A quantitative study of adsorption in solution and at interfaces of sugars, dextrin, starch, gum arabic and egg albumen, and the mechanism of their action as emulsifying agents: GEORGE L. CLARK and Wм. A. MANN. By the most accurate method known-the use of the stalagmometer for measurement of interfacial tension and the Morgan drop weight apparatus for measurement of surface tension of single liquids-the adsorption of sugar in solution has been determined to be negative, the adsorption increasing in ratio with increase in concentration of sugar, while the adsorption of sugar at the interface is positive and increases in the same ratio as in solution. Dextrin and starch, in very dilute solutions, are negatively adsorbed in solution but, in more concentrated solutions, are positively adsorbed while at the interface they are negatively adsorbed-starch more so than dextrin. Gum arabic behaves, in this respect, in the same manner as starch. A favorable surface tension, seemingly, is not a prerequisite for a good emulsion with any solution studied, although viscosity is an important factor. Better emulsions have been obtained with egg albumensaturated solutions than with any other solution under inspection.

The preparation, properties and molecular volume relationships of the ammines and hydrates of cobaltous fluoride, bromide, nitrate, carbonate and citrate: GEORGE L. CLARK and H. K. BUCKNER.

The molecular volume compression of the subsidiary valence groups, NH, and H2O, were studied in the same manner as in the previous work on the chloride and sulphate of cobalt (J. Am. Chem. Soc., 42, 2483 (1920)) in light of difference in volumes between anion and cation and possibilities of space cavities. It was confirmed that the larger the difference between anion and cation the greater the possibility of cavities, and the less the compression of NH, and H2O required the more stable the compound and the greater the possibility of holding a maximum number of groups. The various compounds prepared and studied are distinctly new with the exception of cobaltous bromide hexammine and cobaltous nitrate hexammine, the following being representative: cobaltous fluoride hexammine, cobaltous nitrate di-hydrate, ammines of cobaltous nitrate indicating more than six subsidiary valence groups and ammines of cobaltous carbonate and citrate.

Emulsification with soaps of linoleic and ricinoleic acids: GEORGE L. CLARK and H. K. BUCK

NER.

This work has involved the quantitative study by means of surface energy measurements for solutions and at the interface between aqueous solution and a pure oil by the drop weight method of the following points: the solubility in water of the free acids from the log. concentration-surface tension curve; the adsorption at the surface and the volume occupied by each molecule in the surface; the surface tensions of various concentrations of the soaps in solution, the effect upon the interfacial tension between water and pure toluene, the effect of the hydrolysis of the soap and the prevention of hydrolysis by the addition of various concentrations of NaOH to a given concentration of the soap; a test of the antagonistic effect of sodium and calcium soaps by testing the effect upon the interfacial tension between water and toluene containing a small amount of the free acid of adding various salts of sodium and calcium, resulting in the fact that the sodium salt promotes emulsions of oil in water and calcium salts promote emulsions of water in oil; and finally a thorough comparison of all the data so obtained with previous work on soaps of palmitic, stearic and oleic acids. All results of this purely quantitative study are necessarily numerical and can not be included in a short abstract.

Notes on the preparation of pure platinum: EDWARD WICHERS. The paper briefly states the method used in preparing platinum sponge free from other platinum metals and base metals, and describes in more detail the work that has thus far been done on the conversion of this sponge to contamination. compact metal with minimum

E.M.F. tests show the resulting platinum to be of higher purity than the best previously obtainable, i.e., the thermoelement wire of Heraeus, and show the difference in purity to be a matter of calcium content. Results of experiments with magnesia refractories are also given. There is a brief outline of further work to be done on this subject.

Modified method for the determination of iron and vanadium after reduction by hydrogen sulphide: G. E. F. LUNDELL and H. B. KNOWLES. Published methods for the determination of iron and vanadium after reduction by hydrogen sulphide ordinarily yield high values. A modified method which is sufficiently accurate for most purposes is presented, and a procedure for accurate analysis is outlined.

The free energy of dilution of hydrobromic acid; the activities of its ions in very dilute and con

centrated solutions: MILLER SPENCER and ALBERT G. LOOMIS.

