INORGANIC CHEMISTRY. ii. 65 In or gani e Chemistry. A Convenient Method €or the Preparation of Aqueous Hydrobromic Acid of Constant Boiling Point. J. G. F. DRUCE (Chem. News 1923 126 l).-The method described by Pickles (A. 1919 ii 411) gives rather a dilute acid and the dis- tillation is difficult to carry out without " bumping " taking place. The following modification obviates the latter difficulty and gives a distillate containing more than 48% of hydrogen bromide 10 C.C. of strong sulphuric acid are added to a solution of 15 g. of potassium bromide and 0.2 g. of stannous chloride in 25 C.C. of water. After remaining over-night the clear liquid is decanted from the crystals of potassium hydrogen sulphate which haveii. 66 ABSTRACTS OF CHEMICAL PAPERS. sepm&& and distilled the fraction coming over between 120" and 125" being collected. The yield is more than 90% of the theoretical and the product contains only a trace of chlorine.The Structure of the Sulphur Dioxide Molecule. A. 0. RANKINE and C. J. SMITH (Proc. Physical Xoc. 1922,35 33-38).- The authors' method of determination of the molecular dimensions of gaseous molecules from viscosity measurements (A. 1921 ii 192) has been employed to decide between the relative merits of two formulze for sulphur dioxide. That proposed by Langmuir O=S-0 should possess the dimensions of a neon-argon-neon complex with the components arranged in a line or arranged so that the lines joining the neon to the argon atoms make an angle of 135". The calculated mean collision area of this model is con- siderably in excess of that deduced from viscosity measurements 0*94xlO-15 cm.2.On the other hand the calculated value for S<? is 0.99 x 10-15 cm.2 which is in much closer agreement with experiment. W. E. G. Formation of Sulphur Dioxide from the Sulphates of the Alkaline Earths and Iron or Iron Sulphide. F. MARTIN and 0. FUCHS (2. anorg. Chem. 1922,125,307-348).-The authors found that the alkaline sulphates on being heated with metallic iron are completely reduced to the sulphides; this reduction begins at temperatures lower than that required in the case of coal. The reduction is complete in half an hour at 750" for calcium sulphate at 850" for strontium sulphate and at about 950" for barium sulphafe. In the presence of insufficient iron and at about 150" higher a rapid evolution of sulphur dioxide takes place.The best conditions for this are for calcium sulphate the mixture 16CaSO4+15Pe giving 80% of the theoretical yield of sulphur dioxide ; for strontium sulphate the mixture 9SrSOp+SPe giving an SS.9y0 yield. These reactions lead to the formation of well defined ferrites 3CaO,2Fe,O3=Ca3Fe,O ; 2Sr0,Fe20,=Sr2Fe,05 and Ba0,Fe,0,=BaFe20p. Too little iron leads to the formation of alkaline-earth sulphide too much iron gives also iron sulphide. Sulphur dioxide is also evolved by heating the alkaline-earth sulphates with iron sulphide; this can be carried out commercially by using pyrites to which has been added a little iron oxide. Catalytic Hydrogenation of Sulphurous Anhydride. A. R. P. 0 W. T.( MLLE) MARGARET G. TOMKINSON (Compt. rend. 1923 176 35-36).- When a mixture of dry sulphur dioxide and dry hydrogen is passed over finely divided nickel or over nickel sulphide a t 400-450" a mixture of water vapour hydrogen sulphide and sulphur is obtained and if the gases are passed sufficiently slowly the whole of the sulphur dioxide is decomposed. Cobalt sulphide and to a lesser degree ferrous sulphide may also be used as catalysts for the hydrogenation. W. G.INORGANIC CHEMISTRY. ii. 67 Preparation of Sulphuryl Chloride. L~ONCE BERT (BUZZ. Soc. chim. 1922 [iv] 31 1264-1270).-Ruff's method (A. 1902 ii 13) consisting in boiling chlorosulphonic acid with about 1% of mercury or mercuric sulphate is the most practical process for the laboratory preparation of sulphuryl chloride the operation being conducted under a reflux condenser kept at about 70° in order to return unchanged chlorosulphonic acid to the flask.For the successful conduct of the process attention must be directed to the following important but hitherto unpublished points. To prevent risk of fracture the tube of the reflux condenser should be of iron connected with the neck of the flask with a stopper of asbestos paper and silicate cement which resists the action of chlorosulphonic acid admirably. The top of the iron tube is con- nected by means of asbestos thread and silicate to a short glass tube leading to a condenser and receiver for the sulphuryl chloride. The reaction does not proceed so rapidly as Ruff asserted (bc. cit.) and slows down considerably in the latter stages.In one hour a 40%. yield in three hours 60% and six hours 75.8% yield was obtained and for this reason when considerable quantities of sulphuryl chloride are required it is preferable to collect for one hour only and then change the flask for another containing a fresh charge of chlorosulphonic acid. The latter is conveniently obtained by saturating 70% fuming sulphuric acid with hydrogen chloride and distilling. G. I?. M. Phosphorus. ALFRED STOCK (2. anorg. Chem. 1922 125 228-234).-A theoretical paper in which the paper of Marckwald and Helmholtz (A 1922 ii 845) is criticised. The author points out that the melting point of phosphorus depcnds to a very large extent on the rate of heating. MAURICE L. HUGGINS ( J . Physical Chem. 1922 26 833-835).-The author criticises the view held by Stock that boron in the boron hydrides is quadrivalent and puts forward structures for the hydrides B,H B4HI0 B6H12 Bl0HI4 in which boron is regarded as tervalent.These structures are strikingly like those adopted for ethylene butadiene benzene and naph- thalene respectively. The hydrogen atoms are held by means of four-electron bonds such as often exist a t least temporarily in unsaturated organic compounds. Each such four-electron bond is surrounded by four atoms. C. MAZZETTI and F. DE CARLI (Atti R. Accad. Lincei 1922 [v] 31 ii 119-120).