when the poles are kept far apart. It is then a dense hazy spark, with an aureole enveloping the usual spark. This hazy spark or aureole I take to be an evidence of quantity, as it never takes place but where the resistance of the circuit is small as compared with the electro-motive force. Now, in building up a coil, when you examine the spark that is got after a fourth say of the wire is wound, you find that when the aureole ceases the spark itself can no longer be got; whereas when the whole wire is on, although sparks of several inches in length are got, the aureole can be got only when the poles are near. In the first case tension and quantity march together; in the latter, tension goes on ahead, but quantity lags behind. The same thing is illustrated when the same coil occupies different positions on the same electro-magnet. When the secondary coil is at the centre, the hazy spark and the clear spark are more nearly of the same length than when the coil is shifted towards the end.

These being the principles on which I have constructed my coil, I shall now briefly detail how they have been carried out in the coil before you. In my first attempt to construct it I had three double bobbins of the kind I shall afterwards describe, and the wire was carried out nearly to the end of the primary coil, and in external diameters in accordance with Lenz and Jacobi's observations. This construction, which promised and acted exceedingly well when two bobbins were filled, broke down in consequence of its very power; when all three were filled, the electricity broke through the bobbin into the primary coil. A vulcanite bobbin would have prevented this. In the coil before you there is only one large double bobbin employed. A section of it is shown in Fig. 2. The wooden tube AB contains the primary coil, which is two feet in length. Three discs of gutta percha, G1 G2, and G1, are fixed to this tube, having been melted on to it by a hot iron. The two side discs, G, and G,, are six inches in diameter and half an inch thick. The central disc, G2, is a inch thick. The ends of the primary coil are left free; the secondary wire only occupying 17 inches of its length. The outer diameter of the secondary coil is 5 inches, and it in

creases to 6 inches at the centre. The length of wire on the secondary coil is about 7 miles. It is of the finest silkcovered copper wire. The winding of the secondary wire is different from that found in most if not in all other coils. In the secondary wire there is no electricity at the middle, which is the neutral ground between the opposite electricities of the ends. The power increases as the ends are approached, and reaches a maximum at the terminals, where the two opposite electricities unite. The insulation is made with a two-fold object, to insulate the coils from each other

and the ends from external conductors. In the ordinary construction the middle of the coil lies half-way through the bobbin and well insulated, while one end lies in dangerous proximity to the primary coil. The result is, that with the best insulation the pole that comes from the inner end of the coil has much less power than the outer one. This comes either from leakage, or from the inmost coil acting as the coating of a Leyden jar. Now, in this coil the middle is placed next the primary coil, and the two poles stand in precisely the same relation to the whole. The bobbin to effect this is divided at the middle by a diaphragm, which is pierced near its attachment to the wooden tube so as to allow both halves to communicate with each other. The secondary wire on each side leaving this passage as a starting point, is coiled in opposite ways, so that the whole forms one continuous coil cut in the middle. Both poles thus possess equal power, and the tension is removed from the central tube to the outside, where it is best insulated.

For this division of the bobbin I am indebted to a suggestion from Dr Strethill Wright, whose contributions to this department of science have more than once been acknowledged by the Society to be both valuable and original. The great difficulty in this construction lies in keeping the spark from travelling through or over the disc. It was only by thickening the disc, and enlarging its diameter beyond the coil, that proper insulation was secured.

With eight Bunsen cells the coil gives sparks of from 63 to 7 inches in length. The aureole can be distinctly seen in sparks 4 inches in length. Both the tension and quantity are thus, considering the length of the secondary wire, highly satisfactory.

In concluding, allow me to express my obligations to Mr Hart, who has laboured most enthusiastically to bring the construction of the coil to a successful issue. Whatever be your opinion of my plans as "architect" of the coil, I am sure you will not have two opinions as to his merits as "builder."

On a New Current-Interruptor for the Induction Coil. By ROBERT M. FERGUSON, Ph.D.

It is a well-known fact in the science of electricity, that when a spiral of very fine wire is made to dip at its lower end into a cup of mercury so as thereby to complete a galvanic circuit, the spiral coils up and shortens. If the end of the spiral be made to dip very little, the shortening of the spiral will lift the end out of the mercury and open the circuit. The weight of the spiral brings the end down again so that it again dips. The action of the current once more draws it out, and thus an alternate coiling and uncoiling of the spiral keeps up a continuous interruption of the circuit. It struck me that this action, which had hitherto only be used as an illustration of the principle that currents in the same direction attract each other, might be usefully

* Read before the Society and Current-Interruptor exhibited in action, 9th April 1866.

applied to act as a current-interruptor for the induction coil. In this form I found it was quite inapplicable. The great resistance offered by the thin wire reduced the strength of the current so much, that the spark got from the coil lost greatly in power. Again, the spiral hanging only by one end had a very unsteady motion. Lastly, the action was far too languid. With a view to intensify the action, I introduced a thin bar of iron into the spiral; the motion of the spiral was thereby much accelerated. In this case the shortening of the coil illustrates the principle, that in a magnetic field (ie. the region in the neighbourhood of a magnet) currents are urged in a direction at right angles to the lines of magnetic force. These lines at the poles of the magnet proceed right outwards from its sides, hence the turns of the spiral conveying the current are urged upwards by the action of the iron rod. The iron rod becomes a magnet each time the current passes in the spiral, and ceases to be so when the end of the spiral leaves the mercury. The magnetic action of the bar, and the electric action of the various convolutions of the spiral, act together in producing the coiling up of the spiral. The whole constitutes, in fact, an electro-magnet with its coil movable at one end. If the coil of an electro-magnet were free to move at both ends, it would coil up at each of the poles to the middle as a neutral point. To remedy the other defects already mentioned, I took a No. 14 wire of copper, which being tolerably thick, would in the small amount taken of it offer little or no resistance to the passage of the current, and coiled it into a spiral about one inch in diameter, and six inches in length. Such a spiral would not, in consequence of its stiffness and the small number of turns in it, give any result in the way described. To make it serviceable, I fixed its ends to corks sliding with some friction on a rod of iron, and drew it out to about ten inches. I had previously soldered a wire rather below the middle of it, coming out at right angles from the spiral, which being turned down at the end, should act as a dipper when the spiral wire was placed upright; one pole of a Bunsen cell was connected with the top of the spiral, and the other with the cup of mercury. The projecting wire

just barely touched the surface of the mercury. On being made gently to oscillate by the hand, it was kept in a constant state of motion. The play of the dipper was between a quarter and a half of an inch, more than sufficient to make a sudden and complete break of the circuit. When used with the induction coil it produced excellent results. A spiral fixed in this way is a most delicate apparatus. When made to oscillate with the hand it continues in motion for a long time afterwards under its own elasticity, and little is needed to keep it in motion. This little the electric and magnetic action just described more than supplies, and the spiral keeps moving up and down with the steadiness of a pendulum. The apparatus I use here (fig. 1) is of the same size as the spiral just described, the turns, for the sake of simplicity, being shown few in number. The core cc consists of

W

d

Fig. 1.

an iron rod fitting into a brass tube below, the junction of the two being rather lower than the point where the wire projects. The upper end of the spiral is fixed upon a ring r which slides on the core, and may be fixed by a binding screw to it. The lower end is insulated from the core by the gutta percha ring r to which it is fastened. The current ascends the core from the binding screw a, descends the upper part of the spiral, and reaches the cup d by the projecting wire w, thence it passes to the pillar p and the