14 ABSTRACTS OF CHEMICAL PAPERS, Inorganic Chemistry. Origin of A€mospheric Hydrogen. By ARMAND GAUTIER (Compt. rend., 1900, 131, 647-652).-The hydrogen in the atmo- sphere may be derived, not only from volcanic action and other well known sources, but also from the action of water on ancient granitic rocks a t 4emperatures considerably below a red heat. Granite from the interior of a fairly large block, powdered in an atmosphere of carbon dioxide and afterwards heated with dilute acids at 100’ or with water alone a t 280°, yields a considerable quantity of gas, consisting mainly of hydrogen, together with hydrogen sulphide, ammonia, carbon dioxide, nitrogen, and traces of unsaturated hydro- carbons, and sometimes traces of methane. A granite from Vire when heated with phosphoric acid diluted with its own volume of water, gave 1400 C.C.of gas per kilogram and 916 C.C. of this was hydrogen. With water only a t 280-300°, the proportion of gas liberated is smaller. The hydrogen and the ammonia that accom- panies it are probably due to the action of the water on nitrides and chiefly on iron nitrides, although possibly some of the hydrogen may be derived from its action on carbides. Higher Hydrogen Peroxides. By A. BACH (Bey., 1900, 33, 3111--3118).-This paper is largely a reply to criticisms by Arm- strong (Proc., 1900, 16, 134) and Baeyer and Villiger (Abstr., 1900, ii, 719) of a previous paper (Abstr., 1900, ii, 470). Although Caro’s acid when diluted does not reduce potassium permanganate, yet, undiluted, it rapidly decolorises a solution of the latter in concen- trated sulphuric acid (permanganic oxide) ; the titration is sharply defined, and the whole of the oxygen of the persulphate is evolved together with that due to the permanganate.The view that a catalytic decomposition of the Caro’s acid here occurs, appears to be excluded by the fact that the latter yields no oxygen either with manganous sulphate or with the liquid obtained after titrating Caro’s acid with permanganic anhydride. The existence in Caro’s reagent of a ‘‘ higher persulphuric acid,” (S0,H*O*O*S03H),, is therefore assumed, which on dilution yields the acid, SO,H*O*OH, incapable of reducing potassium permanganste. In the higher acid an ozonoid grouping is probable, since Cnro’s acid is formed on dissolving sodium dioxide in C. H.B.INORGANIC CHEMISTRY. 15 well cooled sulphuric acid, although, on warming, ozone is evolved (compare Bamberger, Abstr., 1900, ii, 536). W. A. D. Action of the Silent Electric Discharge on Sulphur Per- fluoride. By MARCELLIN P. E. BERTHELOT (Ann. Chirn. Phys., 1900,[vii], 21,205).-Sulphur perfluoride when submitted to the silent electric dis- charge remains unaffected, its stability under these conditions being comparable with that of the fluorides of boron and silicon ; the gas is not absorbed by bromine, thiophen, or acid solutions of cuprous chloride. G. T. M. Distribution of Sulphuric Acid in the Atmosphere. By HERMANN OST (Chem. Centr., 1900, ii, 733; from Chern. Ind., 23, 292-296).-In order t o compare the quantity of sulphuric acid in the atmosphere at various places, rectangular pieces of cotton wool or floconn6, 20 x 30 em., saturated with barium hydroxide solution, were attached to the branches of trees in such a way that the material was kept taut and these were then left exposed for a considerable period.On the Suutel mountains, far from any commercial industry and where scarcely any coal is burnt, only 0*0555-0*180 gram of sul- phur trioxide was found to have been absorbed in 153 days, whilst on the plain north of Hanover between Fuhrberg and Celle and 8 kilometres from the latter, 0*118-0.323 gram was absorbed in 193 days ; the maximum amount was obtained by exposing the material in an open position unprotected on every side and away from trees, and the minimum in forests. I n the neighbourhood of Hanover, 0*534-0.790 gram was absorbed in 175 days.Any sulphur pre- sent in mineral dust was excluded from the tests. E. W. W. Persulphates. By RODOLFO NAMIAS (L’Orosi, 1900,23,218-223). -Persulphates rapidly absorb moisture from the air with the forma- tion of the corresponding sulphate and sulphuric acid accompanied by the evolution of oxygen. Solutions of the persulphates are, however, very stable and may be employed in volumetric analysis for the estimation of substances having a reducing action. The best method of estimating the persulphates is to add excess of potassium iodide to a cold solution of the persulphate having