The ultra-violet arc spectrum of yttrium: L. F. YNTEMA and B. S. HOPKINS. The ultra-violet are spectrum of Y was measured, using the yttrium oxide prepared by one of us for the determination of the atomic weight value accepted by the International Committee. The spectrograph used is an autocollimating quartz prism machine manufactured by Adam Hilger, of London. A current of 4 amperes at an E.M.F. of 220 volts was passed between vertical copper electrodes, the lower of which held the Y2O, in its crater. Several prominent lines, attributed to Y in the literature, were absent, and several new lines were measured. The results given are the mean of five determinations.

On the viscosity of gelatin sols: ROBERT HERMAN BOGUE. Experiments were carried out upon gelatin sols to accurately determine the relation between viscosity and concentration. The data have been applied to Hatschek's formula for the viscosity of emulsoids and the value A1/A, representing the volume occupied per unit weight of dispersed phase, was shown not to be constant with varying concentration, but to rise to a maximum and thereafter regularly decline. A tentative explanation is presented based upon the effect which increasing concentrations of dispersed phase will have upon the surface tension of the dispersion medium. An empirical expression defining the departure of the values obtained under varying conditions for Hatschek's constant is given.

The structure of molecules of water: IRVING LANGMUIR. Dennison has recently shown by Xray crystal analysis that ice consists of molecules of the formula H,O. In view of our knowledge of the structure of atoms it is not possible to account for the existence of this molecule on the basis of quadrivalent oxygen atoms. If pairs of electrons (duplets) constitute the valence bond, there can be no such bond between the two oxygen atoms. A structure for this molecule is therefore proposed in which the four hydrogen nuclei bind the two oxygen atoms. The duplet held by each hydrogen nucleus has one of its electrons in each of the oxygen octets, instead of the more usual arrangement in which both electrons of a duplet form part of the same octet.

The purification of helium by means of charcoal: L. FINKELSTEIN.

The importance of diffusion in organic electrochemistry: ROBERT E. WILSON and MERRILL A.

YOUTZ. The authors, in searching for definite evidence of depolarization in the electrolytic halogenation of organic compounds, found that either slow reaction rate or slow diffusion, or both, prevented any marked depolarization. Experiments on the oxidation of FeCl, in HCl gave surprisingly accurate information as to the rate of diffusion under a variety of conditions, and showed that without stirring there is a film of stationary liquid about 0.5 mm, thick through which diffusion must take place. This emphasizes the great importance of violent stirring and of the use of rapidly dif fusing inorganic carriers to extend the sphere of the organic reaction from a surface to a volume. Observations on the drying and swelling of gelatine gels: S. E. SHEPPARD and F. A. ELLIOTT. Attention is drawn to the importance of capillarity in the first phase of drying of jellies. It is shown that this, in conjunction with factors depending upon the shape of the jelly, causes the formation of an exo-skeleton tending to conserve or increase the original external surface extension. It is chiefly due to this, rather than to any internal supermolecular structure, that dried jellies on reswelling in water tend to return to their original concentration.

Note on the influence of silver salts in catalyz ing the decomposition of ammonium persulphate solutions: S. E. SHEPPARD and A. BALLARD. The influence of silver salts in facilitating the decomposition of ammonium persulphate in solution, first observed by H. Marshall and J. Inglis, has been confirmed, and quantitative data on the rate of change in relation to the silver content obtained.

Further developments of the hydrogen elec trode: FELIX A. ELLIOTT. Two new forms of hydrogen electrode were described representing probably the limits of simplicity and ruggedness without reducing accuracy and rapidity of operation. Especial attention has been given in designing this new apparatus to reduce the internal resistance to the lowest possible value so that a less sensitive and hence cheaper, simpler and portable electrical measuring instrument might be used. Such an instrument was described, working on the potentiometer principle and employing a pivot type movement. Examples of results obtained with the two types of apparatus indicate that potentials are reproducible to about 0.1 mv.

