INORGANIC CHEMISTRY. Inorganic Chemistry. ii. 117 Condition of Dissolved Iodine. PERCY WAENTIG (Zeitsch. physikal. Chem. 1909 68 513-571. Compare Beckmann Abstr. 1907 ii 340).-The investigation of solutions of iodine in a large number of solvents has led to the conclusion that in all the solutions iodine is partly combined with the solvent according to the reversible equation SoI f So + I (So = solvent) ; for the violet solutions the spectra of which approximate to that of iodine vapour the amount of combination is much less than for the brown solutions. These conclusions are mainly based on spectroscopic observations and in particular the effect of change of temperature on the spectra has been fully investigated. On heating the violet solutions the absorption band becomes displaced towards the red end of the spectrum in other words towards that of iodine vapour and is displaced in the opposite direction on cooling.The spectra of many of the brown solutions tend to become permanently altered on heating a result which speaks in favour of considerable association between iodine and solvent. The view that there is only a difference in degree between violet and brown solutions is further supported by the fact that brown solutions in thiophen and in sulphur dioxide become violet on heating and regain the original colour on cooling. The conclusion to be drawn from the displacement of the equilibrium with temperature that the heat of formation of the compound must be positive is supported by observations on the effect of temperature on the solubility of iodine and on its heat of solution in different solvents.Although the heat of solution is negative in all the solvents examined except pyridine the heat absorption is much less for brown than for violet solutions. Moreover cryoscopic investigations with iodine and a ‘ 6 solvent ” dissolved together in an indifferent solvent show that the VOL. XCVIII. ii. 9ii. 118 ABSTRACTS OF CHEMICAL PAPERS. depression is so much the smaller the greater the degree of combin- ation according to the above considerations (compare Hildebrand and Glascock Abstr. 1909 ii 225). A cornpound with pyridine of the formula PyI has been isolated. The influence of dilution on the absorption spectra has been measured for two kinds of light and it was found that contrary to the requirements of the law of mass-action the ratio of the absorption for both kinds of light increases with dilution.This is probably con- nected with ionisation of the additive compound as many of the solu- tions have a considerable molecular conductivity which increases with the dilution. The observations on the nature of the absorption bands and of their displacement with temperature are satisfactorily accounted for on the assumption that there is a maximum of absorption for the additive compound in the ultraviolet and that the absorption due to the compound is more or less affected by the absorption band due to free iodine. The partial vapour pressures of the components in boiling solutions of iodine in ether carbon disulphide chloroform carbon tetrachloride and benzene have also been determined and the results considered from the point of view due to Dolezalek (Abstr.1909 ii 22) support the above conclusions. Even at the boiling point there is considerable association between iodine and solvent in the brown solutions. The molecular freezing-point depression for carbon tetrachloride is 299. G. S. Production of Ozone by Ultraviolet Light. EDMOND VAN AUBEL (Compt. Tend. 1910 150 96-9S).-The following correction should be made in the previous paper (this vol. ii 2s). I n the experiments where water was used to absorb the ozone produced the resulting solution contains hydrogen peroxide the presence of which was detected by its action on a photographic plate. The presence of hydrogen peroxide in the water proves the formation of ozone in the air by the action of the ultraviolet light.T. S. P. Dissociation Isotherms of Sulphur between 3003 and 850". GERHARD PREUNER and W. SCHUPP (Zeitsch. physikal. Chem. 1909 68 129-156).-By means of the quartz-glass manometer described by Abegg and Johnston (Abstr. 1908 ii 157) the authors have determined the density of sulphur vapour at nine temperatures between 300° and 850' and within wide limits of pressure. The results cannot be reconciled with the assumption that only S and Sa molecules are present but indicate that above 30 mm. pressure only S S and S molecules are present. A t pressures lower than 30 mm. it is possible that S molecules are also contained in the vapour. From the displacement of the different equilibria with temperature it is calculated that in the reaction 35,=4S 29,000 calories are absorbed and in the other reaction S = 3S2,64,000 calories are absorbed.Similarly the change of gaseous S8 .to 45 absorbs 95,000 calories. From the most trustworthy results on the variation of the vapour pressure of sulphur with the temperature i t is calculated t h a t the heatINORGANIC CHEMISTRY. ii. 119 of vaporisation 8S[solid] -+ S&gaseous) is 20,000 calories hence for the reaction 8S,[solid] -+ 4S2(gaseous) 115,000 calories are absorbed. Otherwise expressed it requires about 115,000/4 = 28,800 calories t o transform 64 grams of solid sulphur into gaseous S,. On the assumption that S is identical with Sp the modification of sulphur insoluble in carbon disulphide the proportion of Sp in saturated sulphur vapour is calculated from the density results at different temperatures and compared with the values determined directly by Gal (Abstr.1893 ii 455) and by Kruyt (Abstr. 1908 ii 1028). The authors’ results are in excellent agreement with those of Gal but not with those of Kruyt. G. S. Action of Hydrogen on Sulphur or Selenium in Presence of Another Element. HENRI P~LABON (.I Cl&~t,. Phps. 1909 7 447-463).-Excess of pure sulphur or of sulphur containing less than 2/3 atomic proportion of arsenic when heated a t 610’ in hydrogen gives a gaseous mixture containing about 98% of hydrogen sulphide. When the proportion of arsenic is increased until the sulphur is saturated at approximately the composition SAs the amount of hydrogen sulphide decreases regularly to 78%.Further addition of arsenic which remains as a separate phase does not decrease the amount of hydrogen sulphide produced. Tho uaiphase solutions of arsenic in sulphur give a greater proportion of hydrogen sulphide the less the pressufe of hydrogen but mixtures containing a free arsenic phase always give 78% of hydrogen sulphide. The results obtained with mixtures of arsenic and selenium are analogous but the proportions of hydrogen selenide are in all cases much lower. Addition of selenium or tellurium to mixtures of sulphur and arsenic promotes the formation of hydrogen sulphide but to a less extent than an equivalent amount of sulphur itself. Selenium produces a similar effect on mixtures of sulphur and antimony b u t tellurium is here without influence.The addition of antimony to sulphur makes no change at first in the amount of hydrogen sulphide (98%) but when 43-90 atoms of antimony are present per 100 of sulphur the liquid consists of two phases antimony sulphide saturated with antimony and antimony saturated with its sulphide. Whilst these two phases are present the proportion of hydrogen sulphide produced is. about 60%. With larger proportions of antimony the hydrogen sulphide rapidly decreases towards zero. The phenomena with mixtures of selenium with antimony bismuth tin and thallium are similar to those obtained with selenium and arsenic. Silver and copper on the other hand which form stable selenides solid at the temperature employed entirely prevent the formation of hydrogen selenide when their proportions exceed Ag,Se and Cu,Se.R. J. C. Preparation of Colloidal Solutions of Selenium. ALFREDO POCHETTINO (Atti R. Accad. Lincei 1909 [v] 18 ii 544-551).