2 70 ABSTRACTS OF CHEMICAL PAPERS. I n o r g a n i c C h e m i s t r y. Preparation of Pure Hydrogen Peroxide Solution. By L. CR~SMER (BUZZ. Soc. Chim. [3], 6, 24--25).-The solution of hydrogen peroxide which results from the action of hydrochloric acid (sp. gr. 1.1) on barium dioxide is extracted by shaking with ether, and the ethereal solution is agitated with distilled water, to which it yields the dissolved hydrogen peroxide. By repetitions of this process, a pure, neutral solution corresponding with 0.8-0.9 per cent of hydro- gen peroxide is obtainable, from which the dissolved ether may be eliminated by distillation under reduced pressure. Preparation of Hydrobromic Acid. By G. S. NEWTH (Chena. News, 64, 215).-By means of the following arrangement a large quantity of bromine can be rapidly converted into hydrobromic acid :- A glass tube, 7 inches long and 8 inch in diameter, is fitted at each end with a cork carrying a piece of small tubing and a piece of fitout wire.The ends of these pieces of stout wire, within the longer tube, T. G. N.INORQANlO OHEMISTRY. 271 are joined hy a spiral of platinum wire 1 inch long, and after expelling the air the spiral is heated to bright redness by Rn e1ecOric current; a stream of hydrogen, impregnated with bromine by bubbling through that liquid, which may be heated at GO", is passed through the longer tube, and, as long as a Blight excess of hydrogen is maintained, hydro- bromic acid quite free from bromine issues from the other end, and is collected in water. There is very little danger of explosion, but t o render it impossible, the small supply tube may be plugged with a little glass -vr7001.Solubility of Gases in Water. By I;. W. WINKLER (Ber., 24, 3602-3610; compare Abstr., 1891, 384).-In this paper the author gives in tabular form the results of experiments on the solubility of nitrogen and of oxygen in water, a t temperatures ranging from 0" to 80"; the calculated values for the solubility of the two gases at temperatures ranging from 80--100" are also given. As regards the solubility of nitrogen, the author's values are considerably larger than those given by Bunsen. The Densities of Sulphuric Acid Solutions. By S. U. PICKER- ING (Chem. News, 64,31l).-Lunge's doubts as t o the accuracy of the author's density determinations are, in the opinion of the latter, due to a misapprehension as to the method employed to determine the strength of the acid, and as to the experimental error involved.D. A. L. E'. S. K. s. u. P. The Contraction on Mixing Sulphuric Acid and Water. By S. U. PICKERING (Chem. Netus, 64, 14--15).-From his own results the author has calculated the strength a t which the maximum contraction occurs, and finds that this maximum, when calculated for unit weight of solution, shifts from 67 per cent. at 8" to 70.1 per cent. at 38", whereas the maximum contraction calculated per unit volume remains practically constant at 76 per cent. throughout this range of tempera- ture. Neither of these maxima occurs at the composition of the di- hydrate (73.1 per cent.) or of that of any other hydrate of which indications have been obtained.The rate atr which the amount of contraction is influenced by the temperature varies irregularly with the actual value of the temperatnre ; tlhus with solutions from 64 to 80 per cent. st,rength, the contraction diminishes at nearly the same rate for the intervals 8" to 18", and 28" to 38", whereas for the inter- mediate interval, 18" t o 28", the rate of diminution is about 25 per cent. smaller. s. u. P. Density and Composition of Dilute Sulphuric Acid. By A. W. R ~ C K E R (Phil. Mug. [ 5],32,304-313) .-Pickering (Trans., 1890, 64) deduced the existence of various hydrates in sulphuric acid solu- tions from sudden changes of curvature in the curves representing the variation of the density, &c., of such solutions with the composition. The author considers the " first diflerential " curve for 18" between 46 per cent.and 80 per cent., in which there are supposed to be f o u r breaks, and shows that it is possible to find an equation, and, therefore, a continuous curve, which will represent the results within the limits of the experimental error. The equation is of the form u 2272 ABSTRACTS OF OHEMIOAL PAPERS. and contains fieven ai-bitrary constants. The author thus doubts the value of '' differentiation " or of the bent ruler, as used by Pickering, in discovering changes of curvature in curves drawn to represent experimental numbers. J. W. The Densities of Sulphuric Acid Solutions. By S . U. PICKERING (Phil. Xag. [ 5 ] , 33, 132).-The curve used by RiicBer (see preceding abstract) to bridge over four of what the writer con- sidered to be breaks in the figure foymed by the first differentials of the densities of the acid can, he maintains, prove nothing about one of these breaks, and very little about another, for it extends too short a, distance beyond them.The formula for the curve suggested by Rucker consists of a combination of an exponential curve and a straight line, on to part of which (as this curve did not agree with the results) a hump was engrafted by means of a complex fourth term, tlhe whole forming a curve for which, as an expression of physical facts, there would seem to be no precedent. The points, moreover, at which the term expressing the hump begins to be appreciable and again becomes inappreciable correspond exactly with the positions assigned by the author to two of t,he breaks, thus confirming, rather than disproving, the existence of these as points where some new conditions in the solutions become sensible.The only break which Bucker's curve does successfully bridge over is that particular one which the author pointed out to be especially doubtful: and even in favour of this break the evidence is not thereby entirely negatived. The author maintains that even the most successful attempt to offer an alternative explanation of a small portion of some of his results could not upset his conclusions which were based entirely on the cumulative evidence derived from many sources, and of which he here gives a summary. s. u. P. Boron Phosphides. By H. MOISSAN (Compt.rend., 113, 726- ~29).--Bo~on phosphide, PB, is obt,ained by reducing the phospha- iodide in hydrogen at 450-500". The product is powdered, and again heated in hydrogen at the same temperature in order to remove excess of iodine, and the process is repeated two or three times if necessary, care being taken that the temperature does not exceed 500". I t is a very light, amorphous, maroon-coloured powdel., insoluble in the chlorides of arsenic, phosphorus, carbon, and anti- mony, and in all solvents, organic and ino~ganic, that were tried. It is n o t volatile in a, vacuum at 500". At 200", in presence of oxygen, it burns and yields boric and phosphoric anhydrides ; when thrown into fused alkaline nitrates, there is incandescence and deflagration ; fused sulphur has no action, hut sulphur vapour converts it into boi-on and phosphorus sulphides.Chlorine converts boron phosphide into boron trichloride and phosphorus pentachloride, with incand- escence ; bromine has no action in the cold, but combination takes place if the temperature is raised. Vapour of iodine, arsenic, orINORQANIC CHEMISTRY. 273 phosphorus has no action at a dull-red heat. When heated at 5OO"in a current of nitrogen, boron phosphide yields no nitride, and, although a,t higher temperatures it loses phosphorus (as it does in a vacuum), no nitride is formed. When a mixture of boron phosphide and sodium is gently heated in a current of hydrogen, it rapidly becomes incandescent, with formation of sodium phosphide and boride. Potas- sium yields the same products at a lower temperature.A mixture of the boron phosphide with powdered magnesium becomes incand- escent at about 500", but aluminium has no action, except at a much higher temperature. Finely-divided silver, copper, and platinum react with the phosphide when gently heated, biit mercury has no action at its boiling point. When boron phosphide is thrown into the strongest nitric acid, it takes fire, even in the vapour, and burns brilliantly on the surface of the liquid; on slightly heating, it dis- solves immediately and completely. Concentrated solutions of hydro- chloric and hydriodic acids have no action on the phosphide, and sulphuric acid has no action in the cold, but is reduced on heating. Concentrated hot solutions of potash and soda dissolve i t slowly, whilst fused potash dissolves it completely with formation of hydrogen phosphide and potassium borate.Gaseous hydrogen fluoride attacks it below dull redness, with formation of boron fluoride, hydrogen, and phosphorus. Gaseous hydrogen chloride behaves similarly at R higher temperature. Boiling water has no action on the phosphide, but water vapour decomposes it at 400" with production o€ boric acid and hydrogen phosphide. Hydrogen sulphide at A dull-red heat yields boron sulphide and hydrogen phospliicle. I n ammonia at about 300" the phosphide burns, with formation of boron nitride and libera- tion of phosphorus. Boron phosphide, B5P3, is obtained by heating the preceding com- pound at 1000" in a current of hydrogen. It has a paler colour than the phosphide PB, does not inflame in chlorine or nitric acid, and is not attacked by the latter, even when boiling.It is insoluble in all inorganic and organic solvents, burns with some difficulty in oxygen, and is attacked by