axle parallel to the axis of rotation of the experimenter, which, of course, is vertical. We have supposed that the experimenter happened to revolve himself in the same direction as that in which the fly-wheel of the gyrostat was rotating. If he should revolve himself in the opposite direction, the gyrostat would immediately turn a somersault, and thus place its fly-wheel with its direction of rotation corresponding with the direction of revolution imposed upon it as a whole. It is quite obvious that a gyrostat with only two degrees of freedom behaves in a different manner to a gyrostat having three degrees of freedom.

The gyro-compass, as already indicated, is a gyrostat with only two degrees of freedom, so that it seeks to bring its axle parallel to the earth's axis, with its direction of spin the same as that of the earth. But the gyro-compass is floating in a bowl of mercury, and the axle of its fly-wheel will remain horizontal. It is clear that under these conditions the gyro-compass could not have its axle parallel to the earth's axis unless the apparatus happened to be at the equator. Although the gyrostat would tend to do so when carried north or south, it could not, as its axle would be pulled into a horizontal position by gravity. Those accustomed to experiment with gyrostats know what this means; the applied forces tend to set the axis of the gyrostat parallel to the earth's axis, causing the gyrostat to 'precess' or wheel round in a direction at right angles to the direction in which the applied forces tend to turn it.

This phenomenon of 'precession' is beautifully demonstrated by the well-known Wheatstone compound gyrostat, in which we may apply a weight to one end of the horizontal axle. Instead of tilting the axle, as would be the case if the fly-wheel were at rest, the applied force causes the gyrostat to wheel round, and it will continue this 'precession' so long as the force is applied. We have a demonstration of 'precession' in the school-boy's spinning top, and we know that the earth precesses, principally because of the attractive pull of the sun and moon on the oblate portion at the equator. We know that it is because of this precession that the north pole of the earth has not pointed always to the Pole Star, but that the earth's axis of revolution describes a curve, which is very nearly circular, about the pole of the ecliptic.

Even when one holds a toy gyrostat in one's hand, say with the axle horizontal, when one seeks to depress one end of the axle, there is not only a feeling of resistance, but a distinct twisting motion of the gyrostat. This we recognise as the wheeling round of the gyrostat at right angles to the applied force, or in a single word—its 'precession.'

It is this precessional motion which turns the gyro-compass into the desired position. The applied forces, due to gravity, keeping the axis of the gyrostat horizontal, cause the gyro-compass to wheel round until it gets into the position with its axle parallel with the meridian, in which position the gyrostat may be carried round by the earth without resistance. Should the axle swing beyond the meridian on the other side, the pull of gravity will be on the end of the axle opposite to that which we have been considering, and the precession will be in the opposite direction, bringing the axle of the gyro-compass back to the meridian.

One difficulty, in giving the gyro-compass only two degrees of freedom, was that such a gyrostat would be affected by other forces which might be brought to bear on it, as, for instance, by the movements of the ship. Such forces would set the gyrostat swinging, and render its indications unreliable. It became evident that if a gyro-compass were to be of practical value it would require to possess a very large gyroscopic resistance, strongly opposing any attempt to tilt its axle to an angle. Also that it would be necessary to have the friction of the suspension system as small as possible. But when these objects have been attained there remains a serious difficulty. If for any reason the gyrostat should be deflected a long way out of the meridian line, its swinging motion to and fro will last a very long time. This would render the gyroscope useless as a compass, unless it were possible to damp out this swinging motion. The first idea of overcoming this difficulty was to use a second gyrostat to damp out the oscillations, but it was soon found that a reliable damping could be obtained by a much simpler arrangement, which will be described when we consider the construction of the practical compass.

It need hardly be pointed out that although Foucault set forth the laws describing the actions of the gyrostat, it would have been quite impossible for him to produce a gyro-compass, as he had no means of supplying the necessary continuous motion to the fly-wheel, the rotary speed of which must be very high. Besides, in his day, there was no necessity for such a compass, as the simple magnetic compass would serve all practical requirements. Indeed, until recently it was found quite satisfactory to apply proper compensation to the magnetic compass to balance the attractive force of the iron in the ship. But with the increase in size of warships, and the great masses of moving steel in use in modern guns and their shields, this magnetic compensation became a very serious problem. Then, again, the submarine with its great number of electric motors, producing magnetic fields of their own, and thus affecting the compass, created a demand for a non-magnetic compass.

