INORGANIC CHEMISTRY.IN preparing the Report for 1927 it has been found, as in previousyears, that much work of interest and importance has perforce t obe omitted. Three subjects have been chosen for special con-sideration in the earlier paragraphs, and at the end of the Reporta list of the systems investigated during the year has been given.Otherwise, the scheme used in previous years has been followed.Valency.An important contribution to the theory of chemical combinationhas been made by N. V. Sidgwick in his Presidential Address toSection B of the British Association Meeting at Leeds. In additionto the generally accepted forms of linkage (a) electro-valent, inwhich electrons are transferred from one atom to another, and ( b )co-valent, in which the unit linkage consists of two shared electronsderived one from each of the linked atoms, he postulates a thirdform of linkage (c) co-ordinate, in which the unit linkage consistsof two electrons, shared in the manner characteristic of the co-valentlinkage, but both derived from one only of the linked atoms.Thissimple extension of accepted views explains all the peculiarities ofco-ordination compounds, of which the most important are theability of apparently saturated molecules such as water or ammoniato combine further, the existence of a valency limit (the co-ordinationnumber) independent of the Periodic Group to which the atombelongs, and the peculiar change in electro-valency that attendsthe replacement of a univalent radical such as chlorine by a wholemolecule such as ammonia.Although space precludes their detailed consideration here, it isextremely interesting t o follow these explanations and their furtherimplications, e.g., with reference to the mechanism of hydrolysisof the chlorides of the non-metals and the great stability of carbontetrachlpride, sulphur hexafluoride, etc.; these are fully and clearlyset out in the address and in the author’s book,l to which the readeris referred.N. V. Sidgwick, “ The Electronic Theory of Vdency,” Clarendon Press,Oxford, 192738 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Molecular Association.Remarkable results have followed from a new attack on theproblem of molecular association by the investigation of the effectof catalysts on the equilibrium between the complex and simplemolecules in a liquid.Acetic acid in contact with charcoal,platinum-black, or thoria had a vapour pressure greater than itsnormal value by 1-2 mm. of mercury. This difference in vapourpressure was increased after a period of heating and diminishedafter a period of cooling. The vapour pressure of benzene in con-tact with nickel was increased about 3 mm., that of methyl alcoholwith charcoal by 6 mm., and that of ether with charcoal by as muchas 40 mm., and other results of the same type have been obtained.Values for the molecular weights, calculated from surface-tensionmeasurements for liquids in contact with catalysts, seem to showthat, except in the case of acetic acid, the molecular complexity isincreased by heating the liquid for a short period, but is decreased bya longer period of heating.2 Similar changes have.been observed inthe density of water and ether when in contact with such catalystsas carbon and t h ~ r i a .~ There is no quantitative relation betweenthe vapour-pressure and surface-tension data, but this is hardly tobe expected, for the former depend chiefly upon the presence oflighter molecules in the liquid, while the latter tend rather tomeasure the complexity of the molecular species which accumulatesin the surface.In extension of the work on intensive drying, it has been foundthat nitrogen tetroxide, dried by repeated distillation over phos-phoric oxide, for a period of 4-6 months, has a vapour pressuregreater than the normal value by as much as 25 mm.After asudden change in the temperature of the dried material, thecorresponding change in vapour pressure was established slowly,several hours being required to attain a steady value.4 Newdata have been obtained for the vapour pressures of intensively-dried hexane, carbon disulphide, ethyl bromide, and nitrogenper~xide.~ Dried ammonium chloride has a vapour pressure lowerthan its normal value, and although internal transformations arenot inhibited, the equilibrium is shifted, the shift being greaterthe higher the temperature. These effects vanish above about310", and as the temperature is lowered association in the vapourbecomes appreciable at 286" and increases progressively.H. B. Baker, J., 1927, 949.8 J.B. Peel, P. L. Robinson, and H. C. Smith, Nature, 1927, 120, 614; A.,1019. 