fused nitrates with incandescence, but only with difficulty by metals and non-metals. It does not burn in chlorine below a dull-red heat. I f the hydrogen used for reduction contains water or oxygen, a white phosphoboric acid, also known as boron phosphate, is formed, and the same compound is obtained by the action of nitrogen oxides on a mixture of the phosphides. Boron Phosphide. By A. BESSON (Compt. Tend., 113, 772-7731, -The author draws attention to his previous note on the subject (Abstr., 1891, 1418), and describes 'an additional property of the phosphide.It is oxidised by dilute nitric acid to a substance, prob- ably phosphoboric acid, which is left in nacreous plates on evaporating the solution to dryness, is soluble in water, and gives a white, gelatin- ous precipitate with excess of ammonia. Moissan (ibid., 787-788), commenting on this and the previous note, draws attention to the absence of numerical data justifying the formula, BP attributed to the phosphide, and claims priority in the C. H. B.274 ABSTRACTS OF CEEMTOAL PAPERS. Rystematic study of boron phosphides (Compt. rend.! 112, 71 7, and 113, 19). JN. w. The Influence of Steam and other Gases on the Combustion of Carbonic Oxide and Oxygen. By N. BBK~TOFF (Ckenz. Centr., 1831, ii,449-450 ; from Bull. Acad. St. PrStersbourg [2], 2, 175-179). -Having repeated Dixon's experiments (Trans., 1885,94), the author has obtained the same results, namely, that, a mixture of carbonic oxide and oxygen, when dried by nieans of phosphoric anhydride, is not exploded by the passage of an electric spark, and further, that if the gases be dried by sulphuric acid, which appears to leave a trace of moisture in the gas, the combustion proceeds so slowly that it may be followed with the eye.The presence of other gases, sulphurous anhydride, o r nitrous oxide, had not a similar effect. Cyanogen, on the other hand, when present to the extent of 10 per cent., exerted an influence similar to that of steam, and caused an immediate explosion on the passage of the spark. BBkbtoff suggests that the action of cyanogen may be explained on the assumption that the beak liberated in the decomposition of the cyaiiogen is added to the heat of conibustion of the carbon, and he further assumes that the dissociation tempern- ture of the water niolecule being lower than that of the oxygen molecule is the proper explanation of the fact that the presence of water in a mixture of carbonic oxide and oxygen assists the com- bustion of the latter gases.J. W. L. Reactions of Carbonic Anhydride at High Pressures. By A. D'ARSONVAL (Conzpt. rend. SOC. Biol., 1891, 320--321).--Liquefied carb- onic anhydride is a powerful antiseptic. It does not coagulate albu- min. At high pressures it can displace both organic and mineral acids. When urine is subjected to a pressure of 40 atmospheres of carbonic, anhydride, crystals of uric acid are deposited.A dilute solu- tion of potassium silicate similarly treated becomes solidified from the deposition of silicic acid. Carbonic anhydride at the same pressure is also able t o liberate hydriodic and hydrobromic acids from solu- tions of potassium iodide and bromide respectively. By N. BEKBTOFF (Chenz. Centr., 1891, ii, 450-451 ; from Bull. Acad. St. Pe'tersbourg [2], 2, 169--17G).-The author has prepared considerable quantities of cssium by redncing the hydroxide with aluminium. The reaction proceeds well, and is as readily carried out as the reduction of rubidium. The author refers t o the results, which Winkler has recently published, on the heat of combination of the alkali metals, and points out that the deductions which Winkler has made are diametrically opposed to his own.The author finds that the combining heat of the alkali metals varies inversely with their atomic weights. He further urges that Winkler has not employed the oxides, but the hydroxides and carbonate5, for his determinations, which are not suitable for the purpose. Properties of Czesium and its Hydroxide. By N. BEKETOFF (Chem. Centr., 1891, ii, 451 ; from Bull. Acad. St. Pe'teysbourg [SJ, 2, W. D. H. Reduction of CEsium. J. W. L.INORGANIC CHEMISTRY. 275 171--173).--The author has used a very pure specimen of caesium sulphate as material on which t o work. A determination of the sulphnric acid gave almost the theoretical amount, and by means of the spectroscope only a trace of rubidium could be detected.The hydroxide was obtained from the sulphate by precipitation with barium hydroxide, and concentration of the filtrate in a silver dish, which was placed in a metal retort. In this manner carbonic an- hydride was excluded. After concentration in the retort, the solution was transferred to a small silver dish, the remainder of the water evaporated, and the hydroxide finally fused. The silver dish was slightly attacked, and the silver oxide dissolved in the fused caesium hydroxide ; it was again precipitated, however, as the latter cooled. The quantity of silver oxide was inconsiderable. Cmium hydroxide has a sp. gr. 4.0178, compared with water at 4" ; the molecular volume = 57.3 (sodium hydroxide = 18 ; potassium hydroxide = 27 ; rubidium hydroxide = 32 ?).The molecular volume of the hydroxide corresponds with a very considerable diminution of volume during its formation from the elements. The heat of solution of the hydroxide in water is 15,876, which is higher than f o r all other alkali hydroxides. The heat of neutralisation of the hydroxide with hydrogen chloride (dilute) is 13,790, or nearly the same as that found by Thomsen for potassium hydroxide, or, indeed, for the other alkali hydroxides. Metallic cEsium was obtained by heating 114 grams of the hydroxide in a nickel retort wit,h 27 grams of aluminium, and collecting the distilled metal in glass receivers; 20-25 grams of the metal were thus obtained. The heat of combination of cmium with water was found to be from 50 to 53 Cal. J. W. L. Action of Sodammonium and Potassammonium on Metals.By JOAXNIS (Cornpi!. rend., 113, 795-798).- Sodammonium and potassammonium are decomposed by mercury, lead, and antimony, but are not affected by aluminium, silver, zinc, or copper. When a solution of sodammonium in liquefied ammonia is allowed to fall drop by drop on mercury, it is decomposed with the formation of a crystalline amalgam, Hg,Na, which can be washed free from excess of the reagent with liquefied ammonia. Potassammonium under similar conditions yields an amalgam of the composition Hgl8K. When pure lead is brought into contact with sodammonium, the reddish-brown liquid turns blue, and then green, and a little hydrogen is disengaged, owing to the spontaneous decomposition of the sod- ammonium. Finally, however, the metal is partially converted into an indigo-blue mass, which dissolves in liquefied ammonia to a bottle- green solution, and has the composition Pb4Na,2NH,.This substance decomposes at ordinary pressures, leaving a grey mass resembling spongy platinum, and oxidises rapidly on exposure to air, with evolu- tion of heat. When thrown into water, the first portion dissolves completely, owing to the combination of the lead with the dissolved oxygen, and the subsequent solution of the oxide in the sodium hydroxide simultaneously formed, but as soon as the oxygen is used up, metallic lead is thrown down as a black, curdy precipitate. JN. W.276 ABSTRACTS OF OHEMICAL PAPERS. Influence of Ammonia on the Solubility of Ammonium Chloride. By R. EXGEL (BUZZ. Xoc:Chim.[3], 6, 17).-The pre- sence of ammonia diminishes the solubility of ammonium chloride in water at 0' at first, but as the quantity of ammonia is increased, a corresponding increase in the chloride dissolved obtains. This result is probably due to the formation of compounds of ammonium chloride and hydroxide. Tables of values are given. T. G. N. Precipitation of Copper by Iron and the Action of Iron on Ferric Solutions. By J. C. Essmit (BdZ. SOC. Chin%. [ 3 ] , 6, 147- 148) .-In the wet method of copper extraction, very mixed qualities of acrap iron are employed. The reduced copper occurs in powder, grains, and filaments which cannot be readily washed free from the mud of ferric hydroxide formed. The author finds that the structure of the iron used exerts a marked influence on that of the copper obtained.By selecting the iron to be used, it is possible to obtaiii the reduced copper in a fibrous or granular condition, when i t admits of being readily washed free from ferric hydroxide. The occurrence of the mud of ferric hydroxide is due to the form- ation of a basic ferric sulphate, Fez(OH)aS04. This salt is decom- posed by iron as follows :-6Fe2(0H)4SOa + Fez = 6FeS04 + 4E'e,(OK),. On reduction with iron, dilute Rolutions of ferric sulphate give ferrous sulphate only. The addition of R little sulphuric acid to the copper solution prevents the formation of the mud, and enables a clean deposit of copper to be obtained. W. T. Mercurammonium Compounds. By E. BALESTRA (Gatxettn, 21, ii, 294-305) .-The author has examined some of the ammonia- cal mercury compounds prepared by Millon (Ann.Chim. Phys. [3], IS), and also described by Gmelin. A. NHzHgC1,HgC1,.