[Illustration] from Twentieth Century Inventions by Charles Gibson

The practical construction of the gyro-compass will be understood from the accompanying diagram (Fig. 4), which represents a vertical section of the instrument. It will be observed that the gyrostat G runs on bearings fixed in the case C, which is suspended from a hollow steel ring F. This hollow vessel floats in a bowl B, also made of steel and filled with mercury as at M. It will be observed that the gyrostat is mounted at the lowest point of the moving system, with the spinning axle horizontal. The centre of gravity is below the meta-center (that point on the position of which its stability depends). The circular hollow float has a dome to which the compass card CC is rigidly fixed. The axle of the gyrostat is directly under the 'north' and 'south' of the card, so that when the gyrostat places its axle north and south, the compass card takes up the same position.

The mercury bowl is carried on gymbals in the well-known manner, and the outer gymbal ring is borne by springs from the binnacle case, and so protecting it to a great extent from damage due to violent shocks. In order to keep the whole floating system central, a steel stem S is fixed centrally in the cover glass GL, and the lower end of the stem dips into a small mercury cup carried on the top of the float. This serves also as one of the contacts to conduct the electric current to the gyrostat motor. Around the steel stem S is a steel tube, there being an insulating material between them. It will be observed that the lower end of the tube is widened and dips into a second mercury cup. This serves as a second electrical contact. Both these contacts are, of course, insulated from the general metal portions of the apparatus. As the gyro-motor is a three-phase one, it requires a third electrical contact, and this is obtained through the mercury bowl, the mercury, and the float. This arrangement necessitates the whole instrument being insulated from the binnacle.

The motor of the gyrostat consists of a very small three-phase motor, the stator (fixed coils into which the live current enters) is mounted inside the case C, so that all the connections can be rigidly made. The rotor (the conductor in which only induced current flows) is rigidly fixed into the inside of the gyro-flywheel itself. This rotor has no coils of wire, only copper bars let into the fly-wheel. It is necessary to have the fly-wheel and axle made from one solid piece of special nickel steel, as the speed of rotation is to be about 20,000 revolutions per minute. This means that the peripheral speed of the gyrostat is almost six miles per minute. Its enormous speed may be realised if we think of the fly-wheel making 333 revolutions in each second. The stress to which the rim is subjected amounts to ten tons per square inch.

The axle is provided with ball bearings, which have to be so exactly gauged that special precision appliances have been devised to examine the spherical condition of the balls. The axle is of the de Laval type, forming a flexible axis, so that the centre of gravity of the whole rotating mass coincides with the rotation axis as soon as a critical speed is exceeded. Even although the axle is relatively weak, the gyrostat, while running, is not sensitive to shocks, because while even the very shortest possible shock lasts, the gyrostat has made many revolutions (333 revolutions per second), and therefore any bending tendencies neutralise one another.

It is interesting to note that about ninety-five percent of the motor energy is absorbed in overcoming the resistance of the air. The air friction on the surface of the gyrostat is so great that after the gyrostat has run for a few thousand hours its surface becomes noticeably smoother, having been polished by friction with the gaseous air particles.

It is this air disturbance which the inventor uses to damp the pendulum motion of the gyro-compass. The high speed of the gyrostat within its enclosing case practically forms a centrifugal blower. A hole near the centre in each side of the case C serves to admit air, and a hole in the periphery at the lowest point of the case acts as the outlet, and through this a constant stream of air issues. This constant flow of air through the case serves to keep the gyrostat cool, but what is of greater importance is that the energy of the escaping air may be used to damp out the pendulum motion of the instrument. One method employed by the inventor is to divide the outlet into two compartments having a movable division between them. This division practically forms the bob of a pendulum, which moving to one side decreases the size of that compartment and enlarges the other. The pendulum, which carries this division of the air outlet, is so balanced that when the axle of the gyrostat is horizontal, the division separates the air outlet into two chambers of equal area. In these circumstances the air is forced out with equal energy through both chambers, one part on each side of a vertical centre line through the whole moving system.