-1n preparing colloidal solutions of selenium by the method of Muller 9-2ii. 120 ABSTRACTS OF CHEMICAL PAPERS. and Nowakowski (Abstr. 1906 ii lS) the author finds that at the same time as a red coloration forms in the liquid near the cathode the current intensity exhibits a continuous increase which begins immediately the circuit is closed. If the liquid in the cell is shaken the current intensity falls sharply but rises to the original value as soon as the shaking ceases and then continues t o rise to a certain limiting value which may be as much as ten times the initial current ; if the liquid is kept continually agitated a longer time is required for the current to increase to the limiting value.By this method if the duration of the current is sufficient to allow of the liberation from the cathode of more than about 0.24 gram of selenium per litre of water present a deposit of red selenium begins to form at the bottom of the cell. If a solution prepared in this way is filtered and subjected to an E.MF. of 40 volts between polished platinum electrodes the current increases gradually from about 40 milli-amperes at the beginninglto about 60 milli-amperes after two hours; if the current is interrupted for some time and then re-applied the initial value is again about 40 milli-amperes and the same gradual increase takes place. The increase of current seems t o be due to a diminution in the resistance of the liquid present in the cell.When a dilute solution of selenions anhydride is electrolysed with platinum electrodes the phenomena observed vary with the E. M.F. employed. With 3 volts a red deposit forms on the cathode and the current gradually falls; with 17 volts the deposit forms almost instantaneously and increases in thickness the current remaining sensibly constant ; whilst with 48 volts the current gradually increases whilst the liquid appears red in reflected and blue in transmitted light a colloidal solution of selenium and a black cathodic deposit of selenium being formed. The internal friction of such colloidal selenium solutions is sensibly identical with that of the distilled water from which they are prepared the same being the case with solutions prepared by Gutbier’s method (Abstr.1902 ii 652). T. H. P. Chemical Reactions in Gases Submitted to very High Pressures ; Decomposition of Nitric Oxide ; Formation of Nitrosyl Chloride. E. BRINER and A. WROCZYNSKI (Compt. rend. 1909 149 1372-1374. Compare Abstr. 1909 ii 557).-Nitrosyl chloride is formed when a mixture of nitric oxide and hydrogen chloride is submitted to a pressure of 300 atmospheres. When nitric oxide is allowed to remain a t high pressure in a sealed tube i t is colourless at first but after a day appears bluish-green. If the tube contains a large quantity of the gas drops of a blue liquid appear. The reaction is supposed to be represented by 6N0 = 2N20 + N ; the synthesis of nitrosyl chloride would then be explained by the action N203 + 2HCl= 2NOCl+ H20.No action occurs unless the nitric oxide is above a certain minimum pressure. I n experiments in which the pressure in the tubes was gradually increased the blue gaseous phase was first noticed a t 28 atmospheres. The rate of formation of nitrosyl chloride increases with the pressure. w. 0. w.INORGANIC CHEMISTRY ii. 121 A New Chloride of Phosphorus. ADOLPHE BESSON and A. FOURNIER (Compt. rend. 1910 150 102-104).-0n submitting a mixture of phosphorus trichloride and hydrogen to the action of an electric discharge (compare Abstr. 1909 ii 663) a colourless liquid holding in suspension a yellow solid is produced. After filtration and purification by distillation under diminished pressure in an inert atmosphere the liquid has a composition corresponding with that of phosphorus dichloride P,~l,.It is a colourless oily and strongly fuming liquid. The fuming is not only caused by the action of moisture but also by oxidation and under certain conditions the liquid takes fire spontaneously. It has b. p. about 180°/760 mm. (decomp.) 95-96’/20 mm. without decomposition m. p. - 28O. It is decomposed by water with the formation of phosphorous acid and a yellow solid of indefinite composition. It decomposes slowly at the ordinary temperature and more quickly when heated to phorphorus trichloride and a yellow to red solid of indefinite composition which is possibly a mixture of amorphous phosphorus with other chlorides. Attempts to prepare phosphorus dibromide in a similar manner and also by the action of hydrobromic acid on the dichloride were not successful. A yellow to red solid of indefinite composition mas obtained which in the light of the results obtained with the dichloride may have resulted from the decomposition of a dibromide.T. 5. P. Phosphorus Suboxide. ALFRED STOCK (Chem. Zeit. 1909 33 1354. Compare Burgess and Chapman Trans. 1901 79 1235).- Gutbier has stated (Sitxungsber. physik.-med. Soc Erlccngen 1909 40 176) that the existence of phosphorus suboxide (P,O) may be con- sidered as proved ,by the work of Weidner (Znaug. Diss. Erlangen 1909) who hsts repeated Michaelis and Pitsch’s experiments (Abstr. 1900 ii 137) and found in agreement with them that the substance in question does not contain hydrogen which is attached to phosphorus any hydrogen present being due to moisture.The author points out that Michaelis and Pitsch’s experiments are not conclusive and that Weidner has adduced no fresh experimental evidence in support of the existence of phosphorus suboxide. T. S. 9. Formula of Hypophosphoric Acid. I. and 11. E. CORNEC (Bull. SOC. chim. 1909 [iv] 5 1081-1084 1121-1126).-1n the first paper the author reviews the evidence so far brought forward in favour of the simple H,PO and double H4P,06 formulze for this acid. It is pointed out that the five sodium salts prepared by 8alzer (Abstr. 