WHOLESALE PRICES IN FRANCE, GERMANY, BELGIUM AND ITALY. The index numbers for April revealed a cheapening of commodities in all four countries under review, the greatest fall (3.8 per cent.) occuring in Italy. Belgian wholesale prices declined by 1.4 per cent. and French wholesale prices by 0.8 per cent., while in Germany the decline was of little significance. Owing to the employment of different base periods, the index numbers are not strictly comparable in magnitude as between one country and another, but some comparision can be made regarding the relative movement of prices. addition of silica or other constituents to the water, while at the same time being as practically chemically reactive as artificial zeolite, and allowing regeneration by means of salt solution in 10 minutes after an 8-hour run on hard water. No base exchange compound is of course perfect, and the main difficulty with natural minerals is lack of reactivity and with artificial zeolites solubility in water, although the advantages of the method in general are obvious. MEASUREMENT OF MEAN SPHERICAL CANDLE-POWER. Modern development in lamp design and the increasing use of shades and reflectors in good illuminating engineering candlepower as a basis for describing the lightgiving power of lamp, and to-day the use of luminous flux or mean spherical candlepower is almost universal. In the present British Standard Specification for electric lamps, for instance, values of lumens or mean spherical candle-power are scheduled, and in the testing of lamps to this specification it is therefore necessary to use some type of photometric integrator. A recently published paper describes the underlying principles of this apparatus and the method of making measurements with it in language as free from technicalities as possible, and is the first simple exposition of the subject to be published for the guidance of practical engineers. It is hoped that it will facilitate the more widespread use of this type of measurement, for not only electric lamps, but incandescent gas and other illuminants also may be rated in terms of luminous flux. The paper has been prepared by the staff of the Photometric Section of the National Physical Laboratory and embodies the results of the experience gained there. (a) On cellulose acetate in powdered form, when imported after the thirty-first day of October, 1927, and prior to the first day of May ,1928, by manufacturers and further manufactured in their own factories before the first day of June,1928, there may be paid a drawback of ninety-nine per cent. of the duties paid on the cellulose acetate in powdered form so used; (b) On yarns produced from cellulose acetate,dry spun, not more advanced than singles, when imported after the thirty-first day of July, 1927, and prior to the first day of February, 1928, by manufacturers and further manufactured in their own factories before the first day of March, 1928, there may be paid a drawback of eighty per cent. of the duties paid on such yarns so used. 2. Provided, however, that the said drawback shall not be paid unless the customs duty has been paid on the materials so used as aforesaid within the time specified in (a) and (b). ABRASIVES. REPORT ON ABRASIVE MATERIALS AND THEIR USES OF UNUSUAL INTEREST. Abrasive materials play a very important part in modern industry and in the modern home. Metal polishes for silver, for nickel, or for brass are familiar to every one; French polishing as applied to pianos and fine furniture is accomplished by the use of certain kinds of abrasives; buffed and suede leathers owe their characteristic surface finish to the correctly applied tearing action of minute particles of an abrasive such as garnet; the accurate grinding and honing of engine cylinders that is necessary in the production of the modern gasoline engine as used in motor cars, in motor boats, or in air planes is possible by applying suitable abrasives. The great importance of these non-metallic mineral products led the Mineral Resources Division of the Mines Branch, Canada, to investigate the production and use of abrasive materials in Canada. As a result of these investigations the Mines Branch has recently issued a monograph on abrasives by V. L. Eardley-Wilmot. It was found necessary to subdivide the report into four parts, which are to be published. separately. The first part, which deals with siliceous abrasives of all kinds, including such materials as quartz and pumice, will probably be issued in August; the second part, dealing with corundum and diamond, is now ready for distribution; the third part, treating of garnet and its uses, is also ready for distribution; the fourth part, dealing with artificial abrasives and manufactured products, is now in press but will not be issued for some months. This series of valuable monographs describes in considerable detail all the principal deposits of natural abrasive materials in Canada which have been discovered and developed; treats of the characteristics and trade requirements of abrasive products; and describes methods of preparation for markets and the various industrial applications of each material. This group of reports constitutes a very valuable and useful addition