The Alternate Current Transformer Vol. 1: The Induction of Electric Currents (1896) — part 18 of 35

These, and many other experiments of a similar sort, indicate that we may regard the inductance of a conductor as an effect which is due to the fact that the current takes time to pene-

• Inaugural Address, Journal Soc. Tfl. £ng., 1886.

MUTUAL AND SELF INDUCTION. 305

trate into the conductor, and that a reduction of the time required to arrive at an equal current density in all parts of the conductor can be effected by any change of form which brings the inner parts of the conductor nearer the surface, or makes them more get-at-able from the dielectric, The better the conductor the slower is the rate of equalisation of current density over its cross-section — in other words, the less rapid is the rate of diffusion of the current inwards from circumference to centre; and the " time constant " of the circuit, or the time in which, under the operation of a constant electromotive force, the current will rise to a definite fraction of its maximum value, is a quantity proportional to the con- ductivity of the circuit, and to another factor (the formal inductance), which may be considered as expressing the accessibility of the conductor as regards geometrical form to the entrance of the current into it ; and finally, in the case of magnetic conductors, to a quantity (the permeability) deter- mined by the capacity of the conductor to utilise part of this incoming energy in producing magnetisation of its substance.

We are indebted to a Paper read before the Austrian Academy by Prof. Stefan for a simple and intelligible analogy helping to comprehension of the electrical distribution of current in a conductor. Imagine a cylinder or cylindrical wire heated throughout to a uniform temperature ; let it be suddenly brought into a chamber where the temperature is higher. The outer layers of the cylinder will rise first in temperature, and gradually convey the heat to the successive interior layers. Precisely the same order of phenomena occurs if an E.M.F. is suddenly set up between the ends of the wire or cylinder. The current during the variable state passes first through the outer layers alone, and gradually penetrates the inner layers. When the external E.M.F. is suddenly removed, the action, of ceasing in the current resembles the cooling of the cylinder. The current ceases first, or, rather, most quickly, in the outer layers.

Now, let us imagine the cylinder transferred to and fro from a very hot place to a cool one. It is easy to see that waves of heat will pass in and out radially, and also that the condition at any instant will depend largely upon the rate of transference.

x

306 MUTUAL AND SELF INDUCTION.

When the rate of motion is sufficiently slow, the waves of heat passing any given point in the radius of the wire follow exactly with the periodic changes of position. The amplitudes of these variations have values which decrease from the sur- face inwards. "When the rate of change is increased, the amplitude of the waves gets shorter and shorter, and at an infinite velocity of transference the wire would acquire an equable temperature throughout. In the electrical analogue the rate of transference corresponds to the inverse of the periodic time of an alternating current. The heat conducting power of the material corresponds to electrical resistance.

Prof. Stefan gives some numerical illustrations which are useful. If an alternating current have a frequency of 250 per second and is passed through an iron wire of 4mm. diameter, the amplitude of the waves of current density is about twenty- five times greater upon the surface than at the axis of the wire. For double the number of vibrations per second the external amplitude becomes only six times as great. The difference of phase is one-third the duration of the vibration in the first case and one-half in the second. The latter statement implies that the external current is at a given moment actually in the reverse direction to the internal current.

For non-magnetic wires the difference is not nearly so marked, and it decreases as the specific resistance increases. For a copper wire of 4inm. diameter, with a periodic time of one 500th second, the difference between the current density at the surface and at the centre is only 14 per cent. If, how- ever, the copper wire Jbe increased to 20rnm. diameter, then we should get the same difference as in the particular iron wire quoted.

. It is obvious that this non-homogeneous distribution of current must increase dissipation of energy, which is, of course, proportionate in each transverse section to the square of the current strength, at that spot. In the case of the iron wire quoted,. the increase of resistance is 48 per cent, at the 250 per second frequency, and 100 per cent, at the higher speed. As the frequency of alternation is increased, the resultant self-induction of the circuit is lessened, but although the true resistance is increased, the impedance may be diminished on the whole.

