-10° and +75° by a right line with a coefficient 0.1470, and between 75° and 180° by a second right line, which has a coefficient of 0.0793, and a limiting point at 913°, or considerably above the melting point of the salt. In presence of sodium chloride, the curve of solubility of potassium chloride between -20° and +75° is a right line y = 103 +0.0962t; from 75° to 120° the solubility increases rapidly, and above 120° it is represented by a right line with the same coefficient as between -20° and +75°. Its limiting point is 913°, and hence the curves of solubility of potassium chloride alone and in presence of sodium chloride are not parallel, but converge to 913°. At the limiting point for the mixed salts, 738°, the proportion of the two salts would be 16.7 per cent. of sodium chloride and 83.3 per cent. of potassium chloride. The total quantity of chlorine is practically equal to the sum of the metals. The curve representing the quantity of chlorine in solution is a right line; that representing the sum of the salts is also a right line; and hence the sum of the metals is likewise represented by a right line. The curve of the chlorine and the curve of the sum of the metals intersect at 738°. C. H. B. Determination of Molecular Weights of Substances from the Boiling Points of their Solutions. By H. W. WILEY (Chem. News, 60, 189–190). "The apparatus employed consisted of an ovalround bottom flask of about 200 c.c. capacity," with a side tube from the neck connected with a condenser to keep volume of liquid constant. A thermometer graduated to tenths, but capable of being read to 0.02 of a degree was employed, the bulb being enveloped in fine copper foil to prevent interference of bubbles of steam. Sodium chloride was used to determine the factor, the number obtained, 8.968, was used for calculating the results in the following table; the volume of water being in all cases 150 c.c.; the temperature of boiling water was 99.50° except during the experiments with sodium nitrate, when it was 99.44°. It will be noticed that the two organic compounds give double the theoretical molecular weight by this method. The results obtained with salts containing water of crystallisation do not agree with the molecular weights with or without this water. These results were obtained quite independently of those of Beckman. D. A. L. Behaviour of Colloïd Substances with Respect to Raoult's Law. By E. PATERNO (Zeit. physikal. Chem., 4, 457-461).-The reduction of the freezing point by colloïd substances in water is very slight, and therefore leads to very high numbers for the molecular weights of such substances (Brown and Morris, Trans., 1889, 462). This, the author has observed, is the case with gallic and tannic acids; which behave like colloïds in aqueous solution and give molecular weights many times greater than those ordinarily accepted for these substances. If, however, solutions in acetic acid are taken, the behaviour is found to be perfectly normal, and the reduction of the freezing point is that corresponding with the ordinary simple molecular weights. Hence substances only behave as colloids towards certain solvents, and the author holds that when a solid dissolves as a colloïd, the laws of freezing are not applicable to its solutions. H. C. Can Raoult's Method distinguish between Atomic and Molecular Union ? By R. ANSCHÜTZ (Annalen, 253, 343—347 ; compare Anschütz and Pulfrich, Abstr., 1888, 1273). The depression produced by naphthalene picrate in the freezing point of benzene corresponds with that which would be produced by its constituent parts present together in an uncombined state. The author concludes therefore that the combination of the components of naphthalene picrate and analogous substances such as dimethyl diacetylracemate is not dependent on atomic union in the sense of the valence theory, but on molecular union. If Raoult's method is capable of deciding between atomic and molecular union, it could be employed for determining the valency of elements. F. S. K. Kinetic Nature of Osmotic Pressure. By G. BREDIG (Zeit. physikal. Chem., 4, 444-456).