to the series of monographs on industrial minerals that has been issued by the Mines Branch. Copies of Part II. (Report No. 675), Carborundum and Diamond; and Part III (Report No. 677), Garnet; may be obtained now upon application to the Director, Mines. Branch, Department of Mines, Ottawa, Canada. IONISATION BY COLLISIONS OF THE SECOND KIND IN THE RARE GASES. By GAYLORD P. HARNWELL. Abstract from the Physical Review, U.S.A. A positive ray apparatus was used to investigate the products of ionisation by electron impact in mixtures of helium, neon, and argon. The variation with pressure of the ratio of the two types of positive ions present was investigated in detail for three cases. Case 1: A mixture of half helium and half neon was investigated up to 0.15 mm pressure. The ratio He+/Ne + was found to decrease regularly between 0.03 mm and 0.15 mm. At 0.03 mm the mean free path is approximately equal to the dimensions of the apparatus. The suggested reaction is: Ne + He+Ne++ He. Case 2: A mixture of half neon and half argon was investigated throughout the same pressure range. The ratio Ne+/A+ decreased regularly between 0.05 mm and 0.15 mm, but this decrease was less rapid than that of the ratio in Case 1. The suggested reaction is: A+ Ne+ A++ Ne. Case 3 This mixture was half helium and half argon, and the pressure range was the same as in the first two cases. The ratio He+/A+ remained constant within the limits of experimental error. Case 3a: A mixture of 15 per cent. helium and 85 per cent. argon were also investigated as there were theoretical grounds for believeing that the rate of variation of He+/A+ with pressure would be greater in such a mixture. In that case the ratio He+/+ was found to decrease slightly. The suggested reaction is: A+ He+ A++ He. The observed effects are best explained by a type of collision of the second kind which is equivalent to ionisation by positive ions. To account for the results obtained an electron must be transferred from an atom to an ion as a certain fraction of the collisions between an atom and an ion of higher ionising potential. The results obtained at low pressures corroborated the values for the probability of ionisation in these gases obtained by K. T. Compton and C. C. Van Voorhis. CATALOGUES. British Drug Houses catalogue in alphabetical order, with prices of the various pure chemicals for research and analysis. Donovan & Co. Manufacturing electrical engineers, Cornwall Street, Birmingham. Illustrated catalogue and price list of 56 pages. MAGNETIC RELUCTIVITY RELATIONSHIP. By RAYMOND L. SANFORD. An experimental and theoretical study of the magnetic properties of pure iron near saturation has led to the conclusion that the so-called reluctivity relationship of Kennelly does not truly represent the properties of pure and homogeneous materials, and that the constants in the reluctivity equation are without physical significance. The Kennelly formula and one recently proposed by Gokhale have practical value, each within its own range, for interpolation and extrapolation. The conclusion that the constants in the reluctivity formula are without physical significance makes the problem of correlatmg the magnetic properties more difficult. A new basis for correlation must be sought. SUMMARY AND CONCLUSIONS. To recapitulate briefly: The problem of the correlation between the magnetic and other physical properties would be much simplified if a formula were available for expressing the magnetic properties in terms of constants having a definite physical significance. It has for some time been considered that the Kenelly law of magnetic reluctivity represented the magnetic properties of magnetically homogeneous materials, and it was hoped that the constants might have some definite physical significance. Gokhale has called this view into question and proposed a substitute formula. While his experimental evidence did not seem to be conclusive and his proposed formula has no better theoretical foundation than the Kennelly law, it appeared worth while to make a further study of the subject. Such a study was carried out with the following conclusions: 1. The characteristic curve between magnetisation and permeability for pure materials near saturation has a point of inflection and double curvature, and is not straight as called for by the Kennelly relationship. This view is upheld by experimental evidence and theoretical considerations. 2. Since the reluctivity curve for pure and homogeneous materials is not straight, but only apparently so, on account of the insensitive nature of the curve, the true saturation value is not indicated by the reciprocal of the slope of the curve and the value of the intercept has no physical significance. 3. While the Gokhale formula fits the observed curve near saturation better than the Kennelly formula, it is developed from a basic assumption having no better physical foundation and it is improbable that its constants have a physical significance. 4. Both formulas are useful, each in its own range, for interpolation. The Gokhale formula is better for extrapolation, provided data are available for values of H greater than 500 gilberts per centimeter. 