INORGANIC CHEMIS'I'RY. 15 Inorganic Chemistry. The Demity of Chlorine Gas. FREDERICK P. TREADWELL and W. A. K. CHRISTIE (Zeit. anorg. Chem. 1905 47 446-454).-The measurements were made by a method described by Bunsen. Two bulbs of equal capacity and of approximately equal weight were filled at conatant temperature with chlorine and air respectively and the difference of weight determined; the density of air being known the density of chlorine can then be calculated readily. The chlorine prepared from potassium dichromate and hydrochloric acid and care- fully purified still contained a small proportion of air for which a correction was applied. As a mean of three determinations at 20° and 730 mm. the value 2.488 (air = 1) was found and two determinations a t 10" and 725 mm.gave the value 2.489; according to Moissan and Binet du Jassoneix (Abstr. 1904 ii 114) the density at Oo and 760 mm is 2.490. G. S. Action of Hydrochloric Acid on Potassium Chlorate. A. KOLB (Zeit. angew. Chena. 1905 18 1693-1694).-Polemica1 (compare Ditz Abstr. 1905 ii 760; Kolb and Davidson ihicl. 59). H. M. D. The System Bromine and Iodine. P. C. E. MEERUM TERWOGT (Zed. anorg. Chem. 1905 47 203-243).-The freezing-point and boiling-point curves of mixtures of bromine and iodine have been determined and vapour pressure and specific gravity measurements have been carried out chiefly with the object of deciding whether a compound IC1 exists in the solid liquid and gaseous systems. Mixtures containing 50 atom. per cent. of each element solidify at constant temperature but with mixtures in any other proportion the temperature falls a few degrees from the commencement to the completion of solidification ; this indicates according to Hoozeboom's theory the existence of a compound IC1 which when the elements are present in other than equivalent proportions separates out in mixed crystals with bromine or iodine.The boiling-point curve was constructed by finding the initial boiling points of the various mixtures and the composition of the vapour at these temperatures and then plotting the composition of the liquid both against the initial boiling points and the boiling points of the evolved vapour on the same diagram. The two curves bend towards each other when the two elements are present in equivalent proportions which may indicate the presence of ICI in a highly dissociated state.The vapour pressure curve of the .system was investigated by a static method up t o 58 atom. per cent. iodine at 50*2" and up to 80 ut,om. per cent. at 92.8"; larger proportions could not be employed owing t o the formation of mixed crystals. The pressures were plotted The boiling point of pure iodine is 188-189O.16 ABSTRACTS OF CHEMICAL PAPERS. both against composition of liquid and of vapour ; the resulting curves deviated considerably from a straight line. An attempt was made according to a method suggested by van Laar (Zeit. php&-d. Chem. 1904 47 129) to deduce from the extent of this deviation whether IC1 exists in the liquid and vapour phases and the extent of its dissociation but the results are not very conclusive.The specific gravities of liquid and solid mixtures up to 58 atom. per cent. iodine were determined and plotted against the composition. The curves (except in the change from liquid to solid) are continuous but deviate from a straight line showing contraction; this may indicate formation of IC1 or may arise from a physical cause. The author concludes that 101 is present in the solid system; its existence in the liquid and vapour systems is probable but not satisfactorily proved. G. S. Iodic Acid. ERICH GROSCHUFF (Zeit. unorg. Chem. 1905 47 331-352).-Besides iodic acid and iodine pentoxide the existence of a crystalline substance of the formula I,O,,&H,O or HI,O has been proved by dehydration of iodic acid at different temperatures and by direct determination of the transition temperature.An amorphous modification of iodic acid is also described which differs from the ordinary acid in being very hygroscopic. No hydrates of the acid could be obtained. The complete solubility curve of iodic acid at different temperatures has been determined The cryohydric point lies at - 14'; the mixture contains 72.7 per cent. of the acid. The transition tem- perature of iodic acid to the compound HI,O is at 110' ; the transition from the latter t o the pentoxide takes place between 190' and 200'. The saturated solution boils at 11 1' under atmospheric pressure. The solution saturated at 18' contains about 295 grams of the acid to 100 grams of water and has a sp. gr. of about 2.48. Strong solutions of the acid are very viscous and it readily forms supersaturated solutions.The solubility of iodic acid in solutions of nitric acid of different strengths has also been determined. Freezing-point determinations in aqueous solution show that the apparent molecular weight varies greatly with the concentration. In dilute (1 per cent.) solutions the acid is present in single molecules which are highly ionised ; in concentrated solutions on the other hand it seems to be polymerised as suggested by Rosenheim and Liebknecht (compare Abstr. 1899 ii 743). This conclusion is confirmed by the results of electrical conductivity measurements ; the degree of ionisation as calculated from freezing-point determinations agrees better with that deduced from conductivity measurements when the presence of complex molecules in concentrated solution is assumed.G. S. Application of the Partial Liquefaction of Air with Reflux Action to the Complete Separation of Air into Pure Oxygen and Pure Nitrogen. GEORGES CLAUDE (Compt. rend. 1905 141 823-826).-In a previous paper (Abstr. 1904 ii 23) the author described a method for separating the oxygen and nitrogen of air byINORGANIC CHEMISTRY. 17 partial liquefaction ; the separation was not however complete as the final products contained only 92 per cent. of oxygen and 97-98 per cent. of nitrogen respectively. The present paper contains a descrip- tion of a modification of the method whereby the separation is carried to completion. The apparatus consists essentially of a reservoir of liquid oxygen surmounted by a rectifying column ; immersed in but not communicating with the liquid oxygen are two concentric coolers each consisting of a small reservoir communicating with a series of vertical tubes The air cooled and compressed enters the inner cooler where it undergoes partial liquefaction ; the liquid phase containing about 48 per cent.of oxygen falls back into the inner reservoir the gaseous phase rich in nitrogen is partially liquefied in passing through tbe vertical tubes of the outer cooler and nearly pure liquid nitrogen collects in the outer reservoir ; by allowing the contents of the two reservoirs to percolate through the rectifying column the liquid rich in nitrogen entering the column from the top whilst the liquid rich in oxygen enters at a slightly lower level the descending liquids effect a complete rectification of the ascending gases and pure liquid oxygen collects in the large reservoir whilst pure gaseous nitrogen escapes from the top of the rectifying column.M. A. W. The Liquefaction of Air by Expansion with Performance of External Work. GEORGES CLAUDE (Compt. rend. 1905 141 762-764).-A description of a modification of the apparatus employed in producing large quantities of liquid air under pressures of 23 to 40 atmospheres (compare Compt. rend. 1902 134 1568 ; 1903 136 1659 ; Abstr. 1904 ii 23 ; preceding abstract). AT. A. W. Molecular State of Water ; its Chemical Constitution and the Relative Value of the Two Valencies of the Oxygen Atom. LOUIS HENRY (BUZZ. Acad. Boy. BeZg. 1905 377-393).