constants of cyanin, deduced indirectly by means of Cauchy's formulas for metallic reflection, agree very well with the values observed directly, except for the red part of the spectrum. For this the values of n agree with those observed, but the values of κ come out larger. If these larger values of are used, the discrepancy above alluded to disappears.

G. B. M.

Kathode and Röntgen Radiations. A. A. C. SwINTON. (Royal Institution Discourse, 4 February, 1898; Electrician, vol. xli., 1898, pp. 246–248, and pp. 317-319.)

A pear-shaped Crookes' tube is suspended over a straight electro-magnet, towards one pole of which the kathode beam is projected. When the magnet is excited, the beam is drawn down to a fine point, which rapidly erodes the internal glass surface of the tube. By moving the tube or the magnet, any desired pattern can be engraved on the glass. In another experiment two concave kathodes are arranged to focus on a small piece of quicklime, placed between them. The kathodes are supplied with alternating current, at about 20,000 volts, and a very brilliant and beautiful light is produced by the incandescence of the quicklime. The light is found to fluctuate in a curious manner, and after a short time the kathode rays bore perfectly straight and very minute holes, right through the lime. It may eventually be found possible to produce commercial high-voltage lamps on this principle of much higher efficiency than filament lamps, and possibly even rivalling arc lamps. The luminous material in this case need not be a conductor, and there is therefore a much wider range of available refractory substances. An electric furnace might also be constructed on these lines for delicate chemical investigations.

A tube is described in which, in addition to a concave kathode and anode arranged as in a focus tube, there is a delicately pivoted wheel, with mica vanes, which can be moved bodily either out into the centre of the tube, so that the kathode stream impinges upon the vanes, or back into an annex, when the vanes are quite outside the kathode line of fire. In the former position, as in Crookes' experiments, the wheel revolves very rapidly in a direction that indicates a stream of particles proceeding from the kathode, but when placed in the latter position, the wheel is found to rotate more slowly in the opposite direction. It appears, therefore, that in a focus tube, while the kathode stream proceeds at great velocity through the centre of the bulb, there is also a slower reverse stream of particles returning from the anode to the kathode, round the outside of the kathode stream.

When the stream from a concave kathode is caused to impinge upon an anti-kathode of carbon, the latter becomes luminescent

where struck. If the anti-kathode be so placed as to intersect either the convergent or divergent cones of rays, these, instead of producing a uniformly luminous patch upon the carbon, produce a bright ring with a dark interior. The diameter of the ring is smaller the higher the exhaustion and the nearer to the focus the point of intersection. From this it appears that the convergent and divergent cones of kathode rays are hollow in section.

Birkeland's kathode-ray spectrum, produced by deflecting a thin kathode stream by a magnet, and then allowing it to fall upon the glass walls of the tube, can be photographed by binding a strip of sensitive photographic film round the tube, and making a single discharge through the latter. Further, by inserting between the glass and the film a piece of black paper, so placed as to cover only one-half of the spectrum, a photograph can be obtained, one-half of which is due to the visible fluorescence of the glass and the other half to the invisible Röntgen rays. Such photographs show that the bands in the spectrum produced by the Röntgen rays are co-terminous with the fluorescent bands. It is suggested that the bands are due to kathode rays of different velocities, due to the oscillatory character of the discharge, those that travel fastest being the least deflected, and the most active in producing Röntgen rays.

When a focus-tube is fitted with two or more kathodes of different diameters, but all arranged to focus upon the same antikathode, it is found that, for any given degree of vacuum, the smaller the kathode put into use the greater is the electromotive force required to cause a discharge to pass through the tube, and the more penetrative are the Röntgen rays produced. With a tube fitted with a movable anti-kathode, half of platinum and half of aluminium, either of which can be brought opposite to the kathode at will, platinum is found to give the most rays; while similar experiments with other metals show that the metals of the highest atomic weight form the best anti-kathodes. This is in accordance with what would be expected, if the Röntgen rays are due to sudden change in velocity of the kathode-ray atoms by collision with the anti-kathode.

The resistance of a tube, and the penetrative quality of the Röntgen rays produced, can be varied by making the anode (which also forms the anti-kathode) movable, when the nearer it is placed to the kathode the higher will be the resistance and the more penetrative the rays. According to another plan, the anode is fixed and the kathode is made movable relatively to the glass walls of a conical annex to the tube, when the nearer the kathode is to the glass the higher is the resistance and the more penetrative are the rays; or instead of making the kathode movable, a conical shield of glass is arranged so as to be adjustable relatively to the kathode, with a similar result. It is suggested that the increase of resistance in each case leads to a greater velocity of the kathode stream, and thus causes the Röntgen rays to be more penetrative. Curves showing the variations in resistance, and photographs

illustrating the effect on the penetrative quality of the rays are given in the Paper.

By means of pin-hole photography, the exact position, dimensions, and shape of the active area of the anti-kathode, from which the Röntgen rays proceed, can be investigated, and such photographs show that in a focus-tube the active area is a small spot, either alone or surrounded by a hollow elliptical ring of larger or less dimensions, depending upon the distance beyond the focus at which the anti-kathode intersects the kathode stream.

The photographic effect of the most powerful Röntgen rays that can be produced is relatively very feeble. A comparative test of a very good Röntgen-ray tube screened with black paper, as against a naked standard candle, showed that the candle was sixty times more active, photographically, than the tube.

