ABSTRACTS.

S. J. LOCHNER-ON THE ELONGATION PRODUCED IN SOFT IRON BY MAGNETISM.

(Philosophical Magazine, Vol. 36, No. 223, p. 498.)

The author used for his researches an instrument resembling in its essential features a Michelson interferential refractometer, of great sensibility, the elongation being measured by means of a mirror, and the limit of sensibility being lower than one-millionth of an inch. He. found that the route by which a certain magnetising force had been arrived at affected the amount of expansion for that force, the expansion varying for rising and falling currents in a curve very similar in shape to the hysteresis cycle.

His general results may be summed up as follows:-They indicate that the expansion is a function of the ratio between the diameter and the length, and that the elongation varies directly as, possibly, the square root of this ratio. Also, that the expansion varies directly as the permeability. That the amount of magnetising force required to produce the maximum expansion, and the point of no expansion, depends on the ratio between the diameter and the length. That there are two maxima-one produced by increasing the current, and another by decreasing the current from the point which produced the first maximum. That the first contact gives more expansion than the second and following contacts; and, further, that hese appear to disagree among themselves, the expansion falling off with subsequent contacts. The paper is fully illustrated with tables and diagrams.

R. HENNIG-THE PERMEABILITY OF OXYGEN.

(Wiedemann's Annalen, Vol. 50, No. 11, p. 485.)

This is an account of a new determination of the relative permeability of oxygen and atmospheric air by Toepler's method, the principle of which is as follows:-A glass tube of small internal diameter has a slight bend in it, and is placed symmetrically in a magnetic field, the bend being downwards and filled with a short thread of some liquid, the arms on either side being filled with the gases whose permeability is to be compared. If one gas is more permeable than the other, it is attracted with greater force into the magnetic field, and this action is balanced by the force of gravity against which the thread of liquid has to be raised. Any want of uniformity in the field can be detected by filling both limbs with the same gas and observing the deflection, and formula are worked out for the difference in permeability, with this correction. Methods of measurement of the field are discussed, and Weber's bifilar method is preferred, owing to the small size of the requisite apparatus and the long range of the values which may be accurately determined. The mean error of the author's observations is about 3 per cent., and the probable error is 1 per cent.

The following table gives the author's result, compared with other observers':

CURRENT STATIONS.

(Elektrotechnische Zeitschrift, No. 32, 1893, p. 466.)

The author remarks on a small correction necessary in voltmeters containing iron when used with alternating currents. He points out that the reading of the voltmeter depends, not merely on the frequency, but also on the shape, of the alternate pressure wave. He starts with the general equation,

=

5

E1 E, sin m t + E, sin (3 m t + a) + E, sin (5 m t + B) + representing any curve which is symmetrical as regards the axis of abscissæ, and works out values for the effective current; from which he draws the general conclusion that in a certain circuit of given resistance and self-induction the strength of current for a given effective pressure varies with the shape of the pressure curve. In particular, where the self-induction is relatively small the strength of current is independent of the pressure curve; when the curve is nearly a sine function, and the resistance of the circuit comparatively large, no very great difference occurs in the value of current with different E.M.F. curves.

But if the resistance and self-induction are of the same order of magnitude, the differences in current-strength may be very noticeable. And in taking the particular case of a Ganz machine the author finds for a voltmeter of resistance 1,000 and coefficient of self induction 1 that the error is about 8 per cent. The author therefore advises that all iron instruments should be calibrated on the alternator for use with which they are designed.

Lord KELVIN-ON A PIEZO-ELECTRIC PILE.

(Philosophical Magazine, Vol. 36, No. 221, p. 342)

This is a description of an apparatus constructed by the author, briefly as follows:-Twenty-four double plates of zinc and copper soldered together are taken, and the corners of the zinc plates cut off, in such a manner that one side of the plate is a square of 8 centimetres of copper, and the other a similar sheet of zinc, but with the corners removed and showing the copper. These plates are piled together with distance pieces in such a manner that an air space of 3 or 4 mm, is formed between

the zinc of one and the copper of the next. Connect the bottom and top plates to the quadrants of an electrometer, and discharge all parts of the ay paratus; and then let a weight of a few pounds fall from an inch or so on to the top plate and rest there. A startlingly great deflection is produced, and if the zinc is at the top it becomes positively charged. A similar instrument was invented by Messrs. J. & P. Curie, and described in 1881.

J. TROWBRIDGE—ON THE OSCILLATIONS OF LIGHTNING DISCHARGES AND OF THE AURORA BOREALIS.

(Philosophical Magazine, Vol. 36, Nc. 221, p. 343.)

When air is subjected to a sudden strain due to an electrical discharge, it acts like an elastic solid, and cracks in zigzag fissures, and through the crack thus produced by the first oscillation the subsequent ones take place. The author has made some new photographs of sparks by the rotating mirror method, using an alternator and a step-up oil transformer and a Ruhmkorff coil; an idea of the power employed being given by the fact that when the sparks were passed through the secondary of a transformer (turns not given), the primary circuit of two layers of thick wire lighted three 50-volt Edison lamps in parallel. The photographs are given, and are exceedingly interesting. From 10 to 12 oscillations can be traced, about three in the same path, the time interval being about one hundred-thousandth of a second. For three hundred-thousandths of a second, therefore, the air remains passive, and the heat is not conducted away.

With reference to the question of lightning, the author refers to its inductive effects on electric circuits: he notes that lamps often "blink," and sometimes go out, when a lightning flash takes place in the neighbourhood. He considers it dangerous to put electric supply wires on gas fittings for this reason, and has known a gas leak to be thus ignited.