-T his compound was obtained by Millon by adding small quantities of ammonia to a large excess of corrosive sublimate. The author succeeded in preparing a compound having the same physical and chemical characteristics by gradually adding to a solution of corrosive sublimate half the quantity of dilute ammonia required for the complete precipitation of the mercury. It does not evolve ammonia when boiled with potash, but a large quan- tity is given of€ on treating it with a cold concentrated solution of ammonium bromide, according to Pesci's reaction (Abstr., 1890, 1211), showing that. the nitrogen is wholly contained in the mercur- ammonium radicle.Its composition was found to be NHg$1,2HCI. When it is suspended in water and exactly neutralised with potash, it is converted into dimercurammonium chloride, NHgzCl. B. 4NH,HgCl,NH,(Hg,O) Cl.-Millon obtained a yellow powder of the above composition by pouring a boiling solution of corrosive sublimate into a large excess of ammonia and thoroughly washing the product. Under hhese conditions, however, the author finds that a white precipitate is formed which loses ammonium chloride, on re- peatedly washing with water, gradually acquiring a yellowish tinge ; i t then approximates in composition to dimercurammonium chloride.INORQANIC CHEMISTRY. 27'7 The unaltered product has the composition of infusible white pre- cipitate, NHg2C1,NH4C1. C. NH2HgCI,2NH2(Hg,0) C12.-This compound was obtained by Millon in two ways : (1) by washing A completely with water, and (2) .by adding small quantities of ammonia to a large excess of R boiling solution of corrosive sublimate.By washing A with water, the author, however, obtained a bright yellow powder of the compo- sition 2NHg$1,HgC1,,H20, which loses + mol. H,O at 105-110". By Millon's second method (the author added to a boiling 7 per cent. solution of corrosive sublimate half tho quantity of ammonia required for neutralisation) ; a, pale-yellow, amorphous powder was obtained having the composition NHg,Cl,HCl. On neutralising this compound with potash, dimercurammoniuin chloride is formed. When it is washed with water, it is converted into the compound 2NHg2CI ,HgC1,,H20.S. B. A. A. Solubility of Glass in Cold Water. By 17. KOHLRAUSCII. ( B e y . , 24, 3560--5575).-The author has made experiments on the solu- bility of various kinds of glass by measuring the electrical con- ductivity of the solutions obtained on treating the glass with distilled water at 18". He finds that the quantity of glass dissolved increases with the time, but not proportionately, the solvent action being more rapid at first ; in the case of some of the best samples of glass, no appreciable action takes place even on prolonged contact. It was also found ttliat of two solutions, prepared in like manner from two kinds of glass having approximately the same composition, the one may have almost twice the conductivity of the other ; this result may be due t o the two samples having undergone different treatment in their manufacture.The composition of the various kinds of glass cinployed and the numerical results of the experiments are given in tables. The author also points out. that the best and most rapid test for inorganic salts in distilled water is a measurement of its electrical Conductivity ; the presence of carbonic anhydride can be ascertained in the same way. IF. S. K. Reaction between Potassium Permanganate and Hydrogen Peroxide Solution. By R. EXGEL (BUZZ. XOC. chi^. [3], 6, 17-19) -It often happens that when potassium permanganate solution i s xdded to hydrogen peroxide solution, decolorisation does not at once ensue, but thst when once the reaction has begun, decolorisatioii proceeds rapidly.Brodie attributed this result to the degree of dilution of the hydrogen peroxide, and Schone to the action of light on this compound. The author regards neither of these hypo- theses as tenable, and shows that the onset of the reaction is deter- mined by the presence of r?, trace of a manganous salt in the mixed solutions; this results from the action of traces of sulphurous o r nitrous compounds in the hydrogen peroxide solution on the per- manganate ; the maiiganous salt is then oxidised t o manganic sulphate, which is unstable in presence of hydrogen peroxide. The addition278 ABSTRACTS OF CHEMICAL PAPERS. of 8 trace of mnnganons sulphate t o the hydrogeu peroxide solution before running in the permanganste solution determines decolorisa- tion at the onset.T. G. N. Action of Ferric Chloride on Metallic Sulphides. By CAMMERER (Chem.. Centr., 1891, ii, 525; from Berg. Hiitten Zeit., 50, 295--298).-(For the first part of this work, see this vol., p.' 18.) l!'ei*ric chloride reacts wTith stannic sulphide forming stannic chloride, sulphur, and ferrous chloride. Mercuric sulphide is converted into mercuric chloride, sulphur and ferrous chloride being formed at the same time. The mercuric chloride at first formed reacts again with two more molecules of mercuric sulphide, with production of Heumann's double salt, 2HgS,HgCl2. This is a white substance, which is blackened by alkalis, but appears to be regenerated by the subsequent action of nitric acid. With silver sulphide, ferric chloride reacts, forming silver chloride, sulphur, and ferrous chloride.With the snlphides of lead, bismuth, cobalt, and manganese, ferric chloride reacts, the corresponding chloride of the metal is formed togethey mit'h ferrous chloride, and the sulphur is set free. Most of the reactions take place readily and completely . J. W. L. Coloration of Solutions of: Cobalt and the State of the Salts By A. ETARD (Compt. rend., 113, 699-701).- Its solubility a t in the Solutions. Cobalt iodide yields red, green, and blue solutions. various temperatures is as follows :- t .. .. -22" -8" -2" +go 14" 25" 34" 46" Sol.. . 52.4 56.7 58.7 61.4 61.6 66.4 73.0 79.0 t .... 60" 82" 111" 156" Sol.. . 79.2 80.7 80.9 83.1 Cobalt chloride yields rose-coloured or blue solutions, and its solu- bilities are as follows :- t ....-22" -4" +7" 11" 12" 25" 34" 41" Sol. . . 24.7 28.0 31.2 31.3 32.5 34.4 37.5 398 t . . . . 45" 49" 56" 78" 94" 96" 112" Sol.. . 41.7 4!6.7 48.4 48.8 50.5 51.2 52.3 The garnet-red, hexahydrated cobalt iodide yields a dull red solu- tion between -222" and about +20", the solubility between these limits being represented by a right line. Above 20", the liquid becomes brown, then olive, and finally at 35" deep chrome-green, this crolour persisting even up to 320". The deep green liquid yields green, lamellar crystals of the composition Co12,4H20. The formation of this salt begins at 20", and since it is more soluble than the red salt, and the two solubilities are superposed, there is a gradual increase in the total solubility between 20" and 35", the curve being convex towards the axis of temperature.Above 35", the green salt alone exists in solution, and its solubility is represented by a rightINORGANIC OHEMISTRY. 279 line. If it were possible to make experiments above 320", it is prob- able that the green liquid would become blue, and would contain a lower hydrate analogous to CoCl2,2H20. Such a blue solution is obtained when a solution of the cobalt iodide is poured into a satu- rated solution of magnesium chloride. In the case of cobalt chloride, the hydrate, CoCI, + 6H20, dissolves without change between -22" and +25", the solution has a pure rose colour, and the solubility is represented by a right line. At 25", dissociation begins, and the more soluble blue hydrate, CoCl, + 2H20, is formed, the colour of the solution changing to purple, and finally, at 50", to blue, this colour persisting up t o 300".Between 25" and 50", the curve of solubility is convex towards the axis of temperature, but beyond 50", it again becomes a right line. The changes of colour are not due to the presence of free acid or an acid salt, for they can be observed in presence of calcium carbonate or precipitated cobalt carbonate. Oxidation of Nickel Carbonyl. By BERTHELOT (Compt. rend., 113, 679--680).-Nickel carboiiyl can be kept under water in a flask wit,hont undergoing any alteration, provided that air is excluded, but i n presence of air, green nicbelous hydroxide, free from carbon, separates, and, at the same time, some of the nickel carbonyl escapes into the air and is oxidised t o a whit'e powder which, in mass, has a pale greenish tinge.It has the composition C, 5.3; NiO, 53.3; H20, 40.1 = 98.7, whilst the formula C203Ni3,10H20 requires C, 5.6 ; NiO, 52-5 ; H20, 41.9 = 100. The compound would, therefore, seem t o be the oxide of a complex radicle, analogous to croconic and rhodi- eonic acids. It is possible that one part of the nickel may be present in the form of nickel monoxide, mixed or combined with a complex oxide of the simpler formula C20Ni, belonging to the ethylene type or to a more condensed type of the same order, this oxide being formed thus, CaOtNi + 0 = C20Ni + 2C0,. C. H. B. C. H. B. Action of Hydrogen Phosphide on an Ethereal Solution of Bismuth Tribromide. By A. CAVAZZI and D. TIVOLI (Gazetta, 21, ii, 306--308).