MUTUAL AND SELF INDUCTION. 307

§ 12. Electromagnetic Repulsions.— The effects of self and mutual induction in conducting circuits are well illustrated in studying the dynamical actions taking place between con- ductors conveying currents and other circuits. On the 2nd day of October, 1820, Ampere presented to the Royal Academy of Sciences in Paris an important memoir, in which he summed up the results of his own and Arago's investigations in the then new science of electro-magnetism, and crowned that labour by the announcement of his great discovery of forces of attraction or repulsion existing between conductors conveying electric currents.* Kespecting that achievement, when deve- loped in its experimental and mathematical completeness, Clerk Maxwell speaks of it as "one of the most brilliant in the history of physical science." Our wonder at what was then accomplished is increased when we remember that hardly more than two months before that date John Christian Oersted had startled the scientific world by the announcement of the discovery of the magnetic qualities of the space near a current- traversed conductor. Oersted named the actions around the conductor, which we now refer to as the magnetic field, the electric conflict, and in his first Paper, f in describing the newly-observed facts, he says : " It is sufficiently evident that the electric conflict is not confined to the conductor, but is dispersed pretty widely in the circumjacent space." " We may likewise collect," he adds, " that this conflict performs circles round the wire, for without this condition it seems impossible that one part of the wire when placed below the magnetic needle should drive its pole to the east and when placed above it to the west." These words are taken from the original paper, which stimulated the philosophic thought of

• Me"moire pre"sente a 1' Academic Royale des Sciences le 2 Octobre, 1820, ou se trouve compris le re'sume' de ce qui avait ete lu a la meme Academic lea 18me eb 25me Septembre, 1820, sur les effets des courants Electrique, par M. Ampere. See Vol. XV. Annales de Chimie, 1820.

t In the Annals of Philosophy for October, 1820, VoL XVI., p. 274, is to be found an English translation of Oersted's original Latin essay, dated July 21, 1820, describing his immortal discovery. This Paper is entitled "Experiments on the Effects of a Current of Electricity on the Magnetic Needle," by John Christian Oersted, Knight of the Order of Danneborg, Professor of Natural Philosophy in Copenhagen.

x2

308 MUTUAL AND SELF INDUCTION.

Ampere, and finally led him to the valuable discovery of the electro-dynamic actions between conductors conveying currents.

Eeferring the student to text -books on Physics for the complete statement of Ampere's work, we may describe briefly some illustrations of the interactions of two circuits traversed by currents in the same or opposite directions. Holding a circular coil traversed by a continuous electric current near to a similar circuit free to move, we find that when the circuits are parallel to each other there is an attractive force between them if the currents in adjacent parts of the circuits flow in the same direction, and a repulsive effect if they flow in the opposite. This is the electro-dynamic action discovered by Ampere and utilised in the construction of instruments for the measurement of electric currents in practical work. If one conducting coil, such as that of an electro-magnet, is traversed by an alternating current, and the other is simply a closed circuit or coil placed a little distance off, but in its field, it has been previously explained that the closed circuit becomes the seat of an alternating induced current, which, if the inducing current is sufficiently powerful, can be made to render itself evident by illuminating a small incandescent lamp placed in the secondary circuit.* We notice, however, that in perform- ing the experiment the secondary circuit must be so placed that the magnetic induction of the primary coil perforates through the secondary circuit. If the secondary circuit is held in such a position that the reversal of direction of the primary current causes no reversal of direction of the magnetic field traversing the secondary circuit, because it is not linked with any of the lines of induction of the primary, the secondary circuit is no longer the seat of any induced current.

• The experiments described in the following paragraphs can be shown with an alternating current magnet, having a core formed of a bundle of fine iron wire about 3in. in diameter and 12in. long, excited by an alternat- ing-current dynamo, giving a current at an electromotive force of about 100 volts. A small shelf around the core a little above the middle serves as a support for rings, &c., to be projected. The performance of these experiments on a scale suitable for large audiences requires from 10 to 15 horse-power at least, and can hardly be shown well unless the alternator can provide a current of 100 amperes at 100 volts available at the moment of maximum demand.