a concentration not greater than 0.5 per cent. and after the liquid has remained for 10-12 hours, to estimate the free iodine by means of standard sodium thiosulphate solution.In neutral solution, the persulphates have the power of dissolving many metals such as zinc, iron, and aluminium, the two normal sulphates being formed; copper is not attacked by neutral or acid solutions of the persulphates but is rapidly dissolved in presence of ammonia, under which conditions zinc and iron are slowly acted on. Silver undergoes gradual solution by either the acid or ammoniacal solution. Chromium salts, in solutions either neutral or faintly acid with sulphuric acid, are quantitatively converted into chromic acid by excess of persulphate, a reaction useful in the estimation of chromium in steels, the chromium being first converted into chromate by means16 ABSTRACTS OF CHEMICAL PAPERS. of persulphate, the iron then separated by the addition of ammonia, and the chromate finally determined volumetrically. By mixing ammonium persulphate and ammonium hydroxide with the solution of a manganese salt, complete precipitation of a hydrated manganese dioxide takes place, the oxidation being much easier and more convenient than when bromine is employed, The oxidising action of persulphates in either acid or neutral solu- tion can be made use of in the bleaching of such materials as textile fabrics, wood, and ivory, whilst when mixed with sulphuric acid they bave the function of energetic depolarisers and may replace the nitric acid of the Bunsen cell. When heated with a persulphate at 70--80°, ethyl alcohol is rapidly converted into aldehyde, whilst many organic compounds capable of oxidation, especially closed chain compounds, give up hydrogen and yields sulphonic derivatives.Thus quinol reacts with the persulphates, forming an insoluble sulpho-compound, and diaminophenol, pphenylene- diamine, p-aminophenol, &a, give characteristic colour reactions. On adding a persulphate to a cold solution of aniline hydrochloride, R dark green precipitate is formed, which is insoluble in water or alcohol and when treated with sodium hydroxide or carbonate solution turns dark green but remains insoluble. T. H. P. Action of Hydrogen Peroxide on Thiosulphates. By ARNOLD NABL (Bey., 1900,33,3093-3095).- When hydrogen peroxide is added to a solution of sodium thiosulphate, a marked rise of temperature occurs and the solution becomes alkaline to litmus.This reaction is not due to liberation of sodium hydroxide, but to the formation of a new sulphur base, possibly in accordance with the equation 2Na,S,03 + H,O, = Na,SO, + Na,S03 + S,H*OH. The new base is not volatile with steam, is insoluble in alcohol, and its aqueous solution has n strongly alkaline reaction and a green fluorescence. It precipitates the hydroxides of silver, copper, iron, and uranium from their salts, reduces hot Fehling’s solution, and pre- cipitates the sulphide from stannic chloride. It yields a platinichloride which is quickly reduced, especially on heating. The base probably possesses analogies with trimethylsulphine hydroxide afid with hydroxy 1- amine. A. L. Nitrogen Iodide. By OTTO RUFF (Bey., 1900, 33, 3025-3029).-When iodine is added to liquid ammonia cooled to -SO0, a black powder (possibly N13:N13) is first formed, but eventually brownish-red plates with a greenish shimmer are obtained, which have the com position N13,12NH,. At - 35O t o - 40°, olive-green needles with the composition NI,,3NH3 are obtained ; under diminished pressure, these lose ammonia and yield ordinary nitrogen iodide, N13,NH, (compare Hugot, Abstr., 1900, ii, 274). When one atom of iodine is added to a mixture of 1 mol. of sodamide and liquid ammonia, cooled with liquid air, a violent reaction takes place; the product is a black compound, apparently of the com- position Na,N13, in which the nitrogen is quinquevalent. On further addition of a second atom of iodine, or of 1 mol. of ammonium iodide,INORGANIC CHEMISTRY.17 one or other of the compounds NI3,12NH, and N13,3NH, is formed, according to the temperature. The cooling mixtures required were obtained by adding alcohol slowly to liquid air contained in a Dewar’s beaker or in a beaker jacketed with cotton-wool, until the solidified alcohol redissolved to a syrupy liquid ; when the temperature had risen to - TOo, more liquid air and then alcohol were added as before, and the operation was repeated until enough of the mixture had been prepared. The sub- stances prepared were filtered through glass wool and asbestos in a filter tube surrounded with the cooling mixture, which was contained in a Dewar’s vessel. The washing was performed in the same tube, first with cooled liquid ammonia, then with cooled absolute ether, C.F. B. Formation of Nitric Acid during Combustions. By MARCELLIN P. E. BERTHELOT (Ann. Chirn. Phys., 1900, [ vii], 21, 145-201).