4 J. W. Smith, J . , 1927, 867; A., 506.A. Smits, 2. physikal. Chem., 1927, 129, 33; A,, 1027.Idem, Rec. trav. china., 1927, 46, 445; A., 819INORGANIC CHEMISTRY. 39Photosynthesis.The work on photosynthesis of organic compounds from carbondioxide and water, begun some years ago by Baly and his co-workers, has, during the past year, been brought to a stage ofquite extraordinary interest and importance. It has now beenshown that when aqueous carbonic acid is irradiated with ultra-violet light, a photostationary state is established involving, asone component, a complex aldehyde, the reaction being probably6H2CO,~C6H,,O,, + 60,. In attempting to assist the formationof carbohydrates by introducing a reducing agent, it was found thatferrous bicarbonate solution when irradiated deposited ferrichydroxide, and that organic compounds with reducing propertieswere simultaneously formed ; also, it was observed, the reactionoccurred primarily upon solid surfaces.Hence experiments weremade in which various insoluble solids capable of adsorbing carbondioxide (aluminium powder, barium sulphate, freshly precipitatedaluminium hydroxide, and the basic carbonates of aluminium,magnesium, and zinc) mere suspended in water through which astream of carbon dioxide was passed while the suspension, con-tained in tubes of quartz or Uviol glass, was subjected to the lightfrom a quartz mercury-arc lamp.Evaporation of the filteredsolutions yields organic residues, apparently of the nature of com-plex carbohydrates. Rigid proof was obtained that the materialsused were free from organic impurity, and negative results wereobtained in numerous control experiments. The most strikingand conclusive of these depended on the fact that aluminiumhydroxide, after a few hours, becomes unable to adsorb carbondioxide: thus, using the same tube, light, water, and carbondioxide, the same aluminium hydroxide when fresh gave the usualyield of carbohydrate, after 8 hours gave a small yield, and after24 hours gave no trace of organic residue.A discovery of still greater importance is that visible light maybe used in similar syntheses if the solid suspended in the solutionand capable of adsorbing carbon dioxide is coloured.Suitablesolids are nickel or cobalt carbonates, which must be free fromalkali, nitrate, chloride, and sulphate. In this case, one of theproducts is a carbohydrate which reduces Benedict’s solution,gives well-marked Molisch and Rubner reactions, and forms a solidosazone. The total quantity of material photosynthesised, and alsoits percentage reducing power, is greater than with a white surfacein ultra-violet light, probably because, with visible light, there islittle chance of photochemical decomposition of the products.There is, apparently, great similarity between these photochemicalprocesses and those in the living plant, e.g., in that-both sho40 ANNUAL REPORTS ON THE PROGRESS OF CHElKISTRY.similar fatigue effects and both give quantitative yields of the sameorder, and a possible mechanism of the process in vivo is advancedby the authors.' The further developments of this work will beawaited with interest,Atomic Weights.Argon.The most probable weight of a litre of pure argon hasbeen computed to be 1.7833 0.0001 g. ; whence the atomic weightA = 39.94 is deduced. Although appreciably higher than thatadopted by the German Commission (1923), this is regarded as aminimal value.8Potassium. Nephelometric determinations of the ratios KCl : Agand KCl : AgCl give as a mean of 31 concordant results, 0.691147 forthe former and 0.520186 for the latter ratio, whence K = 39-104 &0.0014.9 Conversion of potassium nitrate to the chloride by heat-ing in hydrogen chloride gives values for the ratio KNO,:KCl,from which, by the use of N = 14.008 and well-ascertained valuesfor the ratios AgCl : Ag and KCl : Ag, the following atomic weightsare calculated : K = 39.104 0.0007 ; Ag = 107.879 & 0.0011 ;C1 = 35,456 & 0*0003.10The value for silver (107.864 j, 0.0013) obtained by Bakerand Riley and previously reported l1 leads to values for chlorine(35.452) and nitrogen (13.999) which are regarded as improbable,and it has been suggested that the low value for the atomic weightcould be explained by losses of silver (about 0.1 mg.) occurringthrough the volatility of the metal at 1000°.12 