-Almost quantitative yields of boron chloride were obtained by passing dry chlorine through a porcelain tube containing ferro-boron heated at about 500". Silicon Hydrides.XI. The Action of Oxygen on S s 4 and SiH,. ALFRED STOCK and CARL SOMIESKI (Ber. 1922 55 [B] 3961-3969).-The oxidation of the simplest hydrides of silicon has been studied in the hope that the reaction would provide W. T. Boron Hydrides. J. F. 8. Preparation of Boron Chloride. T. H. P.ii. 68 ABSTRACTS OF CHEMICAL PAPERS. a more convenient method of obtaining compounds of the types SiH,*OH SiH,O than is afforded by the hydrolysis of the halides. As oxidising agents oxygen air and mixtures of nitrogen and oxygen containing less of the latter than air have been used. The violence of the reaction is moderated by very slowly adding the oxidising agent to an excess of the gaseous liydride under greatly diminished pressure and at a temperature varying between -70" and -1140".Measured volumes of the reagents are employed and the nature of the volatile products is elucidated by the authors' vacuum process. The necessary apparatus is fully figured and described in the original text. Reaction invariably occurs immedi- ately and in spite of the precautions described above is frequently explosive. The products consist of unchanged hydride (including SiH during the oxidation of Si,H,) nitrogen (and argon) when air or artificial mixtures of nitrogen and oxygen are used (the oxygen is invariably consumed completely) a Considerable quantity of free hydrogen water and in certain cases small quantities of volatile silicon compounds formed during the oxidation. In addition a colourless non-volatile residue which generally con- tains brown particles is produced.This consists partly of compact crusts formed near the point of entry of the oxidising agent partly of a fog in the interior of the reaction vesspl and partly of a uniform vitreous deposit on the whole wall of the vessel. The colourless components are fairly readily dissolved by very dilute alkali and appear to consist of substances such as polymeric prosiloxan (SiH,O) silicoformic anhydride [( SiHO),O] etc. ; the brown components contain silicon and possibly highly condensed silicon hydrides poor in hydrogen; they appear to be iormed chiefly when the action is accompanied by considerable local heating and a deficiency of oxygen. I n spite of variation in the temperature and in the degree of dilution of the oxygen the reaction occurs fairly uniformly.The oxidation proceeds beyond the stage SiH,*OH since the presence of (SiH,),O in more than traces could not be established and the stability of the gas is such that it could not have escaped detection. The main derivatives of silicon produced during the oxidation are the polymeric forms of SiH,O OH*SiHO SiO(OH) etc. The water is regarded as produced from hydroxylated silicon compounds which become decomposed as the temperature is allowed to become atmospheric thus SiH,(OH) -+ SiH,O+H,O. The liberation of free hydrogen is surprising. I n those cases in which elementary silicon is deposited a part of it is doubtless due to the thermal decomposition of monosilane. Its production is not due to a secondary hydrolysis of silane or its primary oxidation products by water formed during the reaction.It appears to be a primary product of the oxidation SiH,+O=SiH,O+H,; a reaction of this type is unusual and apparently due to the preponderating affinity of silicon for oxygen. The production of hydrogen accounts for the peculiarities observed during the combustion of silicon hydrides which differs markedly in its violence from similar cases of spontaneous ignition such asii. 69 INORB-4NIC CHEMISTRY. the quiet inflammation of phosphorus hydride. The process con- sists of the following successive steps inception of oxidation liberation of hydrogen formation of a mixture of oxygen and hydrogen and ignition of the latter when the temperature has been raised sufficiently by the heat of the reaction. H. W. Theory of the Structure and Polyrnorphism of Silica.ROBERT B. SOSMAN ( J . Pranklin Inst. 1922 194 741-764).-A theory of the molecular structure of the various forms of silica is advanced. It is suggested that the silica atom-triplet maintains a certain degree of individuality in its amorphous and crystalline states as well as in its compounds and the freedom of the oxygen atoms to change their positions with respect to the silicon is re- stricted. The triplets are assumed to assemble into chains or threads in the liquid and glassy states and a thread structure persists in the crystalline states. The three principal crystalline modifications (cristobalite tridymite quartz) are built up by com- bining the threaded triplets in three different ways the connexion between the threads being through the oxygen atoms and it is maintained by the sharing of electrons. The high-low (a-p) inversions in all the forms are thought to be due to the same under- lying mechanism namely a change in the state of motion of certain electron orbits resulting from increased thermal vibration of the atoms whereby the shape of the silica triplet and the relative positions of the two oxygen atoms are suddenly altered.The theory gives a satisfactory explanation for many silica phenomena such as the minimum in the temperature-volume curve of silica glass the relation between the specific heats of the crystalline forms the fact that natural quartz formed at a high temperature is always homogeneous whilst low-temperature vein quartz is right- and left-twinned etc.G . F. M. The Constitution of the Silicates. G. TAMMANN (2. anorg. Chern. 