MUTUAL AND SELF INDUCTION. 309

This electromagnetic induction thus taking place across space is not stopped by the interposition of a non-conducting screen. The magnetic induction passes freely through a deal board or a plate of glass, but if we interpose a thick sheet of copper (Fig. 112) we thereby screen the secondary circuit from the inductive action of the primary. The rapid heating of this copper screen makes us aware that the secondary currents are induced in the copper sheet in the form of eddy currents, and it therefore screens the secondary circuit, as already explained, because the inductive action of these eddy

FIG. 112.— Copper Plate interposed between a Primary and a Secondary Coil and shielding the Secondary from Induction.

currents on the side remote from the magnet is exactly equal and opposite in inductive effect to that of the primary circuit on the secondary coil.

If a continuous current is sent through the coils of an electro-magnet, and magnetises its iron core very powerfully, it is found to be impossible to strike the pole of the magnet a sharp blow by means of a sheet of copper. Holding a sheet of copper over such a magnetic pole, and exciting the magnet, the hand holding the copper sheet feels a repulsive action at

310 MUTUAL AND SELF INDUCTION.

the moment when the current is put on and an attractive action when it is cut off. If we try to slap the magnet pole sharply with the copper sheet, it is found that this repulsive force prevents anything like such a sharp blow being given to- the pole when the current is on as can be given when the current is off. Moreover, when a very powerful electro- magnet is employed, it is found that a disc of copper let fall over the pole does not fall down sharply and quickly on to it when the current is flowing through the coils of the magnet, but settles down softly and slowly as if falling through some viscous fluid. The correct explanation of these facts is to be found in the statement that the motion of the conductor towards the magnetic pole causes eddy electric currents to be generated in it by electro-magnetic induction, and that these, being in the opposite direction to the exciting current of the magnet, cause a repulsive force to exist between the inducing and secondary circuits, which creates the apparent resistance we feel.

In order to exhibit the stress brought into existence between an electro-magnet and a metal sheet held near it when induced currents are set up in the disc, we may arrange the following experiment : — Over the pole of a powerful electro-magnet we balance a small disc of copper, the size of a penny, carried on one end of a delicately-balanced bar. A mirror attached to that bar serves to reflect on to a screen a ray of light indicating the smallest motion of the copper disc. On magnetising the magnet the copper is suddenly repelled, but comes to rest again immediately in its initial position. When the magnet is demagnetised the copper experiences a momentary attraction. Or we may illustrate the same action in another way. Consider, for instance, a ring of copper hanging in front of the pole of an electro-magnet (see Fig. 113), having the plane of the ring perpendicular to the lines of magnetic force proceeding out from the pole. Let the magnet be an electro-magnet, and let the pole be suddenly made a north or marked pole. Lines of magnetic force are thrust into the aperture of the ring. This magnetic flux, in accordance with a well-known law, generates an inductive electromotive force, which causes a transient current to flow round the ring in a counter-clockwise direction, as looked at

MUTUAL AND SELF INDUCTION. 311

from the north magnetic pole. The ring becomes virtually a magnetic shell, having a north pole facing the north pole of the exciting magnet. By the fundamental laws of action between currents and magnets established by Ampere, the ring experiences a slight repulsive force, due to the electro- dynamic action between the current in the ring and the magnetic pole. The generation of the momentary induced current in the ring is accompanied by an electro-dynamic impulse tending to thrust it away from the pole.

Suppose, next, that the electro-magnet is demagnetised. The ring has generated in it a reverse induced current flowing in the direction the hands of a clock move when looked at from the magnetic pole. This is also accompanied by an electro-dynamic attraction of the ring towards the pole, but

FIG. 113. — Copper Ring hung in the Field of an Electro-magnet, and Repelled or Attracted when the Current is put on or cut off.

which is much more feeble than the previous repulsion. These attractions and repulsions are obviously due to the Amperian stress set up between the magnet and the metal by reason of the induced currents set up in the latter. It has been pointed out by Prof. S. P. Thompson that Ampere himself probably observed an effect of this kind (Proc. Phys. Soc. of London, Vol. XIII., p. 493, "Note on a Neglected Experiment of Ampere."). Impulsive effects of this nature have been also studied by Prof. Vernon Boys (see Proc. Phys. Soc. of London, Vol. VI., p. 218, " A Magneto -electric Phenomenon ").