In reply, Bakerand Riley have pointed out that their experimental conditionspreclude loss of silver, that there is no evidence of deposition ofsilver in the cooler parts of the apparatus, and that the silver,after repeated fusion within the apparatus, attained constancyin weight within 0.014-02 mg.They have also obtained directexperimental confirmation that loss of silver by volatilisation didnot occur.13The atomic weight of silver has been determined by reducingsilver nitrate with hydrogen. Errors due to adsorption of air on thesilver nitrate and on the reduced silver were eliminated by effecting7 E. C. C. Baly, 5. B. Davies, M. R. Johnson, and H. Shanassy, Proc. Roy.Soc., 1927, [A], 116, 197; -A., 1040; E. C. C. Baly, W. E. Stephen, andN. R. Hood, ibid., p. 212; A., 1041; E. C. C. Baly and J. B. Davies, ibid.,p. 219; A., 1041.Silver.* E.Moles, Ber., 1927, 60, [B], 134; A., 182.lo E. Zintl and J. Goubeau, ibid., p. 302; A., 806.l1 Ann. Reports, 1926, 23, 50.la B. Brauner, Nature, 1927, 119, 348, 526; A., 289, 493.la H. B. Baker and H. L. Riley, aid., p. 348, 703; A,, 289, 493.0. Honigschmid and J. Goubeau, 2. anorg. Chern., 1927,163,93; A,, 806INORGANIC CHEMISTRY. 41all weighings and the final fusion of the silver in a vacuum. Theatomic weight of nitrogen being taken as 14.008, ten highly con-cordant determinations give the value Ag = 107.879 & 0.0014.14In continuation of previous work on the variation in theatomic weight of boron derived from different thedensities of three samples of boron trichloride, portions of theoriginal materials used in the determination of the ratio BCI, : 3Ag,after further purification were compared by the flotation methodand the relative atomic weight of the boron in each was calculated.The collected results for the three samples of boron from California,Tuscany, and Asia Minor, respectively, are : by analysis, 10.841,104325, and 10.818 ; from densities of boric oxide, 104347, 104323, and10418 ; and from densities of boron trichloride, 10.830, 104325,and 104317.l6Scandium. Scandium chloride, prepared by passing carbontetrachloride diluted with nitrogen over the oxide at 750-850",purified by fractional sublimation, and weighed with special pre-cautions for the exclusion of moisture, has been used for a determin-ation of the ratio ScCl, : 3Ag.The mean of 9 analyses gave the valueSC = 45.160.l'Yttrium. Yttrium chloride, spectroscopically free from otherrare-earth metals, has been used for determinations of the ratioYCI3:3AgCl. The mean of 10 analyses gave the value Y =88.925 & 0-002.18Dysprosium. Similarly, dysprosium chloride, containing onlyabout 0.1% of holmium, has been used for 7 determinations of theratio to silver chloride, giving as a mean, after correction for theknown impurity, Dy = 162.459 & 0-004.19Lead. The constitution of ordinary lead has been investigatedby using lead tetramethyl in the mass spectrograph ; three principallines, 206, 207, 208, of intensities, respectively, 4, 3, 7, are in goodagreement with the accepted atomic weight, 207.2, and there is,also, a faint line at 209, and some evidence of lines at 203,204, and205.2QNitrogen.The average values for the densities of nitrogen a t 0"Boron.14 0. Honigschmid, 33. Zintl, and P. Thilo, 2. anorg. Chem., 1927, 163, 6 5 ;15 Ann. Reports, 1925, 22, 43; 1926, 23, 49.16 H. V. A. Briscoe, P. L. Robinson, and H. C. Smith, J., 1927, 282; A,,17 N. H. Smith, J . Amer. Chem. SOC., 1927,49, 1642; A., 806.18 0. Honigschmid and A. von Welsbach, 2. anorg. Chem., 1927, 165, 284;l@ Idem, ibid., p. 289; A., 915.80 F. W. Aston, Nature, 1927, 120, 224; A., 806.A,, 806.392.A., 915.B 42 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.and 253.33, 506.67, and 760 rum. pressure (at sea-level, in latitude45') have been accurately determined, and are 0.41667, 0433348,and 1.25036, respectively.If the value PV a t 1 atmosphere beunity, the values at 8 and Q atmosphere are 1.00011 and 1.00028,respectively, the average value of the coefficient of deviation fromBoyle's law between 0 and 1 atmosphere being -0.00045, a value inagreement with existing data. These results lead to the value14.007 for the atomic weight of nitrogen.21In a complete synthesis of silver chloride, chlorine,weighed as the liquid, was reduced to hydrogen chloride by meansof ammonium arsenite, and precipitated with a weighed quantity ofpure silver dissolved in nitric acid, and the silver chloride so formedwas weighed. As a mean of 9 determinations the ratio C1: Ag wasfound to be 0.328668, whence, if Ag = 107.880, the atomic weight ofchlorine is C1= 3.5~457.