- The experimental results embodied in these five memoirs have already been published (compare Abstr., 1899, ii, 648; 1900, ii, 475, 476, 538). G. T. M. By R. DUPOUY (Chem. Centr., 1900, ii, 8 3 6 ; from Bull. Trav. Soc. Pharrn. Bordeaux).-A sample of officinal arsenic triiodide, consisting of a crystalline powder, when treated with water left a yellow, insoluble residue of antimony oxyiodide, SbOI, this compound being formed by the action of water on antimony triiodide contained in the drug. The insoluble residue from a second sample consisted of antimony oxyiodide and free arsenic, whilst a third samplg left a residue of free arsenic alone.Another preparation containea an excess of free iodine and formed a clear brown solution which gradually became colourless. Arsenic triiodide should be tested by titrating with mercuric chloride, silver nitrate, or iodine solution ; in the first case, the compound Hg12,2HI is formed, whilst in the last two silver iodide and hydriodic acid are formed respectively. Arsenic Triiodide. E. W. W. Constitution of the Perborates. By EMIL J. CONSTAM and J. CORA BENNETT (Zed. anorg. Chem., 1900, 25, 265--269).-The mode of preparation of the perborates and the fact that the ammonium salt contains +H,O appear to indicate a doubled formula for these com- pounds. The authors, however, find that the difference between the electrical conductivities a t v = 1024 and B = 32 is 9 4 units, indicating a monobasic acid, so that the simple formula must be assumed.At- tempts to produce a perborate by the electrolysis of orthoborates were unsuccessful, and hence did not confirm Tanatar’s observations (Abstr., 1898, ii, 427). Reactions of Oxygen and Carbon Monoxide in the Presence of Alkalis. By MARCELLIN P. E. BERTHELOT (Arm. Chim. Phys., 1900, 1 vii], 21, 205-206).-After leaving a mixture of equal volumes of carbon monoxide and oxygen at the ordinary temperature for three months over a concentrated solution of potassium hydroxide, it was found that the former gas was absorbed by the alkali to form potass- ium formate, whilst the whole of the latter remained unchanged ; a L. M. J. VOL. LXXX. ii. 218 ABSTRACTS OF CHEMICAL PAPERS, similar result was attained in 14 hours by conducting the experiment at looo. When ammonia is substituted for potassium hydroxide either at the ordinary temperature or at looo, there is no formation of am- monium carbonate, and the absorption of carbon monoxide is very slight.G. T. M. Aotion of Sodammonium and Potassammonium on certain Metalloids. By CHARLES HUGIOT (Ann. China. Phys., 1900, [ vii 3, 21, 5-87).-This memoir contains an extensive bibliography and a detailed account of the preparation of compounds previously described (compare Abstr., 1896, ii, 20 ; 1898, ii, 573 ; 1899, ii, 151, 650, 747 ; 1900, ii, 14, 274). Hydration Processes. By PAUL ROHLAND (Ber., 1900, 33, 2831-2833. Compare Abstr., 1899, ii, 590).-It has been found that certain substances exert an influence on the hydration of Portland cement exactly similar to that which they have on the hydration of calcium oxide, Thus calcium chloride, aluminium chloride, and sodium carbonate accelerate, potassium dic hromat e, calcium chromate, boric acid, borax and potassium sulphate retard hydration, and sodium or lithium chloride has no action.The same generalisations do not hold good for the hydration of calcium sulphate ; in this process, sodium chloride, aluminium chloride, potassium dichromate, calcium chromate, and potassium sulphate accelerate, boric acid, bomx and sodium carbonate retard hydration, and calcium chloride is indifferent. The acceleration or retardation is roughly proportional to the con- centration of the solution of substance employed.If the velocity of hydration is great, then the added substance has a greater influence, either accelerating or retarding, than when the velocity is small. G. T. M. J. J. S. Constitution of Hydraulic Cements. By ORAZIO REBUFFAT (Gaxxetta, 1900, 30, ii, 177-182).-A correction of misquotations of the author’s views on this question. Action of Sea-Water on Pozzuolana Mortar. By ORAZIO REBUFFAT (Gaxzelta, 1900, 30, ii, 157--164).