-When a solution of bismuth tribromide in dry ether is allowed t o fall gradually into a vessel containing dry hydrogen phosphide, a lustrous, black substance is formed which strongly adheres to the sides of the vessel.It becomes dull and hard on dry- ing in a vacuum, and probably has the composition PBrH(BiBr,),, being formed according to the equation SBiBr, + PI-& = 2HBr + PBrH( BiBr2)3. It is very hygroscopic, and is gradually decomposed by cold water and rapidly by boiling water, hydrogen phosphide, hydrobromic and phosphoric acids being formed and bismuth liberated. On heating it with a solution of potash, potassium bromide and phosphate are formed with evolution of hydrogen and hydrogen phosphide. Concentrated sulphnric acid has no action 011 it at the ordinary tem- perature, but, on boiling, bismuth sulphate, phosphoric acid, hydro- bromic acid, bromine, and sulphurous anhydride are formed.Concen- trated nitric acid also acts very violently on it. When triturated280 ABSTRAOTS OF CHEMICAL PAPERS. with fragments of sodium amalgam, it ignites and decomposes with more or less violence, according to the richness of the amalgam in sodium. When heated at 220" in an airnosphere of dry carbonic anhydride, it has the same composition as when dried in a vacuum. When heated in the air, it becomes viscid and decomposes with some violence and with evolution of fumes of bromine, bismuth bromide, and phosphoric anhydride. S. B. A. A. Auric Sulphide. By U. ANTOSY and A. LUCCHESI (Gazzetta, 21, ii, 209-212) .-Auric sulphide, prepared by the authors' method (Abstr., 1891, 526), is a graphitic, amorphous powder having a sp.gr. of 8-754 compared with water at 0" ; it decomposes into its elements at 197-200". Hydrochloric acid has no action on nuric sulphide, concentrated nitric acid oxidises it in the cold with separation of gold, whilst a,qua regia readily dissolves it. Caustic alkalis (15 per cent,) decompose the sulphide on heating, metallic gold and alkali sulphide and ttiio- sulpbate being obtained. Ammonia solution, on prolonged contact, decomposes it., yielding snlphuric acid, free sulphur, and a. little hydrogen sulphide. The action of potassium cyanide solution on auric sulphide differs from its action on aurous sulphide and auroso- auric snlphide, both of which dissolve and are reprecipitated by acids, Auric sulphide dissolves in aqueous potassium cyanide (25 per cent.) to a yellow solution, which, on boiling, becomes colourless and deposits xurous potassium cyanide on cooling, AuzS3 + 6KCN = 2AuKCzN, + K,S + SKCNS.This reaction shows the substance to be homo- geneous, the auric sulphide prepared by Berzelius having been shown by Kruss and Hoffmann (Abstr., 1887,1019 ; 1888,28) to be a mixture of aurosoauric sulphide with sulphur. Hydrosulphide and polysulphides of ammonia dissolve auric sulph- ide readily on warming, gold being deposited. The sulphides and polysulphides of the alkali metals dissolve the substance slowly ; whilst the alkali hydrosnlphides dissolve it i n the cold, yielding a deep reddish-brown solution, which, on boiling, becomes pale yellow and deposits gold. These solutions, on treatment with hydrochloric acid, give a flocculent, yellow precipitate, probably consisting of auric hydrosulphide, which soon turns brown with formation of hydrogen sulphide and auric sulphide.On precipitation with alcohol at -lo", a concentrated solution of nuric sulphide in sodium hydrosulphide gives a yellowish precipitate ; this, when filtered through asbestos at the same temperature in an atmo- sphere of nitrogen and washed wihh alcohol, yields a white, crystalline product which soon darkeiis in colour. The freshly prepared sub- stance is very soluble in water and is probably fiodium sulphaurate, but it decomposes so rapidly that tiaustmorthy analyses could not be made. W. J. P. Iodonitro- and Bromonitro-platinum Compounds.By M. VEZES (Compt. rend., 113, 696--698).-The relative stability of the nitro- and hdoid groups in the iodonitro-platinum compounds is the inverse of that observed in the case of the chloronitro- and bromo-INORQANIC OEEMISTRY. 281 nitro-compounds (Abstr., 1891, 807). The iodine is displaced by nitrogen oxides, and the stable term of the series is the platonitrite and not the platoiodide. Nitrogen oxides readily decompose potassium platniodide wit>h liberation of iodine ; whilst, on the other hand, iodine in the forin of vapour or in solution in hydriodic acid or a solution of potassium iodide does not decompose potassium platonitrite. The action of an alcoliolic solution of iodine on a warm solution of potassium platonitrite, however, yields large, brilliant prisms of t'he compound Pt,2N02,K,T, + 2Hz0, described by Nilson, but this, ttlthough stable in warm solutions, is decomposed by nitrogen oxides o r potassium nitrite, with liberation of iodine and formation of the platonitrite.The author has so far failed to obtain a compound Pt,4N0,,KzI,, analogous to the platichloronitrite and the platibromo- nitrite previously described. If, however, an excess of an alcoholic solution of iodine is added to a warm concentrated solution of the platonitrite, and the mixture is concentrated as rapidly as possible at a gentle heat, potassium nityoso- plati-iodide, PtT3,N0,K212, separates on cooling in small, brilliant, black crystals, which remain unaltered a t 100". When heated in a current of hydrogen, this compound yields water, ammonium iodide, iodine, hydrogen iodide, and a, residue of potassium iodide and platinum.If the liquid is slcwly concentrated (without ebullition) a different compound is obtained on cooling. T t is potassium platitetraiodonitrite, Pt14,2NOz,Kz, and separates in small, well-defined, black crystals with a greenish lustre. Like the preceding compound, it is only slightly soluble in water, yielding a deep brown solution, and it undergoes no change at 100". When heated in a current, of hydrogen, it yields water, ammonium iodide, iodine, and hydrogen iodide, with a residue of potassium iodide and pIatinum. When the mixture yielding these two salts is submitted t o pro- longed ebullition, the excess of iodine is expelled and potassium platoiodonitrite is formed.When bromine-water is added t o a solution of potassium platoiodo- nitrite until a11 the iodine is expelled, the brown liquid, when con- centrated in a dry vacuum, yields yellow, tabular crystals. The same crystals are obtained by heating n solution of potassium platibromo- nitrite, Pt,4N0,,K(,Rr2, with alcohol at 80", aldehyde and other gases being evolved. This new cornpound is potassium platobronionitrite, Pt,2NO2,K,Br, + H,O, and ib very soluble in water. At loo", it becomes anhydrous and biaiglit yellow ; at a higher temperature, it blackens with evolution of nitrogen oxides, a residue containing potassium bromide and platinum in the proportions Pt + 2KBr being left. When heated in hydrogen, it yields the same residue without any formation of ammoni urn bromide, bromine, o r hydrogen bromide, By R.SCIINEIDER (J. pr. Ckem. [ 21, 44, 507--512).-Potassium plati?ao~eleilostannate, K,Pt,SnSe,, is prepared by melting together 10 parts of platinum sponge, 6-8 parts of tin selen- ide, 30-40 parts of potassium carbonate, and 30-40 parts of selenium in a covered porcelain crucible SO that the mass shall remain liquid for 8-10 minutes after the frothing has ceased. The melt is extracted C. H. B. Two New Seleno-salts.282 ABSTRACTS OF CHEMICAL PAPERS. with water, and the undissolved residue digested with concentrated potassium hydroxide solution, whereby the selenostannate is left un- attacked. It forms small, well-defined, hexagonal tables which are leaden-grey by reflected light and have a strong metallic lustre ; in thin layers they are reddish-brown by transmitted light.In its general habitus it is very similar to the author's potassium platino- sulphostmnate (Ann. Phys. Chem., 138, 612). At the ordinary tem- perature, it is stable in air ; when heated in air, it loses all its selenium, the residue consisting of a mixture of platinum with potassium stan- nate. It is not attacked by water, ammonia, potash, 01- hydrochloric acid, either hot or cold. Sodium pZat inoseZenostnnnate, Na,Pt,SnSe,, cannot be obtained by merely substituting sodium carbonate for potassium carbonate in the above prescription. It is readily formed, however, when 10 parts of platinum sponge, 5-6 parts of tin selenide, 40 parts of potassium carbonate, 5-6 parts of sodium carbonate, and 40 parts of selenium are fused together, and the melt treated as described above.It forms leaden-grey, microscopical, hexagonal lamin* with a brilliant, metallic lustre ; the remarks concerning the potassium salt apply also to this. A. G. B. Saline Compounds of the Lower Ruthenium Oxides with the Higher Oxides. By A. JOLY (C'ontpt. rend., 113,694-695.)-When the products of the sudden decomposition of potassium perruthenate at 440" are kept for a long time at this temperature, interaction takes place with formation of a black, crystalline compound of the composi- tion K20,6Ru205. Sodium permthenate at 440" loses oxygen and water, and after treatment of the residue with water, which removes sodium oxide and the orange sodium ruthenate, a black crystalline powder is left with a, composition approximating very closely to that for Na20, 3Ru205. Barium ruthenate, IBaRuO4, at 440" loses oxygen, and yields a ruthenite, BaRuO,, different in properties from a mixture of barium monoxide and ruthenium dioxide.The author directs atten- tion t o the analogy between these compounds and the products obtained by Rousseau from the permanganates, and the product K,O,Oe,O,, obtained by himself from potassium osmiamate. Action of Light on Ruthenium Peroxide. By A. JOLY (t%~~pt. rend., 113, 693--694).