1886 420) can all be regarded as derived from an acid represented by the simple formula and that evidence furnished by (1) the decomposition of the salts by heat (Salzer and Joly) and (2) the electrical conductivity (Rosenheim Stadler and Jacobson Abstr.1906 ii 744 and by Parravano and Marini ibid. 744 848) is not conclusively in favour of either although the latter by analogy with the case of sodium pyrophosphate to a certain extent supports the double formula. Ebullioscopic determinatioiis of the molecular weight of the methyl ester are in favour of the simple formula (Rosenheim Stadler and Jacobson Zoc. cit.).ii. 122 ABSTRACTS OF CHEMICAL PAPERS. I n the second paper the author gives the results of measurements of the lowering of freezing point of solutions of hypophosphoric acid its potassium salts and of solutions of the acid progressively neutralised with potassium hydroxide sodium hydroxide or ammonia (Abstr.1909 ii 972). These all afford evidence in favour of the double formula H4P206 for this acid (compare Parravano and Marini Zoc. cit.). The only valid argument for the simple formula H,PO still remaining is drawn from the esters of this acid and these will now be further investigated. Extinction of Flames. WILLEM P. JORISSEN and N. H. SIEWERTSZ VAN REESEMA (Chern. Weekblad 1909 6 1053-1062).-An investiga- tion of the power of various gaseous mixtures to extinguish flame and a review of previous work on this subject. Inner Cone of the Bunsen Flame. FRITZ HABER and BURRITT S. LACY (Zeitsch. physikal. Chenz. 1909 68 726-’762).-The paper consists largely of a recapitulation and discussion of results already published by Haber’s students and others (compare Haber and Richardt Abstr.1904 ii 166; Davidson Abstr. 1906 ii 325 ; Tufts Physical Review 1906 22 193 ; Lacy Abstr. 190s’ ii 1033). The experiments of Davidson and of Tufts on the electrical con- ductivity of flames have been repeated by Lacy but are only briefly described as Epstein and Krasoa have since made extended observa- tions by an improved method; the results of which are shortly to be published. All the observations indicate that with a plentifuI supply of air the electrical conductivity of the inner green zone of the Bunsen flame is relatively high and much exceeds that in the neighbouring non- luminous regions. Further the velocity with which the water equiiibrium is established is much less outside than inside the luminous zone. Reasons are advanced in favour of the view t h a t these two phenomena are connected both being due to the influence of gas ions in the green luminous zone.I n this case the ions are produced as a consequence of the chemical changes taking place in the luminous zone. The effect of ionisation in accelerating the establishment of equilibria has already been investigated for the carbon monoxide flame by Haber and Coates (compare Abstr. 1909 ii 997). G. 8. LUDWIG WEISS and THEODOR ENGELHARDT (Zeitsch. anorg. Chem. 1909 65 38-104).-A review of previous work on the ill-characterised silicon nitrides is given. Pure silicon is best prepared by reducing potassium silicofluoride with massive aluminium a better regulus being obtained than when aluminium powder is used. The regulus is crushed and extracted successively with hydro- chloric concentrated sulphuric and hydrofluoric acids.The product even after repeated boiling in a state of fine powder with hydrofluoric acid contains 0-3-0*5:! Fe 0.1% Cu and 0.72% SiO,. It forms brown crystalline particles 13 2.30 and is not oxidised by oxygen at 700’. Heated in pure nitrogen in a porcelain tube combination begins near 1240” the velocity of reaction increasing rapidly with the tempera- T. A. H. A. J. W. Nitrogen Compounds of Silicon,INORGANIC CHEMISTRY. ii. 123 ture. The product is amorphous and bulky resembling cork and is seen under the microscope t o consist of several different substances. About 4% of the silicon volatilises and forms a sublimate. The composition of the residue varies with the means of purification adopted.If boiled with potassium hydroxide solution followed by hydrofluoric acid a product having the formula Si,N is obtained mixed with silica which it is impossible to remove. If the free silicon is removed by boiling with a mixture of nitric and hydrofluoric acids and the residue ignited and washed with hydrochloric acid the compound SiN is obtained. By heating silicon in nitrogen at 1300-1400" until saturated a nitride of the approximate formula Si,N is formed and is only slightly decomposed by treatment with potassium hydroxide and hydrofluoric acid or with a mixture of nitric and hydrofluoric acids. The nitride SIN is white and has D 3.17; the nitride Si,N has D 3.64 and the nitride Si,N D 3.44 after allowing for the silica present. All the compounds are more or less decomposed by alkalis and by hydrofluoric acid.Heating silicon in the flame of a coke fire gives a product containing carbon and nitrogen corresponding approximately with the formula Si,C,N. The estimation of nitrogen in the products is performed by heating with a mixture of equal parts of lead oxide lead chromate and lead peroxide in a porcelain tube in an atmosphere of carbon dioxide increasing the temperature from 600 t o 1000". Kjeldahl's method gives less than half the total nitrogen. Silicon is estimated by fusion with potassium and sodium carbonates as in the analysis of silicates. Commercial silicon is best analysed by fusion with potassium and sodium carbonates and potassium nitrate in a platinum crucible which is first coated with a lining of the salt mixture before introducing the silicon.Any csrborundum present is completely dissolved by these means. C. H. D. Action of Metals on Fused Sodium Hydroxide. I. MAX LE BLANC and L. BERGMANN (Bey. 1909 42 4728-4747).-The action of various metals on fused sodium hydroxide at temperatures ranging between 400' and '720" has been investigated all the experi- ments being carried out in an atmosphere of nitrogen. Preliminary experiments having shown that gold is the only metal which is no5 attacked by anhydrous fused sodium hydroxide the latter was always contained in a gold crucible which was placed at the bottom of a silver tube surrounded by a porcelain tube. This tube could be heated to any desired temperature by an appropriate furnace arrangement.Pure dry nitrogen was passed through the reaction tube and the extent of any reaction taking place was measured by determining the amount of water and hydrogen present in the issuing gas. Sodium hydroxide can be readily dehydrated at a temperature of 400° and undergoes no further loss in weight on heating to 720° so that a dissociation corresponding with the equation 2NaOH = Na,O + H,O does not take place between these temperatures. Silver and sodium react with fused sodium hydroxide underii. 