1922 125 301-306).-The author finds that the mole- cular specific heats of the silicates are additive and concludes that the molecules are independent and at temperatures far re- moved from their melting point do not vibrate appreciably. This conclusion is supported by the fact that only in the neighbourhood of the melting point do isomorphous complex substances diffuse into one another visibly. In solution silicates undergo hydrolysis and this therefore gives no definite information as to the molecules which exist in the solid The structure of molten silicates cannot be investigated. The complex silicates differ from the carbon compounds in that they are decomposed in solution and in the fused state and the molecular theory of organic chemistry finds no application in the chemistry of silicates.The [Suggested] Formation of Helium and Neon in Dis- charge Tubes containing Hydrogen. A. PIUTTI and E. Boaa~o- LERA ( B e r n . Accud. Lincei 1921 [v] 13 657-701).-Full details of work a summary of which has already appeared (cf. Piutti this W. T.ii. 70 ABSTRACTS OF CHEMICAL PAPERS. vol. ii 20 and also Piutti and Cardoso A. 1920 ii 311; Baly Annual Report 1914 45; 1920 29 30). T. H. P. The Rectilinear Diameter of Neon. E. IMATHIAS C. A. CROMMELIN and H. KAMERLINGH ONNES (Compt. rend. 1922 175 933-935).-The density of neon was studied between its critical temperature (-228.71') and its normal boiling point (-245.92"). Prom the experimental values obtained the angular coefficient of the diameter was found to be a=-0.00716146 a notably high figure At the critical temperature the critical density is calculated to be d,=0*4835 whilst the critical coefficient is (RTcd,)/p,=3.249 the smallest value yet obtained experimentally with the exception of that for helium.The general conclusion is drawn that neon follows the law of rectilinear diameter. H. J. E. Electrical Preparation of Solid Alkali Amalgams in Quantity. PAUL M. GIESY and JAMES R. WITHROW (J. Ind. Eng. Chem. 1923 15 57-60).-Kerp's method (A. 1898 ii 516; 1900 ii 656) for the electrical preparation of amalgams can be much simplified and improved by allowing the mercury at the bottom of the cell as well as that from the flowing jet to receive current.I n this way the current can be increased to a very great extent with a corresponding reduction in manipulation and in the time of the run. If the current be led out at the tip of the jet tube a single jet tube with a stop-cock can be used instead of the changeable jets used by Kerp. It was also found that it is not necessary to work in an atmosphere of hydrogen in order to produce a pure product. Smith and Bennett's method (A. 1909 ii 663; 1910 ii 500) can be improved by strongly cooling the electrolyte when used for the preparation of sodium amalgam and also by forcing the current. If a suitable diaphragm could be found which was unacted on by solutions of alkali hydroxides or by the amalgams themselves and whilst offering a low electrical resistance was a non-conductor Shepherd's method (A.1903 ii 210) would undoubtedly be the best but no suitable material could be found. It is pointed out in conclusion that all solid alkali metal amalgams are lighter than mercury and not heavier as recorded by Kerp. Isomorphism of Potassium Fluoborate and Permanganate. FERRUCCIO ZAMBONINI (Atti R. Accad. Lincei 1922 [v] 31 ii 67- 73).-The author claims priority over Barker (T. 1912 101 2484) for his discovery of the isomorphism between potassium fluoborate and perchlorate (2. Kryst. Min. 1905 41 60). In order to ascer- tain whether alkali fluoborates are isomorphous with or merely exhibit crystallographic relations towards alkali perchlorates and permanganates the author has examined the crystals deposited from a solution containing equal weights of potassium fluoborate and permanganate. Mixed crystals of the fluoborate type con- taining at most 0.4% of the permanganate are formed but on the permanganate side the miscibility is practically zero.G. F. M. T. H. P.INORGANIC CHEMISTRY. ii. 71 Crystalline Form of Trihydrated Sodium Chlorite. ETTORE ARTMI (Atti R. Accad. Lincei 1922 [v] 31 ii 65-67).-This salt NaC10,,3H20 (cf. Levi A. 1922 ii 567) crystallises in the triclinic system a b c=0.4041 1 0.6353 a=103' 43~5'~ p=119" 3' and y=81" 46.5'. Purification of Sodium Hyposulphite A Modification of Jellinek's Method. WALTER G. CHRISTIANSEN and ARTHUR J. NORTON ( J . Ind. Eng. Chem. 1922 14 1126-1128).-An apparatus is described by means of which sodium hyposulphite of 98.3% purity can be rapidly obtained in 42% yield from the commercial article by a modification of Jellinek,s procedure.It consists essentially of a source of carbon dioxide connected on the one hand with a bottle cemented with litharge and glycerol into a Biichner filter the bottom of the bottle being removed and replaced with canvas and hardened filter-paper and on the other with a litre bottle standing in a 2-litre beaker which serves as a water-bath and provided with a combined stirring and filtering device. The latter consists of an inverted glass funnel sealed on to a glass tube which acts also as the shaft of the stirrer. The whole apparatus is exhausted and filled with carbon dioxide the solution of commercial hyposulphite in freshly distilled water is introduced into the Buchner filter and the clear yellow filtrate is passed by the gas pressure into an intermediate receiver and thence into the litre bottle previously charged with the requisite quantity of sodium chloride.The stirrer is set in motion for about thirty minutes to ensure complete solution of the salt and pre- cipitation of the hydrated hyposulphite. The washing dehydration and drying of the precipitate is carried out as in Jellinek's method (A. 1911 ii 278). The ferricyanide method was found superior to either the copper or silver methods for the estimation of the purity of the product. Sodium Chromite. ERICH MUUER (2. angew. Chem. 1922 35 557-558).