-When sealed tubes containing perfectly dry ruthenium peroxide are kept in the dark, no alteration takes place, but on exposure to sunlight, the walls of the tube become coated with a pale brown layer, which gradually increases in thickness and eventually transmits only red light.Beyond this point, the peroxide behind undergoes no further change. The brown deposit dissolves at once in potassium hydroxide solution, forming a yellow liquid without any trace of green, and in hydrochloric acid with evolution of chlorine and formation of a solution of ruthenium sesquichloride. It seems, therefore, t,hat when exposed to sunlight, ruthenium per- oxide is reduced to the trioxide, Ru03. C. H. B. C. H. B.2 70 ABSTRACTS OF CHEMICAL PAPERS.I n o r g a n i c C h e m i s t r y.Preparation of Pure Hydrogen Peroxide Solution. By L.CR~SMER (BUZZ. Soc. Chim. [3], 6, 24--25).-The solution of hydrogenperoxide which results from the action of hydrochloric acid (sp. gr.1.1) on barium dioxide is extracted by shaking with ether, and theethereal solution is agitated with distilled water, to which it yieldsthe dissolved hydrogen peroxide.By repetitions of this process, apure, neutral solution corresponding with 0.8-0.9 per cent of hydro-gen peroxide is obtainable, from which the dissolved ether may beeliminated by distillation under reduced pressure.Preparation of Hydrobromic Acid. By G. S. NEWTH (Chena.News, 64, 215).-By means of the following arrangement a largequantity of bromine can be rapidly converted into hydrobromic acid :-A glass tube, 7 inches long and 8 inch in diameter, is fitted at eachend with a cork carrying a piece of small tubing and a piece of fitoutwire. The ends of these pieces of stout wire, within the longer tube,T.G. NINORQANlO OHEMISTRY. 271are joined hy a spiral of platinum wire 1 inch long, and after expellingthe air the spiral is heated to bright redness by Rn e1ecOric current; astream of hydrogen, impregnated with bromine by bubbling throughthat liquid, which may be heated at GO", is passed through the longertube, and, as long as a Blight excess of hydrogen is maintained, hydro-bromic acid quite free from bromine issues from the other end, and iscollected in water. There is very little danger of explosion, but t orender it impossible, the small supply tube may be plugged with alittle glass -vr7001.Solubility of Gases in Water. By I;. W. WINKLER (Ber., 24,3602-3610; compare Abstr., 1891, 384).-In this paper the authorgives in tabular form the results of experiments on the solubility ofnitrogen and of oxygen in water, a t temperatures ranging from 0" to80"; the calculated values for the solubility of the two gases attemperatures ranging from 80--100" are also given.As regards thesolubility of nitrogen, the author's values are considerably larger thanthose given by Bunsen.The Densities of Sulphuric Acid Solutions. By S. U. PICKER-ING (Chem. News, 64,31l).-Lunge's doubts as t o the accuracy of theauthor's density determinations are, in the opinion of the latter, dueto a misapprehension as to the method employed to determine thestrength of the acid, and as to the experimental error involved.D. A. L.E'. S. K.s. u. P.The Contraction on Mixing Sulphuric Acid and Water. ByS.U. PICKERING (Chem. Netus, 64, 14--15).-From his own results theauthor has calculated the strength a t which the maximum contractionoccurs, and finds that this maximum, when calculated for unit weightof solution, shifts from 67 per cent. at 8" to 70.1 per cent. at 38",whereas the maximum contraction calculated per unit volume remainspractically constant at 76 per cent. throughout this range of tempera-ture. Neither of these maxima occurs at the composition of the di-hydrate (73.1 per cent.) or of that of any other hydrate of whichindications have been obtained. The rate atr which the amount ofcontraction is influenced by the temperature varies irregularly withthe actual value of the temperatnre ; tlhus with solutions from 64 to80 per cent. st,rength, the contraction diminishes at nearly the samerate for the intervals 8" to 18", and 28" to 38", whereas for the inter-mediate interval, 18" t o 28", the rate of diminution is about 25 percent. smaller.s. u. P.Density and Composition of Dilute Sulphuric Acid. By A.W. R ~ C K E R (Phil. Mug. [ 5],32,304-313) .-Pickering (Trans., 1890,64) deduced the existence of various hydrates in sulphuric acid solu-tions from sudden changes of curvature in the curves representing thevariation of the density, &c., of such solutions with the composition.The author considers the " first diflerential " curve for 18" between 46per cent. and 80 per cent., in which there are supposed to be f o u r breaks,and shows that it is possible to find an equation, and, therefore, acontinuous curve, which will represent the results within the limits ofthe experimental error.The equation is of the formu 272 ABSTRACTS OF OHEMIOAL PAPERS.and contains fieven ai-bitrary constants. The author thus doubts thevalue of '' differentiation " or of the bent ruler, as used by Pickering,in discovering changes of curvature in curves drawn to representexperimental numbers. J. W.The Densities of Sulphuric Acid Solutions. By S . U.PICKERING (Phil. Xag. [ 5 ] , 33, 132).-The curve used by RiicBer(see preceding abstract) to bridge over four of what the writer con-sidered to be breaks in the figure foymed by the first differentials of thedensities of the acid can, he maintains, prove nothing about one ofthese breaks, and very little about another, for it extends too short a,distance beyond them.The formula for the curve suggested byRucker consists of a combination of an exponential curve and astraight line, on to part of which (as this curve did not agree withthe results) a hump was engrafted by means of a complex fourth term,tlhe whole forming a curve for which, as an expression of physicalfacts, there would seem to be no precedent. The points, moreover, atwhich the term expressing the hump begins to be appreciable andagain becomes inappreciable correspond exactly with the positionsassigned by the author to two of t,he breaks, thus confirming, ratherthan disproving, the existence of these as points where some newconditions in the solutions become sensible.The only break whichBucker's curve does successfully bridge over is that particular onewhich the author pointed out to be especially doubtful: and evenin favour of this break the evidence is not thereby entirely negatived.The author maintains that even the most successful attempt to offeran alternative explanation of a small portion of some of his resultscould not upset his conclusions which were based entirely on thecumulative evidence derived from many sources, and of which hehere gives a summary. s. u. P.Boron Phosphides. By H. MOISSAN (Compt. rend., 113, 726-~29).--Bo~on phosphide, PB, is obt,ained by reducing the phospha-iodide in hydrogen at 450-500". The product is powdered, andagain heated in hydrogen at the same temperature in order toremove excess of iodine, and the process is repeated two or threetimes if necessary, care being taken that the temperature does notexceed 500".I t is a very light, amorphous, maroon-coloured powdel.,insoluble in the chlorides of arsenic, phosphorus, carbon, and anti-mony, and in all solvents, organic and ino~ganic, that were tried.It is n o t volatile in a, vacuum at 500". At 200", in presence of oxygen,it burns and yields boric and phosphoric anhydrides ; when throwninto fused alkaline nitrates, there is incandescence and deflagration ;fused sulphur has no action, hut sulphur vapour converts it intoboi-on and phosphorus sulphides. Chlorine converts boron phosphideinto boron trichloride and phosphorus pentachloride, with incand-escence ; bromine has no action in the cold, but combination takesplace if the temperature is raised.Vapour of iodine, arsenic, oINORQANIC CHEMISTRY. 273phosphorus has no action at a dull-red heat. When heated at 5OO"ina current of nitrogen, boron phosphide yields no nitride, and, althougha,t higher temperatures it loses phosphorus (as it does in a vacuum),no nitride is formed. When a mixture of boron phosphide andsodium is gently heated in a current of hydrogen, it rapidly becomesincandescent, with formation of sodium phosphide and boride. Potas-sium yields the same products at a lower temperature. A mixtureof the boron phosphide with powdered magnesium becomes incand-escent at about 500", but aluminium has no action, except at a muchhigher temperature. Finely-divided silver, copper, and platinumreact with the phosphide when gently heated, biit mercury has noaction at its boiling point.When boron phosphide is thrown into thestrongest nitric acid, it takes fire, even in the vapour, and burnsbrilliantly on the surface of the liquid; on slightly heating, it dis-solves immediately and completely. Concentrated solutions of hydro-chloric and hydriodic acids have no action on the phosphide, andsulphuric acid has no action in the cold, but is reduced on heating.Concentrated hot solutions of potash and soda dissolve i t slowly,whilst fused potash dissolves it completely with formation of hydrogenphosphide and potassium borate. Gaseous hydrogen fluoride attacksit below dull redness, with formation of boron fluoride, hydrogen,and phosphorus.Gaseous hydrogen chloride behaves similarly at Rhigher temperature. Boiling water has no action on the phosphide,but water vapour decomposes it at 400" with production o€ boric acidand hydrogen phosphide. Hydrogen sulphide at A dull-red heatyields boron sulphide and hydrogen phospliicle. I n ammonia at about300" the phosphide burns, with formation of boron nitride and libera-tion of phosphorus.Boron phosphide, B5P3, is obtained by heating the preceding com-pound at 1000" in a current of hydrogen. It has a paler colour thanthe phosphide PB, does not inflame in chlorine or nitric acid, and