When the axle of the gyrostat is not horizontal, which is the case when the instrument is precessing to or from the meridian, the small pendulum swings to one side, and the two air passages are no longer equal; the difference of their reaction forms a turning couple round the vertical axis of the system. This couple produces a motion opposed to the precession, in such a direction as to bring the axle of the gyrostat to the horizontal once more. In this manner the oscillations of the gyrostat on either side of the meridian are powerfully damped.

Corrections require to be made for changes of latitude and for changes in the speed of the ship. These, however, are simpler than the corrections which have to be made in the readings of a magnetic compass, such as the continual secular change in 'variation.' The gyro-compass has a great advantage in pointing to the true north, and not to a certain spot known as the magnetic pole, from the direction of which the true north may be deduced.

So far as change of speed is concerned, it is obvious that this will cone into play only when the ship is steaming in a northerly or southerly direction. In such cases if the ship were stopped suddenly, there would be a pendulum motion of the suspended system; the gyrostat would swing forward by its own inertia. This tilting of the axle would cause a precession, producing an error. This is called the ballistic deflection, and is corrected by means of a simple table of figures. When a ship is going in an easterly or westerly direction, the pendulum motion of the suspended system is then about an axis parallel to the axle of the gyrostat, and therefore the gyro-axle moves parallel to itself, and there is no precession.

The directive force on the gyro-compass is fifteen times as great as in the case of a good magnetic compass, even when placed quite free of any surrounding iron. The value of the directive force of the magnetic compass decreases rapidly with the vicinity of large masses of iron and the presence of electric circuits, whereas the directive force of the gyro-compass remains constant under these circumstances.

Another advantage is that the compass card of the gyro-compass cannot roll about in any direction as can the card of the magnetic compass, whose point of support is at its centre only. The gyro-compass resists any alteration of its axle, both in a horizontal and vertical plane, and the compass card, therefore, can only see-saw up and down with the north and south line as an axis. Consequently, if an electrical contact point be placed at east and at west of the card, these points will not be rolled about which would alter their compass direction; they will move up and down only, remaining due east and west. This phenomenon makes it possible to construct a system of transmission mechanism which would be impossible with a magnetic compass in which the card can roll about in any direction.

Just as we have a master clock with a number of electric dials placed at a distance, so we may have a master compass with receivers placed in convenient positions on board a ship. One advantage in this arrangement is that the master compass with its gyrostat may be placed in some well-protected position near the bottom of the ship.

In the master compass, no actual work is put upon the gyrostat; it merely carries the contact points, and these control a reversible electric motor which turns the whole mercury bowl, causing it to follow the movements of the gyrostat. This reversible motor remains out of action so long as the ship keeps on a steady course, but any alteration in the course causes the gyrostat to switch on the current, in one direction or the other, to the motor. A commutator is mounted on the axle of the reversible motor, and this distributes electric energy to the mechanism of the receiving instruments, according to the position of the mercury bowl of the master compass, so that the receivers always turn in synchronism with the transmitter.

The receiving mechanism is connected by means of gearing to a compass card, and it is on these dials that the directions are read. In the centre of the receiving card is a second card, which makes one complete revolution for ten degrees of alteration of the course. This inner card is divided in such a manner that an alteration of a few 'minutes' is at once apparent. The employment of this fine adjustment is of the greatest possible assistance in steering. When the ship is in motion, the small central card is constantly on the move to and fro, on account of the ship continually departing from an absolutely straight course. So long as the movement is merely to and fro, the steersman knows that all is well, but a continued movement in either direction shows that an alteration of the course has taken place.

The high speed of the gyro-motor is obtained in this way. The three-phase portion of the motor generator has sixteen poles, and runs at a normal speed of 2500 revolutions per minute. As the motor in the gyrostat has only two poles, it follows that the speed of this latter is 16 x 2500 = 2 = 20,000 revolutions per minute. This is the limit of speed in practice, but for the purpose of testing that the factor of safety in the gyrostat is sufficiently high, a special motor generator was built to give an enormously increased periodicity. To make an exceptionally high speed possible, the gyrostat was run in a vacuum. We have seen that even at 20,000 revolutions per minute the centrifugal force is so great that the rim of the fly-wheel is subjected to a stress of ten tons per square inch. In the super-high speed trials it was found that the gyrostat could withstand not only double the strain, but the normal power had to he increased five times before any yielding of the material took place.