124 ABSTRACTS OF CHEMICAL PAPERS. evolution of hydrogen ; platinum copper iron nickel aluminium zinc and magnesium enuse an evolution of hydrogen and a t the same time water is eliminated. The simplest explanation of this double reaction is that the compound M(ONa) which is formed according t o the equation M + xNaOH = M(ONa) + xH where M is a metal forms a more or less complex compound with Na20 resulting from the loss of water from two molecules of the sodium hydroxide.The valencies of the metals mere calculated from the amount of hydrogen evolved and found to be normal in the case of sodium magnesium copper iron aluminium and zinc. Silver nickel and platinum gave abnormal values as for example tervalent silvbr. When the gold crucible containing sodium hydroxide and copper turnings was heated a t 700° i t absorbed a not inconsiderable quantity of copper forming an alloy. This alloy is not formed at this tempera- ture i n the absence of sodium hydroxide. When silver was used in place of copper the silver became alloyed and at the same time the silver tube containing the gold crucible took up some of the gold.The latter phenomenon was also noticed with nickel but with none of the other metals. Magnesium also formed a gold-magnesium alloy. T. S. P. Alkali Hydrogen Garbonates. ROBERT DE FORCRAND (Compt. rend. 1909 149 825-829).-Hydrogen carbonates of the formula R,CO 2RHC03,xH20 such as natural trona have only occasionally been obtained in the laboratory. Dry potassium carbonate when left exposed to the air attains the composition K2C03,2KHC0,. When dilute solutions of potassium rubidium and czesium carbonates are exposed at room temperature for several weeks carbon dioxide is absorbed until the composition corresponds with the formule SK2C0,,2KHC0 ; 3Rb2C0,,2RbHC03 and 4Cs2C03,2CsHC0,.Although the amount of cdrbonic acid absorbed depends on the dilution temperature and pressure of carbon dioxide in the air the results are held to indicats the formation of definite compounds. When solutions of hydrogen carbonates are boiled for prolonged periods they are found to contain hydrogen carbonates of the same composition as the above but the crystals deposited on evaporation have the composition 5(K2C0,,1~H20),4KBC0 ; 3(hb2Co~,l~H20),2RbHC0 ; 5(Cs,C0,,3~H20),2CsHC0,. The last is converted on the water-bath into 5(Cs2C'0,,2H20),2CsHC0,. The heats of formation of these hydrogen carbonates are negative and small in value. R. J. C. Hydrates of Rubidium and Caesium Kydroxides. ROBERT DE FORCRAND (Conapt. rend. 19U9 149 13 41-1344. Compare Abstr.1906 ii 445).-When an aqueous solution of rubidium hydroxide is allowed to evaporate at 15O crystals of the hydrate RbOH,2H20 are deposited ; these have m. p. 45-46' and heat of solution - 0.646 Cal. at 15". The thermal properties of rubidium hydroxide and its hydrates are closely analogous with those of the corresponding potassium compounds.INORGANIC CHEMISTRY. ii. 125 The monohydrate of czesium hydroxide already studied (Zoc. cit.) is more stable than those of the rubidium and potassium hydroxides. Although thermochemical measurements indicate the probable exist- ence of a dihydrate this has not yet been isolated the crystals deposited when a solution of cssium hydroxide is allowed to evaporate being those of the monohydrate contaminated by mother liquor.A saturated solution of rubidium hydroxide a t 15' contains 64.17% of RbOH whilst a saturated solution of the czesium compound contains 79.41% of CaOH. w. 0. w. Acid Sulphates. V. JOH. D'ANs (Zeitsch. anorg. Chem. 1909 65 228-230. Compare Abstr. 1909 ii S85).-Determinations of the solubility of ammonium sulphate in mixtures of sulphuric acid and water at 25" indicate the separation of three solid phases ammonium sulphate and the acid salts (NH,),H(SO,) and NH,HSO,. The com- position of the second salt was confirmed by isolation and titration. Washing with alcohol or ether is impracticable on account of decom- position of the salt and mechanical separation of the acid must there- fore be employed. Unlike sodium and potassium sulphates the solubility of which increases with increasing sulphuric acid concentration that of ammonium sulphate falls to a minimum and then rises only very slightly to the limiting solution (NH4)&30 - (NH,),H(SO,),.The curve for (NH,),H(SO,) also has a minimum. C. H. D. Action of Heat and Light on Silver Sulphite and its Alkali Double Sulphites. Amount of Dithionate Obtained. HENRI BAUBIGNY (Compt. rend. 1909 149 858-860. Compare Abstr. 1909 ii 1004).-Analysis of the products obtained by decomposing silver sulphite under various conditions shows that as much as 89.84% may be converted into dithionate whilst the remainder is converted into sulphate. Sodium silver sulphite may give as much as 9'7.5% of sodium dithionate. Silver sulphite is perfectly stable in the dark but in diffused light it is slowly decomposed giving dithionate and a small proportion of sulphate.R. J. C. [Action of Heat on Sulphites.] HENRI BAUBIGNY (Compt. rend. 1909 149 1378).-Polemical against Colson (compare this vol. ii 34) pointing out that nowhere in the literature is there the statement ascribed by Colson to Berthier that the double sulphites of the alkali metals and of silver decompose with the formation of sulphate. Necessity f o r Exactness in Describing Reactions. T. S . P. New Preparation of the Second Anhydrous Modification of Calcium Sulphate. PAUL ROHLAND (Zeitsch. anorg. C'hem. 1909 65 10b-l07).-When either gypsum or the hemihydrate is dissolved in hot concentrated sulphuric acid an anhydrous salt separates on cooling which proves t o be the second anhydrous calcium sulphate known as Estrich gypsum.There are in addition to the naturally occurring anhydrite four anhydrous modifications of calcium sulphate.ii. 126 ABSTRACTS OF CHEMICAL PAPERS. (1) Kraut's anhydride obtained a t 100' ; (2) van't Hoff's anhydride also prepared at looo and hydrating very rapidly; (3) dead-burnt gypsum prepared above 130" incapable of hardening; (4) a modif~ca- tion prepared by heating above 530° or by the action of sulphuric acid hydrating very slowly. C. H. D. Decomposition of Calcium Carbonate. ERNST H. RIESENFELD (J. Chirn. phys. 1909 7 561-569).-New measurements of the dissociation pressure of calcium carbonate hare been made with the object of explaining the discrepant results obtained by Brill (compare Abstr. 1905. ii 522). 'l'he following values are recorded in mm.of mercury 700" 50 ; 750" 99 ; 800° 195 ; 850'' 370; 900° 700. These data are in good agreement with the thermodynamic equation of Nernst and by means of this equation the dissociation pressures are calculated for every 100' between 600' and 1500O. H. M. D. Solubility of Cadmium Sulphide in Light Petroleum Con- taining Oil. G. C. A. VAN DORP and J. RODENBURG (Chenz. Weekblad 1909 6 103S).-A colloidal solution of cadmium sulphide is obtained by triturating this substance with oil and adding light petroleum. A. J. qT. [Formation of Alloys by Pressure.] WALTHERE SPRING (Zeitsch. Elektrochem. 1909 15 984).-The author agrees with Tammann'a view (Abstr. 1909 ii 669) that the formation of alloys under pressure is a result of diffusion which is not accelerated by the pressure.T. E. Constitution and Heat Contents of Lead-Tin Alloys. W. GIJERTLER (Zeitsch. Blektrochern. 1909 15 953-965).-The recent investigations of Rosenhain and Tucker (Abstr. 19OS ii 1038) and of Degens (Abstr. 1909 ii 888) leave the question of the cause of the development of heat at about 150' undecided; it may be due to decomposition of mixed crystals of lead and tin (Guertler Abstr. 