battery in quantity and discharged in tension. Using ten of these condensers charged with a secondary battery of 800 cells, as many as 15 brilliant sparks from 13 to 14 millimeters long were obtained per second, giving the same noise, and otherwise resembling closely the condensed spark from an induction coil. The difference between the negative and the positive spark is more marked in this machine than in the ordinary ones. Its length appears to increase with the number of condenser plates, 40 condensers giving a spark 4 or 5 centimeters long, and 51 of 6 centimeters. In vacuo the light is brighter than that of the coil, though no stratifi'cation is observable with it in tubes in which the induction spark gives it distinctly. Nor is there any difference between the poles in these tubes, the purple glow at the negative pole being absent. In proof of the minute quantity of electricity in each spark, a secondary battery of 40 cells. was charged by 15 seconds contact with two Bunsen cells, and was then connected to the machine. It illuminated a Geissler tube for more than 15 minutes. Contact for 10 minutes would therefore illuminate a tube for more than 10 hours.

Jablochkoff has discovered and utilized the fact that if one of the surfaces of a large condenser be connected with one electrode of a to-and-fro current machine, an alternating current is obtained between the second condenser surface and the other electrode, much more powerful than the current given directly by the magneto-machine itself.

Sabine has investigated the remarkable motions which are produced by placing a drop of very dilute acid upon the clean surface of a newly filtered and rather rich amalgam of some metal which is positive to mercury. The drop does not lie still, as it would do on pure mercury, but sets itself into an irregular jerky motion. This is true of copper, tin, antimony, zinc, and lead amalgams. If, however, amalgams of metals negative to mercury be used, such as gold, platinum, and silver, the drop lies quite still. Sulphuric, hydrochloric, oxalic, and acetic acids were used, and all produced the result, but in different degrees. In oxygen the movements are increased; in hydrogen they are arrested. The author hence infers that the motions result from alternate deoxidation of the mercury beneath the acid by electrolysis,

causing the drop to contract by an altered surface tension; and reoxidation outside of the drop, causing it to expand again over the surface.

Beetz, from his experiments on the comparative value of the thermopiles of Noë and of Clamond, as modified by Koch, concludes in favor of the latter. The objection that it requires a little time to be heated he regards of little importance, but once in action it continues remarkably constant, both as regards its electromotive force and its resistance. Since with the same number of pairs the Noë pile gives the greater electromotive force, the utility of the Clamond battery is quite equal, from the facility with which the number of pairs may be increased. He suggests that the burner of the latter should be improved.

Gore has communicated to the Royal Society a paper on the thermo-electric properties of liquids, in which he describes a new form of apparatus he has contrived for the investigation, and gives the results of his observations made with it. He shows that the electric currents obtained were produced neither by chemical action nor by a temporary dissociation of the liquid, nor by the action of gases occluded in the metals; but they owe their origin solely to the heat, which disappears in producing them, and are thus true thermo-electric currents of liquids. He concludes as probable that when a metal is immersed in a liquid, heat results. His experiments suggest the construction of a new thermo-electric motor.

3. Electrical Measurements.

Gaiffe has contrived two simple forms of galvanometer, one for measuring electromotive force directly, and the other for measuring current strength. In the former the coil has a high resistance (about 3000 units of the British Association committee), so that the resistance of the rheomotor may be neglected in comparison, and the deflections of the needle be sensibly proportional to the electromotive forces. By means of two additional resistance coils the resistance may be increased 10 and 50 times. The circle is graduated empirically into 60 divisions, each of which represents one tenth of a volt when the galvanometer resistance alone is used. Electromotive forces from 0.1 to 150 volts may thus

be measured. The latter galvanometer has a coil of low resistance, with shunts by which its delicacy may by still further reduced. Using the galvanometer alone, one division on the scale represents one ten-thousandth of a B. A. unit. With the first shunt the divisions represent hundredths, and with the second whole units. Current strengths from 0.0001 to 200 units may thus be measured. These instruments are accurate to one per cent., sufficient for testing currents used in medicine, for which they were devised.

Foster has exhibited to the London Physical Society a very simple form of the trap-door form of Thomson's absolute electrometer. According to Nature, one arm of a balance has suspended to it by silk fibres a zinc disk hanging horizontally in the plane of a sheet of the same metal, which acts as a guard plate. Below the disk about one inch is a second horizontal sheet of zinc. The guard plate and disk are electrically connected by a bridge of very fine wire. To use the apparatus it is first accurately counterpoised, an excess weight-say, of one grain-is added, the guard plate and the lower attracting plate are connected with the electrodes of the electromotor-a Holtz machine, for example-a spark-measurer being introduced into the same circuit. If the machine be put in action, and the knobs of the sparkmeasurer gradually separated, a point will be reached where the attraction of the suspended disk just balances the excess weight. Reading off the length of spark, the data are obtained for calculating the difference of potential required.