Prevention of Ships Rolling


Another gyrostatic invention of general interest has been made in connection with the rolling of ships. Of the three motions, heaving, pitching, and rolling, it is the last-mentioned which is feared most by those who dislike sea passages. Most of us do not worry about the heaving of a ship as it rises and falls bodily keeping its deck in a horizontal position; the greater the displacement the lower is the frequency. But the motion which is caused by pitching is not so pleasant, one may be moved up and down through a distance of thirty feet in a few seconds, but we can escape this motion almost entirely by getting amidships. There is no doubt that it is the rolling of a ship which brings about the discomforts of sea-sickness most easily. The vertical motion of the sides may be avoided by getting to the centre of the deck, but the angular movement cannot be evaded, a it affects equally all parts of the ship.

Several inventors have suggested anti-rolling devices, but by far the most ingenious of these is the gyrostatic invention patented by Dr Otto Schlick, of Hamburg, an eminent marine engineer, who gave successful demonstrations of his apparatus both in his own and in this country.

It was the gyrostatic action of paddle wheels in a steamer which led Dr Schlick to study this subject. He had observed that when a steamer is heeled over by a wave, the course of the steamer is altered slightly, and conversely that when the course of a steamer is altered suddenly the steamer heels over. With an ordinary paddle steamer these phenomena are not very apparent, for the speed of rotation of the paddles is comparatively slow.

In order to study the subject, Dr Schlick used a model with two solid discs to represent the paddles. To increase the gyrostatic action, these were driven at a high speed, and the model ship was pivoted to permit of its turning freely upon a vertical axis. When this model was heeled over to starboard by the addition of a weight on that side, the how of the model turned to starboard. When the weight was transferred to the port side, the vessel turned to port. The same holds good in the case of an actual steamer on the water, but the amount of turning is very slight. The action is not what one would expect, for when a vessel is heeled over to starboard, the paddle on that side will get a bigger grip of the water, and one would expect the vessel to turn to port, but not so.

It was when studying these phenomena that Dr Schlick was led to the invention of his anti-rolling apparatus. In this case the gyrostat is not placed with its fly-wheel in a vertical position, as are the paddles of a steamer; the gyrostat revolves in a horizontal plane, with its axle vertical. It may be remarked that a gyrostat, if placed with its fly-wheel in a vertical plane, and its axis of rotation transverse to the ship, could be used in an anti-rolling device, but not so conveniently.

In Dr Schlick's invention the gyrostat is mounted in a frame which has trunnions or bearings at the sides, and is so placed that its pendulum motion will be fore and aft. Any attempt to tilt the vertical axle of the gyrostat from side to side will cause the gyrostat to swing at right angles to that, which will be lengthwise in the ship.

In connection with the gyro-compass, we have considered how the gyrostat's resistance is at right angles to the applied force. But the enforced pendulum motion of the gyrostat, swinging fore and aft, will not prevent the rolling of the vessel; it is necessary to oppose this force. This Dr Schlick has done by applying brakes to the swinging movement of the gyrostat frame. These brakes damp the pendulum motion and thus absorb the energy of the waves which tend to tilt the ship. The brakes may be either hydraulic or friction, but they must be automatic in their action.

Dr Schlick made an experiment with a German torpedo boat measuring 117 feet long by 12 feet 6 inches broad, and displacing 65 tons on a draft of 3 feet 6 inches. The meta-centric height was 1.3 feet, and the natural period of rolling was about 2.1) ?> seconds from side to side. Such a vessel would prove a very bad roller, as its natural period would be very similar to the period of the waves; the test, therefore, would be a severe one.

The gyrostat wheel was driven by steam, turbine blades being fixed on its circumference. The frame enclosing the gyrostat formed a steam-tight cast-iron casing, receiving and exhausting the steam through the trunnions on which the frame oscillated. The diameter of the steel fly-wheel of the gyrostat was 39 inches, and its speed of revolution was about 1600 revolutions per minute.