The photographs show that after a few oscillations the discharge ceases to be disruptive, and becomes a brush glow-as in the case of a vacuum tube; and the author considered the latter case to be allied to Aurora Borealis. He therefore experimented with stratified discharges, and found that period of oscillation had no effect on the distance of the stratifications apart. Self-induction also has no effect, but resistance has. On the other hand, transitory stratification can be produced by applying the finger or other earth connection to a vacuum tube. The author therefore considers that the waving of the Aurora Borealis is due to intermittent or moving earth connections due to damp air or cloud. He sees no reason to believe in the rapid oscillation of the Aurora.

J. SAHULKA-EXPLANATION OF THE FERRANTI PHENOMENON. (Elektrotechnische Zeitschrift, 1893, No. 52, p. 739.)

One of the most remarkable of the so-called Ferranti effects is the rise in the ratio of transformation of a transformer whose primary is coupled to an

alternator, and whose secondary is connected to a condenser of capacity not exceeding a certain value; under which circumstances the primary volts rise, and the primary current falls.

The author shows in the first instance, from theoretical considerations, that the explanation of the phenomenon is caused by so-called "drop," or leakage of magnetic induction, and proceeds to describe some experiments bearing out his theory.

1

=

Supposing L, and L, the coefficients of self-induction of the two circuits, and M their mutual induction, and if N, and N, are the respective number of turns, then, if there be no magnetic leakage, M2 L L2, and the transformation ratio is V = N, N1; but if there is leakage, then M2 is less than L, L., and V is less than N, N1, or, as is well known, the ratio of transformation is reduced. Now, if the secondary be coupled to a condenser of capacity C, L2 becomes

2

2

1

2

(2 πn)2 С'

2

where n is the frequency; and in this case, where there is magnetic leakage V becomes greater than N, N,-that is to say, the transformation ratio is raised, and depends on the capacity of the ccndenser. The apparent resistance of the primary, and the primary volts of the transformer are raised, while the primary current falls.

In the remaining part of the paper are described experiments made with a transformer having a cylindrical core made of iron wires, and wound with four layers of 73 turns on one quarter of its length, and on the remaining three-quarters with four layers of 219 turn". If these be used as primary and secondary respectively, the magnetic leakage is naturally very great, and the transformation ratio was only 1.38 on open circuit; while, if a condenser was in the secondary, of about 5 microfarads capacity, the transformation ratio rose 18 per cent.

On the other hand, if the magnetic leakage were reduced by completely superposing the primary and secondary, the ratio rose from 27 to 274, or only one and a half per cent. The author concludes that there is no reasonable doubt that in magnetic leakage lies the whole explanation of this phenomenon.

R. J. HOLLAND-ALTERATION OF THE CONDUCTIVITY OF A SOLUTION BY ADDITION OF SMALL QUANTITIES OF A

NON-CONDUCTOR.

(Wiedemann's Annalen, Vol. 50, No. 10, p. 261.)

The author tried solutions in methyl alcohol of various salts, such as the nitrates of potassium, sodium, and ammonium, and several chlorides. The non-conductors were such as did not affect these salts-benzol, toluol, xylol, oil of turpentine, glycerine, &c.—and the determinations of resistance were by the Kohlrausch telephone method. The following were, briefly, the conclusions arrived at:-The electrical conductivity of a methyl-alcohol solution of an electrolyte diminishes

with the addition of a non-conductor by an amount varying with the nature of the non-conductor and its strength; the above-mentioned materials-benzol, toluol, xylol, and oil of turpentine-affect the conductivity in that order, the last being most powerful. The temperature coefficients of the alcoholic solutions become greater as the solution is made weaker; they are about half as great as those of the corresponding aqueous solutions, and are influenced to a very small extent by the presence of a non-conductor.

F. SCHULZE-BERGE-ROTATING AIR PUMP.

(Wiedemann's Annalen, Vol. 50, No. 10, p. 368.)

To combine the advantages of the high vacua produced by the Sprengel and Geissler pumps with a greater rapidity of working, the author has devised a rotating mercury air pump, here fully described. The pump vessel is a curved tube returning on itself, attached to an axis on which it can rotate. A mass of mercury flowing in the interior of the tube, and partially filling it, is arranged to produce the vacuum on one side and expel the air on the other, the connection to the pump vessel or the atmosphere being effected by suitable cocks or valves. The best form into which this pump has been developed is the one called by the author "double"ring pump," in which the air expelled by the first tube has to pass the second before reaching outer air. It is of simple construction, and produces a very high vacuum by simple rotation, and can be belt-driven. When precautions are taken to have the apparatus and all its parts thoroughly dry, vacua can be obtained which surpass the limits of accurate measurements on McLeod's apparatus. The pump already made, having a capacity of about 1 litre in the ring (pump vessel), can be rotated about 15 times per minute.

G. DETTMAR

EXPERIMENTS ON "KRUPPIN," A NEW
RESISTANCE MATERIAL.

(Elektrotechnische Zeitschrift, 1893, No. 50, p. 710.)

This is a new metallic alloy produced by Messrs. Krupp, of Essen, and distinguished by a high specific resistance. The accompanying tables and values give the qualities of the new material.

This value is about 50 times that of pure copper.

Temperature coefficient = 0.13% per 1° C

The following is a table of carrying power. The limit of temperature for the spirals is taken as that at which the wire "spits" when touched with the moistened finger. The spirals are in every case wound with that amount of spacing which is most favourable for the cooling of the wire.