5. In the light of these conclusions the reluctivity relationship has a limited value, if any, in connection with the problem of the correlation between the magnetic and other physical properties of ferromagnetic materials, and a new basis of correlation must be sought. + Babcock's value for the wave-length of the auroral line being 5577.350 0.005, there would thus seem to be no doubt as to the identity of the lines. It would thus appear that oxygen as well as nitrogen must be present in those regions of the upper atmosphere whence the auroral light is transmitted. The Combination of Nitrogen and Hydrogen Activated by Electrons. By A. CARESS AND E. K. RIDEAL. Communicated by T. M. Lowry,F.R.S. A study of the combination of nitrogen and hydrogen to form ammonia in a triode valve has been made. It is shown that hydrogen atoms formed by thermal dissociation at a hot tungsten emitter will combine with nitrogen at platinum and nickel surfaces to form ammonia. Hydrogen atoms excited by collision with ca. 13-volt electrons will react with molecular nitrogen in the gas phase to form ammonia. In the absence of hydrogen atoms (by using a mixed barium calcium oxide emitter) no formation of ammonia is observed until a voltage of 17 volts is attained! a further rise in rate of formation is noted at 23 volts. These two rises are attributed to the chemical reactivity of N2+ and N+. 2 It is shown inter alia that H+ ions will reduce the oxides of the alkaline earths. From a study of the rate of formation of ammonia with different types of grids, it is suggested that ammonia is produced by interaction of N,+ and N+ with hydrogen to form primarily excited hydrogen atoms. The electron-efficiency of ammonia-formation from nitrogen ions has been shown to be at least one ammonia molecule produced for the passage of four electrons ;in the case of excited hydrogen atoms the efficiency appears to be even higher. The Total Ionisation Due to the Absorption in Air of Slow Cathode Rays. By J. F. LEHMANN AND J. H. OSGOOD. Communicated by Sir Ernest Rutherford, P.R.S. Experiments were undertaken to determine ionisation produced by complete absorption in air of slow cathode rays. Electrons ejected from a hot tungsten filament were accelerated to an anode, by a potential difference of 200 up to 1,000 volts. A portion of these electrons passed through a carbon capillary into an ionisation chamber. The air pressure in this chamber could be maintained at any required value, below 2 mm. Hg., by means of a capillary leak from a gas pipette. Air leaking from the ionisation chamber to the filament chamber, was removed by high-speed diffusion pumps. The electron current entering the ionisation chamber and the positive ion current due to the absorption of these electrons, were measured alternately. Ratio of ionisation current to electron current gave average ionisation per electron. Using an electron bea mof given initial energy, ionisation per electron was approximately proportional to pressure of absorbing air, provided this pressure was less than a certain "critical pressure." For greater pressures ionisation. per electron was constant. The magnitude of the critical pressure was determined by the initial energy of the electron beam. Ratio of ionisation current to electron current, at pressures greater than critical pressure, represents average total ionisation due to complete absorption of an electron. This was found directly proportional to initial energy of electron, and the ratio, initial energy of electrons average per ion pair was 45 electron-volts, whereas the ionisation potential of air is 17 volts. It is evident that in air a considerable portion of initial electronic energy is dissipated in processes other than simply separatin gthe molecule into positive and negative ion. THE FARADAY SOCIETY. The following paper was read at the meeting on April 23, 1927 : The Activity of Zinc Chloride in Concentrated Solution. By F. FOXTON AND W. J. SHUTT. The electrochemistry of fused salts is essentially simpler than that of aqueous electrolytes in that one component (water) is eliminated from the system. Thus it might be expected that the properties of molten electrolytes would ultimately prove simpler than those of the corresponding aqueous solutions. On examination, however, the behaviour of fused salts presents the anomaly of indicating a high degree of molecular association, while at the same time all the phenomena usually coupled with high or complete electrolytic dissociation are observed. In a review of the accumulated data regarding the constitution of fused salts, R. Lorenz showed that the application of the many fused salts indicated an association to the extent of at least four molecules to each aggregate. This conclusion was further borne out by the work of Walden on the latent heats of fusion and vaporisation of such melts. A similar anomaly has been observed by McBain and co-workers in the case of concentrated solutions of soaps. Their results have indicated the presence of highly charged aggregates "ionic micelle ") consisting of large complexes of fatty ions appreciably hydrated. The high electrostatic charge of the micelle ensures it a high mobility despite its relatively large dimen sions. It seemed that it would be of interest to examine the properties of that type of electrolyte which lies between the conditions of aqueous solutions and those of fused salts. This paper is a description of the investigation of the electromotive forces of cells containing an electrolyte in aqueous solutions so strong as to approach the conditions of molten salts. The electrolyte chosen was zinc chloride. This salt has, at ordinary temperatures, a high solubility which increases continuously as the temperature rises; it also furnishes ions which are reversible with respect to massive electrodes both in aqueous solution and in the fused state. EXPERIMENTAL. Apparatus.