-Chromium hydroxide was precipitated from a solution of a chromic salt by ammonia dried over sulphuric acid shaken with sodium hydroxide of different concentrations and the concentration of chromium determined from time to time.The presence of hydroxide in solution begins to be noticeable with SN-sodium hydroxide solution. The hydroxide used contained about 6 mol. of water to 1 mol. of chromium trioxide. The amount of hydroxide dissolved was found to be dependent on the time the mixture was agitated rising to a maximum and after this decreasing to an almost constant value. The minimum solubility set in after from ten to fourteen days. The chromium seems to exist in solution as a complex ion and not as a colloid. It is con- sidered that a tervalent chromium salt of sodium is formed. H. M. T. H. P. G. F. M. Ammoniacal Silver Fluoride. DERVIN and OLMER (Compt. rend. 1922 175 1058-1061; cf.Gore Chem. News 1870 21 28; Bruni and Levi A. 1916 ii 482 617).-Anhydrousii. 72 ABSTRACTS OF CHEMICAL PAPERS. silver fluoride dissolved in cold concentrated ammonia solution and the solution evaporated first in a vacuum then at normal pressure over concentrated sulphuric acid yielded colourless ortho- rhombic prisms having the composition AgF,2NH3,2H,0. Light slowly decomposes this substance but it may be preserved un- changed in the dark. It is very soluble in water the solubility increasing on addition of ammonia and although the solubility in alcohol is increased by the presence of water addition of alcohol to the aqueous solution does not result in precipitation. Dry ammonia gas is without action on the crystals; if placed in a vacuum over sulphuric acid they lose water and when the an- hydrous state is reached loss of ammonia takes place.On being heated ammoniacal silver fluoride detonates the products including nitrogen finely divided silver and ammonium fluoride. The authors suggest that the decomposition may be represented by the equation 3(AgF,2NH,,2H20) -+ Ag3N+3NH4F+2NH3+6H,0 with subsequent breaking up of the silver nitride into its elements. H. J. E. The Dissociation of Calcium Sulphate at High Temper- atures. P. P. BUDNIKOV and J. K. SYRKIN (Chem. Ztg. 1923 47 22).-Samples of calcium sulphate were heated t o constant weight at temperatures of from 800" t o 1375" and the proportion of calcium oxide in the product was determined in each case. After heating a t lOOO" only 0.21% of calcium oxide was present a t 1300° 3.00% and a t 1375" 98.67%.Above 1300" the salt melts and a vigorous evolution of fumes takes place corresponding with almost complete decomposition. A. R. P. Valency. XIX. The Ammonia Compounds of Barium Halides. GUSTAV P. HUTTIG and WILHELM MARTIN (2. anorg. Chem. 1922 125 269-ZSO) .-The following ammonia compounds were found to exist BaC1,,8NH3 ; BaBr,,8NH3 ; BaBr,,4NH3 ; BaBr2,2NH3 ; BaBr,,NH ; Ba12,10NH3 ; BaI,,9NH3 ; Ba12,8NH3 ; Ba12,6NH3 ; Ba12,4NH3 and Bal,,2NH3. The method of investig- ation was the same as before (cf. A. 1920 ii 293 and 318). The dissociation temperature a t which the pressure equals 100 mm. is given in each case and also the heat of formation as calculated by Nernst's formula. W. T. Complex Magnesium Salts.111. G. SPACU and R. RIPAN (Bulet. Xociet. qtiinje Clwj 1922 1 267-283; from Chern. Zentr. 1922 iii 1046-1047 ; cf. this vol. i 96).-Monoaquopentammine- magnesium chloride [&Ig(NH3),,H,0,]C12 prepared by passing ammonia into a solution of magnesium chloride in absolute alcohol at the ordinary temperature is a colourless crystalline powder. The following are also menticned hexamminemagnesium bromide a white crystalline powder ; tetra-aquodiamminemagnesium sulphate [Mg(NH,),(H,O),]SO a crystalline powder ; triaquotriammine- magnesaum sulphate [Mg(NH,),(H,O),]SO colourless crystals ; diaquodiamminemagnesium sulphccte [Mg(NH3),(H,0),]S04 whiteINORGANIC CHEMISTRY. ii. 73 powder ; triaquopiperidinemagnesium sulphute a slightly green substance ; penta-aquobenxylamin emccgnesium sulph- a-te a white crystalline powder.[Mg (w-%) (H,O),IS04 G. W. R. The Solidification of the System MgCl,-KCl-BaCl,. J. VALENTIN (Compt. rend. 1922 175 1061-1063).-A thermal study of the three salts was made and the results are shown in a series of diagrams. Two compounds MgCl,,KCl and BaC12,2KC1 are formed so that the complete triangular diagram is divided into five regions each corresponding with the deposition of a pure substance and contains three triple points. H. J. E. A Simple Method for the Preparation of Highly Phos- phorescent Zinc Sulphide. JULIUS SCHMIDT (Ber. 1922 55 [B] 3988-3989).-Equal weights of pure zinc sulphate and sodium acetate are dissolved in water and the hot solution is treated with hydrogen sulphide until the precipitation of the zinc is complete. The zinc sulphide is washed by decantation and finally completely on the filter after which it is dried on the water-bath.The product (200 g.) is moistened uniformly with a solution of mag- nesium chloride (20 g.) calcium chloride (10 g.) strontium chloride (10 g.) and crystalline barium chloride (10 g.) in water (200 c.c.) to which 0.04 g. of ammonium tungstate dissolved in 10 C.C. of water has been added. The mixture with the zinc sulphide is evaporated to dryness with occasional stirring on the water-bath. The product is slowly heated to redness in a crucible (the operation should require about an hour) and maintained at this temperature during thirty to forty minutes after which it is allowed to cool in the furnace. The soluble salts are removed by water and the residual zinc sulphide is dried on the water-bath.It exhibits a strong green fluorescence. Mixtures of salts other than those described have little influence on the strength but considerable effect on the colour of the phos- phorescent light and the readiness with which the phosphorescent effect is excited by illumination Rijntgen rays or radioactive materials. Most metallic salts induce a green phosphorescence but that caused by manganese is yellow (cf. Hofmann and Dacca A. 1904 ii 690). Phosphorescence appears to be weakened or completely inhibited by metallic salts which give dark coloured or black sulphides (cf. MacDougall Stewart and Wright T. 1917 111 663). H. W. Action of Natural Waters on Lead. JOHN C. THRESH (Analyst 1922 47 4 5 9 4 6 8 500-505) .