isnot attacked by the latter, even when boiling. It is insoluble in allinorganic and organic solvents, burns with some difficulty in oxygen,and is attacked by fused nitrates with incandescence, but only withdifficulty by metals and non-metals.It does not burn in chlorinebelow a dull-red heat.I f the hydrogen used for reduction contains water or oxygen, awhite phosphoboric acid, also known as boron phosphate, is formed,and the same compound is obtained by the action of nitrogen oxideson a mixture of the phosphides.Boron Phosphide. By A. BESSON (Compt. Tend., 113, 772-7731,-The author draws attention to his previous note on the subject(Abstr., 1891, 1418), and describes 'an additional property of thephosphide. It is oxidised by dilute nitric acid to a substance, prob-ably phosphoboric acid, which is left in nacreous plates on evaporatingthe solution to dryness, is soluble in water, and gives a white, gelatin-ous precipitate with excess of ammonia.Moissan (ibid., 787-788), commenting on this and the previousnote, draws attention to the absence of numerical data justifying theformula, BP attributed to the phosphide, and claims priority in theC.H. B274 ABSTRACTS OF CEEMTOAL PAPERS.Rystematic study of boron phosphides (Compt. rend.! 112, 71 7, and113, 19). JN. w.The Influence of Steam and other Gases on the Combustionof Carbonic Oxide and Oxygen. By N. BBK~TOFF (Ckenz. Centr.,1831, ii,449-450 ; from Bull. Acad. St. PrStersbourg [2], 2, 175-179).-Having repeated Dixon's experiments (Trans., 1885,94), the authorhas obtained the same results, namely, that, a mixture of carbonicoxide and oxygen, when dried by nieans of phosphoric anhydride, isnot exploded by the passage of an electric spark, and further, that ifthe gases be dried by sulphuric acid, which appears to leave a trace ofmoisture in the gas, the combustion proceeds so slowly that it maybe followed with the eye.The presence of other gases, sulphurousanhydride, o r nitrous oxide, had not a similar effect. Cyanogen, onthe other hand, when present to the extent of 10 per cent., exerted aninfluence similar to that of steam, and caused an immediate explosionon the passage of the spark. BBkbtoff suggests that the action ofcyanogen may be explained on the assumption that the beak liberated inthe decomposition of the cyaiiogen is added to the heat of conibustionof the carbon, and he further assumes that the dissociation tempern-ture of the water niolecule being lower than that of the oxygenmolecule is the proper explanation of the fact that the presence ofwater in a mixture of carbonic oxide and oxygen assists the com-bustion of the latter gases.J. W. L.Reactions of Carbonic Anhydride at High Pressures. By A.D'ARSONVAL (Conzpt. rend. SOC. Biol., 1891, 320--321).--Liquefied carb-onic anhydride is a powerful antiseptic. It does not coagulate albu-min. At high pressures it can displace both organic and mineralacids. When urine is subjected to a pressure of 40 atmospheres ofcarbonic, anhydride, crystals of uric acid are deposited. A dilute solu-tion of potassium silicate similarly treated becomes solidified from thedeposition of silicic acid.Carbonic anhydride at the same pressureis also able t o liberate hydriodic and hydrobromic acids from solu-tions of potassium iodide and bromide respectively.By N. BEKBTOFF (Chenz. Centr., 1891, ii,450-451 ; from Bull. Acad. St. Pe'tersbourg [2], 2, 169--17G).-Theauthor has prepared considerable quantities of cssium by redncingthe hydroxide with aluminium. The reaction proceeds well, and isas readily carried out as the reduction of rubidium. The author referst o the results, which Winkler has recently published, on the heat ofcombination of the alkali metals, and points out that the deductionswhich Winkler has made are diametrically opposed to his own. Theauthor finds that the combining heat of the alkali metals variesinversely with their atomic weights.He further urges that Winklerhas not employed the oxides, but the hydroxides and carbonate5,for his determinations, which are not suitable for the purpose.Properties of Czesium and its Hydroxide. By N. BEKETOFF(Chem. Centr., 1891, ii, 451 ; from Bull. Acad. St. Pe'teysbourg [SJ, 2,W. D. H.Reduction of CEsium.J. W. LINORGANIC CHEMISTRY. 275171--173).--The author has used a very pure specimen of caesiumsulphate as material on which t o work. A determination of thesulphnric acid gave almost the theoretical amount, and by means ofthe spectroscope only a trace of rubidium could be detected.The hydroxide was obtained from the sulphate by precipitation withbarium hydroxide, and concentration of the filtrate in a silver dish,which was placed in a metal retort.In this manner carbonic an-hydride was excluded. After concentration in the retort, the solutionwas transferred to a small silver dish, the remainder of the waterevaporated, and the hydroxide finally fused. The silver dish wasslightly attacked, and the silver oxide dissolved in the fused caesiumhydroxide ; it was again precipitated, however, as the latter cooled.The quantity of silver oxide was inconsiderable. Cmium hydroxidehas a sp. gr. 4.0178, compared with water at 4" ; the molecularvolume = 57.3 (sodium hydroxide = 18 ; potassium hydroxide = 27 ;rubidium hydroxide = 32 ?). The molecular volume of the hydroxidecorresponds with a very considerable diminution of volume during itsformation from the elements.The heat of solution of the hydroxidein water is 15,876, which is higher than f o r all other alkali hydroxides.The heat of neutralisation of the hydroxide with hydrogen chloride(dilute) is 13,790, or nearly the same as that found by Thomsen forpotassium hydroxide, or, indeed, for the other alkali hydroxides.Metallic cEsium was obtained by heating 114 grams of the hydroxidein a nickel retort wit,h 27 grams of aluminium, and collecting thedistilled metal in glass receivers; 20-25 grams of the metal werethus obtained.The heat of combination of cmium with water was found to befrom 50 to 53 Cal. J. W. L.Action of Sodammonium and Potassammonium on Metals.By JOAXNIS (Cornpi!. rend., 113, 795-798).- Sodammonium andpotassammonium are decomposed by mercury, lead, and antimony, butare not affected by aluminium, silver, zinc, or copper.When a solutionof sodammonium in liquefied ammonia is allowed to fall drop by dropon mercury, it is decomposed with the formation of a crystallineamalgam, Hg,Na, which can be washed free from excess of the reagentwith liquefied ammonia. Potassammonium under similar conditionsyields an amalgam of the composition Hgl8K.When pure lead is brought into contact with sodammonium, thereddish-brown liquid turns blue, and then green, and a little hydrogenis disengaged, owing to the spontaneous decomposition of the sod-ammonium. Finally, however, the metal is partially converted into anindigo-blue mass, which dissolves in liquefied ammonia to a bottle-green solution, and has the composition Pb4Na,2NH,.This substancedecomposes at ordinary pressures, leaving a grey mass resemblingspongy platinum, and oxidises rapidly on exposure to air, with evolu-tion of heat. When thrown into water, the first portion dissolvescompletely, owing to the combination of the lead with the dissolvedoxygen, and the subsequent solution of the oxide in the sodiumhydroxide simultaneously formed, but as soon as the oxygen is usedup, metallic lead is thrown down as a black, curdy precipitate.JN. W276 ABSTRACTS OF OHEMICAL PAPERS.Influence of Ammonia on the Solubility of AmmoniumChloride. By R. EXGEL (BUZZ. Xoc:Chim. [3], 6, 17).-The pre-sence of ammonia diminishes the solubility of ammonium chloride inwater at 0' at first, but as the quantity of ammonia is increased, acorresponding increase in the chloride dissolved obtains.This resultis probably due to the formation of compounds of ammonium chlorideand hydroxide. Tables of values are given. T. G. N.Precipitation of Copper by Iron and the Action of Iron onFerric Solutions. By J. C. Essmit (BdZ. SOC. Chin%. [ 3 ] , 6, 147-148) .-In the wet method of copper extraction, very mixed qualitiesof acrap iron are employed. The reduced copper occurs in powder,grains, and filaments which cannot be readily washed free from themud of ferric hydroxide formed. The author finds that the structureof the iron used exerts a marked influence on that of the copperobtained.By selecting the iron to be used, it is possible to obtaiiithe reduced copper in a fibrous or granular condition, when i t admitsof being readily washed free from ferric hydroxide.The occurrence of the mud of ferric hydroxide is due to the form-ation of a basic ferric sulphate, Fez(OH)aS04. This salt is decom-posed by iron as follows :-6Fe2(0H)4SOa + Fez = 6FeS04 +4E'e,(OK),. On reduction with iron, dilute Rolutions of ferric sulphategive ferrous sulphate only.The addition of R little sulphuric acid to the copper solutionprevents the formation of the mud, and enables a clean deposit ofcopper to be obtained. W. T.Mercurammonium Compounds. By E. BALESTRA (Gatxettn,21, ii, 294-305) .-The author has examined some of the ammonia-cal mercury compounds prepared by Millon (Ann.Chim. Phys. [3],IS), and also described by Gmelin.A. NHzHgC1,HgC1,.