-A theo- retical paper in which the author quotes evidence in favour of the mole- cular association of water the chemical identity of the two hydrogen atoms of the molecule and the consequent equality of the two valencies of the oxygen atom (compare de Forcrand Abstr.1905 ii 696). M. A. W. Formation of Hydrogen Peroxide at High Temperatures. WALTHER NERNST (Zeit. Elektrochem. 1905 11 710-713).-The formation of hydrogen peroxide cannot be observed when a mixture of steam and oxygen is passed through a hot platinum or iridium tube and then rapidly cooled. As this is possibly due t o the great velocity with which the decomposition of hydrogen peroxide takes place experiments on the rate of dscomposition were made by passing a current of air containing a known quantity of hydrogen peroxide through a heated glass tube and then cooling it rapidly.The quantity of undecomposed hydrogen peroxide was estimated by adding a solution of titanium dioxide in concentrated sulphuric acid and measuring the absorption of light a t the blue end of the spectrum by means of t h e spectro-photometer. It appears that the reaction is bimolecular 2H,0,~-2H20 + O and that its velocity constants are very nearly the VOL. SC. ii. 218 ABSTRACTS OF CHEMICAL PAPERS. same as those of the decomposition of ozone (hbstr. 1904 ii 479). Hydrogen peroxide is not formed when electric sparks pass through a mixture of oxygen and water vapour (this is evidence in favour of the view t h a t the formation of ozone is due t o ultra-violet light) ; it is formed however when sparks are passed through liquid water the rate of cooling being then sufficiently great t o prevent its complete decom- position.T. E. Chemical Oxydases acting in the Presence of Hydrogen Peroxide. G. BAUDRAN (Gompt. rend. 1905 141 891-892. Compare Abstr. 1905 ii 407,632).-Chlorittes bromates iodates and the alkali hypochlorites hypobromites hypoiodites phosphates and sulphates give a yellowish-green colour and precipitate with a 1 per cent. solutionof guaiacol in the presence of hydrogen peroxide the colour becoming red on the addition of hydrochloric or sulphuric acid ; the salts of the aliphatic and aromatic acids give a similar reaction which is accelerated by the addition of a sinall quantity of mercuric iodide solution. ill. A. W. Preparation of Colloidal Solutions of Selenium and Sulphur by Electrical Pulverisation.ERTCH MULLER and ROMWALD NOWAKOWSKI (Ber. 1905 38 3779-3’781. Compare Muller and Lucas Abstr. 1905 ii 672; Gutbier Abstr. 1902 ii 652).-A colloidal solution of selenium is prepared by the electrical pulverisation in pure water of a cathode prepared by fusing a small piece of selenium on to platinum foil. This takes several hours with a n E.H.F. of 20 volts or is effected more quickly with a greater E.5I.B’. The hydrogen selenide formed with a n i7.M.R of 20 volts is oxidised by the anodic and atmospheric oxygen; but with a n E.B.F. of 220 volts hydrogen selenide is evolved. The colloidal solution is fiery yellowish-red in thick layers or a dirty yellow i n thin layers by transmitted or yellowish-red by reflected light and deposits selenium only slowly except on the addition of a n electrolyte.A milky-white colloidal solution of sulphur having a strong odour of hydrogen sulphide is formed by the cathodic pulverisation of a sulphur-pl*ttinum electrode in pure water with a n iY.;lL.F. of 220 volts. 0. Y. Preparation of Nitrogen from the Atmosphere. GEORGE A. HULETT (J. Amer. Chenz. SOC. 1905 27 1415-1418).-The following method is described for the preparation of nitrogen from the air. I n a piece of combustion tubing of not less than 2 cm. diameter is placed a roll of copper gituze 20 cm. long followed by a layer of copper oxide of the same length. The tube is heated in a combustion furnace both the copper and copper oxide being maintained at a red heat.S i r and hydrogen are led directly into the tube and allowed to mix just before reaching the hot copper. The resulting gases are passed through solution of potassium hydroxide in order t o rerriove the carbon dioxide. For details of the method the description and diagram in the original must be consdteil. K. (2.INORGANIC CHEMISTRY. 19 Atomic Weight of Nitrogen [and Silver]. PHILIPPE A GUYE (Arch. Sci. phys. nat. 1905 [iv] 20 351-380).-A detailed account and discussion of work previously abstracted (Abstr. 1904 ii 557; 1905 ii 442 506 702). The value 14.009 for the atomic weight of nitrogen appears to be completely confirmed and the effect of this alteration is discussed. Three atomic weights are regarded as correct to 1 in 10,000 namely C = 12.002 H = 14.009 and H = 1.0076. From these the author revises the value of atomic weight of silver from the ratio Ag I AgNO the value 107.882 is obtained from Ag I CH,*CO,Ag 107.886 and from Ag I C,H50,Ag 107*888 with mean 107.885.From the ratios of Ag I NH,CL and Ag 1 C1 employ- ing Dixon and Edgar's value for chlorine 35.463 (Abstr. 1905 ii 696) the numbers 107.871 and 107.895 are obtained ; other values are Ag I Ag,S 107W34; Ag I Ago 107,928; Ag 1 Ag,P04 107.888 with a general mean of 107.885 and the author considers it certain therefore that it does not exceed 107.89. The lowering of this atomic weight will affect numerous other values L. M. J. Oxidation of Ammonia by Alkali Persulphates in Alkaline Solution. R. KEMPF (Ber. 1905 38 3972-3974. Compare Marshall Abstr.1901 ii 156).-Ttie ammonium salts contained in commercial sodium persulphate are oxidised to nitric acid (some 88 per cent.) when the solution in sodium hydroxide is kept for four days at about the ordinary temperature. When the solution is boiled a considerable portion (some 40 per cent.) is oxidised to nitrogen. J. J. S. Hydroxylamine and its Salts. W. H. Ross (Proc. Trans. Nova Scotian Inst. Sci. 1905,11,95-114).-A number of hydroxylammonium salts have been prepared and examined. Hydroxylamrnonium sulphate melts a t 163" with decomposition ; when heated above the melting point it breaks up according to the equation 3(NH,*OH),S04= (NH4),S0 + 2S0 + 2N,O + 8H,O. Hydroxylammonium chloride melts a t 157"' and the molten substance decomposes according to the equation 30H*NH3C1 = NH,Cl+ 2HC1+ N + 3€€,0. Hydroxylammonium phosphate is only moderately soluble in cold water but readily forms supersaturated solutions.When heated in a vacuum i t decomposes according to the equation 2(NH3*OH)3P0 = 6NH,*OH + H20 + H4P20,. I n an opea vessel it decomposes at 148O into ammonia water phosphoric acid ammonium phosphate and nitrous oxide. Hydroxylammonium nitrate is a viscid colourless liquid which decomposes slowly at 80° more rapidly at higher temperatures with the formation of nitric acid water ammonium nitrate nitrous and nitric oxides. It was obtained by distilling its aqueous solution under a pressure of 20 mm. and drying the distillate over phosphoric oxide. The estimation of hydroxylamine by titration with iodine in presence of disodium phosphate (compare Haga Trans.1887,51 794) was found to be unsatisfactory. The amount of iodine used increases considerably 2-220 ABSTRACTS OF' CHEMICAL PAPERS. when the quantity of diciodiumiphosphate present is increased but is almost independent of the dilution The electrical conductivity of aqueous solutions of the free base and its salts was measured. The following numbers express the equivalent conductivity\ (p x lo7) a t l S o in terms of mercury v being the volume per gram-molecule of salt. v = 2. 5. 10. 20. 50. 100. 200. 500. 1000. 2000. 5000. 10,000 Hydroxylamine ... - 0.4 0.5 0.7 0.9 1.2 1.5 1.9 2'2 - - - OH'NH3CI ......... 