-The preparation of a reliable zinc electrode presented less difficulty than had been expected from the literature. Pure zinc was cast in hot graphite moulds to the form of small cylinders, which were turned on the lathe to give a bright smooth surface. Success depended on the preservation of this surface from any trace of and grease this was attained by immediate immersion of the freshly turned rods in pure benzene. Shortly before use the electrodes washed with alcohol and water and plated in an ammoniacal solution of zinc sulphate, as recommended by Richards and Lewis. The electrodes prepared in this way were placed in half element vessels designed to ensure that the zinc was entirely below the surface of the liquid in the thermostat. were Platinum and carbon presented themselves as suitable media for the chlorine electrode. As it was feared that the former might cease to remain passive at the high temperatures to be employed, it seemed desirable to devote considerable attention to the behaviour of carbon, especially as the latter had been previously studied at high temperatures, notably by Lorenz and Fox, who had shown that the carbon-chlorine electrode is reversible and gives readily reproducible potentials in molten chlorides of heavy metals. The results were disappointing, however. At ordinary temperatures, even with electrodes of the smallest possible dimensions, the time required for attainment of equilibrium was inordinately long, and up to 60° C. it was impossible to reproduce the potential of such an electrode with greater precision than about 10 millivolts. The results seemed to indicate that, at temperatures higher than those employed in this work, carbon might function satisfactorily and replace the platinum electrodes which had perforce to be used for this research. The metal was platinised for use in solutions below about molal strength, but with higher concentrations bright platinum prove dquite satisfactory. The chlorine, prepared by the action of pure hydrochloric acid on cold solid potassium permanganate, was stored over sulphuric acid, and tests by absorption in alkali indicated a purity of over 99.7 per cent. for the gas. VAPOUR PRESSURE MEASUREMENTS. The interpretation of the results obtained. by the use of a chlorine gas electrode involves a correction for the partial pressure of the gas and, since data for the vapour pressures of zinc chloride solutions under the conditions employed were lacking, it became necessary to measure these values to an accuracy of about + 3 mm. of mercury. For vapour pressures at 60.48° C., a modified form of isoteniscope was used. At this temperature little difficulty was experienced in establishing equilibrium between the solution in the bulb and manometer of the isoteniscope. Dilution of the solution by condensation of water during cooling was avoided by sealing the isoteniscope from the vacuum reservoir and gauge at a point immediately above the surface of the thermostat. The bulb was broken from the isoteniscope and its contents analysed after cooling. At 80.4° C. the isoteniscope proved unreliable and a comparison method was adopted. Two solutions, one of zinc chloride and the other of sulphuric acid, were placed in bulbs, connected by wide tubing. After evacuation the system was sealed and placed in the thermostat, where it was gently shaken for at least two hours. The breaking of a fine capillary tube permitted air to enter the apparatus and practically stopped further distillation during cooling. The contents of both bulbs were analysed. Qualitative tests showed that no mixing had occurred. Sulphur.c acid solution's were chosen as reference substances since their vapour pressures have been thoroughly worked out by Burt for a wide range of concentrations and temperatures. Points on the curve were checked at.comparatively low concentrations by the replacement of sulphuric acid by saturated solutions of sodium carbonate and sodium sulphate respectively, whose vapour pressures have been determined by Gerasimow and P. Pawlowitsch. THE PHYSICAL SOCIETY. Proceedings at the Special Meeting held at the Special Meeting held at the Imperial College of Science on Friday, June 3, D. Owen,B.A., D.Sc., in the chair: A Lecture on the Eclipse of the Sun was delivered by E. H. Rayner, M.A., Sc.D. EXTRACTS FROM LECTURE. The initial partial phase and totality take place before 6 a.m. G.M.T. In such circumstances the diurnal rotation causes England to have a small component opposite to the eastward motion of the moon's shadow. The final (fourth) contact will occur after |