ii. 42 ABSTRACTS OF CHEMICAL PAPERS. Inorganic Chemistry. Revision of the Atomic Weight of Bromine. Normal Density of Hydrogen Bromide Gas. CLARENCE KENWORTHY REIMAN ( J . Chim. phys. 1917 15 293-333).-A full account of work already published (compare A. 1917 ii 137 ZOO). El. M. D. Revision of the Atomic Weight of Bromine. Normal Density of Hydrogen Bromide Gas. WALLACE JENNINGS MURRAY ( J . Chim. phys. 1917 15 334-359).-A full account of work previously published (A. 1917 ii 201). Chemical Reactions in the Corona. I. Ozone Formation. F. 0. ANDEREGG ( J . Amer. Chem. SOC. 1917 39 2581-2595).- An account is given of some preliminary observations which have been made on the formation of ozone during the passage of a current of oxygen through a tube in which corona discharge is maintained.The results suggest that a condition of equilibrium is rapidly attained. It is hoped that the study of this and other reactions in the corona discharge will throw some light on the connexion betweeii ionisat'ion and chemical reaction. H. M. D. H. &I. D. Analytical Control of the Ammonia Oxidation Process. G. B. TAYLOR and J. D. DAVIS ( J . Ind. Eiig. CJiem. 1917 9 1106-1110).-1n the catalytic oxidation of ammonia by means of atmospheric oxygen for the production of nitric acid the following reactions t'ake place (1) 4NH3+ 50,=4NO +.6H,O . . . . (2) 4NH3 + 30 = ZN + 6H,O * . . and the possibility of the second reaction being produced by means of an intlermediate reaction 4NH,+6NO=5N,+6H20 . . . . . (3) was pointed o u t by Reinders and Cats (A.1912 ii 248). Evi- dence that tliis third reaction does take place under certain conditions has been obtained in the case of a badly con- st'ructed oxidiser in which irregular cooling of the catalyst was produced locally by currents of the burned gases. Samples taken from points near the cool places contained ammonia whereas in samples taken outeide the catalyst chamber the proportion of ammonia was low. It was probable that most of this ammonia passing such points was sulssequently ' burned " by the hot nitric. oxide. Effective working in a converter depends on maintaining conditions which promote reaction (1) a i d are unfavourdde t o reaction (2). [See further J . Soc. Chem. Ind. 1918 37 54~.1 C. A. M.INORGANIC CHEMISTRY. ii. 43 The Solubility of Silica.VICTOR LENHER and HENRY BALDWIN MERRILL ( J . Amer. Chem. SOC. 191'7 39 2630-2638).-The solu- bility of silica in wat,er and in aqueous solutions of hydrochloric and sulphuric acids of varying concentration has been measured at 35O and 90°. With gelatinous silica a condition of equilibrium is reached after a few hours or days but with ignited silica much longer periods are required for the attainment of saturation. The results seem to show that the solubilit'y of gelatinous silica is in- dependent of the method of preparation of the substance and that ignited silica will give solutions of the same concentration when the equilibrium condition is reached. The solubility increases with the temperature and a t a given temperature decreases with iiicrease in the concentration of the acid.Carbon dioxide has no measurable influence on the solubility of silica. The results are not only of interest in connexion with the estimation of silica but possess considerable geological significance. H. ng. D. Polysulphides of the Alkali Metals. 111. Solidifying Points of the Systems Sodium Monosulphide-Sulphur and Potassium Monosulphide-Sulphur . JOHN SMEATH THOMAB and ALEXANDER RULE (T. 1917 111 1063-1085. Compare T. 1914 105 177 2819).-The freezing-point curves for mixtures of sulphur with sodium monosulphide and potassium monosulphide derived from observations on the rate of cooling of the mixtures in a current of dry nitrogen show the existence of a complete series of compounds of the formula R,S where x is a whole number having the maximum value 5 in the sodium series and 6 in the potassium series.The members of the potassium series of compounds resemble closely the corresponding rubidium and czsium compounds (Biltz and Wilke-Dijrfurt A. 1905 ii 162; 1906 ii 283 611) but the sodium compounds differ from their analogues in appearance and properties. This difference is also shown in the comparative stabilities of compounds of corresponding composition and in the maximum combining power of the metals. The rate at which the polysulphides lose sulphur when heated in a steady stream of hydrogen at regularly increasing temperatures has also been examined. The results obtained indicate that the disulphides of both metals are very stable compounds from which sulphur can only be removed with difficulty a t 70C)-8OO0.This suggests that the polyysulphide molecules contain two atoms of sulphur which are combined differently from the remaining atoms of sulDhur. The disulphides are represented by the formula R-R*S*R and the higher p0lysii1phides by formnlp such as s s R.0 S* S R and .... . S R-S-S-R Molecular weight determinations by the boiling-point method in ethyl alcohol solution gave numbers less than those required by theformula &S,. These results favoui. t lie simple forinul,~ K,SZ as opposed to the doubled formula R,S suggested by Rloxani (T. 1900 77 753). The difference between the actual results and those required by the formula R,S are attributed t o ionisation. No evidence of intermediate compounds such as the ennea- sulphides (Bloxam loc.c i t . ) has been obtained in either series. H. M. D. Ammoniacal Copper-Mercury Derivatives. %. ANDERLINI (Gazsettn 1917 47 ii 171-176. Compare A. 1912 ii 764).