-The presence of varying quantities of silicates carbonates sulphates and salts of organic acids together with the organic acid and free carbon dioxide causes the differences observed in the action of natural waters on lead.Silicates prevent the oxidation of lead iron and aluminium and moorland water after t'reatment with sodium silicate ceases to act on lead w. P. s.ii. 74 ABSTRACTS OF CHEMICAL PAPERS. Complex Salts of Copper and Thallium. G. CANNERI (Gaxzettu 1922 52 ii 266-270).-The following complex com- pounds analogous to the simplest types of those of the alkali elements (cf. Rosenheim and Steinhauser A. 1900 ii 653) are described T12Cu(S0,),,6H20 blue crystals ; Tl,Cu(SO,) yellow powder ; CuS03,3Cu2S03,T12S03 cinnabar-red crystals &ssolvmg in aqueous ammonia to a blue solution; T1,Cu(S20,) a straw- yellow microcrystalline powder.No more than one compound corresponding with one and the same series was obtained in spite of variation of the conditions of formation. T. H. P. The Vapowr Tension of some Copper-Zinc Alloys in the Solid State. L-EoN GUILLET and MARCEL BALLAY (Compt. rend. 1922 175 1057-1058).-The rate of volatilisation of zinc from copper-zinc alloys depends on the temperature and on the nature of the surrounding atmosphere. When the latter can exhibit oxidising properties the surf ace film of oxide mechanically retards loss of zinc from the alloy whilst in an atmosphere of carbon monoxide nitrogen or hydrogen no film is formed and the tem- perature alone is the controlling factor.In such cases the vapour pressures observed cannot be represented as a linear function of the atomic percentage of dissolved copper as the alloys used ( C ~ = 5 5 - 2 0 ~ ~ Zn=44-75%) were not dilute solutions of copper in zinc. Raoult's law therefore does not hold. H. J. E. Solubility of Halides of Univalent Copper in Sodium Thio- sulphate. G. CANNERI and R. LUCHIXI (Guzzettu 1922 52 ii 261-266).-1n the compounds originating when cuprous halides are dissolved in ammonium thiosulphate solution the univalent copper apparently exerts its co-ordinating power not merely towards the thiosulphuric radicle but also towards the halogen. To the compounds formed which have the general formula CUX,NH,X,~(NH~)~S~O~ where X represents a halogen atom or the thiocyanogen radicle (cf.Rosenheim and Steinhauser A. 1900 ii 653) Werner (" Neuere Anschauungen," 1913 168) ascribed the formula [ R (c&3)J(NH4)s R in this case representing Cu. On the basis of this formula the integrity of the nucleus within the brackets should not be diminished by replacement of the ammonium residue by an atom of an alkali element of similar chemical character such as sodium. There seems therefore to be no reason sufficient to explain the difference in behaviour between the ammonium salts and those of other alkali metals other than the varying solvent power on the cuprous halides. The authors find that sodium thiosulphate fused in its own water of crystallisation dissolves considerable proportions of freshly prepared cuprous halides and cuprous thiocyanate the following salts being formed CuCl,5Na2S,0 ; CuBr,5Na2S,03 ; CuBr,Cu2S2O3,9Na2S2O3 ; CuBr SNazS,03.,NaI ; 2CuCNS ,5Na2S,0,. To these compounds which form white crystals unaffected by the action of light and give clear colourless aqueous solutions,INORGANIC CHEMISTPY.ii. 75 the following formulz analogous to those attributed by Werner to the complexes formed by ammonium thiosulphate might be Such hypothetical structural formule cannot however be supported by experimental data other than the colour of the salts this indicating that the cuprous ion forms part of a complex radicle constituted of groupings in such condition that they exhibit increased resistance to react with their specific reagents. The dissociating action exerted by water on these complex compounds prevents physical measurements of any value in this connexion being made.T. H. P. The Preparation of Metallic Phosphides by the Action of Hydrogen Phosphide on Solutions of Metallic Salts. A. BRUKL (2. anorg. Chem. 1922 125 252-256; cf. Moser and Brukl A. 1922 ii 393).-~ercurous phosphide PHg is obtained when hydrogen phosphide is allowed to react on a solution of mercurous sulphate in dilute sulphuric acid as a black amorphous precipitate; it is rapidly oxidised by the oxygen of the air and gradually by dilute nitric acid. Phosphine is gradually evolved by the action of cold concentrated hydrochloric acid but much more rapidly by the action of the hot acid; with concentrated sulphuric acid sulphur dioxide is evolved. The compound is not explosive.Pure mercuric phosphide P2Hg can be obtained by the action of a solution of mercuric chloride in ether on hydrogen phosphide; it is a dark brown solid becoming grey on exposure to air. I n the cold water alkalis and dilute acids are without action but on warming phosphine is evolved; it is oxidised in the cold by dilute nitric acid. A solution of lead acetate in alcohol reacts with phosphine to form a black flocculent precipitate of lead phosphide P2Pb an unstable compound decomposed in the cold by water and alkalis. Cadmium phosphide P,Cd3 obtained as a black flocculent precipitate by the action of phosphine on an ammoniacal solution of cadmium sulphate is unstable and yields hydrogen phosphide when treated with dilute hydrochloric acid. It is very vigorously oxidised by concentrated nitric acid.Its reactions are similar to those of lead phosphide. Alloys of Iron and Aluminium. N. KURNAKOV G. URASOV and A. GRIGORIEV (2. anorg. Chem. 1922 125 207-227 and J . Rziss. Phys. Chem. Xoc. 1918 50 270).