—In replying to certain objections raised by Pupin against the Van't Hoff theory of osmotic pressure (Abstr., 1888, 778), the author develops an equation for the behaviour of a dissolved substance which is similar to that of Van der Waals for the behaviour of gases. A special point of interest is, that account is taken of the presence and specific attraction of the solvent, and in this way an explanation of the mechanism of solution is obtained, which, it is claimed, is of wider application than that of Nernst (this vol., p. 3), in which this attraction is neglected. H. C. Sphere of Action of Molecular Forces. By B. GALITZINE (Zeit. physikal. Chem., 4, 417-426). By a process of theoretical reasoning similar to that already employed by Van der Waals, and using data given by Nadeschdin for several of the ethereal salts of the fatty acids in the critical condition, the author arrives at the conclusion that the sphere of action of the molecular forces is proportional to the masses of the attracting molecules. He also concludes that the attraction is inversely proportional to the square of the distance. H. C. Fluid Crystals. By O. LEHMANN (Zeit. physikal. Chem., 4, 462472).-Under the name of "fluid crystals," the author describes a cholesteryl benzoate first prepared by Reinitzer, which, although apparently melting at 145°, behaves between 145° and 178° towards polarised light as though still having crystalline structure. other respects the substance is in a perfectly liquid condition between H. C. these temperatures. In New Gas Burners. By M. GRÖGER (Zeit. ang. Chem., 1889, 329— 331). These are in general form similar to Bunsen burners, but instead of having any means of regulating the entry of air at the bottom of the mixing tube, the top of the burner is made conical, and there is a screw arrangement by which a solid cone can be raised within, so as partially to close the opening. By this means a flame of any character can be obtained, from a luminous one to one approaching that of a blowpipe, whilst the size of the flame can be greatly reduced without altering its character, and without risk of its flashing down. A burner on the same principle giving a flat flame is also described. M. J. S. Inorganic Chemistry. Hydrogen Peroxide. By G. TAMMANN (Zeit. physikal. Chem., 4, 441-443). The spontaneous decomposition of hydrogen peroxide in alkaline solution is found to be independent of the amount or nature of the base which is present. It appears probable from the author's experiments that it is really caused by the presence of traces of metallic oxides, such as the oxide of iron, dissolved in the alkali. It is shown that the addition of small quantities of such oxides increases enormously the rate of decomposition. The freezing points of aqueous solutions of hydrogen peroxide were determined, and from these a molecular reduction of 8.79 was found. Hydrogen peroxide being a non-electrolyte, this number H. C. would correspond with the formula H.O.. Hydrochlorides of Chlorides. By R. ENGEL (Bull. Soc. Chim. [3], 1, 695—699).-A review of the known hydrochlorides of chlorides, T. G. N. and a discussion of their probable constitution. Iodic Acid. By H. LESCŒUR (Bull. Soc. Chim. [3], 1, 563).—The crystals of iodic acid deposited from its solution in dilute or moderately concentrated nitric acid are monohydrated, whereas those obtained from the solution of iodic acid in concentrated nitric acid are anhydrous. The author thinks that the crystals deposited from solution in nitric acid of intermediate strength are mixtures of the hydrated and anhydrous varieties. T. G. N. Iodic Acid; Double Salts of Iodic Acid with other Acids. By C. W. BLOMSTRAND (J. pr. Chem. [2], 40, 305-340; compare Abstr., 1887, 327).-In the oxygen-acids of phosphorus and iodine only one atom of oxygen is strongly united to the phosphorus and iodine, the radicles being POO and IO-O respectively, thus differing from the oxygen-acids of nitrogen, chlorine, and bromine, where two atoms of oxygen are equally strongly united to the nitrogen, chlorine, or bromine, the radicles being NO2, CIO2, and BrO, respectively. 