with the nature of the gas, being as much more considerable as the gas is a better conductor of electricity.

In the same memoir, we next studied the effect of the magnet upon a discharge which takes place along the line joining its two poles. In this case we observed, on the contrary, a very perceptible augmentation of the intensity of the current. We confined ourselves to establishing that this diminution of resistance, occasioned in the Geissler tube placed axially between the two magnetic poles, is the more marked the better conductor the gas and the less its pressure. With a view to complete our first observations, we have resumed and varied this experiment. May we be permitted to explain here briefly, though they are still far from complete, the results to which our latest researches have conducted.

Instead of arranging our electromagnet in the shape of a horseshoe as in our former experiments, with the two horizontal bobbins in a line with each other and the two opposite magnetic poles, separated by an interval of 10 centims., which compelled the Geissler tube to be introduced into the cylindrical aperture pierced in the axis of each of the two soft-iron cores, we here employed the electromagnet in a column, so as to cause only one of the magnetic poles to act upon the discharge. The apparatus in which the discharge took place (Geissler tube or a large bell) rested on the upper extremity of the soft-iron cylinder, the line of the electrodes being in the prolongation of the axis of the magnet.

or

We commenced by operating with cylindrical Geissler tubes of 30 centims. length and 32 millims. breadth, presenting interior electrodes of platinum wire. One of these tubes contains nitrogen, the other hydrogen, both at a very low pressure, about 1 millim. even less—at least, judging from the appearance affected by the discharge in their interior. The induction-current, furnished by a Ruhmkorff machine of medium size excited by four of Grove's couples, passed through the Geissler tube and then the derivationapparatus employed in our previous researches. It was on a very small portion of the current, derived into a galvanometer placed far enough from the magnet to be uninfluenced by it, that we observed the variations of intensity of the discharge, according as it was or was not submitted to the action of the magnet, which was excited by 20, 25, 30, and sometimes even 40 Bunsen couples.

Traversed by the discharge from the Ruhmkorff machine, each of the two Geissler tubes exhibits around the negative electrode a beautiful blue aureola extending to the sides of the tube, beyond this a long dark interval, and thence to the positive electrode streaks wide apart. The appearance of this discharge is completely changed as soon as it is submitted to the action of the magnet and when the negative electrode is at the bottom or under the immediate action of the magnetic pole. Indeed, as soon as magnetization commences, the negative aureola, which, with a length of about 35 millims., occupied the whole diameter of the tube, is transformed into a cylinder of only 8 or 9 millims. diameter, very luminous, extending to the positive electrode, across the interval previously occupied by the

dark space and the streaked positive jet, presenting, except the streaks and the colour, an appearance analogous to the narrow positive jet observed at about 8 or 10 millims. pressure.

When, instead of a Geissler tube, we employed a large bell or one of the balloons by aid of which the experiment was made of the aurora boreales with a central negative electrode encircled by the positive ring, we still obtained the same effect; that is, the large spherical aureola, developed at very low pressures isolated around the negative electrode, was replaced by a narrow blue jet of vivid splendour, having sometimes the appearance of a brilliant blue flame escaping from the positive electrode. This negative jet is always produced in the continuation of the axis of the electromagnet, even when the positive electrode is a ring situated in the same horizontal plane as the negative electrode. The electricity, which escaped equally in all directions from the negative electrode, now issues only in one direction, as if projected to a distance from the magnetic pole. But it is only at very low pressures (1 millim., and even lower) that the effect is produced with this degree of intensity. The greater the elastic force of the gas, the shorter becomes the negative dart, giving place to the positive jet. It is at about 2 millims. that the repulsion apparently exerted by the magnet upon the negative aureola commences to become sensible.

Such is the modification produced by magnetism in the appearance of the electric discharge. It is accompanied by quite as marked an alteration in the resistance opposed by the rarefied gas to the passage of the discharge. As we had already observed, and recorded in the memoir before cited, the magnet has the effect, in the case of an axial discharge, of notably augmenting the intensity of the current.

