the cub is licked into shape, this unseemly protuberance may be reduced, if not entirely removed. It is much to be regretted, that Government have taken up the idea, that with the exhibition of lights their duty for the protection of the mariner is fulfilled, and that the noble scheme of life-boats is left to languish on public charity, when it, equally with lights, ought to be supported from a toll levied on shipping so long as that principle exists. Fog signals are not less important when required than lights, and a very small addition to the present tolls, or even an annual grant out of the light fund, would suffice to put life-boats on a permanent footing, as well as to execute and make a complete work of the infant now entrusted to your tender consideration.

On a New Method of Constructing Induction Coils. By ROBERT M. FERGUSON, Ph.D., F.R.S.S.A., Edinburgh Institution.*

Some years ago, I got from Mr Hart, of College Street, an induction coil, which gave, when worked with eight Bunsen cells of moderate size, a spark of nearly four inches. The insulation kept well up, so that after the lapse of several years the coil had not fallen much in strength. Wishing, however, to bring it back to its former power, and if possible beyond it, I determined to get it taken down and wound up again with Mr Hart's assistance. It is this process that I purpose detailing to you in the present paper.

The insulating material of my former coil was guttapercha; thick gutta-percha for the frame, and thin guttapercha, usually called gutta-percha paper, for the insulation of the different coils from one another. Mr Hart's method of winding on this paper admits of the secondary wire, the most expensive element in the construction of a coil, to be taken off without loss. In the unwinding, we found that the exterior gutta-percha alone had suffered. I am informed

* Read before the Society, and Induction Coil exhibited in action, 9th April 1866. For this and the succeeding communication the Society awarded the Hepburn Prize, value twelve sovereigns.

that it is a fact well known to practical electricians, that gutta-percha exposed to the air losses its softness and becomes brittle. In this state it losses also its insulating virtues. To guard against this change, they carefully enclose their gutta-percha covered wires in vulcanite tape. In the coil, the central tube on which the secondary wire was wound, and which consisted of gutta-percha a quarter of an inch thick, was wellnigh gone. The sides of the bobbin, which were of the same material, half an inch thick, and had been varnished over, were little the worse. The gutta-percha paper near the outside had suffered, but to a small extent compared with the central tube. But the paper in the inside, lying between the various coils of the secondary wire, was quite perfect. The operation of unwinding proved the excellence of Mr Hart's method of insulating coil from coil. Nowhere could any puncture be discovered in the paper, or any change in its structure, even on close examination.

Having thus discovered the merits and demerits of guttapercha as an insulator, the first object in the reconstruction was to make a bobbin that would insure the former, and avoid the latter. My first idea was to get a bobbin of vulcanite, the well known compound of india-rubber used for making combs, and which is the best insulator known. Mr Hart, with this view, applied to the vulcanite works in this city, but could not persuade those in management to undertake the manufacture of a bobbin. Such apparatus did not lie in their programme of manufacture. He learned, moreover, that should such a bobbin be put in hands, the price of it alone would have far exceeded the sum I meant to spend on the entire reconstruction. A vulcanite bobbin was thus out of the question, and I had therefore to be contented with one of wood and gutta-percha, taking care to protect the latter from the action of the outer air as much as practicable. The tube of the new bobbin was of wood well dried and varnished. To it were fixed discs of gutta-percha, so thick and so well varnished that exposure would have but a very slow action on them. Round the tube were wound several layers of gutta-percha paper, to such a thickness that the tube was virtually one of gutta-percha, but screened by the action of

the varnished wood from the air. By encasing the whole exterior when finished in a layer of paraffin, the action of the air, it is expected, will be reduced to a minimum. Such a composite bobbin did not cost a twentieth of one of vulcanite.