73.3 80.5 E6.6 90.9 95.6 96.9 101'5 104'6 107'6 109'9 111% 112'4 (VH3'0HhSOd ... - 63'6 71.6 80'5 9 2 0 97.7 103.0 109'9 113'0 115'9 118'9 120'3 OH'NE3N03 ..,... - - 106'7 112'6 119.1 122.3 124'7 127-1 129'8 131'8 134.0 138.5 (1L'Hs00R)3P04 ...- 1'3's 22.0 23.6 24.7 25.5 26'0 26.4 26'6 26'8 27'0 27'1 The nitrate has a greater conductivity than the chloride which is not what would beexpected from a comparison with the salts of the alkali metals. In all cams the conductivity of the salt solutions was found to increase with time especially with the more dilute solutions. This is due to the decomposition of the salts under the influence of the platinum black of the electrodes ; with polished electrodes the effect is considerably smaller. The base itself is oxidised even more rapidly than the salts. These phenomena necessitated special care in the determination of the conductivity data and the values for solutions of the base are only regarded as approximate.H. M. D. Density of Nitric Oxide; Atomic Weight of Nitrogen. PHILIPPE A. GUYE and CH. DAVILA (Compt. rend. 1905 141 826-82s. Compare Abstr. 1904 ii 475 557 812; 1905 ii 442 506).-Owing to the discrepancies between the values obtained for the density of nitric oxide (air = 1 ) 1 *041 (Thomson) 1.0888 (Bbrard) 1.094 (Davy) 1.188'7 (Kirwan) and 1.0388 (Lecluc) the authors have redetermined the constant using nitric oxide prepared by three distinct methods (1) decomposition of sodium nitrate or nitrite by mercury in sulphuric acid solution (2) reduction of nitric acid or sodium nitrite by ferrous sulphate or (3) decomposition of a dilute solution of sodium nitrite by sulphuric acid; it was purified by pass- ing it through concentrated sulphuric acid and subsequent repeated fractional liquefaction by means of liquid air.The mean of fourteen determinations of the weight of a litre of nitric oxide at N.T.P. is 1.3402 grams the limiting values being 1.3408 and 1.3398 respec- tively and is identical with the value recently found by Gray (Trans. 1905 87 1601) who prepared his nitric oxide by the reduction of sodium nitrite by potassium ferrocyanide and acetic acid and purified it by treatment with potassium hydroxide and final liquefaction and fractional distillation. The atomic weight of nitrogen calculated from the ratio of the densities of the gases NO/O lies between 14.010 and 14.006 (0= 1 6 ) . M. A. W. Nitrosyl Fluoride. OTTO RUFF and KURT STAUBER (Zeit.anorg. Chem. 1905 47 190-202. Compare Moissan Abstr. 1905 ii 518). -With the object of obtaining a compound of nitrogen and fluorine,INORGANIC CHEMISTRY. 21 Gore's experiments (this Journal 1869 22 391 393) with nitric and hydrofluoric acids and with nitric acid sulphuric acid and sodium fluoride were repeated and the action of nitrosylsulphuric acid on sodium fluoride was also investigated in each case with negative results. It was found however that by the action of nitrosyl chloride on silver fluoride a gaseous substance was formed which proved to have the formula NOF. The following method for preparing nitrosyl fluoride was found to give satisfactory results. A flask containing nitrosyl chloride kept at -So was attached to one end of a long platinum tube containing silver fluoride and kept at 200-250' ; the other end of the tube was connected with a small platinum flask provided with two stopcocks and cooled by liquid air.The nitrosyl chloride distilled slowly over the silver fluoride and the nitrosyl fluoride along with unaltered chloride was condensed in the platinum flask and purified by fractional distillation. Nitrosyl fluoride is a colourless gas condensable to a colourless liquid which boils a t - 5 6 O and solidifies on further cooling to a colourless mass melting at - 134O. I n the cold silicon boron red phosphorus and sodium take fire in the gas and arsenic and antimony are also immediately acted on fluorides being formed in each case and nitric oxide evolved. Lead aluminium bismuth and copper are only slowly acted on whilst iodine sulphur and carbon are not affected even on heating.Nitrosyl fluoride is decomposed by water with formation of nitrous and hydrofluoric acids; it gives brown fumes in moist air. Both in physical and chemical properties it closely resembles the nitryl fluoride NO,F isolated by Moissan but differs from the latter in its behaviour towards water and iodine. The formula NOF is confirmed by analysis and vapour density determinations. G. S. Preparation of Phosphorus Di-iodide. HOWARD W. DOUGHTY (J. Amer. Chern. Soc. 1905 27 1444-1445).-A mixture of 50 grams of iodine and 4 grams of red phosphorus in a 250 C.C. flask is heated with a free flame until completely melted. When the product has cooled to 60° 2.5 grams of yellow phosphorus are added in small pieces.By this method phosphorus di-iodide can be safely and rapidly prepared. E. G. Production of Carbon (Lampblack and Graphite) from Acetylene and Metallic Carbides. ADOLPH FRANK (Zeit. nngew. Chem. 1905 18 1733-1735).-When acetylene under a pressure of five or six atmospheres is exploded the charcoal (lampblack) obtained is contaminated with oily products resulting from the simultaneous partial condensation of the acetylene. If the acetylene is mixed with carbon monoxide or carbon dioxide in proportions given by the equations C,H + CO = 3C -I- H,O and 2C,H + CO = 5C + 2H,O these condensation products are not formed. The pressure of the mixture of acetylene and carbon monoxide before explosion should be at least six atmospheres ; with this initial pressure valne the maximum pressure developed during the explosion amounts to 40-50 atmospheres.On22 ABSTRACTS OF CHEMlCAL PAPERS. account of the high temperature developed the reaction is not complete but 85 per cent. of the theoretical quantity of carbon is obtained. The charcoal so produced has a high specific gravity-1.93 to 2.0- and its electrical conducting power is considerably greater than that of other forms of charcoal. It is very finely divided very black ancl has a high covering power. When calcium strontium or barium carbide is heated in a current of carbon monoxide or dioxide carbon separates in the form of graphite according to the equation CaC + CO = CaO + 3C. Graphite is also obtained when the carbides are heated in a current of chlorine phosphorus or arsenic but in these cases the yield is considerably less than the theoretical. The graphite obtained from calcium carbide and carbon monoxide leaves only about one per cent.of ash on combustion is very hard and has a high sp. gr. (2-2.05). It is eminently suit- able for electrotechnical and electrochemical applications. H. M. D. Attempt to Liquefy Helium. KAm OLSZEWSKI (Bull. Acad. Sci. Cracow 1905 407-411. Compare Abstr. 1897 ii 31 ; Dewar Abstr. 1901 ii 597 ; Travers Senter and Jaquerod Abstr. 1903 ii 9).--Helium obtained by heating thorianite with potassium hydro- gen sulphate contained as impurity only nitrogen which was removed completely by two coolings with liquid hydrogen. The helium so purified was cooled to -252.5' by boiling hydrogen and then to - 259' by hydrogen solidifying under 50 mm.pressure and subjected to a pressure of 180 atmospheres which was reduced slowly or suddenly to 1 atmosphere when the gas showed no signs of iique- faction or the deposit of any solid the temperature being lowered to - 271.3". The boiling point of helium is therefore below - 273.O or -+ 2' absolute. G. Y. Coloration of t h e Halogen Salts of the Alkalis and Alkaline Earths. LOTHAR WOHLER and H. KASARNOWSKI (Zeit. anorg. Chem. 1905 47 353-370. Compare Abstr. 1901 ii 166; Giesel Abstr. 1897 ii 170; Borchers and Stockem,Abstr. 1903,ii 19).