1909 ii 319) or t o the formation of a compound of about the composition Sn,Pb,. Using measurements of the quantity of lieat given out by mixtures of lead and tin in cooling from 380' to 100' made by Spring in lSS6 and the melting-point curve determined by Rosenhain and Tucker and Degens the author has calculated the heats of fusion of different alloys of tin and lead.The values obtained are very much larger than those calculated on the assumption that the heat of fusion of the alloy is the mean of the heats of fusion of its constituents This shows that liquid tin and lead must give out a considerable quantity of heat when they are mixed together. T. E. Peroxidiaed Compounds. LUIGI MARINO (Zeitsch. anorg. Chew&. 1909 65 25-31).-l'he isolation of a salt Pb,Se,07 (Abstr. 1909 ii 575) proved lead sesquioxide to be a true ieebly-basic oxide and not a salt of the dioxide. I n the attenupk to prepare other sesquioxides the liquid obtained by the addition of an acid solutionINORGANIC CHEMISTRY. ii. 127 of potassium permanganate to a mixture of hydrogen peroxide and sulphuric acid a t - 15' supposed by Berthelot to contain the compound H203 has been examined.The solution is capable of oxidising sulphurous acid completely to sulphuric acid. It is probable that the compound present is a higher acid of manganese as the curve connecting volume of oxygen evolved in the absence of sulphurous acid with time has exactly the same form as that obtained with a mixture of chromic acid and hydrogen peroxide in which perchromic acid is known to be present. It is shown that the oxidation of the sulphurous acid is not due to the action of free oxygen in supersaturated solution. Relation of Thallium to the Alkali Metals a Study of Thallium Zinc Sulphate and Selenate. ALFRED E. H. TUTTON (Proc. Roy. Xoc. 1910 83 A 211-226).-A detailed crystallc- graphic examination of the double salts TI,S0,,ZnS04,6H,0 and T1,Se0,,ZnSe04,6H,0 has been made.The various crystallographic data and the morphological angles are tabulated. The double sulphate was found to have D 3-7204 by the author's pyknometer method ; consistent results could not be obtained for the double selenate. From a comparison of the thallium salts with the corresponding potassium rubidium ammonium and czxium salts the author concludes that the morphological and physical properties of the crystals of the thallium double salts are such as quite entitle them to places in this isomorphous series but not to places in the more exclusive eutropic series obeying the law of progression according to the atomic weight oE the interchangeable metals. The position of thallium in either the simple or double salt series is very close to that of arrimonium and therefore also to that of rubidium. This does not hold however for the refractive power.I n respect of this property the thallium double salts are quite exceptional in that they exhibit abnormally high refraction and a larger amount of dispersion and of double refraction. The mean refractive index of thallium zinc sulphate for sodium light is 1.6064 whereas the values for the other four double sulphates range from 1.4859 to 1.5054. Acid Sulphates. VI. JOH. D'ANs and 0. FRITSCHE (Zeitsch. anorg. Chem. 1909 65 231-232. Compare this vol. ii l25).- Thallium sulphate and dilute sulphuric acid form two acid salts TI,H(SO,) and TlHSO the limits of existence of which at 25' have been determined.The curves differ from those of the alkali sulphates the solubility increasing almost continuously with the sulphuric acid concentration. C. H. D. The System Mercuric Chloride and Mercurous Chloride. W. P. A. JONKER (Chem. Weekblad 1909 6 1035-1038).-Mercuric chloride has m. p. 2 7 7 O and b. p. 301'. Addition of mercurous chloride lowers the rn. p. to the eutectic point 271'. A saturated solution of mercurous chloride in mercuric chloride contains 12% of the former and has b. p. 304O. Mercurous chloride sublimes at 373'. C. H. D. H. M. D. A. J. W.ii. 128 ABSTRACTS OF CHEMICAL PAPERS. Resolution of Ytterbium. CARL AUER VON WELSBACH (~Monatsh. 1909 30 695-700 + i-vi).-Polemical (compare Welsbach Abstr. 1908 ii .591). Disputes Urbain's claim of priority (compare Abstr.1907 ii 956; 1908 ii 849). W ITOLD BRONIEWSKI (Compt. rend. 1909 149 853-855. Compare Pushin Abstr. 1907 ii 774).-Determinations were made of the electrical conductivity temperature-coefficient of resistance solution potential and thermoelectric power of a complete series of alloys of copper and aluminium. Alloys containing more than 12% of either metal which are very brittle bad not been investigated previously. The electrical properties in question were often greatly modified by annealing but the inferences drawn from the curves of both tempered and annealed alloys are the same. The annealing was carried out in an electric furnace for four to five hours a t a temperature somewhat below the solidus of each alloy. The curves indicate the compounds Al,Cu AlCu Al,Cu AlCu,.The compound Al,Cu bas not been detected previously but the other three were found by Le Chatelier and by Guillet (Abstr. 1905 ii 712). Carpenter and Edwards (1907) found no evidence for the compound AlCu but on the other hand their compound AlCu could not be detected by the author. The compound AlCu is depolymerised above 500° when the larger crystals are split up into finely crystalline masses and the electrical resistance is doubled. Crystalline Structure of Iron at High Temperatures. WALTER ROSENHAIN and J. C. W. HUMFREY (PTOC. Boy. Soc. 1910 83 A 200-209).-The structural changes which accompany the deformation of iron at high temperatures have been investigated. The material employed for most of the experiments was a pure form of commercial iron of very low carbon content.To remove oxygen the metal was heated to about 900" in a current of pure dry hydrogen. A strip of this iron the surface of which had been polished previously was then placed in a specially designed piece of apparatus in which it could be electrically heated in a vacuum and subjected to stress whilst at a high temperature. By means of comparative observations on specimens which were heated without straining the changes in the structure which resulted from the rise of temperature could be distinguished from those which were a consequence of the applied stress. Under the conditions of the experiments the strips of metal were not uniformly heated and whereas the temperature of the central portion exceeded lOOO" the metal at and near the ends did not reach a visible red heat.Corresponding with the distribution of temperature the micro-structure exhibits three distinct regions when the strips are examined from the central portion towards either end. The structural differences lead to the conclusion that pure iron between the ordinary temperature and 1000" exists in three distinct modifications possessing widely different mechanical properties. The temperature ranges in which these modifications exist are consistent with E. F. A. Electrical Properties of Aluminium Copper Alloys. R. J. C.INORGANIC CHEMISTRY. ii. 129 the view that they are identical with the a- p- and y-forms of Osmond and Roberts-husten as indicated by cooling curves. The deformation observations also indicate that p-iron although existing at a higher temperature is harder and stronger than a-iron and that the reversible transformation of these forms is accompanied by a change of volume.The y-iron obtained in the case of approximately pure iron exhibits the structure and properties of the ‘‘ y-iron ” alloy steels. H. M. D. Passivity of Iron. P. KRASSA (Zeitsch. EZektvochem. 