313.]

ELECTROLYTIC ACTION IN GASES.

643

is barely perceptible. This is, therefore, not the true voltaic arc. On causing the terminals of the Grove's battery in the exhausted tube gradually to approach each other until within about 37mm, the true voltaic arc was suddenly established, and an immense rise of temperature instantly occurred; but the interesting point of the experiment was, that the arc itself was seen to be distinctly stratified.

The passage of the electric spark through compound gases or vapours is attended with a partial separation of their components in the line of the discharge. But the experiments of Perrot (Ann. Chim. Phys. 1861 [3], lxi. 161) appear to have proved that the spark from Ruhmkorff's coil produces in addition a true electrolytic decomposition of the compound vapour. In the case of steam, for example, oxygen appears in larger quantity at the positive wire, and hydrogen in excess is collected at the negative; but a much larger quantity of the two gases is evolved than is due to true electrolysis. Long sparks, if transmitting equal amounts of electricity in equal times, were found to be more effectual in producing decomposition than small ones. In synthetic experiments on the combination of oxygen and nitrogen to produce nitric acid, it was found that long sparks also furnish a larger quantity of acid than shorter ones. If the length of the spark be increased in any given circuit, the gain increases only up to a certain point; the resistance offered by the length of the interposed stratum of air beyond this point diminishes the amount of electricity which circulates, to an extent which more than counterbalances the gain obtained by increasing the length of the spark.

The energy of the secondary induced current in effecting the combination or the decomposition of gases and vapours, is much greater than that of the ordinary cylinder or plate electrical machine. The interposition of a condenser in the induced circuit augments the intensity of the chemical action of the spark; but it decreases the number of sparks in a given time, so that if the spark possess sufficient intensity to pass, no gain in the amount of the body decomposed is effected by the use of the condenser.*

(313) Inductive Action of Currents-Henry's Coils.--When the connexion between the plates of a battery is made by means of a single, long, straight wire, a brilliant spark is seen at the moment that the contact with the battery is broken; but when the con

*Full details of the numerous researches made by Quet and others with Ruhmkorff's coil will be found in Du Moncel's Notice sur l'Appareil d'Induction Electrique de Ruhmkorff, 4th edit.

644

[313.

INDUCTIVE ACTION OF CURRENTS-HENRY'S COILS. nexion is made by means of a short wire, and contact is broken, only a very small spark is produced. When a long wire is employed, the same length of wire, if coiled into a helix, gives a much brighter spark than when it is used merely as a straight conductor. The brilliant spark which is observed when the long wire is used, is produced by the inductive action of the battery upon the electricity of the wire itself. The bright spark obtained from the battery wire on breaking contact arises from a current which is transmitted through the wire in the same direction as that from the battery itself. This inductive action may be entirely diverted, if a second helix, the ends of which are in metallic communication with each other, be placed either within the primary coil or exterior to it.

If the conducting wire be coiled into a helix within which an iron core is placed, the current on breaking contact acquires sufficient intensity to communicate a powerful shock, when the ends of the wire are grasped by the hand at the moment that the wire is disconnected with the battery, although the battery itself may be quite inadequate to produce any shock when its extremities are connected by a short wire. A striking experiment of this kind is related by Prof. Jos. Henry (Phil. Mag. 1840 [3], xvi. 200). A very small compound battery was formed of six pieces of copper bell-wire, each about an inch and a half (4 centimetres) long, and six pieces of zinc of the same size; the current which this arrangement produced was transmitted through a spool of copper wire covered with cotton: this wire was 5 miles (8046 metres) in length, and of an inch (1mm.5) in diameter, and it was wound upon a small axis of iron. The shock, on breaking the connexion with the little battery, was distinctly felt simultaneously by twenty-six persons who had formed a circle by joined hands, and who completed the circuit between the two ends of the wire. The shock which was felt on making contact with the battery was barely perceptible. A current is produced on making contact, but it is feeble, and in a direction the reverse of that emanating from the battery. Even a thermo-electric battery (317), if the current which it yields be passed through the coil, will furnish sparks on breaking contact.

Henry, in the paper above referred to, has made some interesting observations upon the action of the battery current in inducing secondary currents. He employed for transmitting the primary current a flat coil or ribbon of sheet copper about 93 feet (28.3 metres) long and 1 inch (38mm) wide. This ribbon was sometimes coiled in the manner shown at a, fig. 257, sometimes

313.]

INDUCED ELECTRIC CURRENTS-HENRY'S COILS.

645

in the form of a ring as shown at b. This coil was combined under various circumstances with other similar coils, each about 60 feet (18 metres) long, or with helices of fine copper wire of various lengths. The form of ribbon is a very advantageous one, as it offers a large sectional area in the conductor, and thus diminishes the resistance, whilst the different layers of the coil are approximated to each other with the smallest possible intervals between them. When it was coiled as at b, and a helix placed within the ring so formed, each time that the current from the battery through the ribbon was interrupted, a secondary current of considerable intensity was obtained in the helix the helix could be supported upon a plate of glass which rested upon the flat coil, and still the inductive action was obtained; but if a metallic plate were interposed between the coil and the helix, no secondary current was obtained in the helix, because it was transferred to the interposed conducting plate.