- The conzpoz/nd CuHgBr,,4NH3 is obtained in dark blue crystals by mixing boiling solutions of ammoniacal cupric bromide and mercury bromide dissolved in potassium bromide. If this com- pound is dissolved in a warm solution of ammonium bromide and the solution treated with mercury bromide dissolved in potassium bromide free ammonia being present but not in excess the com- pound CuHg,Br,,4NH3 is obtained in bright blue scales or needles. R. V. S. Observations on the Rare Earths. IV. The Purification of Gadolinium. LOUIS JORDAK and B. S m r H HOPKISS (J. 14?ne1*. Chem. Soc. 1917 39 2614-2623).-The results of experiments are described in which the attempt has been made to separate gadolinium from rare earth mixtures by the fractional precipitation of the dimethyl phosphates bromates and glycollates.Fractionation of the dimethyl phosphates rapidly removes europium and samarium completely and gadolinium of considerable purity is obtained from the middle fractions. The terbium which is present in this material can be removed by fractionation of the bromates. The glycollate method rapidly removes samarium from gadolinium but has little effect' on the removal of terbium. The method recommended for the separation of gadolinium from rare earth mixtures is first to remove ceriuni by the bromate method and fractionally crystallise the residue. The less soluble fractions contain only europium samarium gadoIinium and neodymium which are converted into the dimethyl phosphates and again frac- tionated. The least sol~~ble fractions will then consist of high-grade gadolinium.Any remaining traces of samarium can be removed by precipitation with sodium glycollate. H. M. D. . Revision of the Atomic Weight of Samarium. Analysis of Samarium Chloride. 0. J. STEWART ant1 C. J a m s ( J . Anzey. Chem. Soc. 1917 39 2605-2614).-A fractionation method for the extraction of samarium from its admixtures with the other rare earth metals is described. The pure samarium oxide was Converted into the chloride SaCl,,GH,O which was dried finely powdered. and heated below looo until most of the water of crystallisation had been driven off. The temperature was the11 raised to 180° aiicl finally to 300° i h c tuhe lwing traversed by a ciirrent of dryliydrogen chloride.The anhydrous salt was then removed finely powdered and heated t o its melting point in contact with hydrogeii chloride which was then displaced by nitrogen and finally by air. The atomic weight of the saniarium was derived from the ratio SaC13:3Ag. A series of preliminary experiments in which the individual results were not in good agreement gave a mean value Sa=-150*45 and a final series of eight determinations in close agree- ment gave a mean value Sa=150*43. €3. 31. D. Precipitation Stability and Constitution of Hydrated Ferric Oxide Sols. I. MARKS NEIDLE ( J . 1 4 ? ? ~ ~ r . Che??~. S O C . 1917 39 2334-2350).-A systematic study of the precipitatioll of hydrated ferric oxide sols of varying purity and concentration by means of potassium sulphate.The sols were prepared by oxidising a solution of ferrous chloride containing 1 gram equiva- lent of ferrous chloride in 400 C.C. of solution by means of 3% hydrogen peroxide. The solutions were then dialysed and diluted t 9 the required concentrations. The content of iron and chlorine was estimated and the relative stability determined by measuring the volume of a standard solution ( M / S O O or &!/1600) of potassium sulphate required t o effect complete precipitation of the sol. A number of series of such clear sols were prepared and in each series the iron content was constant whilst the chlorine content varied. The results show that for a given iron concentration the stability increases with the chlorine concentration whilst for sols of given purity t h a t is for sols which have t h e same ratio Fe/Cl the stability decreases as the concentration increases this being most pronounced in very piire sols.It is shown that the maximuifi purity of a sol may be ascertained by graphic extrapolation. A general discussion is given of the results of Nicolardot and Duclaiiu (A. 1905 ii 167; 1906 ii 677). The author. advances :I cherrlic.sl theorv of the1 constitution of hydrated ferric oxide sols t o e>cF’liti\l thc relatioiiship between stability and composition J. F. S. A New Oxychloride of Tin. HARRY F. KELLER ( J . L4??zf~*. Chrnt. Soc. 1917. 39 2354-2356).-During the examination of metallic objects found in the aboriginal cemetery on Hogtown Bayou Santa Rosa County Florida the author noticed a number of cavities in a piece of tin which were lined with small shining crystals.The crystals were mostly in the form of thin plates but some were acicular. On heatinc the substniice melts and gives off acrid fumes which form a white sublimate without a trace of moisture. Analysis of a small qtiantity of the material leads to r2 comuosition corresnonding with the formula SnC12,Sn0. rSee also .7. ,COP. f ‘ l r e n i . It?(?. 1918 37 2 9 ~ . 1 J. F. R. Solubility of Bismuth Qxychloride in Hydrochloric Acid and its Relation to Complex Formation. ARTrIuR A. XOYEP FRANK Tv. R A m aiid JAMES A. BEATTIE ( J . Ll???er. (-‘/ic?)?. 8ov. 191 7. 39 3526- 2532).-The elertrical condnctivitly of a soliitlionii. 46 ABSTRACTS OF CHEMICAL PAPERS. of hydrochloric acid is diminished by the addition of bismutb chloride. The observed diminution can be most satisfactorily inter- preted on the assumption that chlorobismuthic acid of the formula H,BiCl is formed. Measurements of the solubility of bismuth oxychloride in hydro- chloric acid solutions of varying concentration a t 2 5 O afford sup- port for this hypothesis in the case of the more concentrated solu- tions whilst for the less concentrated acid solutions the solubility data suggest that the predominant complex acid has the formula HBiCl,. H. M. D.