1909 15 981-984).-8 reply to Muller and Konigsberger (Abstr. 1909 ii 1016). The author maintains his former criticism (Abstr. 1909 ii 738) and insists further that the potentials of Miiller and Konigs- berger’s mirrors were SO abnormal that their behaviour cannot be regarded as proving anything about ordinary iron. T. E. Retardation of the Oxidation of Iron by Chromic Chloride. PAUL ROHLAND (Zeitsch. Elektrochem. 1905 15 865-866).-Chromic chloride retards the oxidation of iron ; an increase of the concentration of the hydrogen ions in the solution causes the protection t o disappear.The action persists in presence of chlorine ions and to a less extent in presence of bromine ions. Ferric chloride is however an exception ; possibly the hjdrogen ions formed by hydrolysis account for this. Chromic ions protect ironin presence of sulphates but not in presence of nitrates. T. E. Carbon Monoxide in Steels. E. GOUTAL (Compt. rend. 1909 ’ 149 1129-1 131)-The author has shown previously (Abstr. 1909’ ii 519) that when steels are dissolved in solutions of copper sabts’ gases are evolved which contain carbon dioxide and monoxide. The percentage of carbon monoxide mas estimated by making use of the oxidising action of iodine pentoxide at 75” and found to have a mean value of 0.014%.These results are now confirmed as follows. The evolved gases were swept by a current of nitrogen through a series of tubes each containing 25 C.C. of a 1% solution of defibrinated guinea- pig’s blood. Each tube could absorb a known quantity of carbon monoxide and by determining the number of tubes which showed the characteristic absorption spectrum of carbon monoxide-hsmoglobin t he percentage of carbon monoxide could be found. The percentages thus determined are independent o f (1) the amount of steeltaken; (2) the duration of the experiment; (3) the acidity of the copper solution used; (4) the substitution of cupric chloride by a solution of iodine in potassium iodide ; (5) the gas (air or nitrogen) employed for sweeping out the evolved gases.Analyses of different steels show that the percentage of carbon monoxide does not appreciably exceed 0.01 4. This percentage (0.014) remains approximately constant for ordinary steels containing more than 0.3% of carbon. Nickel steels contain a much less and chromium steels about the same percentage of carbon monoxide as ordinary steels of the same carbon content. Samples of steel mere taken from a Siemens Martin furnace oneii. 130 ABSTRACTS OF CHEMICAL PAPERS hour and then forty-five minutes before tapping and also just before and just after the addition of ferro-manganese. The percentage of carbon monoxide remains practically constant during the period of decarburisation ; it then falls considerably just before the addition of ferro-manganese t o rise to its final value after the ferro-manganese has been added.This is true for both soft and hard steels in which the percentages of carbon monoxide are respectively 0*0063 and 0.0137. T. S. P. Oxygen Evolved from Ferric Oxide at High Temperatures. SIEGFRIED HILPERT (Bey. 1909 42 4893-4895).-1n -contradistinc- tion to Walden (Abstr. 1908 ii S52) theauthor has not been able to determine the temperature at which the pressure of oxygen over ferric oxide is equal to that of the atmosphere. The results vary considerably with the method of preparation and previous treatment of the ferric oxide and the velocity with which equilibrium is attained is extremely slow. I n general the amount of ferrous oxide formed by heating ferric oxide in the air to 1300' did not exceed 5'; only above 1350' was there further loss of oxygen.In one experiment heating the ferric oxide to 1600' only increased the percentage of ferrous oxide from 2.95 to 3.1%. It is pointed out that thermodynamical calculations according to Nernst's theorem are totally untrustworthy when the dissociation of such substances as ferric oxide is taken as the basis. T. S. P. Phosphorus Compounds of Iron. N. S. KONSTANINOFF (J. Rzcss. Phys. Chem. Xoc. 1909 41 1220-1240).-The alloys of phosphorus aod iron were divided by Stead (J. Iron and Steel Inst. 1900 58 60) into five classes (I) those containing 0-1.7% of phosphorus and consisting of solid solutions of Fe,P in iron; (2) alloys with 1*7-10.2% of phosphorus which form solid solutions of Fe,P in iron and give a eutectic alloy (10.2% P) consisting of a definite solid solution and the phosphide Fe,P ; (3) those with 10.2-15.58% P also consisting of the eutectic and Fe,P the m.p. of the latter (15.58% P) being 1060'; (4) alloys containing 15.58-21*6% P and consisting of Fe,P and Fe,P ; ( 5 ) those with more than 21.6% P and containing Fe,P and another phosphide richer in phosphorus. These results were confirmed by Saklatwalla (J. Iron and Steel Inst. 1908 77 92-103) who gave a melting-point diagram for these alloys. The author's investigations of iron-phosphorus alloys containing up to 2100% (32.4 atom.%) of phosphorus show that the solidification- point diagram consists of three branches (1) AB falling from 1514O the m.p. of iron to 1020' corresponding with 10.2% F e ; (2) BC rising continuously from B t o C and corresponding with the separa- tion of crystals of the phosphide Fe,P. The separate solidification curves exhibit two halts one showing a regular rise corresponding with the separation of crystals of the phosphide and the other at a constant temperature corresponding with separation of the eutectic and gradually diminishing in magnitude as C is approached. Micrographic examination of these alloys shows the presence of rhomboidal plates of Fe,P surrounded by the eutectic ; (3) CD rising at first rapidly andINORGANIC CHEMISTRY. ii. 131 afterwards more slowly to the point D corresponding with the phosphide Fe,P (compare Gercke Abstr. 1908 ii 1041 ; Le Chatelier and Wologdine Abstr.1909 ii 1017). A list is given of the known phosphides arsenides and antimonides of the metals of the iron series of the eighth group in the periodic system. With the typical metals of this series iron cobalt and nickel these compounds are more varied in type than with the initial members manganese and copper. With the metals of the other groups of the periodic system these compounds are in the great majority of cases derived from the hydrogen phosphides PH and PH,. With phosphorus the most stable compounds are those rich in metal such as M,P,. whilst with arsenic and especially with antimony compounds relatively poor in metal such as MSb are the most stable. T. H. P. Iron Phosphides. OTTO KUHN (Chem. Zeit. 1910 34 45-46).