By arranging a series of coils in the manner represented in fig. 257, Henry succeeded in obtaining a succession of induced

FIG. 257.

currents by their mutual action. If a represent the coil in connexion with the battery, b and c are arranged to form a continuous coil, through which, by induction, a momentary current is produced each time that the connexion of the coil a with the battery is broken; the current in b c then being direct, or in the same direction as in a. Now if two wire helices be connected together and placed as at d and e, the induced current in c will produce a second induced current, or current of the third order, in de; but this current will be in the opposite direction to that in b c. If ƒ be a ribbon coil placed above e, with its ends united by a small helix at g, a third current, or current of the fourth order, will be obtained, but it will be in the opposite direction to that in d e. If these different currents be compared together, they will be in the direction following:

646

INDUCED ELECTRIC CURRENTS-HENRY'S COILS. [313.

By acting upon the principle just explained, and carefully insulating the coils, currents even of the seventh order have been obtained, the successive currents being alternately direct and inverse.

Similar currents of equal amount, but of lower tension, are obtained each time that the primary circuit is completed, but the direction of the currents in this case is reversed: so that on completing the primary circuit, the currents would be as follows:

These effects are produced by a series of complicated actions which may be summed up as follows:The primary current has the power of producing two induced secondary currents in opposite directions, one on making the other on breaking contact; these currents admit of being separated from each other. They are equal in amount, but the current on breaking contact has the highest tension, and will traverse the greater distance in the form of a spark. Each secondary current in b c may give rise to two opposite tertiary currents in de, but these currents are separated by an interval of time too small to be appreciated, because the secondary current itself is instantaneous. These two tertiary currents are equal in quantity, but differ in tension; the tertiary current produced by the cessation of the secondary being the stronger. Again, each of these momentary tertiary currents is in its turn capable of developing in fg two opposite quaternary currents, equal in amount but differing in tension. At each interruption of the primary current, therefore, we have one instantaneous secondary current in b c, two tertiary in de, and four quaternary in fg. If all these currents were equal in tension as well as equal in quantity, they would neutralize each other; but since their tension is not equal, a series of phenomena is produced, owing to the alternate predominance of the tension of the currents moving in one direction in one circuit, and in the opposite direction in the succeeding circuit.

Henry has shown that induced currents of several successive orders may also be obtained by the momentary passage of electricity occasioned by the discharge of the Leyden jar.

These induced currents not only give powerful shocks, but they magnetize steel bars and produce chemical decomposition,

314.]

INDUCED ELECTRIC CURRENTS—ARAGO'S ROTATIONS.

647

The latter may be shown by interposing acidulated water or a solution of potassic iodide between platinum wires which are in connexion with the ends of the coil. It is easy to obtain either currents of high intensity such as those required to produce shocks, or currents of large quantity such as would be required for magnetizing steel, or for igniting platinum wire, by varying the diameter and length of the conductor. When a long thin wire was employed, as by uniting the two helices as at d and e, a current of great intensity, producing powerful shocks, was obtained; but this same current could be made to induce in the flat coil ƒ a current of greater quantity, but of less intensity.

Owing to these variations in quantity and intensity, the investigation of the laws of such induced currents is complicated and difficult. Abria (Ann. Chim. Phys. 1841 [3], i. 385, and iii. 5) has published some careful researches upon them, but additional experiments are still needed.

(314) Arago's Rotations.-A remarkable exemplification of the facility with which secondary currents are induced by magnetic influence, and of the mutual action of such induced currents, is exhibited by the following experiments of Arago. If a magnet be suspended freely by its centre in a horizontal direction, parallel to a circular disk of copper which can be made to rotate horizontally beneath the magnet, it will be found, if the centre of suspension for the magnet be directly over the axis of the rotating disk, that when the disk is made to revolve with a certain degree of velocity the magnet begins to rotate also in the same direction as the disk; and the more closely the disk and the magnet are approximated, the more rapid is the rotation, whilst at the same time a repulsive action is exerted upon the magnet in a direction perpendicular to the plane of the disk. This rotation occurs as freely when a sheet of paper or of glass is interposed between the magnet and the metallic disk, as when air only intervenes. Disks of other metals also produce this effect upon the magnet by their rotation, but none of them show it so readily as copper; the facility with which the effect is produced being directly as the power of the rotating disk to conduct electric currents. If a narrow strip be cut out of the metallic disk, extending from its circumference to the centre, no motion will be produced in the magnet when the disk is made to revolve; but if the cut edges of the divided disk be connected by soldering a piece of wire across the division, the rotation will be effected as readily as when the disk was entire. From causes similar to those which produce the foregoing results, it is found that if a