Theory of Deformation of Metals. BRILLOUIN. (Annales de Chimie et de Physique, vol. xiii., 1898, pp. 377–404.) very As a result of microscopical examination and metallurgical investigation, the ordinary metals used in construction are assumed in this memoir to be composed of a great number of small crystalline elements or grains embedded in a network of viscous matter. The crystalline elements are supposed to be possessed of elasticity but to be deprived of viscosity, whilst the surrounding network is deprived of elasticity. The Author then proceeds to show how the laws of permanent deformation follow from this assumption. The present memoir deals with homogeneous deformation. Hook's law is assumed for the elasticity of the crystals, whilst the linear law expresses the viscosity. The deformations due to the crystalline and viscous components of the metal are then calculated, the former being simply the ordinary elastic deformations, and the latter being obtained by integrating the ordinary equations for a viscous fluid for the form of the network in question. The total deformation is thus shown to consist of a term representing the elastic deformation of the crystalline elements, and a term consisting of some function of Xdt (X being the deforming force, and t the time), which represents the viscous deformation. The Author lays great stress upon the importance of taking account of the time in making experimental tests, especially where stops are made with the specimen under load, and states that many of the irregularities in the results obtained in mechanical laboratories, which have usually been put down to faults in the metal, may probably be accounted for by neglect of this precaution. The equation obtained for the deformation is then put into the following form X being the force and x the length, and A being a constant depending on the initial state of hardness of the metal. For a rapid variation of load without shock, the deformation is shown to be elastic, and the coefficient of elasticity is defined as being fixed by the amount of the sudden variation of the deformation when produced without shock. Thus, to obtain the coefficient of elasticity, it is necessary to measure the initial value of the alteration of the deformation. It is then shown that, after a stop of great length under the action of a load, a small deformation effected rapidly without shock is purely elastic. Considering the case of deformation produced at a constant rate, the Author shows that the resulting relations agree with the law discovered by M. Bouasse in his experiments upon the torsion of a platinum wire (Comptes Rendus, Feb. 1897). After discussing the case of a cycle performed at a constant rate, interrupted by stops under load, the Author gives some geometrical illustrations of the foregoing, and concludes by indicating the treatment of torsion and bending, and also the influence of temperature and physical properties of the metals on the function p. A. Gs. Solar Radiation in the High Alps. G. B. Rizzo. (Nuovo Cimento, 1898, pp. 120-130.) The formulas of Pouillet (Q= Ape), Crova Q = A (1 + €)" and Bartoli (Qe"= K), for the quantity of solar heat incident per unit of normal area, though Crova's works best, are all subject to the objection that, when the sun is in the zenith (€ = 0), the values obtained for the solar constant become inadmissible, as they also do with Angström's binomial formula— Q = (2·568 × 0·751′) + (10·335 × 0.049'). Observations made on the Rocciamelone in the Val de Susa, at different heights, gave on reduction the formula Q1 = a - bμ" for the value of the solar radiation considered as zenithal, μ being the mass of air already traversed by that radiation; and the solar constant has a value very near 2.5 small calories per minute and per square centimetre, this value being, of course, apart from that of any particular radiations to which the upper regions of the atmosphere may be opaque. A. D. Properties of Nickel-Steels. C. E. GUILLAUME. (Journal de Physique, vol. vii., 1898, pp. 262-274; also Comptes Rendus, vol. cxxvi., 1898, pp. 738-740.) The characteristics of various compounds of iron and nickel are considered. With regard to magnetic properties, nickel-steels containing besides iron and nickel only small quantities of carbon, silicon, and manganese, fall into two distinct categories. The first, containing from 0 per cent. to about 25 per cent. of nickel, and which seem comprised pretty accurately between the formulas Fe and Fe,Ni, are "irreversible," in the sense that at one and the same temperature they can exist in two essentially different states, according to the preceding cycle of temperatures. When these alloys are heated they lose their magnetism gradually between two temperatures, which are comprised for all the alloys between dull red and cherry red. When they are cooled they pass again through the same temperatures without becoming magnetic, and only re-acquire their first condition at a temperature lower than those between which the loss of magnetism occurs. The return to the magnetic state takes place for an alloy with 24 per cent. nickel a little below zero. The presence of chromium lowers the temperature at which the return takes place. The steel, with 22 per cent. of nickel and 3 per cent. of chromium, remains non-magnetic even in liquid air. Steels with more than 25 per cent. of nickel are "reversible," and possess, at each temperature, magnetic properties which to a first approximation depend only on the actual temperature. The expansion of the material has been observed by comparison with a brass scale. The phenomenon noticed by A. Le Chatelier has been confirmed; the expansion for the same alloy