The Paper concludes with the suggestion that though Lodge was not able to detect any movement of the ether due to dragging a body through it, the very great velocity of the kathode-ray particles may have this effect, with the result that something analogous to the crack of a whip or a clap of the hands is produced as each particle hits the anti-kathode and rebounds; or the effect may be simply due to the enormous temperature attained by the kathode-ray particles on their kinetic energy being converted into heat in their collision with the anti-kathode, the energy, if all converted to heat in the particles themselves, being calculated to give a temperature rise of some 50,000 million degrees Centigrade.

AUTHOR.

Passage of Röntgen Rays along Opaque Tubes. E. VILLARI. (Roma R. Accad. Lincei, Atti, vol. vii., 1898, pp. 225-230.)

When sent along opaque tubes they neither gain nor lose in photographic effect, and accordingly they seem not to be sensibly reflected or diffused within the tube; and they probably do not lose any of their discharging power.

A. D.

Röntgen Radiation and the Luminescence of Gases.

A. VON HEMPTINNE.

(Zeitschrift für physicalische Chemie, vol. xxvi., 1898, pp. 165–169.)

The Author investigates the property of Röntgen rays of enabling a Geissler tube, through which they are passed, to give out light under a much lower vacuum. Therry suggested that the effect of Röntgen rays is not due to their ionising effect (at

any rate primarily), but to the rays acting upon the ether in such a way as to facilitate electric discharge-sparks between molecules, the production of milliards of which sparks is the cause of luminescence of the gas. Luminescent gas is very like a metallic conductor in many ways, even including a screening effect; but it does not absorb Röntgen rays, as metals do.

A. D.

Thermometry. C. CHREE.

(Philosophical Magazine, vol. xlv., 1898, pp. 205–227, 299-325.)

A discussion of the measurement of temperature by means of glass thermometers is here presented, in which are treated the zero-difficulties and lag due to the behaviour of glass, the methods of measurement with fixed and movable zeros, and the limits of accuracy of temperature-determination. In connection with these the subject of calibration is touched upon, and the special difficulties in the determination of the fixed points, and the allowances to be made for the emergent column and for external and internal pressure are discussed at length, as is also Welsh's method of graduation. This Paper should be studied by those who rely upon glass thermometers for accuracy, and are not acquainted with Guillaume's "Thermométrie de Précision," or the work of the Bureau International, or of the Charlottenburg Reichsanstalt.

R. E. B.

Currents measured by Magnetization. F. POCKELS.

(Annalen der Physik und Chemie, vol. lxv., 1898, pp. 458-475.)

The Author endeavours to determine the maximum dischargecurrent by the remanent magnetism induced in magnetic substances acted upon by the current. Ballistic galvanometers only give the average intensity of a current, whereas the remanent magnetism would indicate the maximum current, whenever attained. To avoid eddy-currents, the Author uses bars of basalt as the magnetic substance. The results are very favourable. A magnetic field lasting only about one-millionth of a second shows the same remanent (and probably also the same temporary) magnetization as is induced by a field kept up indefinitely at the same strength. Hence the magnetization of pieces of basalt may be used to determine the current strength of lightning discharges.

E. E. F.

Magnetic Properties of Nickel-Steels. E. DUMONT.

(Comptes Rendus de l'Académie des Sciences, Paris, vol. cxxvi., 1898, pp. 741-744.)

The Author has continued the work of C. E. Guillaume,1 and has examined the magnetic behaviour of nickel-steels, following the method of Prof. Ewing and Miss Klaasen (Electrician, May 15th, 1891). He has caused various alloys to traverse magnetic cycles, and has drawn curves showing the change of the permeability with change of the proportion of nickel, at various temperatures, and with various strengths of field. He finds that at an equal distance from the point of total loss of magnetism, all the reversible alloys have the same magnetic permeability. Moreover, at every temperature the permeability of alloys containing 27 to 44 per cent. of nickel increases with the proportion of nickel.

J. J. S.

Acetylene Compound with Cuprous Oxychloride.

R. CHAVASTElon.

(Comptes Rendus de l'Académie des Sciences, Paris, vol. cxxvii., 1898. pp. 68, 69.)

The compound C2H2 Cu2Cl, is slowly decomposed into hydrochloric acid, acetylene, and a violet-coloured substance having the composition C2H, Cu,Cl, Cu,O. The decomposition is incomplete in presence of hydrochloric acid.

T. E.

Sulphating of Negatives in Lead Accumulators. L. JUMAU. (L'Éclairage Électrique, vol. xvi., 1898, pp. 133-136.)

The evolution of hydrogen and formation of lead sulphate at the negative plates while the element is at rest is due to two principal causes-(1) the chemical action of the sulphuric acid. upon spongy lead; (2) the electrochemical action of the couple formed by the spongy lead of the active material and the lead (often containing antimony) of the support. The accidental presence of particles of lead peroxide from the positives, or of copper from the connections, also favours sulphating, whilst the action is in all cases much increased by the use of acid of high density. The results of experiments are given showing the influence of acid-concentration, and of time on the sulphating of charged plates at rest. In many cases the extent of the action is greater than can be attributed to the above-mentioned causes, and is found to be due to the presence in the active negative material

1 Comptes Rendus de l'Académie des Sciences, Paris, vol. cxxiv., 1897, p. 1515, and Physical Society's Abstracts, 1897, No. 517.