free from impurities; and its color, physical characters and blowpipe reactions are the same as in the Canadian sodalite. A thin section showed the presence of very perfect cleavage, and the commencement of decomposition, especially along the cleavage cracks. Microlitic inclusions of hornblende and a few grains of what appeared to be elæolite were also noticed. Between crossed nicols it was perfectly isotropic. Two sides of the specimen were polished. Sodalite from the Congo State, Africa. The specimen examined was collected by Brazza, the explorer, and, so far as is known, is the only noted occurrence of this mineral in Africa. The cleavage is not very apparent macroscopically, but its color, physical characters and blowpipe reactions are the same as in the Canadian sodalite. A decomposed iron mineral (chiefly limonite) and a decomposed feldspar or clay are associated with the sodalite. A thin section. showed rather a more advanced state of decomposition than in the Ural specimen, and the presence of only imperfect cleavage.. Little patches of oxide of iron were noticed, and between crossed nicols the section was completely isotropic. One side of the specimen was beautifully polished. The specimens from the Ural Mountains and the Congo were loaned for examination by Tiffany & Co., through the courtesy of Mr. G. F. Kunz; but unfortunately no details could be obtained as to exact occurrence, associated minerals, E. Bamberger and K. Feussner note the occurrence of sodalite in Tiahuanaco, Bolivia. Zeit. f. Kryst., 1881, v, 580. Mineralogical Laboratory, Columbia College, March 27th, 1894. SCIENTIFIC INTELLIGENCE. I. CHEMISTRY AND PHYSICS. 1. On Solution and Pseudo-solution.-Some years ago LINDER and PICTON Concluded from their examination of various grades of arsenous sulphide solution, that there is no defined boundary line between suspension on the one hand and perfect solution on the other; the difference being one of degree of aggregation only. They have now added another grade of this solution, having found that on pouring a two per cent arsenous oxide solution into hydrogen sulphide water, the mixture is not only dif fusible but can be filtered through a porous pot. Of the As,S, solutions already prepared therefore, grade (a) is made up of aggregates visible under the microscope, (B) is invisible but not diffusible, () is diffusible but not filterable and (6) is both diffusible and filterable, although it scatters and polarizes a beam of light. Experiments with the higher grade solutions chiefly (1) show that as regards their power of coagulating these solutions, metallic salts can be divided into well defined groups depending upon the valency of the metal; trivalent metals having the highest coagulative power, bivalent metals only one tenth of this power and univalent metals, including hydrogen and ammonium, less than one five-hundredth. Moreover these differences are shown by the same metal when its valence varies. And the authors have observed that silver and thallium (in its thallous salts) fall in the same group as copper and the bivalent metals, while mercury and lead belong in the trivalent group with aluminum and iron. From a table giving the relative quantities needed for coagulation it appears that one molecule of aluminum chloride AICI, possesses the same coagulative power as 16:4 molecules of cadmium chloride or 750 molecules of sulphuric acid. As to the nature of coagulation it was observed that when effected by barium chloride the arsenous sulphide contained barium not removable by water, though exchangeable for another metal when digested with a cold solution of it, such as calcium nitrate. Since coagulation is due to the positive constituent of a salt, the authors were led to inquire whether the coagulative power of salts of the same metal is proportional to the number of free positive ions in the solution. And a comparison of the molecular conductivities of the chlorides, bromides, iodides, nitrates. and sulphates of potassium, hydrogen, sodium and ammonium, which are due to the free ions present, with the coagulative power, appears to indicate that this power is entirely controlled by the number of free positive ions present.-J. Chem. Soc., lxvii, 63, February 1895. G. F. B. 2. On the Fluidity of Metals below their Melting Points.-It has been pointed out by SPRING that many metals exhibit properties characteristic of the liquid state, even when at temperatures much below their melting points. In his experiments, the metals were in the form of cylinders with perfectly plane ends, placed end to end in an iron holder, and forced together by means of a screw, while heated in an air bath or in a bath of an indifferent gas. The metals used were aluminum, bismuth, cadmium, copper, tin, gold, lead, zinc, antimony and platinum. In the earlier experiments both cylinders were of the same metal, and the temperature was kept at from 200° to 400° for from four to eight hours. It was then found that, with the exception of the platinum and antimony, the cylinders had alloyed so perfectly that when one end was fixed in a lathe the entire cylinder could be turned, and when broken in a vise the fracture was not through the line of separation. When different metals were employed, as copper or lead with certain others, an alloy of a considerable thickness was produced, 18mm in the case of zinc and copper and 15mm in that of cadmium and copper. When lead and tin were used a cavity was made at one end of the cylinder and filled with mica, in order that contact should take place only at the edge. The alloy formed had a thickness of 15mm, nine millimeters being in the tin and six in the lead. With cylinders of copper and zine having a central cavity at the ends in contact, the surface of the copper next to the cavity was colored yellow, resembling the alloy formed when copper is exposed to zine vapor. The author explains these results upon the assumption that the molecules of solids, like those of fluids, have not all the same velocity.-Zeit. physikal. Chem., xv, 65, September 1894. G. F. B. 3. On the Light emitted during Crystallization.-The emission of light during the crystallization of certain salts has been examined by BANDROWSKI, who considers it to be in all probability electrical and to be due to the union of electrified ions. If this is the case it should be most decided in the sudden crystallization of strongly dissociated compounds. He suggests the following experiments in proof of this, which are suitable also for the lecture table. A glass cylinder is half filled with a warm saturated solution of sodium chloride and into it is poured an equal volume of hydrochloric acid of specific gravity 112, the whole being mixed by means of a glass rod. A bluish green light fills the entire cylinder. The experiment may be modified by pouring in the two liquids separately and carefully and then strongly shaking the cylinder. A flash of light occurs. In place of the acid, alcohol may be used and the results may be obtained with potassium bromide or chloride in place of the sodium salts. When potassium chloride was used with alcohol the effect was very marked, the light being stronger and greener than that given by sodium chloride.