-The system iron- aluminium was studied by (1) thermal analysis ( 2 ) the micro- structure of the alloys and ( 3 ) the clectrical conductivity and hardness. In the interval 32-1-39.5 atom. yo iron the endothermic compound Al,Fe crystallises from the fused mass ; below 1100" this decomposes into the a- and p-solutions of aluminium in iron. Between 24.2 and 34 atom. yo iron there exists a stable W. T.ii. 76 ABSTRAC,TS OF CHEMICAL PAPERS. phase which the authors call the 6 phase having properties very different from those of the adjoining phases.This crystalline phase corresponds with the maximum change in properties of the system iron-aluminium ; i t possesses the minimum electrical conductivity acd temperature coefficient the maximum hardness and i t is extremely brittle. The authors regard i t as a chemical individual of variable composition (cf. Kurnakov and Shemt- SUSUMU M~YAMOTO (Japan J. Chein. 1922 1 57-80 and Sci. Papers Inst. Phys. Chem. Research 1922 1 31-65).-An English translation of the papcr previously published in Japanese (cf. A. 1922 ii 647). The expression K=ed-B/p( 1 +O.7BCK,,) in the earlier paper should be replaced by K=ed-B/T(l +0.70C‘,,H). The Higher Oxide of Cobalt. OWEN RHYS HOWELL (T. Triple Nitrites of Bivalent Cobalt. V. CUTTICA and M.PAOLETTI (Gazxetta 1922 52 ii 279-286 ; cf. following abstract). -The analytical data given by Erdmann ( J . pr. Chem. 1866 97 385) for the three triple nitrites Co(N0,),,Ba[Ca,Sr](N02),,2KN0 being incomplete the authors have prepared and analysed these compounds as well as various other triple nitrites of cobalt. The above three complex nitrites form dirty yellow powders stable in the air and may be regarded as salts of one and the same complex quadrivalent ion Co(NO,),. The latter is however not stable in presence of water in contact with which the salts yield an insoluble golden-yellow compound K,Co(NO,) the stable cobaltonitrous ion being hence [Co(NO,),]””. The high value of the active mass of the ion [NO,]’ present in the conditions of the reaction in which the triple salt originates tends to displace the equilibrium [Co(N0,),]”+2N02’ Z [Co(NO,),]”” towards the right a dis- placement in the same sense that is towards the formation of undissociated molecules of the triple nitrite being determined also in the equilibrium [Co(N02),]””+2K”+Ca”(or Ba,Ca) Z These cobalt compounds are analogous in corn- position to the corresponding nickel compounds but exhibit greater stability towards the dissociating action of water the co-ordin- ation valency representing a far stronger linking in this case.If the potassium is displaced by thallium a salt of analogous com- position is obtained but when sodium is introduced the composi- tion is changed. Triple cobalt zinc (cadmium mercury) potassium nitrites have also been obtained these being stable in dry air and of the formula Co(NO2)G .The salts described and analysed are cobalt calcium potassium nitrite ; cobalt barium thallioLm nitrite ; cobalt barium sodium nitrite Co(N0,),,2Ba(NO2),,NaNO a dirty green powder ; cobalt zinc potassium nitrite 2Co(NO2),,Zn(NO,),,6KNO dirty yellow ; cobalt schuschny A. 1913 ii 190). w. T. Reducing Action of Ferrous Hydroxide. K. K. 1923 123 65-71). Co(N02)G Ca. [ ]Iz2 [ I,”:INORGANIC CHEMISTRY. ii. 17 cadmium potassium nitrite pale chestnut and cobalt mercuric potassium nitrite rather deeper chestnut precipitate. Triple Nitrites of Nickel. V CUTTICA and G. CAROBBI (Gazzetta 1922 52 ii 270-279; cf. A. 1922 ii 448).-Both the double and triple nitrites of nickel undergo profound ionic scission immediately on contact with water the supposed ion [Ni(NO,)J” being hence classified with the “ imperfect ” complex ions.Hence in the passage of these salts into solution the mobility of the con- stituent simple ions predominates over the co-ordinating action exerted in such conditions by the nickel. In general accumulation of electronegative radicles in a complex ion is accompanied by increase in the extent of the ionic dissociation of its salts. With the complex nitrites the large number of NO radicles displaces the electrolytic equilibrium pi(N02)6]K4 t 4K’+ [Ni(NO,),r”’ almost completely towards the right and the equilibrium [Ni(NO,),]”” Ni(N0,)2+4N0,’ must also undergo displace- ment in the same sense. The results of cryoscopic measurements in water of some of the double and triple nitrites of nickel show indeed the almost exclusive presence of simple ions the nickel nitrite molecules alone remaining undissociated.The following triple nitrites have been prepared. Nickel mercuric potassium nitrite Pi(NO,),]% forms dark green almost black crystals and gives a green aqueous solution. Cryoscopic measurements indicate decomposition into seven ions. Two nickel mercuric thallium nitrites of the respective composi- tions 3Ni(N02)2,Hg(N0,)2,6TlN0 orange-yellow crystals and 2Ni(N02),,Hg(N0,),,8TlN02 are obtained according to the experi- mental conditions employed. Nickel zinc potassium nitrite 2[Ni(N0,),]K,,Zn(N02)2 forms a flesh-red microcrystalline mass. A chestnut-brown nickel uranyl thallium nitrite was prepared but no simple formula could be derived from the results of its analysis.Nickel cobalt potassium nitrite is a greenish-yellow salt of the formula T. H. P. Nickel cobalt thallium nitrile is a dark red salt which when treated with water leaves undissolved the orange-red complex Tl,Co(NO,),. Unsuccessful attempts were made to prepare a nickel magnesium potassium nitrite. Some New Series of Phosphotungstates. F. KEHRMANN and R. MELLET (Helv. Chim. Acta 5 942-944).