The hydraulic brake for controlling the oscillatory movement of the frame consists of a cylinder, with a piston forcing the fluid through a valve, the opening of which can be regulated from deck. The whole apparatus could be thrown in or out of action by means of a friction hand-brake, by which the gyrostat frame may be held in a fixed position or released as may be required.

All the trials at sea were most successful. The little vessel could be taken out into a rough sea, with the apparatus clamped out of action, so that the natural rolling might be witnessed, but as soon as a certain wheel on dock was turned, the rolling stopped almost immediately. So long as the gyrostat was left free to act, the vessel would defy the rolling motion of the waves.

When experiments were made in the Highlands of Scotland, on one of the steamers running between Oban and Tiree, where rolling is an almost constant complaint, the demonstration was entirely successful. On a sea which caused the ship to roll through an area of 32 degrees (while the gyrostat was in check), it was found possible to reduce the roll to four degrees whenever the apparatus was released. A rolling motion through only four degrees is quite inappreciable on any vessel.

When Dr Schlick propounded his theory in 1904, there was considerable discussion as to whether the steadying up of the ship would not be too great a strain on her. Some naval authorities declared that if a ship were stopped from rolling in a beam sea, the next wave would come on board. One speaker went so far as to declare that 'the rolling was provided by Nature to save the ship.' The reply to these statements was that no greater strain would be put upon the ship than was the case when a ship rolled through an angle of twenty-five degrees. Further, that the tendency to swamp the decks would be reduced instead of increased. This reply was endorsed by the great naval authority, Sir William White, I.C.B. It will be understood that a vessel equipped with the gyro-apparatus does not offer resistance such as a rock would do; the vessel is free to rise and fall in the water.

No anti-rolling device is necessary in very large steamers, such as Atlantic liners, for the bilge keels in these serve to prevent any cumulative rolling, and whatever pendulum motion there might be could not synchronize with the much shorter period of the waves. This is not the case with smaller steamers, and it remains to be seen in what direction the shipbuilders who own the Schlick patent will develop the invention.


Brennan Mono-Rail Car


Several systems of mono-rail tracks have been invented, such as a trestle arrangement upon which a divided car rides astride like the packs on the back of a donkey. One of the inventors spent as much as 40,000 on an experimental demonstration track, but so far none of these inventions have passed into the commercial world. There is a mono-rail track in Germany at Elberfeld, but in this case the rail is overhead and the ears are suspended from it. More interesting from the invention point of view is the gyrostatic car invented by Louis Brennan.

The inventor gave a demonstration, with a small ear, before the Royal Society (England) in 1907. Two years later he had an experimental track erected in the War Office grounds at Gillingham, Kent, where demonstrations were given with a full-sized car. This car measured 40 feet long, 10 feet wide, and 13 feet high. It weighed 22 tons, and was capable of carrying 40 passengers on an open platform. A little later Brennan gave some public demonstrations with this car, which naturally attracted a good deal of attention.

While it looked strange to see a heavy car running round a circular track on a single rail, and negotiating the curves as comfortably as a cyclist would, it was more surprising to see the loaded car stop and yet remain upright, even when its forty passengers all crowded to one side of the car. It goes without saying that the mere presence of a rotating gyrostat on board the car would not enable it to behave in this manner; it would only delay but not prevent the fall. It is necessary to have some means of raising the car automatically when it is tilted, and therein lies the ingenuity of Brennan's invention.

He employs two gyrostats mounted side by side in frames, each rotating about a horizontal axis. Each fly-wheel measures three and a half feet in diameter, and weighs about three-quarters of a ton. They are revolved in a vacuum at a speed of 3000 revolutions per minute, each revolving in an opposite direction to the other. The horizontal axles of the gyrostats are transverse to the ear, and are carried in gymbal frames. Each frame turns on a vertical axis, and attached to each of the two vertical spindles is a toothed sector. These two sectors engage with one another, so that as one frame turns about its vertical axis the other is forced to turn a similar amount, but in the opposite direction; if the one turns to the left, the other turns to the right. The object of this is to prevent any tendency of the gyrostats to check the free motion of the vehicle in making a sudden turn round a curve.