With the above-described Geissler tube containing hydrogen, placed vertically on the upper extremity of the soft-iron cylinder, the negative electrode beneath the positive, the galvanometer placed in the derived current marked 20° when there was no magnetization, and 40° when the electromagnet was excited by 25 Bunsen couples. The nitrogen-tube, placed in the same conditions, gave 20° without magnetization, and 30° with. In another case, on throwing into the electromagnet the current from 40 Bunsen couples, we saw the deflection of the galvanometer increase, with the hydrogen-tube, from 12° (before magnetization) to 55°, with the nitrogen-tube from 10° to 35°. These examples, taken at random from a great number of analogous results, show that the intensity of the discharge transmitted through the Geissler tube may be quadrupled by the action of an electromagnet sufficiently powerful. They show, moreover, what we have already recognized, that the effect on hydrogen is more marked than on air, that the augmentation of intensity of the current is greater with the gas which is more conductive than with that which is less conductive of electricity.

This augmentation of intensity is perceived by simple inspection of the tube, from the fact that the negative electrode becomes red-hot and shows traces of fusion as soon as magnetization commences.

When it is the positive electrode that is submitted to the immediate action of the magnet, there is scarcely any appreciable modification in the appearance and intensity of the discharge. The effect, however, is exactly the same, whatever the direction of the tization.

magne

When the circuit contains several consecutive Geissler tubes all placed in the same way upon the upper extremity of the soft iron, each having its negative electrode below, the effect upon the intensity of the current which traverses them all is still greater. But if, in addition to the tube or tubes placed under the action of the magnet, there is one in the circuit out of this action, no effect upon the intensity of its current is produced by the magnet, although the modification which the appearance of the discharge undergoes in the other tubes, placed over the magnetic pole, remains the same. seems, then, that it is a special and peculiarly intense resistance, having its seat at the issue from the negative electrode, which is thus overcome by the intervention of the magnet.

It

A final series of experiments support this view, and have shown us that the dimensions of the negative electrode, which notably influence the dimensions of the aureola and the resistance to the passage of the electricity, influence also the augmentation of intensity produced by the magnet in the case of an axial discharge. Working with the large bell, we had a very great, less, or almost no augmentation of intensity, according to whether we employed as negative electrode a platinum point or wire, a small ball, or a ball of 4 centims. diameter.

We confine ourselves here to briefly recording these few observations, without pretending to deduce from them, at least for the present, any theoretical consequence.--Bibliothèque Universelle, Archives des Sciences Phys. et Nat. May 15, 1874, vol. 1. pp. 41–48.

EXPERIMENTS ON APPARENT ADHESION. BY M. STEFAN.

By the name of apparent adhesion the author designates the phenomenon that, when two flat plates are laid one upon the other, they cannot be again separated without the expenditure of a force. This phenomenon has been hitherto conceived as conditioned by adhesion—that is, by the molecular forces between the particles of the two plates; and experiments have been made for the purpose of determining statically its amount.

In this phenomenon, however, we have not to do with a static, but with a dynamic problem. The experiments made by the author showed that the separation of the two plates can be effected by any force whatever; only the time in which the distance of the plates is changed a measurable quantity by the action of such a force is the greater the smaller the force.

Simultaneously with the commencement of the action of a separating force the distance of the plates commences also to increase; yet the motion is very slow, and grows ever quicker with increasing distance. The apparent adhesion is much greater when the plates are under water or another liquid instead of in air. The distance Phil. Mag. S. 4. Vol. 47. No. 314. June 1874. 2 H

of two plates of 155 millims. diameter, immersed in water, amounting at first to 0-1 of a millimetre, increases, in consequence of the continuous pull of 1 gramme, 0.01 millim. in 1 minute, 0.1 millim. in 7 minutes. From this it is intelligible how, limiting the observation to a short time, one may be misled to the assumption of a static equilibrium.

The author measured, in his experiments, the times which elapsed while a given initial distance, measured by a wire placed between the plates, increased by a certain quantity. Between these times and the other quantities which varied with the experiments the following relations were found. The times are inversely proportional to the separating force; they are, but not exactly, inversely proportional to the square of the initial distance; for plates of different sizes they are to one another as the fourth powers of the radii of the plates; for different liquids, as the times in which, under equal pressure, equal volumes of the liquid flow through a capillary tube.