-When a very concentrated solution of a mixture of sodium tungstate and phos- phate is acidified faintly with acetic acid a t the end of a few days a mass of crystals separates containing two. salts. One of these Its formula is probably Co(NO,) Nf 4T1PU’o2. T. H. P. I lT1 VOL. CXXIV. ii. 4ii. 78 ABSTRACTS OF CHEMICAL PAPERS.can be freed from the other by washing with cold water in which the first i s sparingly soluble. This salt which cannot be recrystal- bed has the composition corresponding with the formula 14Na20,5P205,19W03,aq or 29Na,0,10P20,,39W03,aq. The second salt is readdy soluble in cold water and forms supersaturated solutions which may not crystallise for months. The salt corre- sponds with the formula 3Na,O,P2O5,6WO3,aq. representing the simplest series of phosphotungstates. A sodium salt described by Scheibler (Ber. 1872 5 801) was thought to belong to a P,0,,6W03 series but is now shown to belong to the series 3M,0,P20,,7W0,,aq. It is proposed to continue the study of these three new series of phosphotungstates. E. H. R. Properties of Tin Especially the Density. R.HOFFMANN and W. STAHL (Metall. u. Erx. 1923 20 5-8).-Tin of 99.8% purity produced from Bolivian ore at the Wilhelmsburg Works was found to have d! 7.312 in a vacuum and go 7.311. [Cf. J.S.C.I. 1923 Feb.] A. R. I?. The Equilibrium in the Reduction of Tin Dioxide by Carbon Monoxide. W. FRAENKEL and K. SN~PISCHSHI (2. anorg. Chem. 1922 125 235-252).-The equilibrium Sn02+2C0 =+ Sn+2CO was studied. It was found that the same equilibrium was reached from both sides and it was independent of the amount and composition of the solid phases. The equilibrium concentration of carbon monoxide increases gradually with in- creasing temperature up to 900" and then suddenly increases very rapidly. The temperature coefficient was studied; at 500" equilibrium sets in after some weeks whereas at 600" equilibrium is reached in a few days.An improved method for the preparation of stannous oxide consists in dissolving stannous chloride in the least amount of hot concentrated hydrochloric acid and gradually adding sodium hydroxide solution until the solution reacts alkaline towards phenolphthalein ; the resulting milky liquid is then kept boiling . in a bath of saturated sodium chloride solution. In a few hours a blue lustrous powder is obtained; this is then washed and dried. Composition of Micella?. I. Stannic Acid Peptised by Alkali. ROBERT WINTGREN (2. physilcal. Chem. 1922 103 238-259) .-Electrical conductivity and transport measurements have been carried out with stannic acid sols which have been peptised by alkali and also with the filtrates from these obtained by ultrafiltration.From these measurements by Ducleaux's method (A. 1905 ii 432) the portion of the conductivity and of the transport which is due to the micellar portion of the sol has been calculated and from Zsigmondy's conception of the qualitative composition of the micellae an attempt has been made t o apply Kohlrausch's law of the independent migration of ions to inorganic colloid ions by regarding the micellar ions as ordinary multivalent ions. From the conductivity values the niean quantitative com- W. T.INORGANIC CHEMISTRY. ii. 79 position of the micella equivalent has been approximately deter- mined that is the number of molecules of stannic oxide contained in one equivalent of colloidal stannic acid. This quantity has been termed the equivalent aggregation.Further from the conductivity determinations it is shown in agreement with Varga's calculations from transport measurements (Kolloid Chem. Beihefte 1919,11,25) that a measurable large quantity of the alkali used for peptisation is enclosed in the micellze. The amount of enclosed alkali and the equivalent aggregation may be obtained indirectly from the trans- port measurements and the quantifies so obtained agree within a very little with those obtained from conductivity measurements and with those obtained by Varga. That considerable quantities of alkali are enclosed is shown by the fact that a t the cathode instead of an increase in alkali concentration occurring a decrease is actually found. From the decrease in alkali concentration the amount enclosed can also be calculated and the quantity so obtained agrees well with the quantity obtained by the other methods.With increasing alkali content the equivalent aggregation decreases as also does the percentage of enclosed alkali. On account of the small vaIue of the conductivity of the sols with small alkali content the composition can only be regarded as correct in respect of its order and not as to its actual value. An example of the composi- tion of a stannic acid sol examined is expressed as follows zH,O:SnO,H'). _ - +n;K'. J. F. S. Magnetic Analysis of the Stannic Acids. PAUL PASCAL (Compt. rend. 1922 175 1063-1065; cf. this vol. ii 25).- The theoretical values of the molecular magnetic susceptibilities of the acids Sn(OH) and SnO(OH) calculated from those of stannates methylstannonic acid and metallic tin were compared with those obtained experimentally from stannic acids prepared by various methods.The results indicate that the acids are not definite compounds but mixtures of anhydrous stannic oxide with water. H. J. E. Zirkite Ore. J. G. THOMPSON ( J . Physical Chem. 1922 26 812-832) .