It was not, however, so much in the material of the bobbin as in the shape of the secondary coil that I wished to make alterations. The secondary wire in the first coil was wound in the usual way, beginning at the centre and leaving off at the outside, the poles thus coming, one from the outside, the other from the inside of the coil. The shape was, as is usual, cylindrical, the same amount of wire being wound on every part. This uniform thickness, the form universally adopted, seemed to me to be inconsistent with what had been revealed of the inductive action of an electro-magnet. Jacobi and Lenz, upwards of twenty years ago, showed that an electro-magnet had its inductive power greatest in the middle, and that this power dwindled down to very little at the poles. Fig. 1 shows how this was ascertained. NS

The

is an electro-magnet, SS, a flat spiral sliding on it. ends W,W, of its wire are connected with a galvanometer. Each time the primary circuit was closed or opened, a current was induced in the wire of the spiral S,S2, which produced a momentary deflection of the galvanometer needle. Now, it is well known that the strengths of momentuary

currents are proportional to the sines of half the angles of deflection. The figure also shows the curve of power they obtained by this means, the distances such as aa between the curve and the magnet NS, give the relative powers at each point along the magnet. The electro-magnet was four feet long. At 23 inches from the middle, this line (ordinate), is, what we may call to simplify results, 1. At 15 inches it is 64, at 7 inches 7·12, at 3 inches 7-5, and at the centre 7.6. From these observations it seemed to me, that in order to obtain the best result from a given quantity of wire, an amount of it should be coiled on each part of the primary coil, proportionate to the inductive strength of the core at that part. The coil should thus be thickest at the middle and taper off at the ends. It is now generally admitted, that the electro-motive force of the secondary coil is proportionate to the number of turns in it. The resistance that this has to encounter depends on the length of the wire. The length of the spark is proportionate to the electro-motive force; but the quantity of electricity passing in the spark increases with the quotient of the electro-motive force by the resistance. The quantity of electricity in the spark is a most important element in it, as the heating and lighting effects depend upon it; a spark without quantity may be long, but it lacks volume and power. In the usual construction, where equal lengths of wire are wound round every part of the primary coil, there must be a considerable loss in point of quantity. To fix our ideas, suppose the coil divided into three parts, one in the centre, and one on each side of it, each containing the same length of wire, and the same number of convolutions. Let as assume, for the sake of illustration, that the electro-motive force in the central part, corresponding to the inductive power of the electro-magnet, is twice as great as in the lateral parts. The same length of wire or of resistance has to be encountered in all three. The power to overcome it or to generate quantity is twice as great in the central part as in the other two. Hence, not only will the length of the spark in the central part be twice as long, but also twice as dense (provided it could be got separately) as that given by the lateral parts. When all three are joined

VOL. VII.

2 B

together, the electro-motive force would be double that of the central part, and the spark would also be double in point of length, but having thrice the resistance to encounter, it would be only in point of quantity of the central spark.

To understand now the effect of removing the wire from the sides to the middle, another consideration must be taken into account. Each convolution, although it contributes the same amount of force as any other convolution without it or within it at the same part of the coil, has a very different length. In at the central tube the convolution may have a diameter of say an inch, but an inch out from that the convolution would have a diameter of three inches, or be three times as long, without contributing any more force. The quantity given off by the inner one would thus be three times that given off by the outer one, as the length or resistance is only a third. Hence, then, in adding coil to coil, so as to thicken the secondary coil, though we increase the length of the spark we diminish the quantity of electricity put in circulation. Applying, then, this consideration to the case supposed, a convolution of the central part with a diameter of two inches is equal to two convolutions of the side spiral, with a diameter of an inch in point of force. The quantity of the former is also equal to that of the latter two, for the length of wire is the same in both. Hence, then, to get the maximum result, as much must be unwound from the two ends and wound round the middle so as to make the diameter of the middle double the diameter of each of the sides. With such a construction, not only would there be a gain in tension, but also a decided gain in quantity. The fall in power is not so abrupt along an electro-magnet as two to one; but it is quite possible to keep at something like the same principle, viz., on the outside so to shape the coil that the same length of wire possesses the same electromotive force.

That these views are not simply theoretical, I would give the following facts, which have been furnished by my recent experience in coil building :-When the poles of any coil machine are brought near, the spark is not the clearly defined spark of the electric machine, or that produced