-The author’s experi- ments lead to the following conclusions. 1. Sea-water transforms the cement of mortar made from pozzuolana into a hydrated aluminium sili- cate whichcontainssrnall quantities of lime, magnesia, and the alkalis, and is perfectly stable towards the salts of sea-water. 2. Since pozzuolana mortar immersed in sea-water ultimately loses the whole of its lime, the employment of mixed mortars of cement and pozzuolana recom- mended by Michaelis cannot result in the fixation of the free lime of the cement.T. H. P. Artificial Pozzuolana. By ORAZIO REBUFFAT (Gaxxetta, 1900, 30, ii, 182-1 90). -Experiments on artificial pozzuolana lead to the following conclusions. 1. Aluminium silicate of the composition 2SiO,,A1,0,,2H,O, when heated a t 700--800°, loses water and assumes the nature of pozzuolana ; the residue has not the power to take up again the moisture given off. It is to the presence of this silicate that T. H. P.INORGANIC CHEMISTRY. 19 clay owes the pozzuolana properties it acquires on heating. 2. The silicate is not more easy of attack by hydrochloric acid after heating, although it becomes readily soluble in dilute alkali hydroxide solutions.3. The compound 2Si0,A120, combines with lime in the presence of water, giving a substance of the composition 2Si02,A103,3Ca0, 1 OH2?, the formation of which determines the setting of pozzuolana made from kaolin or from clay heated a t '700-800°. 4. This double silicate is readily decomposed by acids or by solutions of ammonium or mag- nesium salts, these salts removing the lime and leaving a hydrated aluminium silicate, 2Si0,,AI2O3,4H20, which in its chemical characters closely resembles certain varieties of the mineral halloysite. T. H. P. Slow Action of Hydrogen Bromide on Glass. By MARCELLIN P. E. BERTHELOT (Ann. Chim. Phys., 1900, [vii], 21, 206).-When glass tubes containing gaseous hydrogen bromide are examined after two years, a portion of the gas is found to have disappeared and drops of its aqueous solution are formed, the water being probably derived from the slow action of the gas on the alkaline silicates and sulphates which enter into the composition of the glass.Radio-active Lead and Radio-active Rare Earths. By KARL A. HOFMANN and EDUARD STRAUSS (Ber., 1900, 33, 3126-3131).- Pure lead sulphate was isolated from pitchblende, the uranium micas, broggerite, cleveite, and samarskite, which was radio-active although containing no trace of bismuth (polonium), barium (radium), titanium, thorium or uranium ; the sulphate was insoluble in dilute sulphuric acid, but easily soluble in basic ammonium tartrate. The derived lead chloride crystallised from hot water, and then showed diminished radio-activity, whilst the crystals from the mother liquors showed in- creased activity.Only traces of radio-active bismuth could be iso- lated from pitchblende, the uranium micas and samarskite. A mixture of cerium and yttrium oxalates, isolated from braggerite, cleveite and samarskite, and freed from thorium and uranium, was found to be radio-active, the oxides derived therefrom being especially so. Uranosouranic oxide,U308, from all the minerals already named, and also from euxenite, is always radio-active, and the same is true of the derived uranium oxalate, although on recrystallising the latter, the most solu- ble fractions are found to be inactive. Thoria from broggerite, cleve- ite, and samarskite is active, although from euxenite, which is closely allied to samarskite in composition it is obtained inactive.W. A. D. Formation of Mixed Crystals of Thallium Nitrate and Thallium Iodide. By C. VAN EYK (Proc. R. Akad. Wetensch. Amsterdam, 1900, 3, 98--101).-The relation between the composition of mixtures of the fused salts and that of the mixed crystals deposited on cooling has been examined. The melting point line is of Roozeboom's type I V (compare Abstr.: 1900, ii, 132), and rises immediately from the solidi- fying point of the nitrate, showing that mixed crystals are deposited from the melt. The mixed crystals deposited on solidification of mixtures with 0-9.9 molecular per cent. of the iodide are white ; those from mixtures with more iodide are red.The white mixed crystals contain GI. T. M. 2-220 ABSTRACTS OF CHEMICAL PAPERS. from 0-8 molecular per cent. of the iodide, and the red from 65-100 per cent. Mixtures containing from 18-65 molecular per cent. of the iodide solidify at 215.5" to a, conglomerate of the limiting mixed crystals, Thallium nitrate, which is rhombic a t the ordinary tempera- ture, is rhombohedra1 above l42', but this transition temperature is lowered by admixture with the iodide ; similarly, the transition point ( 1 6 9 O ) of the iodide is lowered by addition of nitrate. Alterations in the Chemical Properties of Elements Pro- duced by the Addition of Traces of Foreign Substances. By GUSTAVE LE BON (Compt. rend., 1900, 131, 706--708).