When considering the phenomena connected with Dr Schlick's model boat fitted with high speed gyrostats, it was observed that any sudden turning of the bow of the model caused it to tilt over, and it is apparent that this upsetting torque would be disastrous in the case of a gyrostatic car travelling round in a circle. But with the two gyrostats, as arranged in Brennan's car, the precessional couples are equal and opposite, and the trouble vanishes.

The gyrostats are mounted in gymbal frames, and these gyro-frames are carried in a single central frame, in which they both turn or 'precess.' This central frame is solidly mounted on an axis lengthwise of the car, and on the same level as the gyrostat axes, so that the car might tip over from side to side and yet the gyro-flywheels remain vertical. But the axles of the gyrostats project at the sides of the car, so that when the car tilts, a flat plate will come in contact with the axle and act as a friction surface. This pressure causes the gyrostat to precess about a vertical axis. The second gyrostat is forced by means of the toothed sector to turn equally in the opposite direction. The axle, on account of the friction on the contact plate along which it seeks to roll, has a force exerted on it whose moment tends to increase the precession. But the moment of this frictional force about a vertical axis has the effect, not to increase the precession already taking place, but instead to cause precession about the horizontal axis. This gives us two opposite moments acting on oppositely rotating gyrostats, so that they cause precession in the same direction about the horizontal axis of the car. This rights the car and tilts it slightly to the other side, to be brought back again by a second and similar friction plate on the other side of the car. The centre of gravity of the whole car is thus kept oscillating very slightly on either side of the line of upward thrust of the rail.

The gyro-frames, after the displacement, are brought back to their normal position, which is with the axles of the gyrostats transverse with the car. This is accomplished automatically by a second pair of plates, each engaging an idle or frictionless roller attached to the gyro-frames.

In considering the mere tilting of the car when at rest or on a straight line, the presence of two oppositely rotating gyrostats may seem to complicate matters unnecessarily, for a single gyrostat with friction plates would serve to right the car; the two gyrostats merely act in unison. But we have seen that when rounding a curve on the track the two gyrostats become necessary. When rounding a curve, the car is in balance, not when its centre of gravity is vertically over the rail, but when it lies in the line of the resultant of the centrifugal force and gravity.

The linking together of the two gyro-frames, by means of the two toothed sectors, is very ingenious. But for this connection the two oppositely rotating gyrostats would both maintain their original direction in space while the car turned the curve, but unable to do this because of the sectors, they are both forced to process with the car, and then the opposing forces come into play. The two gyrostats develop equal and opposite torques, which are transmitted to the rigid central frame, and they hold each other in equilibrium by means of internal stresses induced in the frame. If these forces were not opposing and neutralizing the torque of each other, the car would be overturned in seeking to round a curve.

The velocity of the gyrostats is high (3000 revolutions per minute), so that the mass of the fly-wheel may be kept small, and yet store the necessary amount of rotary energy.

In the full-sized car the prime mover is a petrol engine coupled directly to a dynamo. The electric current is distributed to two motors of 40 to 50 horse-power, placed on the bogies and conveying power to the wheels by means of an intermediate shaft. Electric current is distributed also to the gyrostats, the field magnets being on the gyro-frames, and the armatures on the gyro-axles.

One of the most interesting points in Brennan's invention is the means by which the car is automatically raised after being tilted to either side. The principle which he applied is the same as is present in the case of a rapidly-spinning top when 'sleeping.' The top rises to the vertical because of the precessional motion due to the friction between the point of the top and the ground. In the mono-rail car we have the friction between the end of the gyro-axles and the plates fixed on the sides of the car.

We are familiar with the fact that a gyrostatic top can balance itself upon an out-stretched length of string, and thus emulate a tight-rope walker. Brennan demonstrated that a passenger in his gyro-car could travel with safety on a suspended rope. One can imagine a very cheap form of suspension bridge if the gyro-car were to come into practical use, but despite the inventor's actual demonstration with a passenger, it might be difficult to persuade the public to cross a river or a deep ravine with a suspended rope as the sole track.