-Many naturally-occurring halogen compounds such as rock-salt and fluor- spar are coloured in various ways and similar colours can be produced by exposing the colourless salts to the vapours of metals or to cathode rays.The origin of these colours has not been satisfactorily elucidated. With the object of throwing light on this question the authors heated natural blue rock-salt and colourless pieces of the same sub- stance in a stream of oxygen and found that both specimens contained about 0.016 per cent. of carbon and 0.008 per cent. of hydrogen indicating the presence of traces of organic matter. For comparison artificially coloured specimens of several halogen salts have been prepared by heating the crystals in the rapour of different metals and it was found that under these conditions the halogen salts of the alkali metals became deeply coloured in fifteen minutes and the colour did not deepen on further heating.Excess of metal could not be detected in these salts by chemical analysis so that the amount1NORGANIC CHEMISTRY 23 present must be very small. On the other hand a specimen of fluor- spar blue by transmitted and green by reflected light prepared by heating colourless crystals of the salt in calcium vapour was found on analysis to contain 2.4 per cent. of the metal in excess. Attempts to prepare coloured salts by electrolysis were not very successful ; only those parts of the fused mass which ware kept fairly cool were found to be coloured. All the coloured salts referred t o lose their colour when heated but this change takes place at a much lower temperature in the case of the naturally coloured salts. Further the latter are neutral in reaction and the colour is unaffected by water whereas the artificially coloured salts are alkaline and become decolorised on treatment with water.It seems likely that the coloration arises from different causes in the two cases ; with artificially coloured salts it is probably due to traces of metal or subchloride in solid solution in the halogen salt in the other cases to traces of organic matter. The colour does not depend on the amount of foreign substance present but chiefly on its mode of occurrence as in coloured glasses. The coloration produced by the vapours of metals and by cathode rays probably arises from the same cause. G. S. Sodium Hyposulphite. 111. ARTHUR BINZ and W. SONDAG ( B e y 1905 38 3830-3834. Compare Abstr. 1904 i 964; 1905 ii 521).-A method is described of estimating in the same solution hyposulphite sulphate thiosulphate and siilphite.By this means it is possible to follow the changes which take place on mixing the hypo- sulphite with thiosulphate. The products are sulphite and sulphide. Na,S,O + Na2S,0 + 4NaOH = 3Na,S03 + Na,S + 2H,O; in the actual experiment 1.00 mol. Ka,S,O with 0.99 mol. Na,S203 gave 2.92 mol. Na,SO and 1-00 mol. Na,S. T. M. L. Lithium Alumino-silicates Z. WEYBERG (Centr. Bin. 1905 646-655. Compare Abstr. 1905 ii 89 98 262).-By fusing kaolin (H,Al,Si,O,,H,O) with potassium chloride Gorgeu (188'7) obtained the product K,Al,Si,O?. This compound is however only obtained in the crystallised condition when an alkali carbonate is added to the mixture or when the free bases and silica are used instead of kaolin.On the other hand the more readily decomposed lithium salts when fused with kaolin give well-crystallised products. Thus by fusing kaolin with lithium chloride or lithium carbonate a white crystalline powder is obtained the crystals being orthorhombic and having the composition Li,A1,Si20,,. With lithium sulphate the cryetallised (orthorhornbic) product has the composition Li2A1,Si,0,. With lithium bromide a lithium brorno-sodalite 7Li,Al,Si20,,2LiBr crystal- lising in rhombic dodecahedra was obtained. The kaolin used in the experiments had the following composition SO,. A1,0,. H,O. Total. From South Russia ... 45.76 39.87 14.44 100.07 Meissen ......,.. 49.41 37.06 13-53 100.00 L J. 8.24 ABSTRACTS OF CHEMICAL PAPERS. Lithium Chromates. FRANS A.H. SCHREINEMAKERS (C'hewz. Centlr. 1905 ii 1486; from Ckenz. Weekblad 2 633-639. Compare Abstr. 1905 ii 8 18 820).-The only substances which can separate in a solid form from a system which consists of lithium oxide chromic acid and water at 30' are lithium hydroxide LiOH,H,O chromate Li2Cr,0,.2H,0 dichromate Li2Cr,07,2H,0 and chromic acid CrO,. Higher chromates than the dichromate cannot exist at 30°. A saturated solution of lithium hydroxide contains 7.09 per cent. ; 100 parts of water dissolve 99.94 and 130.4 parts of lithium chromate Li2Cr0 and 11 thium dichromate Li2Cr207 respectively. The solubility of the chromates of ammonium potassium sodium and lithium increases in the order given whilst for the dichromates the order of solubility is potassium ammonium lithium sodium.E. W. W. Decomposition of Ammonium Sulphate by Sulphuric Acid in the Presence of Platinum. MARCEL DEL~PINE (Compt. rend. 1905 141 886-889).-The low results obtained when the nitrogen of platinichlorides is estimated by the Kjeldahl method (compare Abstr. 18Y5 ii 290 and van Dam Abstr. 1896 ii 218) are due to the decomposition of the ammonium sulphate with loss of nitrogen for if spl.ngy platinurn or platinum foil is boiled with sulphuric acid containing ammonium sulphzlte sulphur dioxide is formed and nitrogen is evolved the quantity increasing with the duration of heating the temperature and the addition of potassium sulphate. It is probable that the platinum acts as a catalyst the sulphate being first formed and then decomposed by the ammonium sulphate according to the eqaations ( 1 ) 4H2S0 + Pt = Pt(SO,) + 250 + 4H20 ; (2) SPt(SO,) + 2(NH4).?S04 = 2N + 3Pt + 8H2S0 for the solution obtained by boiling platinum in sulphuric acid deposits platinum when heated with ammonium sulphate and further when the platinum in the above experiment is replaced by gold or iridium which is not attacked by sulphuric acid there is no loss of nitrogen.M. A. W. Separation of Constituents of Alloys. A. BOCK (Chem. Zeit. 1905 as 1199-1201).-The introduction of a third metal into a molten alloy of two metals causes a partial separation of the two constituents and a consequent loss of homogeneity in the resulting mass on solidification. The extent of the disturbance is less in thin layers of the alloy than in large blocks.The Pattinson desilverising process is attributed to the influence of the iron of the containing vessels on the molten alloy of lead and silver. The author also attributes to the iron of the mould in which they are cast the segregation observed with certain silver-copper and gold-silver alloys. P. H. Oxidations with Silver Peroxide. I. Oxidation of Oxalic Acid. R. KEMPF (Ber. 1905 38 3963-3966).-Silver peroxide is formed when solutions of silver nitrate and potassium persulphate are mixed (Marshall Trans. 1891 59 771). Such a mixhure is capable of oxidising numerous carbon compounds. Quinone yields maleic acidINORGANIC CHEMISTRY. 25 and carbon dioxide together with formic acid and carbon monoxide. Oxalic acid can be quantitatively oxidised t o carbon dioxide.A simple lecture experiment may be conducted as follows 5 C.C. of a 10 per cent. silver nitrate solution are added to a mixture of 20 grams of ammonium persulphate in 100 C.C. of 10 per cent. sulphuric acid and 100 C.C. of N-oxalic acid. A slight precipitate of silver sulphate is formed the solution becomes warm and the evolution of gas is completed in a few minutes. The active oxygen of the persulphate may be estimated by mixing with excess of standard oxalic acid solution adding dilute sulphuric acid containing a little silver sulphate and titrating the excess of oxalic acid with permanganate J. J. S. Oxidations with Silver Peroxide. 