Note on silver soap gels: G. STAFFORD WHITBY. It has been observed that the silver salts of the fatty acids are capable of giving reversible gels in a variety of organic liquids-particularly in the

The

homologs of benzene and in halogenated derivatives of benzene and its homologs. The silver salts form gels at a lower point in the series of saturated fatty acids than do the alkali-metal salts; gels being obtained with silver caproate. A number of regularities were discernible. silver salts of the higher members of the saturated fatty acid series show a greater solvation capacity than those of the lower ones. None of the saturated fatty acid silver salts gave a gel in benzene; but silver oleate did. Speaking generally, in the case of solvents of the same general chemical character, the higher the boiling point of the solvent, the greater appeared to be the solvation capacity of a given salt, and the smaller the extent to which the gel from a given salt suffered syneresis.

Catalytic effect in the reaction between ketones and halogens in aqueous solutions: F. O. RICE. The velocity constant of the reaction between acetone and bromine is independent of the bromine concentration and Lapworth (J. C. S. Trans., 1904, p. 30) explained this by saying that the acetone slowly enolized and the addition of bromine and splitting off of hydrobromic acid were practically instantaneous. This is probably incorrect since higher ketones have the same velocity constant as acetone, and an explanation based on the radiation theory was offered. The reaction is accelerated by neutral salts contrary to Lapworth's statement.

The transference numbers of sulfuric acid by the concentration cell method: A. L. FERGUSON and W. G. FRANCE. A cell combination was used which permitted the measurement of all the required potentials from one set-up. The value obtained for the transference number for the anion in concentrations between M/10 and M/100 at 25° C. was .1868.0007. The method was shown to be highly reliable. In all of the calculations it was assumed that sulfuric acid dissociates into two hydrogen and one sulfate ion. This assumption was substantiated by the results obtained. A formula for boundary potential was developed in which boundary potential may be obtained from potential measurements alone.

The influence of gelatin on the transference numbers of sulfuric acid: ALFRED L. FERGUSON and W. G. FRANCE. The transference numbers were determined by the concentration cell method. Solutions of sulfuric acid containing from 0.5 per cent. to 20 per cent. gelatin were used. The transference numbers of the anion increased from .187 for pure acid to .685 for acid containing 20 per

cent. gelatin. In the 20 per cent. gelatin solution the boundary potential of both concentration cells became zero. The decrease in conductivity was approximately proportional to the gelatin added. The results are probably best explained on the assumption that there is a chemical action between the gelatin and sulfuric acid in which a single compound is formed. In this compound the hydrogen of the acid loses its identity and when the compound dissociates there is formed a complex gelatin hydrogen positive ion and a negative sulfate ion.

The entropy of monatomic gases: GILBERT N. LEWIS.

The electrometric titration of uranium with potassium dichromate and potassium permanganate: D. T. EWING and E. F. ELDRIDGE.

The heat of coagulation of ferric oxide hydrosol by electrolytes: FREDERICK L. BROWN and J. H. MATHEWS.

Some quantitative experiments on coagulation of colloids: RAY V. MURPHY and J. H. MATHEWS. The lowest concentration (limiting concentration) of electrolytes necessary to coagulate hydrous ferric oxide sol has been studied as a function of the purity of the sol (ratio of gram-equivalents of Fe to gram-equivalents of Cl) and of the concentration of FeO. Chloride, chromate and ferricyanide ions were used in the form of the potassium salts. The conclusions drawn are: (1) The limiting concentration decreases with increasing purity in the case of all three ions, the mechanism of the process being evidently similar for the three ions; (2) The limiting concentration decreases markedly with decreasing concentration of the sol in the case of all three ions, but the relation indicated by Burton and Bishop, Jour. Phys. Chem., 24, 701 (1920), for mastic, AsS and Cu sols, holds for FeO, hydrosol only in the case of the trivalent ion.

The alkalinity of Searles Lake brine: ROGER C. WELLS. The title may mean either the titration alkalinity or the hydrogen ion concentrations. The latter may be considered as determined by certain proportions of the four buffer substances NaCO3, NaHCOs, Na,BO1, and Na„B2O. The writer has found for the brine PH 9.48. By determining the PH values of artificial brines containing each pair of buffers separately it is possible to draw curves from which by interpolation the proper proportions of the buffers to yield PH 9.48 may be read. This method serves as a check of the analyti