-The mineral zirkite an impure zirconium dioxide containing silica iron and titanium has been investigated with the object of making it suitable for use as a refractory. It is shown that 90-95% of the silicon may be removed from the mineral by heating a mixture of the ore and carbon to a tem- perature above 2220" in an electric furnace. The best results are obtained by heating a mixture of the ore and coke in an arc furnace the amount of carbon being approximately that required to transform only the silicon into carbide.If carbon in excess of this amount is used the separation of silicon is less complete due to the formation of stable double carbides of silicon and zirconium. It is suggested that if the above treatment is followed by the removal of iron by treatment with chlorine a product impure 4-2ii. 80 ABSTRACTS OF CHEMICAL PAPERS. zirconia is obtained which is sufficiently refractory for most purposes. Zirconium carbide is shown to be exceedingly refractory and infusible in a 40-50 kw. arc but its use as a refractory is restricted because it is necessary to protect it from oxidation. The various methods of analysis of zirconium compounds are discussed and a method for the estimation of zirconium silicon titanium and iron in zirkite is outlined.J. F. S. The Missing Element of Atomic Number 72. D. COSTER and G. HEVESY (Nature 1923 111 79) .-From theoretical con- siderations it is probable that the element of atomic number 72 is not a rare-earth metal but a homologue of zirconium. Further the conclusions of Dauvillier (A. 1922 ii 463) and of Urbain (A. 1911 ii 115) are considered not to be justified since it appears that the quantity of the element in the sample examined by the former by an X-ray spectroscopic method must have been so small that its identity with the element claimed to have been detected by the latter in the same sample by optical and magnetic examination is improbable. The only lines La and L& detected by Dauvillier were extremely faint and the wave-lengths given are about 4X.u.(lX.u.=lO-ll crn.) smaller than those obtained by a rational interpolation in the wave-length tables of Hjalmar and Coster for the elements in the neighbourhood of atomic number 72. The authors have however by investigation of the X-ray spectrum of extractions of zirconium minerals detected the lines La,=1565*5 a2=1576 p,=1371*4 p,=1323.7 pa= 1350.2 y,=1177 X.U. which must be ascribed to the element of atomic number 72. The L a and Lp lines exactly agree with the anticipated mutual distance and relative intensity and the wave- lengths all agree within one X.U. with those obtained by inter- polation. A sample of a Norwegian zirconium mineral was estim- ated to contain a t least 1% of the new element whilst ordinary zirconium probably contains 0*01-0.1 yo.Experiments are in progress to isolate the new element for which the name " hafnium " is proposed and to determine its chemical properties. [Cf. Scott T. 1923 123 311.1 The Anode Oxidation of Gold. FR. JIRSA and OT. BURYBNEK (Chem. Listy 16 299-305; cf. A. 1922 ii 713).-A strong current was passed through a cell having a gold anode and dilute sulphuric acid as electrolyte. After twenty-four hours the current was stowed and the back E.M.F. of the anode measured. For A. A. E. the cell* AAu~Au,0,,NH2S04~NH2S04,H,~Pt this was + 1-24 volt's a t 18". The same voltage was obtained whether the anode was prepared in the dark or-in the light of a mercury-vapour lamp and reduction of the strength of acid from normal to N/1000 made no difference to this result.The rates of polarisation of gold anodes using currents of 0-02 and of 0.5 milliampere/cm.2 were next measured and the results shown on a graph. It was noticed that below 0.02 milliampere the results obtained were untrust- worthy owing to the slow rate of oxidation producing passivity in the gold. The rate of discharge of polarised gold anodes wasANALYTICAL CHEMISTRY. ii. 81 measured under various conditions and with the introduction of different resistances in order to retard the rapidity of discharge. Gold anodes were next depolarised by passing currenta of increasing magnitude through the cathode until the polarisation E.M.P. disappears. The magnitudes of the polarisation currents produced by stationary and rotating oxidised gold anodes using acids of different concentrations were measured and curves drawn showing the relation between the galvanometer readings and the voltage.R. T. The Occlusion of Hydrogen by Palladium. MITUO YAMADA (Phil. Mag. 1923 [vi] 45 241-243).-The crystalline structure of palladium containing large quantities of absorbed hydrogen has been determined by X-ray analysis. Except for slight con- tractions of the intervals between the lines on the rontgenogram there is practically no difference between the figures for pure palladium and for this metal containing occluded hydrogen. The slight contraction corresponds with a uniform expansion of the space lattice. No new lines were observed and hence a compound is not formed between the hydrogen and the palladium. The results are in accord with the view that the hydrogen is adsorbed in the state of solid solution. The expansion of the lattice deter- mined from the displacement of the lines was 2*8% which is in good agreement with the expansion 29% obtained by direct measurement. W. E. G. Mi n e r a1 o g ie a I C h e mist r y. Riebeckite from Vallone delle M i d r e (Valle della German- asca). E. GRILL (Atti R. Accud. Lincei 1922 [vJ 31 ii 137- 142).-Black crystals of riebeckite up to 7 mm. in length and with submetallic lustre and pronounced prismatic habit were found on the eastern slope of Vallone delle Mini&re (Piedmont). They exhibit pleochroism and have d 3.44 ; the angle (100) (110 is 55" 5'455" 20'. SiO TiO Al,O Fe,O FeO MnO CaO MgO Na,O K,O H,O- H,O+ Total 51.53 trace 2.71 16-22 17.11 0.22 0.79 1-01 7-81 1.13 0.31 1.20 100.04. These results are in accord with the empirical formula Analysis gives :