Let us in the next place consider a circuit, say a closed conducting ring, suspended in front of the pole of an electro- magnet, and let the coils of the electro -magnet be trans-

312

MUTUAL AND SELF INDUCTION.

versed by an alternating current of electricity (Fig. 113). The magnetic field of the magnet is then an alternating field. We shall suppose it to vary in strength according to a simple periodic law. The closed circuit is therefore subjected to an inductive action, and we know that the induced electro- motive force in that circuit is at any instant proportional to the rate of change of the magnetic field in which it is immersed. If, therefore, the variation in strength of that field is represented geometrically by the ordinates of a periodic curve, the varying electromotive force acting in the ring circuit is represented by the ordinates of another such curve of equal

FIG. 114. — Diagram showing the Equality of the Attractive and Repulsive Impulses in a Non-inductive Circuit when held in an Alternating Magnetic Field.

wave length, shifted a quarter of a wave length behind the first. In the diagram (see Fig. 114) the variation of the in- duced magnetic field, and the induced electromotive force in the circuit, are represented as usual by two harmonic curves. This induced electromotive force creates an induced current flowing backwards and forwards in the ring, and we shall, in the first instance, suppose that this current flows in exact synchronism with its electromotive force. The induced current and the inducing magnetic field may there- fore be represented as to relative phase and strength by the curves in the diagram (Fig. 114). The dynamical action,

MUTUAL AND SELF INDUCTION. 313

•or the force which the ring experiences, is at any instant proportional to the product of the strength of the magnetic field in which the ring is immersed and to the strength of the induced current created in it. If we multiply together the numerical values of the ordinates of these two curves at any and every point on the horizontal line, and set up a new ordinate at that point representing this product, the extremi- ties of these last ordinates define a curve, which is a curve representing the force acting on the secondary circuit ; and it is seen from the diagram (Fig. 114) to be a wavy curve having a wave length equal to half that of the first two curves. More- over, the whole area enclosed between the outline of this force •curve and the horizontal line represents to a certain scale the time integral of that force, or the impulse acting on the secon- dary circuit, and the theory shows us that, under the assump- tions made, the secondary circuit so acted upon experiences in each period of the current four impulses, two positive or repul- sive, and two negative or attractive. Hence, it follows that such an ideal conducting circuit held in front of an alter- nating electro-magnet should experience a rapid alternate series of equal pushes and pulls, or of little impulses to and from the magnet. These equal and opposite impulses in quick succession would neutralise one another, and our supposed circuit would not, on the whole, be subject to any resultant force.

When we present a real conducting circuit to the pole of an electro-magnet traversed by a powerful alternating current, we find that under the actual circumstances there is a powerful repulsive action between the pole and the circuit. With a powerful alternating current electro-magnet striking effects of repulsion may be thus shown.

If we hold a copper ring over the pole of a powerful vertical alternating electro-magnet, we find at once that there is a perceptible and strong repulsion. Letting the ring go, it jumps up into the air, impelled so to do by the electro-magnetic repulsion acting upon it (Fig. 116). All good conducting rings will execute this gymnastic feat, and rings of copper and aluminium are found to be most nimble of all. Eings of zinc and brass are sluggish, and a ring of lead will not jump at all. Prof. Elihu Thomson was the first to call attention to this strong repulsive action between conducting rings and an

314

MUTUAL AND SELF INDUCTION.