-Zeit. physikal. Chem., xv, 323, November, 1894. G. F. B. 4. On the Two-fold Spectra of Oxygen.-In a paper to the Royal Society, BALY has sought to account for the two-fold spectra of oxygen. These spectra are of a different nature; they behave differently and there are reasons why in all probability they are spectra of different gases. These spectra may be produced by different vibrations of the oxygen molecule, or they may be the spectra of two different modifications of oxygen, or the spectra of two distinct gases resulting from a dissociation of oxygen. In order to test the last hypothesis, oxygen was sparked in an apparatus with hollow platinum electrodes, connected with a Sprengel pump. The distance between the electrodes was 35mm and the highest pressure consistent with the production of the two spectra was initially employed, being 380m. The fractions of the gas obtained from the anode and kathode were weighed and compared with the oxygen before sparking. With long sparks a lighter fraction was obtained at the kathode and with short sparks a heavier fraction. With long sparks the density of the kathode oxygen was 15.78, 15.79, 15.80, 15.79; with short sparks 16:00, 16:01, 16-02, 16.04, 16:06, 16:05. The density of the unsparked oxygen was 15.88, 15.87, 15 89, 15.88, 15.88. The fractions from the anode showed a difference in the same direction, though not as definite. Further results are promised.— Nature, li, 550, April, 1895. G. F. B. 5. Kräfte der Chemischen Dynamik; 3 Vorträge von Dr. LUDWIG STETTENHEIMER. 8vo, pp. 88. Frankfurt-a-M. 1895. (H. Bechhold.)-These lectures appear to be aimed against the molecular constitution of matter, every substance being regarded as homogeneous and its atoms interacting mechanically with all other atoms. The reasoning seems to be loose and the conclusions altogether hypothetical. G. F. B. 6. Physical Constants of Hydrogen.-Professor RAMSAY has received a letter from Professor OLSZEWSKI in which he says: "I have at last succeeded in determining the critical temperature and the boiling point of hydrogen. I have found for the former -233° and for the latter -243°. I have used the dynamical method which I described in the Phil. Mag. A thermal couple proved of no use and I was obliged to avail myself of a platinum wire thermometer, measuring the temperatures by the alteration in resistance of the wire. I have obtained satisfactory results and intend to publish an account of them in English.Nature, March 21, 1895. J. T. 7. Color Photography.-At a meeting of the Physical Society in Berlin, Feb. 8, Dr. NEUHAUS exhibited a series of color photograph's taken by Lippmann's method with prolonged exposure. Spectra show, if the exposure is sufficiently long, a greenish band in the infra red as well as in the ultra violet, in addition to ordinary colors. The colored band was very markedly displaced by both over and under exposure. The photographs of objects with mixed colors, such as fruits, flowers, butterflies, etc., were good: but their production was extremely difficult and only one plate in twenty-five was, on an average, successful. It was found easier to photograph naturally mixed than artificially mixed colors. Some substance such as eosin or cyanin must be added to the films to make them more sensitive to red rays and less sensitive to blue. The theory of the method is still unsettled.-Nature, March 21, 1895. J. T. 8. Silvering Glass.-To a physicist any method of silvering glass which will replace the method with Rochelle salts or the Martin process is of especial interest. M. M. AUGUSTE and LOUIS LUMIÈRE describe the following method: To 100 cubic centimeters of a 10 per cent solution of silver nitrate ammonia is added drop by drop until the precipitate formed is redissolved. Too much ammonia must not be added at first, for this might prevent the formation of the precipitate. The volume of the solution is increased to a liter by the addition of distilled water. This is solution A. Solution B is made by diluting commercial Formaldehyde of 40 per cent with distilled water so as to form a 1 per cent solution. Solution B can be kept for some time. Two volumes of A are rapidly mixed with one volume of B and the mixture is rapidly poured over the glass to be cooled. In five or six minutes, at a temperature of 15° to 19°, all the silver in the solution is deposited in a brilliant layer which can then be washed with water.-Journal de Physique, January, 1895. J. T. 9. A Form of Sensitive Galvanometer.-In a note to the French Academy, presented by Prof. Mascart, M. Pierre WEISS describes a new method of making the suspended magnetic system of a galvanometer. The system is formed of long vertical needles, placed parallel to the axis of rotation in such a manner that they constitute with their opposed poles almost a closed magnetic circuit. Each one of the two systems of poles is placed at the center of suitably constructed bobbins. The almost complete absence of demagnetizing force, allows the maximum magnetization of the steel: and one can by changing the distance of the needle change at will the ratio of the magnetic moment to the moment of inertia. If the sensibility of a galvanometer is defined as the number of divisions which it indicates for one micro-ampere divided by the square root of the resistance, the scale being at a distance from the mirror equal to 2000 divisions and the duration of the oscillation being five seconds, M. Weiss obtains S 1500. This sensibility can be increased by greater care in the mechanical construction of the instrument. The author states that Mr. Wadsworth, Phil. Mag., No. 38, 1894, describes a galvanometer of more difficult construction which gave S 1300.-Comptes Rendus, No. 13, April, 1895. = J. T. 10. On the Diselectrification of Air.-Lord KELVIN has continued his experiments on this subject with the assistance of Messrs. MAGNUS MACLEAN and ALEXANDER GALT. It was found that positive or negative electricity given to air by an electrified needle-point can be conveyed through 3 or 4 meters of small metal tube (1cm diameter) and shown on a quadrant electrometer by a receiving filter. A filter of 120 wire gauges only reduced the electrical indication to a little less than half of what it was |