Edison described, at the St. Louis meeting of the American Association, a new form of voltameter. Into a suitable vessel of acidulated water two electrodes are placed, one of which consists of platinum wire covered with gutta-percha, and perforated with a fine needle near its lower extremity. This electrode is made negative. The evolved hydrogen escapes in bubbles from the minute opening with a sound like the ticking of a watch, audible at the distance of several feet. By placing a rheostat in circuit, and regulating the bubbles to one a second, a constant current is obtained; and by calibrating the instrument by this means, the strength of any given current flowing through the instrument is known in terms of the number of gas bubbles evolved per minute. Should this number rise above 16 per second, a musical note

is produced, by the pitch of which the current strength may be determined. To obtain accurate results with the apparatus, corrections for temperature and pressure must be applied.

Haga at Strasburg and Clark at Heidelberg have investigated the electromotive force produced by the flow of water through capillary tubes, using a quadrant electrometer to measure the difference of potential. Clark finds that the electromotive force is greater the narrower the tube; that in very narrow tubes it is independent of the length; that dif ferent electromotive forces appear if the interior of the tube be coated with different substances; that the electromotive force decreases with the time; and that the seat of the electromotive force is the limiting surface of the liquid and the solid tube wall. Haga finds that the electromotive force is proportional to the pressure, independent of the length of the tubes, dependent on the nature of the inner surface of the tube, increases with the resistance of the water, and probably also with the temperature. The two results agree closely.

Lippmann has contrived an ingenious method of detecting minute quantities of a metal in solution, founded on the principle that when an electrode made of a given metal is placed in a solution, it will be depolarized only if a salt of that metal exists in the solution. Hence, for example, if a copper wire conveying a weak current be made the negative electrode in any solution, it will be polarized if there is no copper dissolved in this solution, but it will not be polarized if the liquid contain one five-thousandth of copper sulphate. The polarization is easily detected by closing the circuit through a galvanometer, the battery being left out. A contrary deflection indicates polarization. For silver, the sensibility seems somewhat greater.

Börnstein has experimented to determine the influence of light upon the electric resistance of metals, and gives the following as his conclusions: 1st, the property of having the resistance to an electric current diminished by the action of light is not limited to the metalloids selenium and tellurium, but occurs also in platinum, gold, and silver, and hence most probably in all the metals; 2d, an electric current lessens the conducting power of a conductor as well as its sensibili

ty to light; but in both cases the former value of these constants is gradually attained after the current ceases to pass through the substance.

Mills has given the name electrostriction to a curious electrical phenomenon, which may be produced as follows: The bulb of an ordinary thermometer is first coated with silver by a chemical method, and then with some other metal by electrolytic deposition. The mercury will traverse some portion of the scale, and finally take up a definite position independent of temperature. Of the metals thus far experimented with, copper, silver, iron, and nickel constrict the bulb, while zinc and cadmium distend it. The author has succeeded in determining the electrostrictive effect in atmospheres of pressure, and shows that, since the metal which has been deposited on the bulb may be removed by a chemical solvent, it is possible to measure chemical action in terms of atmospheres of pressure.

4. Electric Spark and Light.

Cazin has studied the spectrum of the electric spark taken in compressed gas, both directly and by photography; and he concludes that the spark under these circumstances is compound, containing incandescent gaseous particles producing a spectrum of lines, and solid or liquid particles producing a continuous spectrum. The first of these come from the gaseous medium and from the electrodes; the second are torn from the electrodes or from the adjacent walls of the tube. The solid or liquid particles are collected in the central portions, the spark proper, while the aureole is formed of gaseous particles. This aureole is to the total spark what the bluish base of a candle flame is to the entire flame. As the pressure increases, the solid or liquid particles become more abundant, and their continuous spectrum predominates, finally extinguishing by its superior brightness the linear spectrum of the gaseous portion. Hence he regards it as incorrect to say that the gaseous lines widen and unite to a continuous spectrum.

Ayrton and Perry, in a letter to Nature, show that the wire cage proposed by Maxwell as a protection against lightning is not satisfactory, by quoting a case of lightning in a coal-mine in India by which two miners were killed.