~~ Since the weight of silver chloridediffered from the weights of the constituent elements by less thanthe experimental error, doubts that have been expressed 22 as to thepurity of atomic-weight silver seem to be unfounded.It has further been shown that samples of hydrogen chloride,prepared from the extreme fractions obtained in a fractional dis-tillation of carbon tetrachloride, when used for determinations ofthe ratio Ag : AgCl gave substantially identical values, whence itappears that ordinary fractional distillation does not effect anyappreciable isotopic ~eparation.~~Saturated solutions of pure sodium chloride, prepared from fourdifferent samples of Alsatian potash minerals, differed in densityfrom one another and from a similar solution prepared from com-mercial salt by less than the experimental error ; thus any differencein the isotope ratio is too small to be detected by the methods used.24Chlorine.Group I .Much further work has been done on the reduction of aqueoussolutions of metallic salts by hydrogen under pressure.25 Crystallinehydroxides of aluminium and chromium have been prepared byheating solutions of the corresponding nitrate acidified with nitricacid to between 320-360", under a pressure of 200-370 atmospheres21 G.P. Baxter and H. W. Starkweather, Proc. Nat. Acad. Sci., 1926, 12,23 P. A. Guj7e and E. Moles, J. Chim. physique, 1917, 15, 360; A . , 1918,23 0. Honigschmid, S. B. Chan, and L. Birckenbach, 2. anorg.Chem.,24 (Mlles.) E. Gleditsch and L. Gleditseb, J. Chim.. physique, 1927, 24, 238 ;96 Ann. Reports, 1926, 23, 53.703; A., 1927, 194.ii, 40.1927, 163, 315; A,, 806.A., 493INORGANIC CHEMISTRY. 43of hydrogen, for 12-24 hours. The products resemble diaspore andchrome ochre, respectively. If air be used in place of hydrogen, theproduct is less crystalline and the yield is not quantitative, a portionof the oxide being converted into chromic acid.26 The action ofhydrogen on solutions of copper formate and acetate between90-180", a t pressures up to 150 atmospheres, and at acidities upto 12-5N, has proved interesting. I n neutral solution at loo",hydrolysis occurs with the separation of copper oxide, which isonly with difficulty reduced by hydrogen.The presence of acidrepresses this hydrolysis and the dissolved cupric salt is reduced tothe cuprous state. At moderate hydrogen-ion concentration, thecuprous salt is hydrolysed with precipitation of crystalline cuprousoxide, but here a.gain an increase of acidity inhibits hydrolysis andpermits further reduction to metallic copper. At higher temper-atures (130-180"), the anions decompose, yielding hydrogen whichcontributes to the foregoing reactions.27 Stannic hydroxide isreduced to the stannous state by hydrogen a t 300" and 38 atmo-spheres, and to the metal a t 350" and 50 atmospheres. Stannicsulphate at 302" and 162 atmospheres gives successively stannoussulphate and stannous sulphide : the reduction is retarded by thepresence of sulphuric acid or a sulphate.Stannic chloride isreduced to stannic oxide at 280-300" and 110 atmospheres, and tometallic tin a t higher temperatures. Addition of silver chloridegave, a t 380" and 260 atmospheres, mainly basic stannous chloride.28As with platinum, the production of metal by the action of hydrogenunder pressure on chloroiridate solutions a t 100" and 103" is greatestat low concentrations and increases with pressure, temperature, andtime. Under similar conditions, iridium is deposited more com-pletely than platinum. In cases of incomplete precipitation, ablue, unstable solution of colloidal iridium is produced .29A quantitative investigation has shown that the output of activehydrogen in a Siemens ozoniser, as measured by the reducingaction of the gas on sulphur, varies inversely as the rate of flow ofthe gas through the apparatus, varies directly with the pressure, andshows no linear relation to the voltage used, being negligible above70 mm.pressure and below 3000 volts.30 A thorough examinationof the processes previously stated to yield triatomic hydrogen leadsto the conclusion that there is no evidence for the existence of thissubstance. Hydrogen which has been passed through hot palladium,26 V. Ipatiev and B. Mouromtsev, Ber., 1927, 60, [B], 1980; A., 1043.2 7 V. Ipatiev and V. Ipatiev, jun., ibid., p. 1982; A,, 1042.28 V. Ipatiev