11. Formation of Nitric Acid from Ammonium Persulphate. R KEMPF (Ber. 1905 38 3966-3971. Compare Marshall Abstr. 1901 ii 156).-When silver sulphate is added to a solution of ammonium persulphate in dilute sulphuric acid a brown precipitate of silver peroxide is formed but this disappears in the course of several days pTactically no oxygen is evolved the persulphate is destroyed and nitric acid is produced.The reaction proceeds but slowly for example 0.63 gram of ammonia (as ammonium salt) required forty-eight hours. I n the absence of silver salts persulphates are not able to bring about the oxidation which must be due to silver peroxide and not to ozone Caro's acid or hydrogen peroxide. The oxidation is not quantitative. J. J. S. Electrolytic Calcium. JOSEPH H. GOODWIN (J. Amer. Chena. Suc. 1905 27 1403-1415).-A method is described for the pre- paration of calcium by the electrolysis of fused calcium chloride.A hollow cylindrical vessel of Acheson graphite is employed as the anode and an iron rod as the cathode. With a current of 163 amperes about 17.7 volts are required the current efficiency being 29.1 per cent. For details of the apparatus reference must be made to the description and diagrams in the original paper. The bright metal obtained by this method was found to contain 98.0 per cent. of calcium had a sp. gr. 1.5446 at 29.2" a specific electrical resistance of 3.43 microhms per C.C. a t Oo and a tensile strength of 612 kilograms per sq. cm. Calcium is harder than sodium lead or tin nearly as hard as aluminium but softer than zinc cadmium or magnesium E. G . Composition of Metallic Calcium. B. LARSEN (Chem. Centr. 1905 ii 1466 ; from Mitt Teechn.Gewerb.-Nus. Wien [ii] 15 244-246).-A sample of metallic calcium from the Electrochemical Works in Bitterfeld has been found to contain 99.64 per cent. of calcium 0.2 of iron 0.09 of manganese 0.06 of silica and 0.11 of calcium carbide. The metal was broken on an anvil and the man- ganese may possibly have been derived from the tool which was made of manganese steel. The pieces which weighed from 6-8 grams were quickly immersed in naphtha so that the fresh surfaces scarcely became tarnished. After removing the naphtha from the sample for analysis by means of ether the metal was dried in a stream of dry66 ABSTRACTS OF CHEMICAL PAPERS. air and finally dissolved in hydrochloric acid cooled by ice. The quantity of calcium carbide mas determined by weighing the copper acetylide formed from the acetylene which was liberated.E. W. ?V. Tarugi’s View of the Formation and Composition of Bleaching Powder. HUGO DITZ (Zeit. angew. Chem. 1905 18 1690-1 693. Compare Tarugi Abstr. 1905 ii 32).-Tarugi’s view that bleaching powder is a substance of the formula CaO,CI,,H,O corresponding with 44.09 per cent. of active chlorine is inconsistent with the analyses of different samples of bleaching powder. On the assumption of this formula the sum of the percentage numbers for the various constituents of three samples of bleaching powder examined amounts approximately to 110. The excess of this number over 100 corresponds almost exactly with 1 atom of oxygen. B. M. D. Revision of the Atomic Weight of Strontium. 11. Analysis Of Strontium Chloride.THEODORE W. RICHARDS (%it. nnorg. Chem. 1905 47 145-150).-1n a previous investigation (Abstr. 1895 ii 314) the value Sr=87*663 was found from the analysis of strontium bromide. In the present paper the analysis of the chloride by a similar method is described ; the ratio SrClz 2Ag. was found by titrating the carefully purified salt against silver dissolved in nitric acid the end-point being determined by means of the nephelometer (compare Abstr. 1904 ii 287). When the value C1=35*473 found by Richards and Wells is taken as the basis of calculation the value Sr = 87.661 is obtained as the mean of four concordant experiments in satisfactory agreement with the result obtained with the bromide whilst there is a considerable difference when the former atomic weight of chlorine is employed.The new value for chlorine is thus confirmed. The mean of the two series Sr = 87.662 (0 = 16) may be taken as the most probable value. G. S. Barium Oxide and its Hydrates the Preparation of a New Hydrate. 0. BAUER (Zeit. anorg. Ckem. 1905 47 401-420).-The compound Ba(OH),,SH,O melts at 7 8 O and begins to boil at 103’; on continued heating the temperature rises steadily to logo a t which point it remains constant for some time and crystals of a hydrate Ba(OH)2,3H,0 not previously described separate out. The new hydrate occurs in clear lustrous rhombic crystals ; chemically i t behaves in all respects like the octahydrate. Corresponding strontium and calcium compounds could not be obtained. G. S. Alloys of Cadmium and Zinc containing Lead.FRANZ NOVAK (Zeit. unorq. Chew,. 1905 47 421-445. Compare Ericson- Auren and Palmaer Abstr. 1902 ii 64; Brunner Abstr. 1904 ii 315 ; 1905 ii 235).-The investigation was undertaken with the object of determining how the properties of zinc on which its employ- ment for process blocks depend are modified by small amounts ofiNORGANIC CHEMISTRY. 27 cadmium and lead. The metal was used in the form of thin sheets rolled at 120O. The precautions taken to secure uniformity of com- position in the alloys are described. The rate of solution in acids was determined by measuring t h e hydrogen evolved and by finding the loss of weight of the plates; it was found in agreement with Spring (Abstr. 1888 WO) that the reaction velocity only attains its maximum value after some time (induction period).The rate at which zinc containing 0-1-0.8 per cent. of lead is dissolved by &/lo hydrochloric acid is lessened by the addition of cadmium up to 1 per cent. and the induction period is lengthened. On the other hand the rate of solution in 5 per cent. nitric acid is somewhat increased by addition of cadmium. Quantities of lead up to 2 per cent. (the limit of solubility of this metal in zinc) have no effect on the reaction velocity larger proportions lessen it considerably. These results find a simple explanation on the theory that the rate of solution is greatly influenced by local electric currents (compare Ericson-Auren and Palmaer Zoc. cit. ). When zinc and its cadmium alloys are heated at 270" for some time and then cooled they are attacked by acids rather more rapidly than samples which have been rolled at 120' ; this behaviour is connected with a change in the crystalline structure.per cent. of cadmium increases the hardness and breaking stress of zinc but when more than + per cent. of the former metal is present the opposite effect is produced. Photomicrographs of several of the alloys referred to are given in the paper. G. S. Addition of Determination of Atomic Weights of Rare Earths. OTTO BRILL (Zed. anorg. Chem. 1905 47 464-476. Compare Nernst and Riesenfeld Abstr. 1903 ii 57 l).-Further experinients on the rare earths have been made with Nernst's micro-balance by the method already described (Zoc. cit.). A method largely used in determining the atomic weights of rare earths is to convert the oxide into the corresponding normal sulphate or vice versd. According to Kruss (Abstr.1893 ii 283) acid sulphates are completely converted into normal sulphates by heating for some time a t 350'. To test this method oxides of several rare earths were evaporated with sulphuric acid small quantities of the resulting acid sulphates placed in the scale pan of the micro-balance and heated in an electric furnace for ten or fifteen minutes at temperatures increasing by intervals of 50° the weight being determined after each heating. A t 350° the weight did not remain constant so that as Wild has also pointed out (Abstr. 