-Le Chatelier and Wologdine (Abstr. 1909 ii 1017) state that there are only four iron phosphides namely Fe,P Fe,P FeP and Fe2P3 and further that the existence of the last two has not been proved with absolute certainty. I n the course of an investigation on the preparation of copper phosphide by heating a mixture of bone ash powdered quartz wood-charcoal and granulated copper the author obtained a product which on solution in nitric acid left a small residue of glistening needles which were practically insoluble in hot or cold concentrated nitric acid or in dilute sulphuric acid although readily soluble in aqua regia. The results of analysis agree approximately with the formula Fe5P2 the iron coming from the impure quartz used in the preparation. These needles cannot be con- sidered as a solid solution of Fe,P and Fe,P since the former com- pound is readily soluble in nitric acid.Electromotive Forces of Cobalt Alloys. F. DUCELLIEZ (Compt. rend. 1910 150 98-101. Compare Pushin Abstr. 1907 ii 325 618 774 S37).-From the curves showing the relation between the composition of the alloy and the potential with respect to a normal solution of cobalt sulphate the existence of the following compounds has been deduced CoSn CoSb CoSb,. Alloys of cobalt and bismuth form two phases the one rich in cobalt and the other and heavier rich in bismuth. Bismuth lowers the melting point of the cobalt which when liquid is capable of dissolving some bismuth. From liquid alloys containing 0-94 (approx.) % Bi the cobalt separates in a pure state on cooling.The results obtained with lead and cobalt are similar to those with bismuth and cobalt. Cobalt and copper do not form compounds with each other but give either homogeneous solid solutions or two phase systems of solid solutions. T. s. P. T. S. P. Absorption of Carbon by Metals Especially Nickel in the Electrolysis of Aqueous Solutions. GUSTAV LAMBRIS (Zeitmh. Elektrochem. 1909 15 973-981).-The nickel which is deposited from a solution containing ammonium oxalate contains up to 4% of carbon; the smaller the quantity of nickel deposited and the largerii. 132 ABSTRACTS OF CHEMICAL PAPERS. the quantity of ammonium oxalate used the greater is the quantity of carbon *in the deposit. The carbonaceous residue obtained by dissolving the nickel by copper chloride and chlorine contained ammonium chloride which could not be completely removed.The percentage of carbon in the residue increased as greater precautions were taken to free it from ammonium salts. The ammonium salts of glgcollic formic and acetic acids also yield nickel containing carbon but if a diaphragm is interposed between the anode and cathode carbon-free nickel is obtained if the cathode solution contains free ammonia but not if it is kept slightly acid. Carbon dioxide also gives carbonaceous nickel (from nickel ammonium sulphate solution) but in presence of a little free ammonia carbon-free metal is obtained. Since the metal obtained from ammonium oxalate solution contains carbon even in presence of a large excess of free ammonia the anodic carbon dioxide cannot be the source of the carbon.The same applies to carbon monoxide which behaves in the same way as carbon diqxide. Methane ethane and ethylene give carbon-free nickel but acetylene passed into the cathode solution gives carbonaceous nickel both i n neutral and strongly ammoniacal solutions. It is shown that acetylene is formed in small quantities a t the cathode when ammonium oxalate is electrolysed with platinum and nickel cathodes but not with iron copper or t i n ; some ethane is also formed by further reduction of the acetylene. The carbon is contained in the nickel in the form of a carbide ; when the metal is dissolved in hydrochloric acid the whole of the carbon is evolved in the form of hydrocarbons. Iron at which no acetylene is formed is deposited from ammonium oxalate solutions quite free from carbon.It appears therefore quite clear that the carbon is contained in the nickel as carbide formed from acetylene during the deposition of the nickel. Alloys of Nickel and Copper. EMILE VIGOUROUX (Compt. rend. 1909 149 1378-1380. Compare Kurnakoff Abstr. 1907 ii 525 ; Guertler Abstr. 19013 ii 557).-From an examination of the action of acids on alloys of copper and nickel and by determinations of the X.M.F. produced in cells when one of the metals constitutes one of the electrodes and an alloy of the two the second electrode the author is led to the conclusion that no definite compounds of nickel and copper exist. w. 0. w. T E. Two New Nickel Phosphides. PIERRE JOLIBOIS (Compt. rend. 1910 150 106-108).-Twenty-one grams of an alloy of tin and nickel containing 5% of nickel were heated with 1-4 grams of phosphorus in a vacuous sealed tubo to 700'.On cooling and dissolving the lower part of the ingot obtained in warm concentrated hydrochloric acid small prismatic crystals OE metallic appearance were left having a composition corresponding with the formula Nip and DlS = 4.62. They are soluble in nitric acid decomposed by fused sodium hydroxide and lose their phosphorus at 650'. I f the proportion of phosphorus in the reaction mixture is increased to 7-10 grams hydrochloric acid leaves a mixture of tin phosphideINORGANIC CHEMISTRY. ii. 133 (SnP3) with another nickel phosphide having the formula Nip,. These are separated by heating the mixture in a vacuum t o 360" whereby the compound SnP loses phosphorus giving the compotlnd Sn,P (compare Abstr.1909 ii 319) which is then readily soluble in hydrochloric acid. This second phosphide of nickel forms microscopic crystals of metallic appearance which are readily soluble in nitric acid and is decomposed with incandescence by fused sodium hydroxide ; t h e phosphorus is lost at 580" ; DlS = 4.19. T. S. P. Mixed Halogen Compounds of Tin. VICTOR AUGER (Compt. rend. 1909 149 860-862).-Numerous compounds of the types SoCI,Br SoC12BrzI SnClBr have been described all of which are said to decompose on distillation into the simple stannic salts SnX SnY,. The author shows that mixtures of SnEr and SnI give a normal f reezing-point curve with a eutectic at the composition SnBr3.21,,.8.The supposed compound SnEr2T2 melting at 54' is identical with a n equimolecular mixture of SnBr4 and SnI and may be separated by a series of ten fractional crystallisations into fractions melting a t S8O and 2 7 O of compositions SnBr,.212.s and SnBr2.?11. respectively. Similarly the supposed compounds SnBrI (m. p. 103') and SnBr,I may be partly resolved by crystallisation. The cooling curve of SnBr21 from 5 5 O is an unbroken line such as is given by a mixture although the crystals are brilliant and apparently homogeneous. The author suggests that in all molten mixtures of the composition SnBr,T(,-, the halogen atoms are perfectly mobile. The same is suggested for t h e chlorobromides and iodochlorides. R. J. C. Natural Zirconium Dioxide. LUDWIG WEISS [and in part RICHARD LEHMANN] (Zeitsch.anorg. Chem. 1909 65 178-227).