-T his compound was obtained by Millon byadding small quantities of ammonia to a large excess of corrosivesublimate. The author succeeded in preparing a compound havingthe same physical and chemical characteristics by gradually addingto a solution of corrosive sublimate half the quantity of diluteammonia required for the complete precipitation of the mercury. Itdoes not evolve ammonia when boiled with potash, but a large quan-tity is given of€ on treating it with a cold concentrated solution ofammonium bromide, according to Pesci's reaction (Abstr., 1890,1211), showing that. the nitrogen is wholly contained in the mercur-ammonium radicle.Its composition was found to be NHg$1,2HCI.When it is suspended in water and exactly neutralised with potash,it is converted into dimercurammonium chloride, NHgzCl.B. 4NH,HgCl,NH,(Hg,O) Cl.-Millon obtained a yellow powder ofthe above composition by pouring a boiling solution of corrosivesublimate into a large excess of ammonia and thoroughly washingthe product. Under hhese conditions, however, the author finds thata white precipitate is formed which loses ammonium chloride, on re-peatedly washing with water, gradually acquiring a yellowish tinge ;i t then approximates in composition to dimercurammonium chlorideINORQANIC CHEMISTRY. 27'7The unaltered product has the composition of infusible white pre-cipitate, NHg2C1,NH4C1.C. NH2HgCI,2NH2(Hg,0) C12.-This compound was obtained byMillon in two ways : (1) by washing A completely with water, and(2) .by adding small quantities of ammonia to a large excess of Rboiling solution of corrosive sublimate.By washing A with water,the author, however, obtained a bright yellow powder of the compo-sition 2NHg$1,HgC1,,H20, which loses + mol. H,O at 105-110".By Millon's second method (the author added to a boiling 7 percent. solution of corrosive sublimate half tho quantity of ammoniarequired for neutralisation) ; a, pale-yellow, amorphous powder wasobtained having the composition NHg,Cl,HCl. On neutralising thiscompound with potash, dimercurammoniuin chloride is formed. Whenit is washed with water, it is converted into the compound2NHg2CI ,HgC1,,H20.S. B. A. A.Solubility of Glass in Cold Water. By 17. KOHLRAUSCII. ( B e y . ,24, 3560--5575).-The author has made experiments on the solu-bility of various kinds of glass by measuring the electrical con-ductivity of the solutions obtained on treating the glass with distilledwater at 18". He finds that the quantity of glass dissolved increaseswith the time, but not proportionately, the solvent action being morerapid at first ; in the case of some of the best samples of glass, noappreciable action takes place even on prolonged contact. It was alsofound ttliat of two solutions, prepared in like manner from two kindsof glass having approximately the same composition, the one mayhave almost twice the conductivity of the other ; this result may bedue t o the two samples having undergone different treatment intheir manufacture.The composition of the various kinds of glasscinployed and the numerical results of the experiments are given intables.The author also points out. that the best and most rapid test forinorganic salts in distilled water is a measurement of its electricalConductivity ; the presence of carbonic anhydride can be ascertainedin the same way. IF. S. K.Reaction between Potassium Permanganate and HydrogenPeroxide Solution. By R. EXGEL (BUZZ. XOC. chi^. [3], 6, 17-19)-It often happens that when potassium permanganate solution i sxdded to hydrogen peroxide solution, decolorisation does not at onceensue, but thst when once the reaction has begun, decolorisatioiiproceeds rapidly.Brodie attributed this result to the degree ofdilution of the hydrogen peroxide, and Schone to the action oflight on this compound. The author regards neither of these hypo-theses as tenable, and shows that the onset of the reaction is deter-mined by the presence of r?, trace of a manganous salt in the mixedsolutions; this results from the action of traces of sulphurous o rnitrous compounds in the hydrogen peroxide solution on the per-manganate ; the maiiganous salt is then oxidised t o manganic sulphate,which is unstable in presence of hydrogen peroxide. The additio278 ABSTRACTS OF CHEMICAL PAPERS.of 8 trace of mnnganons sulphate t o the hydrogeu peroxide solutionbefore running in the permanganste solution determines decolorisa-tion at the onset.T. G. N.Action of Ferric Chloride on Metallic Sulphides. ByCAMMERER (Chem.. Centr., 1891, ii, 525; from Berg. Hiitten Zeit., 50,295--298).-(For the first part of this work, see this vol., p.' 18.)l!'ei*ric chloride reacts wTith stannic sulphide forming stannic chloride,sulphur, and ferrous chloride. Mercuric sulphide is converted intomercuric chloride, sulphur and ferrous chloride being formed at thesame time. The mercuric chloride at first formed reacts again withtwo more molecules of mercuric sulphide, with production ofHeumann's double salt, 2HgS,HgCl2. This is a white substance,which is blackened by alkalis, but appears to be regenerated by thesubsequent action of nitric acid.With silver sulphide, ferric chloride reacts, forming silver chloride,sulphur, and ferrous chloride. With the snlphides of lead, bismuth,cobalt, and manganese, ferric chloride reacts, the correspondingchloride of the metal is formed togethey mit'h ferrous chloride, andthe sulphur is set free.Most of the reactions take place readily andcompletely . J. W. L.Coloration of Solutions of: Cobalt and the State of the SaltsBy A. ETARD (Compt. rend., 113, 699-701).-Its solubility a tin the Solutions.Cobalt iodide yields red, green, and blue solutions.various temperatures is as follows :-t .. .. -22" -8" -2" +go 14" 25" 34" 46"Sol.. . 52.4 56.7 58.7 61.4 61.6 66.4 73.0 79.0t .... 60" 82" 111" 156"Sol.. . 79.2 80.7 80.9 83.1Cobalt chloride yields rose-coloured or blue solutions, and its solu-bilities are as follows :-t ....-22" -4" +7" 11" 12" 25" 34" 41"Sol. . . 24.7 28.0 31.2 31.3 32.5 34.4 37.5 398t . . . . 45" 49" 56" 78" 94" 96" 112"Sol.. . 41.7 4!6.7 48.4 48.8 50.5 51.2 52.3The garnet-red, hexahydrated cobalt iodide yields a dull red solu-tion between -222" and about +20", the solubility between theselimits being represented by a right line. Above 20", the liquidbecomes brown, then olive, and finally at 35" deep chrome-green, thiscrolour persisting even up to 320". The deep green liquid yieldsgreen, lamellar crystals of the composition Co12,4H20. The formationof this salt begins at 20", and since it is more soluble than the redsalt, and the two solubilities are superposed, there is a gradualincrease in the total solubility between 20" and 35", the curve beingconvex towards the axis of temperature.Above 35", the green saltalone exists in solution, and its solubility is represented by a righINORGANIC OHEMISTRY. 279line. If it were possible to make experiments above 320", it is prob-able that the green liquid would become blue, and would contain alower hydrate analogous to CoCl2,2H20. Such a blue solution isobtained when a solution of the cobalt iodide is poured into a satu-rated solution of magnesium chloride.In the case of cobalt chloride, the hydrate, CoCI, + 6H20, dissolveswithout change between -22" and +25", the solution has a purerose colour, and the solubility is represented by a right line. At 25",dissociation begins, and the more soluble blue hydrate, CoCl, +2H20, is formed, the colour of the solution changing to purple, andfinally, at 50", to blue, this colour persisting up t o 300". Between25" and 50", the curve of solubility is convex towards the axis oftemperature, but beyond 50", it again becomes a right line.The changes of colour are not due to the presence of free acid or anacid salt, for they can be observed in presence of calcium carbonateor precipitated cobalt carbonate.Oxidation of Nickel Carbonyl.By BERTHELOT (Compt. rend.,113, 679--680).-Nickel carboiiyl can be kept under water in a flaskwit,hont undergoing any alteration, provided that air is excluded, buti n presence of air, green nicbelous hydroxide, free from carbon,separates, and, at the same time, some of the nickel carbonyl escapesinto the air and is oxidised t o a whit'e powder which, in mass, has apale greenish tinge.It has the composition C, 5.3; NiO, 53.3;H20, 40.1 = 98.7, whilst the formula C203Ni3,10H20 requires C, 5.6 ;NiO, 52-5 ; H20, 41.9 = 100. The compound would, therefore, seemt o be the oxide of a complex radicle, analogous to croconic and rhodi-eonic acids. It is possible that one part of the nickel may be presentin the form of nickel monoxide, mixed or combined with a complexoxide of the simpler formula C20Ni, belonging to the ethylene typeor to a more condensed type of the same order, this oxide being formedthus, CaOtNi + 0 = C20Ni + 2C0,.C. H. B.C. H. B.Action of Hydrogen Phosphide on an Ethereal Solution ofBismuth Tribromide. By A.CAVAZZI and D. TIVOLI (Gazetta,21, ii, 306--308).-When a solution of bismuth tribromide in dryether is allowed t o fall gradually into a vessel containing dry hydrogenphosphide, a lustrous, black substance is formed which stronglyadheres to the sides of the vessel. It becomes dull and hard on dry-ing in a vacuum, and probably has the composition PBrH(BiBr,),,being formed according to the equation SBiBr, + PI-& = 2HBr +PBrH( BiBr2)3. It is very hygroscopic, and is gradually decomposedby cold water and rapidly by boiling water, hydrogen phosphide,hydrobromic and phosphoric acids being formed and bismuth liberated.On heating it with a solution of potash, potassium bromide andphosphate are formed with evolution of hydrogen and hydrogenphosphide.Concentrated sulphnric acid has no action 011 it at the ordinary tem-perature, but, on boiling, bismuth sulphate, phosphoric acid, hydro-bromic acid, bromine, and sulphurous anhydride are formed.Concen-trated nitric acid also acts very violently on it. When triturate280 ABSTRAOTS OF CHEMICAL PAPERS.with fragments of sodium amalgam, it ignites and decomposes withmore or less violence, according to the richness of the amalgam insodium. When heated at 220" in an airnosphere of dry carbonicanhydride, it has the same composition as when dried in a vacuum.When heated in the air, it becomes viscid and decomposes with someviolence and with evolution of fumes of bromine, bismuth bromide,and phosphoric anhydride.S. B. A. A.Auric Sulphide. By U. ANTOSY and A. LUCCHESI (Gazzetta, 21,ii, 209-212) .