From this it evidently results that with this phenomenon the question is a problem of hydrodynamics; and it is now easy, at least in general, to describe it. When the separating force begins to act, the distance between the plates receives an infinitely small increment. Thereby the space limited by the plates is augmented, and the fluid within undergoes a dilatation, in consequence of which the hydrostatic pressure becomes less. The excess of pressure of the exterior fluid acts in opposition to the separating force. Nevertheless equilibrium does not ensue, because the diminution of the hydrostatic pressure between the plates has for its result a flowingin of the exterior fluid and thereby, again, a diminution of the difference of the pressures. The distance of the plates can be again increased by the separating force, and the same process repeats itself in a continuous manner.

The author gives also an approximative theoretical solution of the problem, starting from the following consideration. The vis viva acquired by the plates through the separating force is, on account of the great slowness of the motion, vanishingly small in comparison with the work of that force. This work must consequently have its equivalent in another work; it has it in that which is necessary for maintaining the flow of the fluid from the outside into the space included between the plates.

The equation deduced from this assumption gives again all the different laws to which the experiments have conducted. It permits us also to derive from the experiments the coefficients of internal friction for the experimental fluids. If the centimetre be chosen as the unit of length, the mass of 1 gramme as the unit of mass, and the second as time-unit, it follows that for water of the temperature of 19° C. this coefficient = 0.0108, for air = 0.00183, which values almost exactly coincide with those deduced from the experiments of Poiseuille, Maxwell, and O. E. Meyer.-Kaiserliche Akademie der Wissenschaften in Wien, Sitzung der mathematisch-naturwissenschaftlichen Classe vom 30. April 1874.

A SPARK-ADJUSTER FOR THE HOLTZ MACHINE.

BY JAMES J. MINOT.

In the Annalen der Physik und Chemie, Bd. cxxxvii. p. 452, and Bd. cxxxix. p. 509, under the title of "Schwache elektrische Funken in Luft, von P. Riess," a method is described of obtaining different kinds of electric sparks from the Holtz machine.

The following method seems to be preferable to that described in the above-mentioned articles:-Having insulated the outer coating of the two Leyden jars which form a part of the ordinary Holtz machine, short thick wires were connected with these outside coatings and terminated in two brass pointers or conductors, so arranged that the distance between them could be varied at pleasure. At first the conductors were placed in connexion with each other; it was then found that a series of sparks were given off between the conductors of the machine. The extreme length of spark obtainable with the machine which was used was 20 centims. The sparks so obtained were large and luminous, passing only at intervals, and requiring a certain electric tension before they would leap across the space. Then the pointers connected with the outer coatings of the Leyden jars were drawn apart about 13 millims.; it was then found that a succession of fine thread-like sparks passed across the space separating the conductors of the machine, whereas there was no such appearance between the pointers connected with the jars. But at intervals a larger spark, not so bright as the normal spark of the machine, would jump across the conductors; and simultaneously with this a similar spark passed between the pointers. This fine line of sparks was found to have a peculiar form, being brightest and largest at the ends of the conductors of the machine, fading away to a lighter and redder tint, and being of a thread-like character in the space between the knobs of the conductors.

It was found, if the distance between the conductors of the machine exceeded a little that between the pointers connected with the Leyden jars, that no large sparks passed between either set of conductors, but only a series of thread-like discharges. When the distance between the conductors of the machine was less than that between the pointers, a similar result was obtained. When the pointers were but a few millimetres apart, a continuous loop-like discharge passed between them, which was not interrupted by the occasional passage of a bright spark, and was not coincident in path with the latter. By varying the distance of the pointers of the Leyden jars, the number and character of the sparks could be changed at will. This method possesses the advantage that by an easy adjustment of these pointers the form of the electric spark can be readily studied. In experimenting upon the passage of the spark through different media, we can by this method diminish the diameter, so to speak, of the spark, and can change quickly from the spark discharge to that of the brush. The change in tone of the sound of the discharge, when the distance between the pointers is varied, is quite marked.-Silliman's American Journal, May 1874.