alternating electro-magnet. He has thus described his first notice of these effects : — " In 1884, while preparing for the International Electrical Exhibition at Philadelphia, we had occasion to construct a large electro-magnet, the cores of which were about Gin. in diameter and about 20in. long. They were made of bundles of iron rod about j^-in. in diameter. When complete the magnet was energised by a current from a continuous-current dynamo, and it exhibited the usual powerful magnetic effects. It was found also that a disc of sheet copper of about ^in. in thickness and lOin. in diameter, if dropped flat against a pole of the magnet, would settle down softly upon it, being retarded by the development of currents

FIG. 115.

in the disc, due to its movement in a strong magnetic field, and which currents were of opposite direction to those in the coils of the magnet. In fact, it was impossible to strike the magnet pole a sharp blow with the disc, even when the attempt was made by holding one edge of the disc hi the hand and bringing it down forcibly towards the magnet. In attempting to raise the disc quickly off the pole a similar but opposite action of resistance to movement took place, showing the development of currents in the same direction as those in the coils of the magnet, and which current, of course, would cause attraction as a result. The experiment could be tried in another way. Holding the sheet of copper by one

MUTUAL AND SELF INDUCTION.

315

edge, just over the magnet pole (see Fig. 115), the current in the magnet coils was cut off by shunting them. At that moment was felt an attraction of the disc, or a dip towards the pole. On starting the current the plate experienced a powerful re- pulsion." The question may then be asked : Why is it the metal rings are always repelled by the alternating magnet ? The explanation is not difficult to find. The real ring possesses a quality, called its inductance, of which we took no account in our examination of the case a moment ago. As a con-

FIG. 116.— Aluminium Kiug projected from the pole of an Alternating Electro-magnet, and floating over the pole when restrained by three strings.

sequence of this inductance we have seen that the current induced lags in phase behind the inducing electromotive force. We have then to correct the diagram considered just now, to make it fit in with the facts of nature, and we must repre- sent the periodic curve which stands for the fluctuations of the induced current in the ring as shifted backwards or lagging behind the curve which represents the electromotive fore.) in the circuit brought into existence by the fluctuating

316 MUTUAL AND SELF INDUCTION.

magnetic field. Making this change (Fig. 117), and forming, as before, a force curve to represent the impulses on the ring, we then find that, owing to the "lag" of the secondary current, one set of the impulses, namely, the positive or repulsive impulses, has been enlarged at the expense of the negative or attractive impulses. Theory, therefore, points out that, as a consequence of the self-induction of the ring, the balance between the attractive impulses and the repulsive impulses is upset, and that the latter predominate. The real ring behaves therefore, very differently to the ideal non-inductive ring. The real ring is strongly repelled, because the resultant action

/"N

FIG. 117. — Diagram showing the Inequality of the Attractive and Repul- sive Impulses in the case of an Inductive Circuit when held in an Alter- nating Magnetic Field.

of all the impulses is to produce, on the whole, an electro- magnetic repulsion. This repulsion is evidence of the self- induction or inductance of the circuit exposed to the magnetic field, and it forms a new way of detecting it. But although this is part of the truth, it is not the whole truth. The lag of the induced current in the ring, and hence the predominance of the repulsive impulses, depends on the conductivity of the material of which the ring or circuit is made ; and the better

MUTUAL AND SELF INDUCTION. 317

this conductivity the greater is that repulsion, because both the induced current and the lag are thereby increased. Hence it comes to pass that there are two factors involved in making this repulsive effect, the conductance of the ring or disc and its inductance. For equal conductivities, the greater the self- induction the greater the repulsion. For equal self-inductions, the greater the conductivity of the circuit so much the more repulsive effect will be produced.