and V. Niklaev, Compt. rend., 1927, 185, 462 ; A., 950.2Q V. Ipatiev and J. Andreevski, ibid., p. 357; A., 844.30 G. A. Elliott, Trans.Paraday SOC., Jan. 1927, advance proof; A., 18744 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.or submitted to a silent discharge in an ozoniser, and then passedover sulphur does indeed blacken lead acetate paper, but thesame effect is obtained if the passage over sulphur is omitted. Anumber of substances, including palladium, platinum, copper, andglazed porcelain, when heated in hydrogen at 600-700", yieldhydrogen sulphide, doubtless derived from a small sulphur contentin these substances : after prolonged treatment they cease to yieldhydrogen sulphide. This accords with the Reporters' experiencethat the most highly purified samples of magnesia contain a traceof sulphur, which is eliminated as hydrogen sulphide when themagnesia is heated to redness in hydrogen.Similarly, a glassozoniser which initially yields hydrogen sulphide ceases to do soafter it has been in use for a time. The ammonia supposed to beformed by the action of nitrogen upon hydrogen generated frommagnesium and acid has been shown to persist if no nitrogen isused and to arise, in fact, from the decomposition of magnesiumnitride present as an impurity in the magnesium. No reduction ofpermanganate or indigo could be obtained with the gas remainingafter exploding oxygen with excess of hydrogen, and it was shownthat hydrogen subjected to a-radiation from thorium-B andthorium4 did not thereafter reduce sulphur.31A new technique for the preparation of hydrogen peroxide hasbeen described, involving the decomposition of sodium peroxidewith 20% sulphuric acid, removal of sodium sulphate by crystal-lisation as decahydrate, and vacuum distillation under specifiedconditions. It gives a fair yield of 85% peroxide.32 It has beenshown that the decomposition of dust -free samples of hydrogenperoxide heated in smooth vessels is very slow, that solutionscontaining dust give linear decomposition curves, and that in-hibitors evidently act by poisoning the surfaces catalysing thedecomposition.33 By studying the change of pressure at constantvolume in a quartz bulb (the results in glass are not reproducible),it has been shown that the hydrogen peroxide vapour, admixed withwater vapour and oxygen, evolved from 60% aqueous hydrogenperoxide at 85", consists of simple non-hydrated molecules.Thethermal decomposition at 85" is a reaction of zero order, inhibitedby molecular oxygen ; so that the decomposition ceases when about20 yo of the hydrogen peroxide is dec~mposed.~~31 F. Paneth, E. KIever, and K. Peters, 2. Elektrochem., 1927, 33, 102;A., 429; M. Scanavy-Grigorieva, 2. anorg. Chem., 1926,159,55; A., 1927, 119.32 M. L. Kilpatrick, 0. M. Reiff, and F. 0. Rice, J . Amer. Chem. SOC., 1926,48, 3019; A,, 1927, 120.33 F. 0. Rice and 0. M. Reiff, J . Physical Chem., 1927, 31, 1352; A., 1035.34 L. W. Elder, jun., and E. K. Rideal, Trans. Paradav Xoc., 1927.23. 546;A,, 1038INORGANIC CHEMISTRY. 45Anhydrous lithium iodide decomposes at high temperatures in astream of oxygen according to the equation lOLiI + 502+-BLiIO, + 4Li20 + 412.35 A mono- and a tri-hydrate of lithiumchlorate are prepared by seeding solutions corresponding withLiC103,H20 and LiC103,3H,0 at -28" and -20", respectively, withcrystals of the hydrate 3LiC103,H20.36The colourless needles of sodium hydride prepared by directunion of hydrogen and sodium belong to the regular system, andX-ray examination by the Debye-iScherrer method shows that thecrystal lattice is probably similar to that of rock-salt. The dia-sociation of the hydride a t reduced pressures and at ordinarytemperatures is that to be expected with a solution of hydrogen insodium rather than a definite compound.At higher temperatures,however, the dissociation is accompanied by the separation ofsodium, free or almost free from hydrogen.,'A simple and convenient method for preparing small quantities ofpure potassium, rubidium, or czsium depends on the fact thatbarium azide, conveniently prepared by distilling azoimide intobarium hydroxide, decomposes a t ZOO"; so that if alkali chloridesare mixed with the azide in solution, and the solution is evaporatedto dryness in a vacuum, the solid residue on heating to a moderatelyhigh temperature yields a distillate of the alkali meta13*It has been found that the colours shown by a copper oxide filmmay be made truly homogeneous