independent of such momentary changes in the strength of the current as affect the readings of a magnetic needle. It is also portable and inexpensive.-Jour. Soc. Teleg. Engineers, 1875, 296.

THE ACTION OF AN ELECTRO-MAGNET UPON THE SPECTRA OF RAREFIED GASES TRAVERSED BY ELECTRIC DISCHARGES.

M. Cheautard states that he has examined the spectra of rarefied gases illuminated by an electric spark, and subjected to the influence of powerful magnets, and finds that these spectra are characterized as to the position, the number, the separation, and the fineness of their lines by very curious traits peculiar to each gas. As regards the metalloids with which only his experiments have thus far been conducted, he states that the light given out by sulphur and by selenium experienced a notable diminution under the influence of the magnet, so that sometimes the spectrum, which was very apparent at first, disappeared for some moments. On the contrary, chlorine and bromine are characterized by an increase in brilliancy, and by the development of fine brilliant rays especially numerous in the green, whose appearance or disappearance at the moment when we turn on or interrupt the current has a truly magical effect. This phenomenon seems to have some importance in consideration of the obscurity which at present characterizes our knowledge of the relation of magnetism and light.-1 B, 1875, 283.

THE FREEZING OF SALT WATER.

Professor Guthrie, in continuation of former researches on the solutions of salt, has endeavored to ascertain the manner in which mixtures of salt act as cryogens, and to study their combination with water at various temperatures and in various proportions. He finds that when two salts, composed of different acids or bases, are mixed, and no precipitation occurs, it is generally considered that partial decomposition takes place, two new salts being formed; but he finds that if the salts a and by be mixed in atomic proportion, and dissolved in the smallest possible amount of water, a mixture is obtained identical with that produced on mixing ay with b x; and the temperature and composition of the resulting cryohydrate are the same in both cases. Thus, a

saturated solution of a mixture of nitrate of potassium and sulphate of sodium solidifies at -5°. A mixture of nitrate of sodium and sulphate of potassium also solidifies at this temperature; but the temperatures never fall as low as the point which could be reached by employing whichever of the salts a x, a y, b x, or by forms a cryohydrate with the lowest temperature. Thus, in the above case, the solidifying point of nitrate of sodium is -17°.—Nature, XI., 440.

ON UNILATERAL ELECTRIC CONDUCTIVITY.

Dr. Arthur Schuster states that in the course of many experiments he has had frequent occasion to remark that electric currents scem to traverse copper wires more easily in one direction than in the other; so that the galvanometer indicates different intensities when we reverse the direction of the current which traverses it. He gives to this phenomenon the name of unilateral conductivity. He first observed it in using the galvanic battery, but was able to make more accurate observations by means of magneto-electric machines. The phenomenon observed with this apparatus led him to the hypothesis that the current induced by one pole of the magnet traverses a circuit more easily than that induced by the opposite pole. In his second memoir Schuster records another phenomenon which is not without analogy with the preceding. He joined the electrodes of a galvanic battery to the apparatus which he had used in his first experiment, and found that, whatever the intensity of the continuous current might be, or the relative positions of the electro-magnet to the battery, it always happened that the initial deviation of the galvanometer needle augmented during the rotation of the magnet. On the other hand, as long as the magnet was immovable, it exerted no influence. upon the deviation produced by the permanent currents. Noting then, at first, the initial deviation of the needle while the magnet was stationary, he interrupted the current before turning the magnet, and observed again the deviation produced at the first passage of the current during the rotation. The difference between these two deviations was sensibly proportional to the intensity of the permanent current, but decreased rapidly with the increase in the electric vibrations produced by the magnet. The cause of this singular

influence of the rotation of the magnet upon the intensity of the permanent current is scarcely less obscure than that of the unilateral conductivity; and the explanation of the one phenomenon involves the apparent contradiction of the other.-Bull. Hebd., 1875, 297.

SINGULAR PROPERTY OF ALUMINUM ELECTRODES.