We can show the effect of the relative conductivity of discs of equal size, and therefore of equal self-induction, by weighing similar discs of various metals over an alternating pole. If we take discs of copper, zinc, and brass of equal form and size, and weigh these discs on the scale pan of a balance placed over the pole of an alternating current magnet, the scale pan being made of a good non-conductor, we can measure the electro-magnetic repulsion on the disc by the loss in weight it experiences.*

The same result can be illustrated by placing over the pole of our alternating magnet a paper tube. Taking one of the copper rings, and first exciting the magnet, we let the ring drop down the tube. It falls as if on an invisible cushion that buoys it up, and it remains floating in the air. If rings of different metals and equal size are placed on the tube, they float at different levels like various specific-gravity beads in a liquid. The greater the conductivity of the ring the greater is the repulsion on it in any given part of the alternating field, and hence the highly conducting rings will be sustained in a weaker field than the feebler conducting ring, assuming the rings to have about equal weights. Moreover, we are able to show by another experiment the fact that these rings are traversed, when so held, by powerful electric currents. If we press down the copper ring upon the zinc or brass ring floating beneath it, the rings are attracted together and the copper ring holds up the zinc. This is obviously because the rings are all traversed by induced currents circulating in the same direction.

It is, of course, an immediately obvious corollary, from all that has just been said, that any cutting of a ring or disc which

• Experiments of this kind have been made by M. Borgman. See Comptes Rendus, No. 16, April 21, 1890, p. 849, and also February 3, 1890V Vol. CX., p. 233.

318 MUTUAL AND SELF INDUCTION.

hinders the flow of the induced currents causes the whole of the repulsion effects to vanish. We illustrate this by causing a ring of copper wire to jump off the pole, and then cutting it with pliers, find it has ceased to be capable of giving signs of life. When the metallic masses or circuits which are pre- sented to the alternating magnetic pole are of very low resistance the electro-magnetic repulsion may become very powerful, many pounds of thrust or push being produced by apparatus of quite moderate size. It is, in fact, quite startling to hold over the pole of a very powerful alternating magnet a very thick plate of high conductivity copper. It would greatly surprise anyone not acquainted with these principles to be told that a massive copper ring weighing eight or ten pounds could

FIG. 118. — Copper Ring "floating" in air over the pole of an Alternating Current Electro-magnet, \vhen restrained by strings.

be made to float in the air, but it is possible to show this easily. The ring needs to be tethered by light strings (Fig. 118) to prevent it from being thrown off laterally, although these strings in no way support its weight.

One of the most beautiful of Prof. Elihu Thomson's experi- ments exhibits this effect of electro-magnetic repulsion on a closed coil, which is buoyed up in water by a small incandes- cent lamp in circuit therewith. In a glass vase is floated a little glow-lamp like a balloon (Fig. 119). The car con- sists of a coil of insulated wire, and the ends of this coil

MUTUAL AND SELF INDUCTION. 319

are connected with the lamp. The whole arrangement is accurately adjusted to just, or only just, float in water. Placing the vase over an alternating magnetic pole, the magnetic induction creates a current in the coil which lights the lamp, and, moreover, the electro-magnetic repulsion on the coil causes the lamp and coil to rise upward in the water. There is also another class of actions — namely, deflections and rotations— produced by electro-magnetic repulsion on highly conducting discs or rings. If the conducting ring or disc which is presented to the alternating magnet is con- strained by being fixed to an axis around which it can rotate,

FIG. 119. — Incandescent Lamp and Secondary Coil floating in water and Repelled by an Alternating Current Electro-magnet placed beneath.

the action may reduce to a deflective force. On presenting a flat suspended disc to the pole, the disc is prevented by its constraint from being repelled bodily ; so it sets its plane parallel to the lines of magnetic induction, and places itself in a position such that the induced currents in it are reduced to a minimum. On this principle, before becoming acquainted with Prof. Elihu Thomson's original work, the author devised,

320 MUTUAL AND SELF INDUCTION.

a little copper disc galvanometer for detecting small alter- nating currents.

This deflection by an alternating current of a copper disc suspended -within a coil with its plane inclined to the plane of the coils was, in March, 1887, noticed independently by the author, who subsequently described a copper disc galvanoscope for alternating currents based on this fact (see The Electrician, May 6, 1887). He did not at the time know how thoroughly Prof. Thomson had explored the phenomena, but the sub- stantial explanation of the facts as above given had already occurred to him.