-When mercury is alloyed with traces of magnesium, it acquires the property of decomposing water, and becomes rapidly oxidised by exposure to the air at the ordinary temperature, a perceptible alteration being caused by the addition of only 1/14,00Oth part of magnesium.Mag- nesium contaminated with mercury rapidly decomposes water in the cold. Aluminium foil, coated with a thin film of mercury, is quickly oxidised on exposure to air, and a strip of the metal placed verti- cally in a vessel containing water and mercury is continuously cor- roded at its lower end until completely disintegrated (compare Hunt and Steele, Abstr., 1899, ii, 33). Behamiour of Mixtures of Mercuric Iodide and Silver Iodide. By H. W. BAKHUIS ROOZEBOC~M (?roc. K. Akad. Wetensch. Amsterdam, 1900, 3, 84--86).-1nvestigation of fused mixtures of these salts shows that the melting point of mercuric iodide is lowered from 257' to 242' by an admixture of 14 molecular per cent.of silver iodide, whilst the melting point of silver iodide is lowered from 526' to'242' by an admixture of 86 molecular per cent. of mercuric iodide. The course of solidification shows that two kinds of mixed crystals are formed; on the mercuric iodide side, crystals of the rhombic mercuric iodide type with 0-4 molecular per cent. of silver iodide, on the other side crystals of the regular silver iodide type with 18-100 molecular per cent. of silver iodide. After solidification, mixtures containing from 4-18 molecular per cent. of silver iodide consist of a conglomerate of the two limiting mixed crystals. Those with 4 per cent.of silver iodide undergo a change at about 127', because the mercuric iodide is transformed from the rhombic into the tetra- gonal form, When the mixtures with 18-100 molecular per cent. of silver iodide are cooled below 157q the mixed crystals of the composition HgIq2AgI are suddenly transformed into a compound of the same composition, the colour changing at the same time from pink to red. This temperature (157') is comparable with the solidi- fying point of a chemical compound deposited from a liquid mixture. From mixed crystals containing less silver iodide than corresponds with the formula Hg12,2AgI, the formation of the latter compound occurs at temperatures from 157-118'; from mixed crystals con- taining more, at temperatures from 157-1 35'. The temperatures, 1 1 8 O and 135O, are analogous to eutectic points, the conglomerates formed consisting of Hg12,2AgI with either HgI, or AgI.The temperature of transformation of regular silver iodide into the hexa- J. C. P. G. T. M.INORGANIC CHEMISTRY. 21 gonal form is lowered by admixture of mercuric iodide from 157' to 135' ; the temperature of transformation of rhombic mercuric iodide into the tetragonal form is lowered by admixture of silver iodide from 127' to 118'. When the conglomerates of double salt with mercuric or silver iodide are further cooled to 45', the double salt changes from red to yellow, whether it is pure or mixed with mercuric or silver iodide. J. C. P. Separation of Cerite Earths from Monazite Sand. By RICHABD Jos. MEYER and E. MARCKWALD (Ber., 1900,33,3003-3013).-The crude material is the commercial #cerium oxalicum oxydul pur,' which is obtained as a residue after the extraction of thorium from monazite sand; it contains 25 per cent. of water and gave 32.5 per cent. of cerium oxalate, 21.5 per cent. of didymium oxalate, and 14 per cent. of lanthanum oxalate. The oxalates are dissolved in nitric acid and separated as ammonium double nitrates. The cerium is pre- cipitated as basic ceric sulphate from a solution of the mixed nitrates by adding ammonium persulphate and chalk (Witt and Theel, Abstr., 1900, ii, 403), and is purified by dissolving in boiled nitric acid (nitric oxide reduces the ceric salt), separating as ammonium double nitrate, and recrystallising from nitric acid. The didymium and lanthanum salts in the filtrate from the cerium precipitation are separated (from ytterbium salts, &c.) by adding potassium sulphate to the boiling solution until it no longer shows the