-Auric sulphide, prepared by the authors' method(Abstr., 1891, 526), is a graphitic, amorphous powder having a sp. gr.of 8-754 compared with water at 0" ; it decomposes into its elementsat 197-200".Hydrochloric acid has no action on nuric sulphide, concentratednitric acid oxidises it in the cold with separation of gold, whilst a,quaregia readily dissolves it. Caustic alkalis (15 per cent,) decomposethe sulphide on heating, metallic gold and alkali sulphide and ttiio-sulpbate being obtained. Ammonia solution, on prolonged contact,decomposes it., yielding snlphuric acid, free sulphur, and a. littlehydrogen sulphide. The action of potassium cyanide solution onauric sulphide differs from its action on aurous sulphide and auroso-auric snlphide, both of which dissolve and are reprecipitated by acids,Auric sulphide dissolves in aqueous potassium cyanide (25 per cent.)to a yellow solution, which, on boiling, becomes colourless and depositsxurous potassium cyanide on cooling, AuzS3 + 6KCN = 2AuKCzN, + K,S + SKCNS.This reaction shows the substance to be homo-geneous, the auric sulphide prepared by Berzelius having been shownby Kruss and Hoffmann (Abstr., 1887,1019 ; 1888,28) to be a mixtureof aurosoauric sulphide with sulphur.Hydrosulphide and polysulphides of ammonia dissolve auric sulph-ide readily on warming, gold being deposited. The sulphides andpolysulphides of the alkali metals dissolve the substance slowly ;whilst the alkali hydrosnlphides dissolve it i n the cold, yielding adeep reddish-brown solution, which, on boiling, becomes pale yellowand deposits gold.These solutions, on treatment with hydrochloricacid, give a flocculent, yellow precipitate, probably consisting of aurichydrosulphide, which soon turns brown with formation of hydrogensulphide and auric sulphide.On precipitation with alcohol at -lo", a concentrated solution ofnuric sulphide in sodium hydrosulphide gives a yellowish precipitate ;this, when filtered through asbestos at the same temperature in an atmo-sphere of nitrogen and washed wihh alcohol, yields a white, crystallineproduct which soon darkeiis in colour. The freshly prepared sub-stance is very soluble in water and is probably fiodium sulphaurate,but it decomposes so rapidly that tiaustmorthy analyses could not bemade.W. J. P.Iodonitro- and Bromonitro-platinum Compounds. By M.VEZES (Compt. rend., 113, 696--698).-The relative stability of thenitro- and hdoid groups in the iodonitro-platinum compounds is theinverse of that observed in the case of the chloronitro- and bromoINORQANIC OEEMISTRY. 281nitro-compounds (Abstr., 1891, 807). The iodine is displaced bynitrogen oxides, and the stable term of the series is the platonitriteand not the platoiodide. Nitrogen oxides readily decompose potassiumplatniodide wit>h liberation of iodine ; whilst, on the other hand, iodinein the forin of vapour or in solution in hydriodic acid or a solution ofpotassium iodide does not decompose potassium platonitrite.The action of an alcoliolic solution of iodine on a warm solution ofpotassium platonitrite, however, yields large, brilliant prisms of t'hecompound Pt,2N02,K,T, + 2Hz0, described by Nilson, but this,ttlthough stable in warm solutions, is decomposed by nitrogen oxideso r potassium nitrite, with liberation of iodine and formation of theplatonitrite.The author has so far failed to obtain a compoundPt,4N0,,KzI,, analogous to the platichloronitrite and the platibromo-nitrite previously described.If, however, an excess of an alcoholic solution of iodine is added toa warm concentrated solution of the platonitrite, and the mixture isconcentrated as rapidly as possible at a gentle heat, potassium nityoso-plati-iodide, PtT3,N0,K212, separates on cooling in small, brilliant, blackcrystals, which remain unaltered a t 100".When heated in a currentof hydrogen, this compound yields water, ammonium iodide, iodine,hydrogen iodide, and a, residue of potassium iodide and platinum.If the liquid is slcwly concentrated (without ebullition) a differentcompound is obtained on cooling. T t is potassium platitetraiodonitrite,Pt14,2NOz,Kz, and separates in small, well-defined, black crystals witha greenish lustre. Like the preceding compound, it is only slightlysoluble in water, yielding a deep brown solution, and it undergoes nochange at 100". When heated in a current, of hydrogen, it yieldswater, ammonium iodide, iodine, and hydrogen iodide, with a residueof potassium iodide and pIatinum.When the mixture yielding these two salts is submitted t o pro-longed ebullition, the excess of iodine is expelled and potassiumplatoiodonitrite is formed.When bromine-water is added t o a solution of potassium platoiodo-nitrite until a11 the iodine is expelled, the brown liquid, when con-centrated in a dry vacuum, yields yellow, tabular crystals.The samecrystals are obtained by heating n solution of potassium platibromo-nitrite, Pt,4N0,,K(,Rr2, with alcohol at 80", aldehyde and other gasesbeing evolved. This new cornpound is potassium platobronionitrite,Pt,2NO2,K,Br, + H,O, and ib very soluble in water. At loo", it becomesanhydrous and biaiglit yellow ; at a higher temperature, it blackenswith evolution of nitrogen oxides, a residue containing potassiumbromide and platinum in the proportions Pt + 2KBr being left.When heated in hydrogen, it yields the same residue without anyformation of ammoni urn bromide, bromine, o r hydrogen bromide,By R.SCIINEIDER (J. pr. Ckem. [ 21, 44,507--512).-Potassium plati?ao~eleilostannate, K,Pt,SnSe,, is preparedby melting together 10 parts of platinum sponge, 6-8 parts of tin selen-ide, 30-40 parts of potassium carbonate, and 30-40 parts of seleniumin a covered porcelain crucible SO that the mass shall remain liquidfor 8-10 minutes after the frothing has ceased. The melt is extractedC. H. B.Two New Seleno-salts282 ABSTRACTS OF CHEMICAL PAPERS.with water, and the undissolved residue digested with concentratedpotassium hydroxide solution, whereby the selenostannate is left un-attacked. It forms small, well-defined, hexagonal tables which areleaden-grey by reflected light and have a strong metallic lustre ; inthin layers they are reddish-brown by transmitted light. In itsgeneral habitus it is very similar to the author's potassium platino-sulphostmnate (Ann. Phys. Chem., 138, 612). At the ordinary tem-perature, it is stable in air ; when heated in air, it loses all its selenium,the residue consisting of a mixture of platinum with potassium stan-nate. It is not attacked by water, ammonia, potash, 01- hydrochloricacid, either hot or cold.Sodium pZat inoseZenostnnnate, Na,Pt,SnSe,, cannot be obtained bymerely substituting sodium carbonate for potassium carbonate in theabove prescription. It is readily formed, however, when 10 parts ofplatinum sponge, 5-6 parts of tin selenide, 40 parts of potassiumcarbonate, 5-6 parts of sodium carbonate, and 40 parts of seleniumare fused together, and the melt treated as described above. It formsleaden-grey, microscopical, hexagonal lamin* with a brilliant, metalliclustre ; the remarks concerning the potassium salt apply also to this.A. G. B.Saline Compounds of the Lower Ruthenium Oxides with theHigher Oxides. By A. JOLY (C'ontpt. rend., 113,694-695.)-Whenthe products of the sudden decomposition of potassium perruthenateat 440" are kept for a long time at this temperature, interaction takesplace with formation of a black, crystalline compound of the composi-tion K20,6Ru205. Sodium permthenate at 440" loses oxygen andwater, and after treatment of the residue with water, which removessodium oxide and the orange sodium ruthenate, a black crystallinepowder is left with a, composition approximating very closely to thatfor Na20, 3Ru205. Barium ruthenate, IBaRuO4, at 440" loses oxygen, andyields a ruthenite, BaRuO,, different in properties from a mixture ofbarium monoxide and ruthenium dioxide. The author directs atten-tion t o the analogy between these compounds and the productsobtained by Rousseau from the permanganates, and the productK,O,Oe,O,, obtained by himself from potassium osmiamate.Action of Light on Ruthenium Peroxide. By A. JOLY (t%~~pt.rend., 113, 693--694).-When sealed tubes containing perfectly dryruthenium peroxide are kept in the dark, no alteration takes place,but on exposure to sunlight, the walls of the tube become coatedwith a pale brown layer, which gradually increases in thickness andeventually transmits only red light. Beyond this point, the peroxidebehind undergoes no further change. The brown deposit dissolvesat once in potassium hydroxide solution, forming a yellow liquidwithout any trace of green, and in hydrochloric acid with evolutionof chlorine and formation of a solution of ruthenium sesquichloride.It seems, therefore, t,hat when exposed to sunlight, ruthenium per-oxide is reduced to the trioxide, Ru03.C. H. B.C. H. B