1904 ii 173) the temperature given by Kruss for conversion into the normal sulphate is too low. At 450° the conversion to normal sulphate was complete at 700" this began to decompose into basic sulphate the change being complete at 900'; at 1150° the latter was completely transformed into oxide. The basic sulphates of yttrium ytterbium erbium lanthanum and samarium thus prepared for the first time have the general formula M20,,S0 and occur in white needles From the order in which they decompose on heating the relative basic character of the oxides can be28 ABSTRACTS OF CHEMICAL PAPERS.determined ; from the weakest to the strongest base the order is Yb Er Y Sa La It is shown that atomic weights can readily be determined by heating a t 480' to form the normal sulphate then at 1150" for some minutes to form the oxide; the method is particularly useful for rapidly determining the atomic weights of fractions. G.S. Victorium and the Ultra-violet Phosphorescence of Gadolinium. GEORGES URBAIN (Compt. rend. 1905,141,954-958). -I€ the phosphorescent spectrum obtained by Crookes with different specimens of gadolinium (Abstr. 1899 ii 751; 1905 ii 250) is correctly ascribed to a new element victorium the intensity of the spectrum should be diminished by diluting the gadolinium. The author finds however that a mixture of 2.8 parts of gadolinium oxide and 97.2 parts of lime gives a phosphorescent spectrum in which the bands are very intense and similar results are obtained with mixtures of lime containing only 2/10,000 parts of gadolinium oxide ; it follows therefore that the phosphorescent spectrum given by gadolinium is a characteristic of that element itself.M. A. W. Preparation of Crystalline Crocoite and Wulfenite by the Action of Atmospheric Carbon Dioxide on Alkaline Solutions of the Lead Salts. G. CESBRO (Bull. Acad. my. Belg. 1905 327).-Many insoluble lead salts dissolve in a solution of potassium hydroxide and these solutions on slow neutralisation by atmospheric carbon dioxide deposit the lead salts in a crystalline form. Such a solution of lead chromate on exposure to the air for four months,. gave a red crystalline deposit which had the appearance and properties of crocoite. A solution of lead molybdate became turbid and blue but after some time gave a very dense grey granular powder with a metallic lustre formed of microscopic transparent tetragonal tablets F. G. C. S. Distillation of Copper.HENRI MOISSAN (Compt. mad. 1905 141 853-857. Compare Abstr. 1893 ii 507 ; 1904 ii 617; FBry Abstr. 1903 ii 293 ; KraEt and Bergfeld Abstr. 1905 ii 144).- Copper is readily distilled when heated in an electric furnace ; with a current of 300 amperes and 110 volts 77.7 per cent. of the metal is volatilised in eight minutes whilst with a current of 800 amperes and 110 volts several kilograms of copper can be distilled in a few minutes. The metallic distillate condenses on a copper cooler placed in the furnace in the form of iridescent filaments 5 to 7 mm. thick having a red to yellow colour a sp. gr. 8.16 (Kalhbaum Roth and Siedler found 8.932 Abstr. 1902 ii 259) and containing 99-76 per cent. of copper the impurities consisting of lime and graphite. The fused ingot of copper remaining in the crucible exhibits the phenomenon of '' spitting," on cooling evolving large quantities of gas and when quite cold is covered with a layer of graphite having a sp.gr. of 2-12 and consisting of 96.25 per cent. of carbon 3.36 per cent. of ash and 0.21 per ,cent. of hydrogen. A microscopicINORGANIC CHEMISTRY. 29 examination of a polished section of the ingot shows that it contains a large number of spherical cavities lined with brilliant crystals of graphite. M. A. W. Copper and Arsenic. K. FRIEDRICH (Metalltwgie 1905 2 477-.195).-The author has studied the alloys of copper and arsenic containing 0-44 per cent. of arsenic. Alloys richer in arsenic than this cannot be obtained under atmospheric pressure and the higher alloys lose most of their arsenic when heated for any length of time at 300'.The freezing-point curve shows a well-marked maximum at 70.88 per cent. As and 830° corresponding with the compound Cu3As and the existence of a compound Cu5Asq dissociating below its melting point is also indicated. Eutectic points occur at 78.5 60.6 and 52.5 per cent. of arsenic respectively the corresponding eutectic temperatures being 683' 71 lo and 603O. A transformation of unknown character was observed at 307'. The microscopic examination of sections etched with copper ammonium chloride or electrolytically in dilute nitric acid confirms the evidence of the freez- ing - poin t curve. The formation of Cu,As and CU~AS occurring in nature as algodonite and whitneyite respectively is not observed in melting together copper and arsenic and the existence of the compound Cu,As is also not confirmed.C. H. D. Commercial Guprosilicon. PAUL LEBEAU (Compt. qlend. 1905 141 889-891. Compare Vigouroux Abstr. 1896 ii 362 and Chalmot Abstr. 1896 ii 362)-A microscopical examination of the polished surface of commercial cuprosilicon shows that it contains large and small crystals of silicon embedded in a matrix of cuprosilicon ; when the uncombined silicon is removed by treatment with 10 per cent. sodium carbonate solution and the cuprosilicon dissolved in nitric acid a residue is left consisting of steel-grey crystals of ferrosilicon FeSi and these three constituents are present in the following proportions free silicon 51.11 cuprosilicon 49.97 ferrosilicon 3.49 per cent.The cuprosilicon contains 10.36 to 10.88 per cent. of silicon and has therefore the composition Cu,Si and not Cu,Si as stated by Vigouroux (Abstr. 1896 ii 362) and de Chalmot (Abstr. 1896 ii 362; 1897 ii 262). M. A. W. A Modification of Mercurous Chloride. JULIUS MEYER (Zeit. anorg. Chem. 1905 47 399-400).-When solutions of mercuric chloride and lithium sulphite are mixed the heavy white precipitate of ordinary calomel filtered off and the clear solution heated to 70-80" a new modification of mercurous chloride separates in lustrous scales. It has a sp. gr. 4.5-5 as compared with 6.5-7 for the ordinary salt ; it changes into the latter on sublimation. It shows no difference of potential when measured against ordinary mercurous chloride in a galvanic cell so that it does not seem t o be an allotropic modification of the latter. G.S.30 ABSTRACTS OF CHEMICAL PAPERS. Action of Silicon on Pure Aluminium and its Action on Impure Aluminium. Silicoaluminides. EMILE VIGOUROUX (Compt. rend. 1905 141 951-953).-Pure silicon and aluminium do not combine to form an aluminium silicide but when heated in the presence of the oxide or a salt of another metal a compound con- taining silicon aluminium and the metal or the silicoalurninide of the metal is formed. The silicoaluminides of iron nickel cobalt chromium manganese molybdenum tungsten vanadium uranium and titanium have been prepared; they are crystalline hard dense brittle and present a metallic appearance. A few are attacked by dilute acids with formation of silica but most of them resist the action of even concentrated acids with the exception of hydrofluoric acid and they are not attacked by solutions of alkali hydroxides. M.A. W. Physical and Chemical Properties of Slags. THOMAS TURNER (J. SOC. Chem. Id. 1905 24 1142-1147).