A singular property of aluminum has been noticed by Ducretet. A voltameter, whose electrodes are respectively aluminum and platinum, allows an electric current to pass, or prevents it, according as the electrodes are respectively positive or negative. If the positive current passes from the platinum to the aluminum, no unusual resistance is experienced. If, on the other hand, a positive current passes from the aluminum to the platinum, the current is nearly arrested, and the needle of the galvanometer marks in the first case, for instance, 22°, and in the second case 2° only. The explanation of this seems to be that in the former, or favorable case, the current disengages oxygen at the platinum and hydrogen at the aluminum pole. But in the opposite case the oxygen is produced at the aluminum pole, and forms a layer of alumina (or the oxide of aluminum), to the presence of which the arrest of the current must be attributed. If the poles are placed in hydrochloric acid, the phenomenon no longer takes place, and similarly does it not follow if we employ any alkaline liquid. In opposition to this explanation, however, it must be granted that the microscopic examination of the aluminum electrode does not reveal any apparent change in its appearance, no matter in which way the current flows. Whatever may be the explanation, the fact remains, and is certainly a very striking one. Of course, if both electrodes are formed of aluminum, the current will not flow in either direction. Ducretet proposes to apply this interesting property of aluminum in the construction of a telegraphic apparatus, which he calls a rheotome of constant direction.-13 B, III., 218.

ON THE ELECTRIC DISCHARGE.

The electrical sparks of a peculiar nature that have been called "feeble sparks" by Riess, by whom they were first discovered, are distinguishable from the ordinary bright

sparks, not only by their form, their light, and their noise, but also by other peculiarities, such that a further investigation into their nature has been made by him. Among the isolated points considered in connection with these feeble sparks, Riess states that Wiedemann and Ruhlmann have from special investigations concluded that the quantity of electricity necessary for a discharge is much greater when the discharge takes place from the negative than from the positive electrode. But his own experiments seem to him not to justify so positive an expression, as they can be explained by means of the feeble glimmer that precedes an eruptive discharge. We have, moreover, no experiments which would lead us to think that there is a greater quantity of electricity present in a positive than in a negative spark. From experiments made with the Holtz machine, he concludes that the greater length of the negative electrode is not an important condition in producing feeble sparks, but that in these experiments electrodes may be employed of any length whatever. The feeble sparks are, not only in reference to their length, but also to their light and brightness, independent of the composition of the arms of the discharger by means of which they take place. - Monatsbericht der Berlin Akad., 1875, 152.

TELEGRAPHIC GROUND CURRENTS.

Speaking of the importance of observing the underground or so-called earth currents on electric telegraph wires, Sir William Thomson states that an observation which would be of value for scientific study is to observe the indication of the electrometer at each end of the telegraph line at any time--whether during a magnetic storm or not-during the day or night. If the line be worked with a condenser at each end, this observation can be made without in the slightest degree disturbing the practical work through the line by simply putting on an electrometer in direct connection with the line, and connecting the outside of the electrometer with a proper earth connection, when it may be observed, quite irrespectively of the signaling, when signaling is done, as it very frequently is, on submarine lines with a condenser at each end. The scientific observation will bo disturbed undoubtedly by the sending of messages; but the

disturbance is only transient, and in every pause at the end of a word there will be a sufficiently near approach to steadiness in the potential at the end of the wire connected with the electrometer to allow a careful observer to estimate with practical accuracy the indication that he would have were there no work of the line going on at the time. A magnetic storm of considerable intensity does not stop the work-does, indeed, scarcely interfere with the work of a submarine line in many instances-when the condenser is used at each end. Thus observations, even when the line is working, may be made during magnetic storms, and again during hours when the line is not working. Any single observation, or any series of observations, that are made on the electric potentials at one end of the insulated line will give valuable results. When an arrangement can be made for simultaneous observations of the potentials of the electrometer at the two ends of the line, the results will be still more valuable. We may substitute, with satisfactory results, for the electrometer, the galvanometer of very large resistance.-Jour. of the Soc. of Telegraph Engineers, III.,

1874, 10.

THE ELECTRICAL VOTING MACHINE.

Monsieur J. Morin has presented for inspection a model voting machine constructed for a deliberative body of twelve voters, in which he proposes to simplify, in a considerable degree, the tedious process of voting, and to economize the time of an assembly. To attain this end the machine ought to be prompt and certain, and free from errors as to the result of the votes. It is composed of a portable table, having twelve circular openings, below which are written the names of the members, each opening corresponding, by invisible wires, to the place of the representative named upon the table. Beneath are placed two small openings closed by small covers, which disappear at the end of the operation, so as to allow one to see the number of votes that have been cast for and against the project. Each of the deputies has also two balls, black and white, corresponding to the opening which belongs to him in the table. The operation of the process is as follows: The deputy, by touching a button, draws before the opening placed upon the table and under his name