2 In support of the above statement, the author has prepared double salts of iodic acid with other acids which may be regarded as condensation-products, requiring for their formation an extra-radicle oxygen atom, analogous to that which is allowed to exist in aldehyde and to be the cause of the easy polymerisation of that substance; thus, C2HO + 0:С2H1 = C2H ̧:02:C2H4. The formula for iodic acid thus becomes HO⚫I0:0. Potassium sulphatoiodate was obtained by mixing potassium pyrosulphate (1 mol.) and iodate (1 mol.) in concentrated solution, and its formula found to be identical with that of Marignac's salt KO.IO(OH) O·SO2·OK. Potassium molybdoiodate, KO∙IO(OH)•O•MoO2·OH + H2O, is obtained as a white precipitate on adding a concentrated solution of potassium nitrate to a solution of sodium molybdate and iodic acid in nitric acid; it crystallises with difficulty in short needles, and is sparingly soluble in water. Ammonium molybdoiodate is obtained in the same way, and has similar properties, but contains no water of crystallisation. The thallium and lead salts were obtained. Molybdoiodic acid is obtained as a yellowish, transparent mass on evaporating the solution formed by the action of dilute sulphuric acid on a mixture of barium iodate and molybdate. Several of its reactions with inorganic and organic salts are given. Potassium tungstoiodate, KO WO2O·IO(OH)OK + H2O, is obtained by adding, by degrees, a solution of iodic acid to one of potassium tungstate; after some hours, a crystalline magma is obtained, more than 90 per cent. of which consists of slender needles of the tungstoiodate, the rest being tabular crystals of acid tungstate. Potassium chromoiodate has been described by Berg (Abstr., 1887, 776); the author's analyses of this salt leave some doubt both as to the amount of water it contains and as to its formula. The author has also obtained an ammonium triiodate, NH2O·IO(OH)·O·IO(OH)·O·IO2, of which the crystallography is given, and a sodium triiodate, NaO.IO(OH).0·10(OH)·Õ·10, + Н2O. A. G. B. Specific Gravity of Ammonia Solutions. By G. LUNGE and T. WIERNIK (Zeit. ang. Chem., 1889, 181-183).-The authors have redetermined with extreme care the sp. gr., referred to water at 15°, percentage of ammonia, and coefficient of expansion of ammonia solutions of 24 different strengths. The following is an abstract of their table: Specific gravity : at 15°. for±1°. at 15°. A Derivative of Boric and Phosphoric Acids. By G. MEYER (Ber., 22, 2919).—When a mixture of boric and phosphoric acids is heated to redness, a very inert, white substance, PÕB, is formed. It reddens moist litmus paper, but seems not to be dissolved by boiling water, and only to be very slowly attacked by boiling aqueous alkalis. Fusion with alkalis or alkaline carbonates causes instant decomposition, and fusion with sodium chloride also yields a soluble melt. L. T. T. Silicon. By E. P. HARRIS (Chem. Cent., 1889, ii, 283-284).—The author has successfully prepared silicon by means of Gatterman's method, ignition of fine sand with magnesium powder, and in addition to the already known halogen-derivatives, he has prepared a silicon nitride, NH, SiN, by acting on silicon chloride or silicon iodide with dry ammonia, whereby a considerable development of heat takes place. It is a snow-white powder. If the flux, obtained in the preparation of the silicon, be treated with dilute hydrochloric acid to dissolve out the magnesium oxide, silicon chloroform is obtained as a light, colourless, inflammable liquid, boiling at 42-44°. J. W. L. Preparation of the Chlorides of Silicon, Aluminium, &c. By H. N. WARREN (Chem. News, 60, 158).—Iron alloys of silicon or aluminium are heated to redness in a clay crucible and a current of chlorine gas is passed into the mass, suitable means being adopted to collect the volatile products. With chlorine and silicon-iron, the ferric chloride is condensed first, then the silicon chloride; if hydrogen chloride is used instead of chlorine, the ferrous chloride formed remains in the crucible and silicon chloroform distils off. The aluminium chloride obtained from aluminium-iron is purified by mixing with iron borings and distilling, or if the aluminium-iron alloy is mixed with common salt previous to submitting it to the action of chlorine, a sublimate of aluminium sodium chloride is obtained. D. A. L. Combining Energy of Rubidium. By N. BEKETOFF (Chem. Centr., 1889, ii, 245, from Bull. Acad., St. Pétersbourg [2], 1, 117–118). -Preparation of the metal.-Rubidium hydroxide is precipitated from |