-The classification of slags their physical properties the crystallisation melting points temperatures of formation and fluidity are discussed. Experiments have been undertaken to ascertain as far as possible the constitution of a typical slag A quantity of blast furnace slag weighing six tons was allowed to cool slowly and samples were taken from six positions chosen so as to represent the relative order of solidification. No evidence of the separation of calcium or aluminium silicates during the process of cooling is afforded by the analytical data for the various samples.The most noteworthy fact elucidated is that the last fluid portion which had a creamy-white colonr contained a distinctly smaller proportion of sulphur than the other samples. Apart from this elimination of sulphur from the liquid portion of the slowly cooling slag there is but little evidence of any alterationin composition. The proportion of aluminium is however highest in the parts which con- tain most sulphur and these parts have the most marked blue colour. This observation supports the view that the blua colour of the blast furnace slag is due to some form of aluminium oxysulphide. With regard to the relationship of magnesium and sulphur the sample richest in sulphur contained the smallest proportion of magnesium and vice versd.Cooling curves of the different samples were taken but these give no evidence of eutectics and do not permit of a distinction between the fluid plastic and solid states. Solidification extends over a wide temperature interval but very different values were obtained by different observers for the temperatures at which the process begins and ends. H. M. D. Volatility of Indium Oxide. JULIUS MEYER (Zed. anorg. Chern. 1905 47 281-286).-The determination of the atomic weight of indium by conversion into the oxide is inaccurate owing to the volatility of the latter at high temperatures. It is shown in the present paper however that the loss of weight on heating the oxide for several hours in a specially constructed platinum crucible at 1100" is inappreciable.I n the conversion of indium nitrate into oxide byINORGANIC CHEMISTRY. ignition a constant weight cannot be attained even a t l l O O o ; this seeins to be due to the obstinate retention of oxides of nitrogen by the oxide. The difficulty can be obviated to some extent by repeatedly treating the oxide with water and evaporating or by converting the nitrate into hydroxide with ammonia and subsequently igniting. G. S. Interaction of Hydrochloric Acid and Potassium Perman- ganate in the Presence of Various Inorganic Salts. JAMES BROWK (Zed. anory. Clzenz. 1905 47 314-330. Compare Abstr. 1905 ii 166).-Standard solutions of hydrochloric acid and of potassium permanganate were heated alone and with known amounts of various inorganic salts for a definite time at 50° a definite amount of standard oxalic acid was then added and standard permanganate run in until the red coloration was permanent.I n a former paper i t was shown that the greater amount of permanganate apparently reduced in the presence of ferric chloride is not due to the catalytic action of the salt as Wagner had supposed (Abstr. 1899 ii 275) but to oxidation of part of the oxalic acid by the chlorine evolved during the action. Corre- sponding experiments with the chlorides of cadmium chromium gold and platinum show n greater apparent reduction of permanganate in presence of these salts in agreement with Wagner but when the chlorine is removed by a current of air before adding the oxalic acid the apparent amount of reduction is not affected by the presence of cadmium and gold salts but is increased by chromium and platinum salts.The author draws the conclusion that the chlorides of the three metals first mentioned have no catalytic effect on the action the difference in tbe case of chromium being due to secondary reactions (partial oxidation to chromate) whereas platinic chloride acts as a catalyst. Since the observed differences in the majority of cases are due to the disturbing action of the evolved chlorine the latter must have less ef-fect in the presence of salts; no very satisfactory explanation of this has been found. Direct experiments show that after digestion there is less chlorine present in solutions containing salts than in the others but as no appreciable amount escapes into the air the suggestion is made that in the fqrnier case it is used up in some secondary reaction.The retarding effect of manganous salts on the reaction in question is probably due to the production of higher oxides of manganese in larger amount so that more oxalic acid is required for complete reduction. G. S. GEORGES CHARPY (Compt. rend. 1905 141 948-951).-Roozeboom's equilibrium diagrams for iron-carbon systems (Abstr. 1900 ii 728 ; 1904 ii 717) were based on the cooling curves for iron-carbon mixtures obtained by Roberts-Austen by allowing the fused mixtures to cool rapidly ; the author finds however that if in the case of mixtures containing from 2 to 4 per cent. of carbon the cooling is gradual the two main branches of the cooling curve represent respectively the separation of mixed crystals and of cementite; at 1150" (not 1050" as is usually stated) the eutectic product consisting of mixed crystals and cementite Equilibrium Diagram of Iron-carbon Alloys.32 ABSTRACTS OF CHEMICAL PAPERS solidifies and below that temperature cementite separates from the solid solution; whilst if the cooling is rapid the eutectic product which solidifies at 11 60-1 165" consists of mixed crystals and graphite mixed crystals or graphite separating above that temperature and graphite separating from the solid solution below that temperature.M. A. W. Compounds of Iron with Silicon. W. GUERTLER and GUSTAV TAMMANN (Zeit. anorg. Chern. 1905,47 163-1 79).-The authors have studied the freezing-point curve of alloys of iron and silicon and have confirmed the conclusions thus arrived at by microscopic observations and in some cases by chemical analysis.The curve falls rapidly from the freezing point of iron to a point at which 33 atom. per cent. of silicon is present; along this part of the curve mixed crystals of Fe and Fe,Si separate out in the form of cubes. The alloy containing 33 atom. per cent. silicon solidifies at a constant temperature like a pure xriatal and may be made up of saturated mixed crystals or may be a definite chemical compound of the formula Fe,Si; the latter view is the more probable since several previous investigators have isolated a compound of this formula by treating alloys poor in silicon with acids but no other case is known in which a chemical compound is the last member of a series of mixed crystals.Beyond this point the curve falls slightly until at 34.7 atom. per cent. silicon it reaches the eutectic point of mixtures of Fe,Si and a new compound FeSi at a temperature of 1235'. This descending part of the curve between 33-3 and 34.7 atom. per cent. silicon cannot be determined directly owing to the small. temperature difference (about So); its existence is deduced from the microscopic observation that the alloy containing 34.7 atom. per cent. silicon is a eutectic mixture. Beyond 34.7 per cent. silicon the curve rises to a well-defined maximum a t a temperature of 1443O when 50 atom. per cent of each element is present corresponding with the compound FeSi; along this part of the curve crystals of FeSi surrounded by masses of the eutectic mixture of Fe,Si and FeSi separate out. Beyond 50 atom.per cent. silicon the curve falls until a t 76 atom. per cent the eutectic point of FeSi and silicon is reached at a temperature of 1245"; it then rises to the melting point of silicon. Between 76 and 100 atom. per cent. silicon long crystals of this element surrounded by the eutectic Si,FeSi separate out. Hot potassium hydroxide rapidly attacks silicon slowly acts on FeSi and has practically no action on Fe,Si and iron. Hydrochloric acid on the other hand acts rapidly on iron slowly on Fe,Si and has scarcely any effect on FeSi and Si. These facts can be made use of in analysing the alloys. The hardness of these substances decreases with increasing percentage of iron. Alloys containing up to 47.5 atom.per cent silicon are magnetic whilst those containing more than 50 atom. per cent. are non-magnetic. The transition temperature of U- into /3-iron does not seem to be appreciably altered by the presence of silicon. G. S. Action of Silicon Chloride on Iron. EMILE VIGOUROUX (Cowpt rend. 1905 141 828-830).