didymium absorp- tion-spectrum ; the double sulphlttes which separate can be dissolved by boiling with nitric acid and pouring into boiling water, and are purified by precipitating as oxalates ; these are then dissolved in nitric acid, and the solution is boiled until free from acid.The separation of didymium from lanthanum is effected by adding mag- nesia until the solution no longer shows the didymium spectrum (Muthmann and Rolig, Abstr., 1898, ii, 518); by repeating this operation twice and precipitating twice with oxnlic acid and ammonia, the didymium is obtained nearly pure and free from all but a mere trace of lanthanum. The lanthanum is finally precipitated with ammonium oxalate in presence of ammonium chloride.Molecular Weight of Aluminium Compounds. By ELMER P. KOHLER (Amer. Chem. J., 1900, 24, 385--397).-The molecular weights of the aluminium derivatives of acetylacetone and ethyl acetoacetate were determined by the boiling point method, carbon disulphide being used as the solvent, and found to accord respectively with the formulae Al(C,H702)3 and Al(C6H,0,),. Determinations of the molecular weights of aluminium bromide and iodide in the same solvent agree with the formulze and A1216. The molecular weights of the compounds of aluminium bromide with benzenesulphonic chloride and nitrobenzene, and of the compound of aluminium chloride with nitrobenzene, are shown by the same method tocorrespond with theformulzeA12Br6,2CGH,=S02C1; A1,Br6,2C6H,*N02, and AI,C16,2C6H,*N02 J in each case, it appears that 1 mol.of the aluminium haloid com bines with 2 mols. of benzenesulphonic chloride or nitrobenzene to form 1 mol. of the product. It follows that the T. M. L.22 ABSTRACTS OF CHEMICAL PAPERS. addition of benzenesulphonic chloride or nitrobenzene to a solution of aluminium bromide of known boiling point should not affect the boiling point so long as the bromide is in excess; on the other hand, the addition of the bromide to a solution of benzenesulphonic chloride or nitrobenzene rJhould immediately lower the boiling point, and the depression should be proportional to the quantity of bromide added ; these conclusions were verified by experiment. A method is thus obtained for determining the changes which occur when an aluminium haloid is brought into contact with a given substance in an indifferent medium.The substance under investigation is added in successive portions to a solution of aluminium bromide of known con- centration and boiling point until it is present in excess, and a further quantity of aluminium bromide is then added ; the readings of the thermometer after each addition of material give an accurate account of what is taking place in the solution. By this means, it was found that the compounds of aluminium bromide with benzoyl chloride, benzo- phenone, and phosphorus oxychloride have respectively the composi- Acetophenone, benzenesulphonacetone, and ethyl benzoate form similar additive products with aluminium haloids ; benzene and naphthalene, however, slowly react to form insoluble compounds, whilst pyridine and quinoline yield crystalline additive compounds also insoluble in carbon disulphide.An experiment with aluminium bromide and p-dibromobenzene showed that the non-applicability of Friedel and Crafts' reaction to dihalogen compounds is not due to a combination of the dihalogen compound with the aluminium haloid. If one of the components of the additive compounds is present in large excess, dissociation occurs ; thus, cryoscopic determinations of the molecular weights of aluminium bromide and chloride in nitro- benzene solution point to the formuls AlBr, and AlC1,. Benzoyl chloride combines with aluminium bromide, in the absence of a sol- vent, to form a compound the molecular weight of which i8 shown by determination in carbon disulphide to correspond with that required for the formula A12Br6,2C6H,*COC1, whereas the determination in nitrobenzene gives an impossible value.The conclusion of Werner and Schmujlow (Abstr., 1898, ii, 214), that AlCl, is the only formula admissible for aluminium chloride is not justified, since they determined the molecular weight in pyridine, which combines with aluminium chloride to form a well crystallised compound. E. G. tion A1,Br6,2C6H,* coc1, AI2Br6,2COPh2, and AI,Br6,2POCl,. Cobalt Selenide. By HENRI FONZES-DIACON (Compt. Fend., 1900, 131,'704--705).