=

cal determinations and yields a logical conventional form for expressing the whole analysis of the brine. The vapor density of technical phosgene: A. F. O. GERMANN and VERNON JERSEY. A 75 gm. sample of technical phosgene was distilled from a large cylinder of the liquefied gas, obtained from the Chemical Warfare Service through Dr. Goss. Pure phosgene has a vapor tension at 0° C. of about 552 mm.; the sample taken showed a vapor tension in excess of one atmosphere at - 80° C., that is, in a bath of solid carbon dioxide and acetone. The sample was subjected to fractional distillation in a vacuum until the vapor tension at 0° C. was approximately correct. Following this, three fractional distillations alternated with three determinations of the vapor density of the gas, until five groups of three determinations each had been obtained for the vapor density, and the nearly pure gas had been fractionally distilled 15 times. Average of 1st group of densities.....4.4708

66

2d 66 "3d 66

4th "6 5th ""

66

...

.4.5060 4.5216 4.5244 .4.5263

The values given are in grams per liter at standard conditions, uncorrected for the compressibility of the gas and for the contraction of the globes when evacuated. Technical phosgene is very impure. The principal impurities are very volatile and of relatively low molecular weight, probably carbon dioxide and hydrogen chloride resulting from hydrolysis. Repeated fractional distillation yields a product whose density tends toward a maximum value, probably the value for pure phosgene.

The cryoscopy of boron trifluoride solutions: V. Systems with methyl ether and with methyl chlo ride: A. F. O. GERMANN and MARION CLEAVELAND. Gasselin prepared the compound BF,. (CH,)2O by mixing the gases; he obtained a liquid boiling at 126° C. This compound has been prepared from liquid boron trifluoride and liquid methyl ether in the course of the determination of the melting point curve of solutions of the two substances. The curve shows a eutectic at 3 per cent. BF, (molecular percentages) and a maximum at 50 per cent. The vapor tensions of solutions containing from 60 per cent. to 90 per cent. BF, were so high, that the form of apparatus used was inadequate. The form of the curve so far as determined seems to indicate the existence of a second compound in this interval. Methyl chloride was prepared from salt, sulfuric acid and wood alcohol; was purified by repeated fractional distillations;

and gave the usual indications of a pure substance, such as constant freezing point for the first and last fractions of the liquid. The melting point curve with BF, shows two maxima, at 15 per cent. BF, and at 33 per cent. BF, and an angular point in the curve at 50 per cent. The form of the maxima at 15 per cent. is identical with that at 50 per cent. for methyl ether; the mixture yields about 15 per cent. of the total volume of a liquid, whose freezing and boiling points are identical with those of BF,. (CH,),O. The logical inference is that methyl chloride as prepared contains methyl ether as an impurity, and that this impurity yields a constant boiling mixture. It would seem that boron trifluoride might be used as a test for the presence of methyl ether in methyl chloride; the same test might be extended to the homologues. The interpretation of the other maxima mentioned must await the completion of further work on the purity of methyl chloride.

The cryoscopy of phosgene solutions: I. System with chlorine: A. F. O. GERMANN and V. JERSEY. The melting point curve of solutions of phosgene and chlorine was determined. The curve is very complex. There is a eutectic at 25 per cent. (molecular percentages) chlorine; and angular points in the curve at 6 per cent., 11 per cent., 50 per cent., 63 per cent., 75 per cent. and 91 per cent. chlorine. The following compounds, which dissociate at the melting point, are indicated: 16COC12. Cl2; 8COC12. Cl2; COC).Ch; 3COC1,.5C1,; COC1,.3C; and COC12.10C12. The second of these, chlorine octaphosgenate, is particularly interesting, as corresponding with the octahydrate of chlorine. The affinity of chlorine for phosgene, and the instability of the compounds would seem to offer an explanation of the mechanism of certain phases of the catalysis of carbon monoxide and chlorine by means of charcoal saturated with chlorine. If we assume a layer of chlorine molecules on the surface of the charcoal, the high concentration of chlorine thus obtained might be supposed in the presence of carbon monoxide to shift the equilibrium in the direction of one of the chlorine-phosgene complexes, which promptly decomposes because of its instability, leaving the surface film of chlorine for further action. It would be desirable to determine the melting point curve of carbon monoxide and chlorine solutions, to clear up some of the points involved in this reaction.

CHARLES L. PARSONS,

Secretary