3.84 per cent., Se 10.84 per cent., C 16.46 per cent., and Br 65.59 per cent. On standing over soda-lime, this salt, like the pyridine compound, slowly loses bromine. When the mother-liquor from which the red soluble crystals were obtained was allowed to slowly evaporate, a black salt appeared, from which, when treated with water, large quantities of selenium immediately separated. Analysis of various samples of this black compound gave determinations of selenium from 11.79-14.39 per cent. Bromine estimations indicated a variation in content from 67.42 per cent. to 70.71 per cent. From the irregularities in composition of different samples and the variations in color from red to black, it would appear that piperidine bromoselenate, when crystallized from solutions containing an excess of selenium tetrabromide, tends to enclose the latter and carry it down with it. THIRD SECTION. SELENIUM MONOXIDE. Different chemists at various times have either affirmed or denied the existence of the monoxide of selenium. Berzelius ascribes the odor of rotten horseradish, noticed when selenium burns in air, to the existence of this oxide. Chabrie,' in his work on the organic derivatives of selenium, claims to have prepared the monoxide by heating selenium to 180° in air. So firmly convinced was he that the monoxide exists that he used this method as a means of estimating selenium. A. W. Pierce has made an investigation of Chabrie's work, and concludes that it is probably not formed by the method of Chabrie, and thinks the odor produced when selenium is burned in air is due to the presence of moisture, which, under the circumstances, would produce hydrogen selenide. While subliming the large quantities of pure dioxide used in the preceding investigation, at no time was there noticed an odor like that produced when selenium burns in air or when a selenide is roasted. On the contrary, the vapor of selenium 1 Ann. chim. phys. [6], 20, 273. 2 Ztschr. anorg. Chem., 13, 121. dioxide which at times escaped into the air, gave an odor much resembling that of sulphur trioxide. An experiment was made to determine whether selenium would reduce the dioxide when heated in an open vessel at the atmospheric pressure. No combination was found to take place. A second experiment was made, in which equivalent parts by weight of selenium and dioxide were placed in a stout glass tube of hard glass, about forty inches long by one-half inch internal diameter. The tube was sealed and heated in a large flame, until the selenium melted. The dioxide under the pressure generated by its own volatilization, melted to a thick liquid. The temperature was raised to the boiling-point of selenium. Should selenium be able to reduce the dioxide, it should do so in a solution of the combined gases. Such, however, is not the case; on cooling, the mass resolves itself into separate layers of selenium and dioxide. When the tube was opened, no pressure was observed (such as would have been the case had a gas been formed), nor could any odor be detected. On treating the mass with water, selenious acid was formed, which dissolved, leaving selenium insoluble. Selenium and tellurium have many derivatives from which well-established analogies can be drawn. Where a derivative of one element exists, a corresponding derivative of the other is usually found. That tellurium forms a monoxide, there can be little doubt, hence it has seemed probable that a monoxide of selenium should exist. However, in considering their halides, we find that while tellurium forms the compounds TeX, and TeX,, directly analogous to the oxides TeO and TeO,, selenium forms only those halides of the types Se,X, and SeX,, the latter corresponding to the well-known SeO,. From the analogy thus established between the oxygen and halogen derivatives, one would rather expect an oxide Se,O than SeO. It was therefore thought that the lower oxide might be prepared by treatment of the monobromide with dry silver oxide. Se,Br, was prepared by rubbing the tetrabromide with selenium. No action could be brought to take place between silver oxide and a solution of the bromide in carbon disulphide. When the liquid bromide is treated at the ordinary temperature with silver oxide, a violent reaction takes place. Much heat is evolved, the reaction being explosive in character. The resulting mass on treatment with water gives a solution containing only selenious acid. If the monoxide is formed it is likely that the heat of the reaction would decompose it. The experiment was subsequently modified by chilling the bromide to -7°. At this temperature, silver oxide swims unattacked on the surface of the liquid monobromide. No reaction takes place until the whole attains a temperature of +20°, when the same violent reaction takes place that was before noted. On conducting the experiment in a closed tube, it is noticed that on opening the tube, no pressure exists, hence it is unlikely that a gas has been formed. Should a lower oxide be formed at all in this reaction, the experiments seem to indicate that it immediately dissociates into selenium and the dioxide. As Pierce has pointed out, it seems very likely that the odor noticed when selenium is roasted in air, is due to the presence of hydrogen selenide, and that efforts to obtain other oxides than the dioxide, thus far have been unsuccessful. UNIVERSITY OF PENNSYLVANIA. C OSMOTIC PRESSURE. BY C. L. SPEYERS. ONSIDER the following arrangement: The lower vessel contains pure solvent; the tube contains a solution of some non-volatile body in that solvent. The tube is open at the P top but closed at the bottom by a diaphragm permeable to the solvent only. At equilibrium, the arrangement is to be so adjusted that the diaphragm is just at the surface of the solvent in the outer vessel. The tube and vessel holding the solvent are covered by a bell jar, the air being removed so that the only aeriform body under the bell jar is the vapor of the solvent. This is the arrangement described by Arrhenius.' At equilibrium, the counter pressure preventing the entrance of the solvent, is commonly given as equal to hs on the unit surface, h being the height of the column of liquid and s the density of the solution whose counter pressure balances the osmotic pressure, π, so that at equilibrium The difference between the pressure of the vapor on the top of the solution and the pressure of the vapor on the pure solvent in the lower vessel is so slight compared with 7 that for our purpose it is negligible. Likewise the coefficient of compressibility of solvent may be neglected by us. The value of h is a priori unknown, but when the system reaches equilibrium, then h must have such a value that the vapor-pressure of the solution equals the vapor-pressure of the pure solvent at height h. Now dp=―s'dh where s' is the density of the vapor compared with hydrogen. Experiments by Ramsay and Young, quoted by Noyes and Abbot, show that the density of ether vapor at 12.9° and under pressure in terms of hydrogen at the same temperature and under the same pressure is 36.08 +0.0581 p. One cc. hydrogen at 12.9° and under pressure weighs (0.0,8987p/76) 273/283.9 grams. Integrating by partial fractions, we get h = 24540 le P.(36.08 +0.0581p,) cm. Substituting in 1 we get π=24540*13.60 sle P.(36.08 +0.0581p) in grams....(2) The next step is to substitute for s, which in this case is to refer to ether. Now this little paper centers around this quantity s. I wish to show that for s we should substitute the value for the pure solvent, in this case ether, and not the value for the solution which is in the tube at the time of consideration. This latter value was used by Arrhenius.' Does not the great advantage of the modern theory of solution lie in the notion that the constituents of a liquid homogeneous mixture, are independent of each other? Is not this far more suggestive than the assumption of a peculiar, characteristic, almost chemical, action between the constituents, which assumption even the originators of the modern theory of solution seem at times inclined to make? For when the constituents are independent of each other, then we must seek the apparently peculiar characteristic action of one constituent upon the other constituent of a homogeneous liquid mixture, in a difference of condition offered by the one constituent to the other. For instance, suppose we have ammonium chlorid in the form of partially dissociated vapor and we introduce some nitrogen, keeping temperature and total pressure constant. Then the ammonium chloride is believed to dissociate still more; but not because the nitrogen acts chemically upon this system or in any other characteristic way upon it. Not at all. But because the introduction of nitrogen necessitates a change in the condition of the system. Of course the change in condition in this case is readily detected, whereas the changes in condition produced when one of the constituents of a homogeneous liquid mixture is changed are not so easily followed up. Nevertheless, it would seem more profitable to seek for the cause of apparently specific 1 Loc. cit |