-The selenide, CoSe, is obtained in an amorphous form by the action of selenium vapour on cobalt, and also by that of hydrogen selenide on heated cobalt oxide or chlorate.The sespuiselenide, Co,Se,, is produced by passing hydrogen selenide over cobalt chloride at moderately high temperatures ; at lower tem- peratures, these substances react with the formation of friable grey diselenide, corresponding with the formula CoSe, ; this substance roadily evolves selenium when heated. The selenide, Co,Se,, prepared by the interaction of hydrogenINORGANIC CHEMISTRY. 23 selenide, hydrogen chloride, and cobalt chloride at a dull red heat, forms violet-grey, regular octahedra, the specific gravity at 15' being 6.54. A subselenide, Co2Se, is obtained as a fused mass with a silvery lustre by reducing the preceding compounds in a current of hydrogen at a bright red heat; when this action is continued for some con- siderable time, the substance undergoes further reduction, but without yielding cobalt.These selenides are only slowly decomposed by hydrochloric acid and by hydrogen chloride at high temperatures. Chlorine and oxygen displace selenium from the heated selenides, the action in the latter case giving rise to the formation of selenium dioxide and cobalt oxide. The selenides readily dissolve in bromine water con- taining excess of bromine. Cobalt selenate, when reduced by hydrogen, yields either oxyselen- ides or a mixture of selenides and cobalt, according to the temperature employed. G. T. M. Ammoniacal Cobalt Arsenates. By 0. DUCRU (Compt. rend., 1900, 131,675-678).-The action of arsenic acid or an arsenate on solutions of cobalt salts containing ammonium rsalts and ammonia yields three insoluble ammoniacal cobaltous arsenates, Co3(As0,),,2NH,,6H,0, and Co3(As0,),,3NH,,5H,0, formed from the hydrated arsenate, Co3(As0,),,8H,0, by the partial displacement of water by an equal number of ammonia molecules.The composition of the product depends on the proportion of ammonia present, and is not affected by the proportion of ammonium salts. By 0. DUCRU (Compt. rend., 1900, 131, 702-704. Compare preceding abstract).-A solution containing a mixture of nickel and ammonium salts and free ammonia when treated with arsenic acid or a soluble arsenate, produces in the cold or on gently warming a gelatinous greenish-white precipitate which deepens in colour and becomes crystalline when the mixture is heated on the water-bath.The crystals are anisotropic and belong to the monoclinic system; they vary in composition according to the amount of ammonia present. The octahydrated nickel arsenate,. produced in the absence of am- monia, crystallises in needles and is identical in composition and crys- talline form with the mineral annabergite. When the solution contains 1 *38 per cent. of ammonia, a monoammonionickel ccrsenate is obtained ; with 6.9 per cent, of this reagent, a diccrnmonnio-salt is formed, the corresponding triammonio-compound being precipitated in more concentrated ammoniacal solutions, These salts behave in a char- acteristic manner when heated at 1 5 5 O , the composition of the com- pounds before and after heating being indicated in the following table : Co3(As04)2,NH3,7H,O, C.H. B. Ammoniacal Nickel Arsenates. Ni3( As0,),,8H20. Ni3( AsO4),1$R,O. Ni,(As0,),,7H20,NH3. Ni,(AsO,),lH,O,$NH,, Ni,( A s0,)2,6H,0,2NH3. Ni,( AsO,)~,~H,O, 3NHr N13( ASO,) ,&H,O,$NH,. Ni,(AsO,), $HI,O,$NH,24 ABSTRACTS OF CHEMICAL PAPERS. The anhydrous arsenate, Ni3(As0J2, is obtained in each case on heating the salt at dull red haat. Periodic Phenomena in the Dissolution of Chromium in Acids. By WILHELM OSTWALD (Zeit. physikal. Chem., 1900, 35, 204-256).-A continuation of the author's earlier paper (Abstr., 1900, ii, 730), in which the previous observations of periodicity by the dissolution of metals are fully discussed, and improvements on the apparatus employed by the author are described, An alloy of chrom- ium and aluminium was employed and the effect of various compounds in bringing about regular periodicity investigated.It was found that dextrin has a remarkably powerful effect in this direction, other colloidal carbohydrates being active to a smaller extent. The periodic time increases as the action progresses and a number of experi- ments testing the regularity of this increase are recorded. Various curves obtained are given in the paper and further investigations are promised. L. M. J. By JACOBUS M. VAN BEMMELEN [and G. M. RUTTEN] (Proc. K. Akad. Wetensch. Amsterdam, 3, 196--203).