-Silicon tetrachloride is completelyINORGANIC CHEMISTRY. 33 decomposed by iron below a red heat or a t 1100' with the formation of the iron silicide Fe,Si and ferrous chloride according to the equation SiC1 + 4Fe = Fe,Si + 2FeC1,. Iron silicide Fe,Si is decomposed by chlorine with incandescence by concentrated hydrochloric acid or by aqua regia; it dissolves completely in dilute hydrofluoric acid and is not attacked by acetic or nitric acid (compare Moissan Abstr.1896 ii 173 ; Lebeau Abstr. 1900 ii 729 ; 1901 ii 317 ; and Jouve Abstr. 1902 ii 595). Attempts to prepare the higher silicide FeSi obtained by Frdmy (Em. Chirn. Art. Per.) by the action of silicon tetrachloride on iron were unsuccessful. M. A. W. The Composition of Colloidal Ferric Hydroxychloride in Relation to the Concentration of Hydrochloric Acid in the Containing Fluid. G. MALFITANO (Compt. rend. 1905 141 660-662 680-683. Compare Abstr. 1905 ii 459).-The colloid is obtained by heating a 0.5 per cent. solution of ferric chloride in an auto- clave at 100-1 15' for fifteen to thirty minutes; after dialysis and filtra- tion through collodion a solid colloidal residue is obtained. A washed preparation contained 1-57' per cent.Fe and 0.16 per cent. C1 and after dialysis a 0.01 per cent. solution of HCl. It is concluded that the amount of HCl separating depends on the quantity and composition of the colloid and is proportional t o the quantity of water and increases with rise of temperature hence the system H(Fe,O,H,),Cl tends to exist in equilibrium with the solution in which it is suspended. By the action of free hydrochloric acid on colloids of the type H(.Fe20,H6),Cl the value of n diminishes H( Fe,O,H,)Cl being the limiting value ; for instance 25 C.C. of a preparation containing (in millionths of a gram atom) 1654H(Fe20,H,),.,C1 + 76HC1 are filtered and heated with 200 C.C. containing 3464HC1 and after cooling and filtration analysis indicates 1828H( Fe,O,H,),.,,Cl + 3290HC1. From numerous similar experiments the author concludes that on increasing the concentration of HC1 the colloid approaches more nearly t o H(Fe,O,H,)Cl.When the colloid is once precipitated it may lose HCl even in a strongly acid-containing fluid. Two hundred C.C. of the colloidal solution were diluted to 1 litre containing 5 grams HC1 and heated at 130' in a Jena flask; on analysis of the total contents of the flask 13H(Pe,O,H,),,Cl+ 1016 Fe,Cl + 28456HC1 was obtained. When once the HC1 which was combined with the Fe,O,H reunites to form a molecule of HCI the stability of the compound disappears and in spite of the concentration of the solution the colloid loses HC1 and the ferric hydroxide is converted int.0 chloride. F. G. C. S. Vanadium Sesquisulphate. ARTHUR ST~HLER and HEINZ WIRTHWEIN (Ber. 1905 38 3978-3980.Compare Abstr. 1905 ii 595).-Varzadiursz hydrogen sulphute Vd,(SO,),,H,SO,,l 2H,O ob- tained by methods similar to those used for the corresponding titanium salt more especially by t h e electrolytic reduction of blue vanadyl sulphate is a green finely-crystalline silky powder and is VOL. XC. ii. 334 ABSTRACTS OF CHEMICAL PAPERS. insoluble in alcohol ether acetic acid or 60 per cent. sulphuric acid. I t s aqueous solution has a green colour. The ammonium salt Vd,(SO,),,( NH,),SO,,l 2H20 and rubidium salt have been prepared. Vanadium sesquasulphate Vd,(SO,) is obtained when the green salt is heated slowly at 180' in an atmosphere of carbon dioxide; it forms a microcrystalline powder and dissolves in hydrochloric acid t o a yellowish-brown or in dilute sulphuric acid to a green solution.J. J. S . Modifications of Antimony. ALFRED STOCK and WERNER SIEBERT (Bey. 1905 38 3837-3844).-R new black amorphous modification of antimony can be prepared either by the action of oxygen or of air on liquid antimony hydride a t about -40° or by rapid cooling of the vapour of ordinary metallic antimony. Special apparatus for these operations is described the last method consisting in distilling the metal from a porcelain tube heated electrically hy a platinum wire to about 3OOo in a good vacuum on to a surface cooled by liquid air. The black antimony has a sp. gr. 5.3 it is chemically more active than the grey form and oxidises and often catches fire in the air a t the ordinary temperature.At 400° it is converted instantly into the stable grey form and suffers the same change slowly on boiling with water ; it may be freed from antimony oxide by washing with hydrochloric acid. This has been so far prepared only in minute quantities either by the action of oxygen on antimony hydride a t - 90° or better by the interaction of antimony hgdride and chlorine dissolved in liquid ethane at - 100' in red light. The yellow modification is very unstable; it blackens above -90" even in the dark. On shaking yellow antimony with carbon disulphide at temperatures below - 50° a suspension of insoluble antimony in the colloidal state is obtained which above - 50" is in a few seconds changed into the black modification. Cohen's a-antimony is probably identical with the black antimony described above.E. F. A. Antimony also exists in a third-a yelIow modification. Melting Point of Gold and Expansion of Gases at High Temperatures. 111. ADRIEN JAQUEROD and F. Louis PERROT (Arch. Sci. phys. nat. 1905 [iv] 20 506-529).-The mean value for the melting point of gold in the authors' determinations was 1067.4' rt 1%". Previous determinations vary from 1061' to 1200" But the three most trustworthy values are 1064.3 (Holborn and Day) 1065.6" (D. Berthelot) and the authors' giving a mean of 1065.S0 or allowing a little more weight to the authors' value 1066' which they consider to be the vaIue which should be accepted for the melting point. From the experiments the relative values of the coefficients of expan- sion of the various gases employed between 0' and 1067.4' may be obtained and taking the value 0.0036643 for nitrogen the other values' are air 0,0036643 ; carbon monoxide 0.0036638 ; oxygen 0*0036652 ; carbon dioxide (1) 0*0036756 ; (2) 0.0036713 (see Abstr. 1905 ii 506 627). L. M. J.MINERALOGICAL CHEMISTRY. 35 Precipitation of Metallic #old. P. E. JAMESON (J. Amer. Chem. Xoc. 1905 27 1444).-By the following method gold can be rapidly precipitated in a convenient state for collection. A stick of potassium nitrite weighing about 5 grams is placed in a solution of 1 gram of gold chloride i n 30 C.C. of water and about 3 C.C. of con- centrated sulphuric acid are added immediately. When the reaction has ceased another piece of potassium nitrite of the same size is added and the sohiion is stirred until a clear pale blue solution is obtained. The precipitated gold collects a t the bottom of the vessel in the form of dark brown nodules which can be easily separated by decantation. E. G. Hydrated Palladium Dioxide. ITALO BELLUCCI (Zeit. a~zorg. Chem 1905 47 287-288).-A reply to Wohler and Konig’s criticism (Abstr. 1905 ii 722) regarding the author’s statements as to the composition of hydrated palladium dioxide (Gaxxefta 1905 35 i 343). G. S.