More interesting than the deflective actions are those which result in the production of continuous rotation in highly

FIG. 120.— Alternating Electro-magnet with Shaded Poles, causing a- Copper Disc placed between the Jaws to revolve.

conducting bodies placed in an alternating field. We employ for this purpose an electromagnet having a laminated iron core (see Fig. 120), the ends of the iron circuit being provided with copper bars, which embrace and cover portions of the polar terminations of the magnet. When the magnet ia excited by a periodic current, these secondary circuits become the seat of powerful induced secondary currents. Taking in hand a large copper disc pivoted at the centre and held in a fork, we hold this wheel so that part of the disc is inserted between the jaws of the electro-magnet. Immediately, rapid

MUTUAL ANP SELF INDUCTION. 321

rotation is produced. The reason is not far to seek. The alternating field creates induced currents, both in the closed coils and in the neighbouring portions of the disc ; and the con- ductors in which they flow are therefore drawn together. If the polar coils are so placed as to partly shield the poles these attractive actions act unsymmetrically on the disc and pull it continuously round. The action is, perhaps, better illustrated by a simpler experiment. If we hold a pivoted copper disc (Fig. 121) symmetrically over an alternating pole, the action of the pole is one of pure repulsion on the disc, and it causes no rotation in it. When a copper sheet is so placed as to shield or " shade," as Prof. Thomson calls

FIG. 121.— Revolution of a " Shaded " Copper Plate held over the Pole of an Alternating Current Electromagnet.

it, part of the magnetic pole, currents are induced both in the fixed plate and in the movable one. The fixed disc shields part of the other from the induction of the pole, and I e i '-e causes the induced currents in that plate and disc to be so located that they are in positions to cause continual attrac- tion between the conductors and to continuously pull round the movable disc into fresh positions, so creating regular rotation.

This principle of " shading " a portion of a conductor from the inductive action of the pole, and so causing the eddy currents in it to be located in a portion of its service and tc cause attraction between that conductor and the shading conductor, is capable of being exhibited in various ways.

322 MUTUAL AND SELF INDUCTION.

We place on a copper plate a light, hollow, copper ball (see Fig. 122), and support it in a little depression in a copper plate. Holding the arrangement over the alternating magnet, the ball begins to spin round rapidly when the magnet is excited. This rotation is caused by the continual attraction of the eddy currents induced in the fixed plate and in that part of the ball which is not shielded from the pole by the plate. We may vary the experiment, and exhibit many more or less curious and amusing illustrations of it. If we float these copper balls in water (Fig. 123), and place the glass bowl containing them over the alternating pole, the interposi-

Fio. 122. — Light, Hollow, Copper Ball standing in a depression on th« edge of a Copper Plate, and set in rotation when held over the Pole of an Alternating Electro-magnet.

tion of a copper sheet between the pole and the balls causes the latter to begin to spin in a highly energetic manner.*

Amongst other illustrations of the principles above described Prof. Elihu Thomson invented a novel form of electro- magnetic gyroscope (Fig. 124). Over the alternating magnet a gyroscope of the usual form is suspended. The wheel of the gyroscope is made of iron, and the tyre of the wheel is a thick copper band. Immediately the iron core is magnetised, the

• For a mathematical discussion of these electro-magnetic rotations, see Phif.. Trans. Royal Soc., Vol. CLXXXIIU., 1892, p. 279, Mr. G. T. Walker on "Repulsion and Rotation produced by Alternating Electric Currents."

MUTUAL AND SELF INDUCTION.

323

gyroscope begins to rotate with great rapidity over the pole. In this case the unsymmetrical disposition of the eddy currents in the copper band around the wheel is sufficient by itself to cause the rotation to occur. The phenomenon, however, which lies at the bottom of all these effects is that the self-induction of the secondary circuit causes the eddy currents to be delayed in phase behind the magnetising field, and hence to persist into the period of reversal of that field, and so produce the repulsion between the primary conducting circuit and that part of the secondary conducting circuit in which the eddy currents are set up.

Fro. 123. — Hollow Copper Ball floating in water over an Alternating Current Electro-magnet, and caused to revolve by the interposition of a " shading " copper plate.