ratio was observed between the electric conductivies without and with exposure to light under conditions similar to hose described above. The authors, however, would not insist on this concordance of results in the two phenomena. H. V. C Effect of Light on the Conductivity of Selenium. By S. KALISCHER (Ann. Phys. Chem. [2], 32, 108). Of the selenium cells constructed by the author, three in which copper and copper-brass electrodes are used, are found to differ from the rest in their behaviour on exposure to light, the resistance rapidly increasing after undergoing a momentary decrease, and the cell only returning to its normal condition on remaining for some time in the dark. The conclusion drawn from this is, that the cells in question contain a hitherto unknown modification of selenium, the conductivity of which decreases instead of increasing under the action of light. As the author's other cells which do not exhibit the peculiarity described, differ from the above in having zinc, copper-zine, and copper-platinum electrodes, it still remains to be ascertained whether the nature of the electrodes has any influence on this behaviour of selenium. The phenomenon in question has also been observed and described by Hesehus (Ern. Rep. d. Phys., 20, 490). H. C. New Galvanic Battery. By F. FRIEDRICHS (Ann. Phys. Chem. [2], 32, 191).—A tube running below the cells of this battery connects each with a common reservoir, by the raising or lowering of which the fluid used can be transmitted to or removed from the cells. A tap attached at the end of the tube opposite the reservoir allows the fluid to be removed when exhausted. An advantage claimed over other batteries is, that spontaneous evaporation of the liquid and consequent crystallisation of salts when the battery is not in use, is avoided. H. C. Galvanic Polarisation. By F. STREINTZ (Ann. Phys. Chem. [2], 32, 116). The author has examined the galvanic polarisation produced on aluminium and silver plates. The results for aluminium have been already given (Abstr., 1887, 415). With silver, the oxygen plate is found to attain maximum polarisation when the E.M.F. of the cell used is equal to that of three Daniells; the polarisation of the hydrogen plates is at a maximum when an E.M.F. of two Daniells is used, it decreases when a greater E.M.F. is employed, but rises again and becomes equal to the first maximum for an E.M.F. of nine Daniells. The explanation given is that the deposition of metallic silver on the cathode, which is greater the greater the intensity of the current, by increasing the surface decreases the relative strength of the current and amount of the polarisation, so that although a small E.M.F. produces maximum polarisation with clean. plates, a very considerable one is required to attain the same maximum with plates thickly coated with silver. H. C. Production of Electricity by the Condensation of Aqueous Vapour. By L. PALMIERI (Nuovo Cimento [3], 22, 34-39).-The occasion of this paper is the confirmation by Firmin Larroque (La Lumière Elect., 1887) of the author's experiments on the production of electricity by the condensation of aqueous vapour. On the other hand, the experiments of Kalischer (Abstr., 1884, 138) led to negative results, but Tait considers that these were conducted on far too small a scale. Accordingly the author has repeated on a large scale his experiments on the condensation of aqueous vapour on a beaker of platinum containing ice, and connected with a condensing electric cup; in all cases, the production of electricity was observed. The author remarks that his observations, extending over 37 years, leave no doubt in his mind as to the production of electricity under these conditions. The potential of atmospheric electricity is conditioned by the state of the weather; the author's observations also have more particularly shown that the potential is affected by the eruptions at Vesuvius. V. H. V. Electrolysis of Water. By H. v. HELMHOLTZ (Ber. Akad. Ber., 1887, 749-757).-Previous experiments made by the author showed that the smaller the amount of dissolved hydrogen and oxygen near the electrodes, the smaller the electromotive force necessary to electrolyse water. The experiments described in the present paper were made with a view to determine the limits for the smallest electromotive force capable of producing fresh gas under a given pressure of the oxyhydrogen mixture on the liquid. In previous experiments, an error in the measurement of the electromotive force of the decomposition of water was caused by hydrogen or other combustible gas being occluded in the platinum anode or in both electrodes, so that the oxygen carried over in the current comes in contact with the gases of the anodes, and