is much more feeble in the magnetic state than in the non-magnetic state. The transformation is gradual as regards expansion as well as regards magnetism, and the same alloy may possess any coefficient of expansion between two determined limits according to its degree of transformation. In one of the extreme states the expansion is a little greater than that of brass; in the other it is lower than that of ordinary steels. At ordinary temperatures the steels of the second category possess a coefficient of expansion which varies continuously with the composition. Steels containing 35 per cent. to 36 per cent. of nickel expand ten times less than platinum. Other mechanical constants and their changes are also referred to. The alloys of the first group in passing from their non-magnetic state, where they are relatively soft and easily deformed, to the magnetic state, when they become hard and very elastic, undergo a diminution of the modulus of elasticity. The nickel-steels possess a high specific resistance, and the transformations which they undergo do not seem to affect the variations of the electric resistance; the curve of change is an ordinary one. The variation of resistance with temperature seems to be independent of the variations of volume. For the alloys of the first sort the passage to the magnetic state takes place with increase of volume. A chemical origin is suggested for the peculiarities of behaviour, and it is supposed that definite chemical compounds tend to be formed. The experiments show that the irreversible nickel-steels can possess between extended limits of temperature an indefinite number of states of equilibrium, which they keep almost without modification so long as the alloy does not cut, at low or high temperatures, two curves of transformation along which all its properties change gradually and simultaneously. They possess, besides, unstable states of equilibrium which can be broken quickly, and to which an almost instantaneous transformation puts an end. See also Phys. Soc. Abstracts, Nos. 469 and 517, 1897. J. J. S. Deformations of Dielectric in Electric Field. P. SACERDOTE. Comptes Rendus de l'Académie des Sciences, Paris, vol. cxxvi., 1898, pp. 1019–1022.) The unitary deformations undergone by a plate of dielectric substance are always proportional, if the temperature be kept KH2 constant, to the electric energy per unit of volume, coefficients are (k1+a) for the unit deformation perpendicular to 8 π . The the field, (k, — a — 2b) for that parallel to the field, and (k + c) for the variation of volume per unit of volume; a and b being the coefficients of longitudinal expansion and transverse contraction, and c that of cubical compressibility, k, the coefficient of variation of K, with traction perpendicular to the lines of force, k2 that with traction parallel to them, and k that with uniform superficial traction. These k's are all small; if negligible, the result is that there is always contraction in the direction of the lines of force, lengthening transversely, and an increase of volume. A. D. Photo-electric Relations of Coloured Salts. (Annal. Phys. Chem., vol. lxii., 1897, pp. 599–602.) Just as chloride of sodium and other salts which have become coloured through exposure to kathode rays have their colour discharged by exposure to sunlight or daylight, so do these salts when they have acquired similar colours through exposure to potassium or sodium vapour. When coloured with a trace of Prussian blue they do not. Everything points towards these coloured products being solutions of traces of the metal in the solid salts. A. D. Relations between Luminous and Chemical Energy. BERTHELOT. (Comptes Rendus de l'Académie des Sciences, Paris, vol. cxxvii., 1898, pp. 143-160; see also Comptes Rendus, vol. cxxvii. p. 84.) This series of experiments were conducted at the ordinary temperature under (1) direct sunlight, (2) diffused daylight, (3) darkness; also with interposition of various liquids and solutions, with a duration of weeks or months. Nitric acid and nitric anhydride, when pure, underwent no decomposition in the dark. But the former, under protracted exposure to solar light, gave amounts of oxygen corresponding to from 12 to 42 per cent. of the total transformation 2HNO, = 2NO2 + 0 + H2O, free nitrogen and nitrous oxide being absent. The effects of light are parallel to those of heat, but produced at a lower temperature. Diluted nitric acid (1.365 specific gravity) was found to be practically stable in solar light, probably because the layers of NO, which might be formed tended to absorb the actinic rays. A solution of potassium dichromate completely arrested the effective solar radiation, whereas ammonio-sulphate of copper in fairly thick layers did not. Numerous trials were made with chloride of silver, hydriodic acid, and other bodies affected by light, which showed that the changes, as in the case of heat, were reversible under exothermic or endothermic conditions. Copious details of the experiments are given in the original Paper. S. R. Sayers' Automatic Third-Brush Regulation of Dynamos. (Electrician, vol. xli., 1898, pp. 358-359.) If a third brush is arranged to bear on the commutator of a dynamo at a position midway between the ordinary brushes, the voltage between this brush and either of the other brushes will be the same when the machine is running light. When load is put on the field becomes distorted, and the voltage between the backward main brush and the third brush is reduced, while the voltage between the forward main brush and the third brush is increased. If the shunt-coils are wound to give the light-load ampere turns required with half the voltage of the machine, |