- Native zirconium dioxide baddeleyite instead of being isomorphous wit ti rutile and cassiterite is monoclinic. If artificially crpstallised from borax however it is quadratic and isomorphous with rutile. Several hundred kilograms of the native mineral from Brazil have been used for the investigation. It contains SS-O9% ZrO 7.39% SiO 0.74% TiO and 3.78% Fe,O,. About one-half of che impurities q a y be removed by boiling with hydrochloric acid and rather more by heating with sodium sixlphate salt and carbon. Pure zirconia is obtained by heating the mineral with potassium hydrogen fluoride and crystallising the double fluoride. For analysis the mineral is fused with sodium hydrogen sulphnte. Iron is separated with aluminium if present in small quantity by precipitation with ammonium sulphide from a hot solution containing ammonium oxalate and tartrate.After removal of organic salts from the filtrate the zirconium hydroxide is precipitated with sodium hydroxide and hydrogen peroxide titanium being left in solution. The following analyses represent dense fragments ZrO,. Fe,O,. Al,O,. TiO,. SiO,. H,O. Total. I. 92.07 2-73 1-17 trace 2'73 0.88 99'58 11. 84'96 7.01 3 '99 trace 1'57 1'04 98-57 VOL. XCVIII. ii. 10Ti. 134 ABSTRACTS OF CHEMICAL PAPERS. Pure zii-conia is obtained by a long process which is desokibed in detail. Repeated fractionation does not produce any indication of a separation of the zirconium into two constituents. 8mall qnan tities of rare earths have been recognised spectroscopically but scandium is present only in faint traces.The specific gravity varies even of pure samples and is apparently dependent on the extent of the previous heating. After compressing to form rods under a pressure of 8000 kilograms per sq. cm. it may be fused by a powerful arc using 500 amperes’or more and then has D1* 50 5.48. The linear coefficient of expansion is a = 0*00000084 very near to that of fused quartz. Crucibles etc. may be made by mixing the zirconia with 10% of magnesia or clay. Platinum may be rendered completely liquid by heating in a zirconia crucible with an oxy-hydrogen flame. Quartz may also be melted without destruction of the crucible. Other technical applications are described. C. H. D. Thorium.H VON WARTENBERG (Zeitsch. Elektrochem. 1909 15 866-872).-Thorium is prepared by electrolysis of thorium chloride dissolved in a mixture of fused potassium and sodium chlorides (Abstr. 1906 ii 678). The action of the electrolyte on the porcelain crucible is avoided to a great extent by using a graphite crucible which fits closely into the porcelain crucible both as anode and as containing vessel for the fused salts. The cathode is a graphite rod and the electrolysis is carried out as rapidly as possible in an atmosphere of nitrogen. The metal is obtained in the form of crystals mixed with carbon from which it is separated by means of methylene iodide. Careful analyses gave 87+6 to 88.9% thorium 0.06% iron 0004% sodium 0.03% silicon and 0.15% carbon (mechanically admixed).The remainder mas oxide. Microscopic examination showed this to be mixed with the metal; it is left in the form of a felted mass when the metal is dissplved in acid Experiments with mixtures of lead and thallium with their oxides show that mechanically-admixed oxides do not affect the ductility of these soft metals very seriously so that the ductility of thorium cannot be used as an argument in favour of its purity (Abstr. 1909 ii 53). The heats of formation of thorium oxide and chloride were determined and corrected for the impurities present Th + 0 =Tho + 326.0 f 0.5% Gal. Th + 2C1 = ThCl + 300.2 & 0.67% Gal. The results are given in kilogram calories. The metal melted at 1700° but the carbon in it combined with it forming about 2% of carbide which would lower the melting point very considerably.Thorium sulphide has been prepared and its formula is shown t o be ThS,. T. E. The Atomic Weight of Vanadium. WILHELM PBANDTL and BENNO BLEYER (Zeitsch. anorg. Chem 1909 65 152-165)- Vanadium oxy-trichloride VOCl is the only vanadium compound capable of being obtained in a state of sufficient purity for atomic weight determinations. Ammonium metavanadate is three times precipitated from its boiling solution with ammonium chloride ignited,MINERALOGICAL CHEMISTRY. ii. 135 heated with hydrofluoric acid and then mixed with carbon and ignited first in a current of hydrogen and then in one of chlorine. The product is fractionally distilled and then has b. p. 124*4O/723 mm. (corr.). When no longer changed by further fractionation it has D:5'5 1.8362.The liquid is distilled into glass bulbs and weighed ; the bulbs are broken in glass flasks containing water and zinc. After acidifying with nitric acid and filtering the chlorine is precipitated with silver nitrate. The mean result of nine determinations obtained is V = 51.0 (Ag = 107.880 ; C1= 35*460) which is lower than the previously accepted value 5!*2. The difference is explained by a criticism of Roscoe's method C. H. D. Solution of Platinum in Sulphuric Acid and the Producta of Reaction. MARCEL DEL~PINE (Compt. rend. 1910,150,104-106).- Quennessen (Abstr. 1906 ii 551) has attributed the solvent action of sulphuric acid on platinum to the intermediate action of oxygen of the air. The author shows that all Quennessen's results can be explained by assuming that the reaction as represented by the equation 2Pt + 7H,S04 20H*Pt(S04H) + 380 + 4H,O is reversible. Sulphuric acid and spongy platinum were boiled up together in a flask whilevarious gases were passed through the liquid. With carbon dioxide the reaction proceeded according to the above equation. When air was used twice as much platinum dissolved and less sulphur dioxide was evolved owing to the oxidation of the sulphur dioxide to trioxide under the influence of the platinum. With oxygen four times as much platinum dissolved and less still sulphur dioxide was evolved. When a mixture of carbon and sulphur dioxides was used no solution of platinum took place and in some cases platinum was precipitated from solution. The compound of platinum mentioned above can be isolated in the form of the potassium salt OH*Pt(SO,H)*SO,K by adding a dilute solution of a potassium salt to the orange-yellow liquid. If the platinum is boiled with sulphnrio acid for a long time the liquid becomes darker and darker in colour finally almost black and after thirty hours may contain more than 20 grams of platinum per litre From this solution a brown compound can be obtained having the formula Pt(OH),*S04H,H,0 and crystallising in rectangular prisms which are very soluble in water concentrated sulphuric acid alcohol and acetone. The compounds can also be obtained by the oxidation of platinum dioxide with nitric acid. The vnpour density proves it to be undecomposed. At 100" it loses 1*5H,O. T. S. P.