thus bubbles of hydrogen will be liberated at the cathodes with a much less expenditure of electromotive force. To avoid this, the current is kept in the same direction for weeks or months. An apparatus is described with sketch, by means of which the gases produced by the electrolysis are removed as soon as formed, and a vacuum is thus kept above the liquid; the flask containing the solution is so inclined that a small bubble of gas is retained; the gas under these conditions occupies a space 1000 times greater than it would under normal pressure, and the diameter of the bubble is measured in order to ascertain whether it remains the same size or whether it increases. To produce a current, three carbon-iron ferric chloride solution elements were used; the electromotive force was diminished daily in order to determine the limit. The limit for the evolution of gas was found to be 1.64 to 1.63 volt, with a pressure of oxyhydrogen gas 10 mm. of water. The influence of pressure on electromotive force is expressed as follows: : Pa = atmospheric pressure, pr and p, are the pressures of hydrogen Ph and of oxygen above the liquid; ah and ao are the atomic weights of the two elements; is the absolute temperature. second by one Ampère. : = where v is the volume of 1 gram of hydrogen; R, the corresponding A, : = 0.018868. log nat. ⋅ (?!). N. H. M. Electrolytic Separation of the Metal on the Free Surface of the Solution of its Salt. By J. GUBKIN (Ann. Phys. Chem. [2], 32, 114).—When an electric current passes from a solution of a salt into the atmosphere of gas or vapour immediately above it, an electrolytic separation of the metal takes place at the surface of the liquid. Apparatus is described by means of which this is made evident, the space above the liquid being either vacuous or exposed to the air in the ordinary way. Silver and platinum are found to separate out in films which float on the surface; zinc oxidises as it separates out, the white flakes of zinc oxide gradually falling to the bottom. H. C. Action of the Solvent on Electrolytic Conduction. By T. C. FITZPATRICK (Phil. Mag. [5], 24, 377-391).-The author continues his researches on the conductivity of salt solutions, the solvents being varied. The salts examined were calcium, lithium, and magnesium chlorides and nitrates, and ferric and mercuric chlorides, the solvents being water and ethyl and methyl alcohols. Tables of conductivities are given. With mercuric chloride, which is the only salt more soluble in alcohol than in water, the conductivities are little more than those of the solvents alone. For aqueous solutions, the chlorides conduct better than the nitrates; magnesium chloride is anomalous, its conductivity being half that of calcium chloride. Ferric chloride in dilute solution shows signs of dissociation. With alcoholic solutions, the conductivity is not proportional to the amount in solution. The conductivity of lithium salts in ethyl alcohol is 10 to 20 times as great as that of the other salts. In all cases, the aqueous solutions conduct better than the alcoholic ones, the character of the solvent appearing to have an influence on the conductivity. This the author considers to be due to the formation of molecular groups in the solutions. He finds that the conductivity of salt solutions at low temperatures points to the existence in solution of cryohydrates at temperatures above their solidifying points, and also that the conductivity of mixed solvents and of salts in mixed solvents differs from the calculated values, showing that an interaction has taken place with formation of new molecular groups. The action then of the solvent is twofold: (1) decomposition of the salt, the amount depending on the temperature, nature of solvent, and state of dilution; (2) the formation of fresh molecular groups in the solution. H. K. T. Influence of a Magnetic Field on the Thermoelectric Properties of Bismuth. By G. P. GRIMALDI (Nuovo Cimento [3], 21, 57). It is well known that a magnetic field influences in a remarkable degree the electric resistance of bismuth; in this paper, the author shows that its thermoelectric force when paired with copper is varied in a similar degree. This pile was placed in the field of an electromagnet, and coupled up with a galvanometer, in which readings were taken without and with a current passing round the electromagnet. After due allowance for induction, it is shown that the thermoelectric force of the bismuth-copper pair is materially decreased in the magnetic field. The experimental enquiry is, however, only in the preliminary stage. V. H. V. Rotation of Isothermic Lines of Bismuth placed in a Magnetic Field. By A. RIGHI (Gazzetta, 17, 359).