-The various solid phases of this system have been ex- amined. The normal salt, Bi20,,3N206,10H20, has not a true melting point, but decomposes at 75.5" into a liquid and the basic salt Bi,O,,N,O,,H,O.Two other normal salts with 4H20 and 3H20 respectively, have been discovered but not isolated ; their composition has been deduced by means of Schreinemakers' graphical method. The basic salt, Bi203,N20,,2H,0, is the first product of the action of cold water or dilute nitric acid (with less than 6 per cent. of N20,) on the decahydrate, or of cold water on n not too acid solution of his- muth nitrate. It forms very thin, crystalline plates, without definite shape and showing double refraction; these cannot be dried over sulphuric acid without decomposition, and in contact with the mother liquor they are converted into another basic salt. The basic salt, Bi20,,N,05,H20, is formed from Bj203,N205, 2H20 when the latter remains in contact with a solution containing more than 1 per cent.of N,O,; the time required for its formation diminishes as the percentage of N20, rises. The crystals are monoclinic, but when the same salt is formed by the decomposition of the decahydrate a t 75*5O, it has the form of thin, hexagonal prisms. The basic salt, 6Bi,0,,5N20,,9(8)H20, is formed at the ordinary temperature from Bi203,N20,,2H20, when the latter remains for some months in contact with a very dilute solution (less than 1 per cent. of N205 and less than 0.33 per cent. of Bi203) ; it is obtained also when the decahydrate is decomposed with water, and the solid salt which is formed is dissolved in much water. The crystals belong to the rhombic system; over sulphuric acid, they lose neither nitric acid nor water.The salt 2Bi203,N,0,,H20 is the final product of the action of boiling water on the normal salt. The compounds Bi203,2N20,,2H,0 and 10Bi,0,,9N20,,7H20 are also described. The basic salts, 5Bi20,, 4N20,, 9H,O, 4Bi203, 3N20,,9H,0, ~Bi,0,,3N,0,,8H20, and 5Bi,0,,3N,05,6H20, described by other lnvestigatoi s, are declared not to exist. Isotherms, giving the composition of the liquid phases which are in G. T. M, The System Bi203-N~05-H20.MINERALOGICAL CHEMISTRY, 25 equilibrium with the different solid phases have been totally or partially determined for the temperatures 20°, 30°, and 65'. The isotherms have been graphically represented in equilateral triangles, and from these tl figure in space has been drawn showing the con- ditions of equilibrium.The form of the triple lines in this system agrees with the form of those for the system HgO-SO,-HH,O (Hoit- sema, Abstr., 1896, ii, 15). By MARCELLIN P. E. BERTHELOT (Ann. Chim. Phys., 1900, [ vii], 21, 202-204).-Analyses of Egyptian gold coins of the earliest epochs indicate that the material employed in their pre- paration consisted of argentiferous alluvial gold ; only in later periods, from the time of Crcesus onwards, does it appear that the metal was refined, the desilverisation process being that described by Pliny. Owing to the rarity of minerals yielding gold free from silver, it is possible, by analyses of the golden articles found in Egyptian tombs, to state approximately the epoch of their manufacture. The gold leaf coverings of the mummies of the 6th and 12th dynasties contain 90-92 per cent. of gold, about 4 per cent. of silver, and 4-5 per cent. of organic matter, whereas those of the Persian epoch, ten centuries later, consist of nearly pure gold. Certain Properties of Alloys of the Gold-Copper Series. By SIR W. C. ROBERTS-AUSTEN and T. KIRKE ROSE (Proc. Roy. Xoc., 1900, 67, 105-1 12).-The authors have investigated the freez- i n g point curve for gold-copper mixtures. The freezing points of the pure metals are respectively, gold, 1063' ; copper, 1083O, and a mini- mum of 905' corresponding with an eutectic mixture was found for the alloy containing about 82 per cent. of gold, or 60 per cent. in atomic proportions. The atomic proportionsof the eutectic mixture agree closely with those for the eutectic mixture of silver-copper,&he freezing point curves also resembling one another closely, Microscopic examination was made of the various alloys, and these confirmed the existence of the eutectic alloy which makes its appearance even in the solidification of alloys containing only 27 per cent. of gold, as gold is com- paratively slightly soluble in copper, Copper is, however, more easily soluble in gold than in silver, so that the character of rich gold-copper alloys is not very marked, the crystals closely resembling those of J. C. P. Egyptian Gold. G. T. M. pure gold. L. 31, *J.