—In the course of experiments on the heat conductivity of bismuth when placed in a magnetic field, it was observed that the isothermic lines were rotated in a direction opposite to that of the magnetising current when a rectangular strip of the metal was placed with its planes normal to the line of force. The phenomenon is analogous to that observed by Hall, namely, the rotation of the equipotential lines when a magnet acts on a current flowing along a thin strip of metal, and may explain the thermomagnetic currents recently discovered by Ettingshausen. V. H. V. Thermic Conductivity of Bismuth in a Magnetic Field. By A. RIGHI (Gazzetta, 17, 358-359).-The author, as well as other physicists, has observed the marked variation of the electric conductivity of bismuth when placed in a magnetic field (Abstr., 1887, 1009), and the production of Hall's phenomenon under these conditions. Considering the correlation of electric and thermic conductivity, the effect of magnetic field was also studied; the results of the experiments showed that with a field of 4570 C.G.S. units the thermic conductivity of bismuth is to that of the metal under ordinary conditions as 1 : 0886. This result must at present be only considered as approximate; further experiments are being made with more refined apparatus. V. H. V. Specific Heat of Superfused Water. By P. CARDANI and F. TOMASINI (Nuovo Cimento [3], 21, 185).—The specific heat of water at various temperatures has been the subject of numerous investigations, although the results obtained are far from concordant. Thus at temperatures 0-10°, Hirn, as also Pfaundler and Platter, has observed a marked increase of specific heat, whilst Rowland on the other hand observed a decrease. In this paper, a description is given of experiments made to determine the specific heat of water in the superfused condition. The method adopted in the investigation is practically an application of the weight thermometer; a known volume of water is enclosed by mercury within a bulb, connected with which is a capillary tube bent twice at right angles. The whole apparatus is completely filled with water and mercury, and the bulb cooled by suitable freezing mixtures, then the mercury driven out by the expanding water is collected and weighed. The apparatus is then agitated, and the mercury driven out by the solidification of the water is also collected and weighed. Then from these data, together with a determination of the temperature at the moment of solidification, and the quantity of heat absorbed by the glass and the mercury contained, the specific heat of the water at the temperature of solidification is ascertained. The various experimental errors are discussed in full, and the data of all the observations given in a series of tables. The following are the main conclusions: the specific heat of superfused water is less than unity; it increases with decrease of temperature from a minimum at a temperature of -6·52° to 0°. The final results are given below. In New Form of Calorimeter. By W. F. BARRETT (Proc. R. Dublin Soc., 5, 13-16).-The instrument devised by the author is a modification of Bunsen's calorimeter. The cup for holding the substance under experiment forms part of a mercurial thermometer. The cup has a capacity of 4 c.c., and is surrounded by a jacket of polished metal. The stem of the thermometer, of which the cup is a portion, is supported horizontally, and graduated from -5° to 80°. Supported immediately above the cup is a small burette, the level of the liquid in which can be accurately read. The neck of the burette may be closed by a short thermometer graduated from 30° to 100°. making a determination of the specific heat of a liquid with this instrument, the weight of the liquid must be found by taking its specific gravity for the temperature at which it was used; the volume of the liquid used having been read from the burette. This inconvenience may be obviated by converting the thermometer into a balance, the stem being supported by knife-edges somewhere near its centre of gravity. From the end of the stem, a pan is suspended, and beyond this a pointer, fixed to the stem, moves over a graduated arc. With a calorimeter balanced in this way, the weight of the liquid at a given air-temperature may be found directly. B. H. B. Determining the Specific Gravity of Small Quantities of Dense or Porous Substances. By J. JOLY (Proc. R. Dublin Soc., 5, 41-47). The method generally employed for determining the specific gravity of small quantities of minerals of low density is by balancing in a liquid of known specific gravity. This method, however, is inapplicable when the substance has a specific gravity over 4, and also when the substance is of a porous nature. Under these conditions, the |