if special precautions are takenin the preparation and activation of the film.The study of suchfilms affords much evidence that the colours are produced byinterference in a thin layer of oxide.The order of production ofthe colours corresponds with the order tabulated by Rollet for theinterference colours of air films of increasing thickness seen bytransmitted light. It is found that the fall in electrical conductivityof the oxidised film supported on china clay is proportional to theequivalent air thickness of the oxide film, i.e., to the thickness of anair film which would producc the same colour as the copper oxidefilm. The equivalent air thickness is approximately proportionalto the quantity of metal oxidised. The wave-lengths of the maximain the absorption or reflexion bands in the spectrum of the light re-flected from the film move towards the red as the thickness increases,85 J. P. Simmons and C.F. Pickett, J . Amer. Chern. SOC., 1927, 49, 701;36 L. Berg, 2. anorg. Chern., 1927, 166, 231; A., 1042.37 G. F. Hiittig and F. Brodkorb, {bid., 1927, 161, 353; A., 529; compare** J. H. de Boer, P. Clausing, and G. Zecher, 2. anorg. Chem., 1927, 160,A., 429.Ann. Reportc~, 1926, 23, 63.128; A,, 32846 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.and the absolute thickness of the oxide film, calculated from thedensity and the mass of oxygen taken up per of the surface,agrees moderately well with the thickness calculated from theposition of the absorption band and the corresponding refractiveindex .39Curves connecting the absorption as a function of the compositionin solutions of cupric ion and ammonia in concentrated ammoniumnitrate, showed marked maxima, corresponding in the case ofequimolecular mixtures to the formula Cu(NH3),.For non-equimolecular solutions, the dissociation constant was calculated,and was found to agree with the result obtained from the absorptiondata. The tetra-ammonia ion, therefore, is probably the onlycupriammonia ion stable a t the ordinary temperat~re.~~ A com-plex thiosulphate, Cu,,S15018,9NH3, has been prepared by mixinga solution of sodium thiosulphate in concentrated ammonia withcupric chloride. The substance forms stable deep blue crystals.41Unsuccessful attempts have been made to discover the missingelement ,87 produced by the loss of an a-particle from mesothorium-2or as a P-product from radon.42Group I I .The process of dehydration of the hemihydrate of calcium sulphateabove 80" is reversible, but the dehydration does not take placespontaneously.Below 80", the water-vapour pressure of thesystem lies below 2 mm. of mercury. The decomposition of thehemihydrate into a mixture of dihydrate and anhydrous salt istheref ore thermodynamically impossible, as the regeneration ofthe hemihydrate could not take place. Such a decomposition ispossible only if two forms of the soluble anhydrous salt exist, vix.,an u-form stable below 72" and incapable of existence in equilibriumwith the hemihydrate, and a p-form stable above 72" and in thepresence of hemihydrate. Dilatometric measurements confirm theexistence of two forms : the transition a e p occurs a t 82", and isrendered evident also from the nature of the heating and coolingcurves, the p-form being produced from the a with the evolution ofheat.These facts constitute a probable explanation of contradic-tions in the literature concerning the crystal form of anhydrouscalcium sulphate .4339 F. H. Constable, Proc. Roy. SOC., 1927, [A], 115, 670; A., 930.40 P. Job, Compt. rend., 1927, 184, 204; A., 205.41 D. W. Horn and R. E. Crawford, Amer. J . Pharm., 1927, 99, 274; A.,42 G. von Hevesy, Kgl. Danske Videnskab. Selsk. math.-fys. Medd., 1926,43 D. Balarev [with A. Spassov], 2. anorg. Chem., 1927, 163, 137; A,, 829.634.7, No. 11, 1; A., 1927, 289INORGANIC CHEMISTRY. 47Large, violet mixed crystals of barium sulphate with potassiumpermanganate, which are not decomposed by oxalic acid in thepresence of sulphuric acid or by sulphurous acid, are obtained byallowing 2N-solutions of sulphuric acid and barium chloride todiffuse slowly into a 10% solution of potassium permanganate.'Doubt has been thrown upon the volatility of barium sulphatepreviously reported 45 on the strength of the observations that thesalt moistened with sulphuric acid colours the flame of a Bunsenburner green and that the flame obtained by igniting the vapoursfrom a heated mixture of barium nitrate, concentrated sulphuricacid, and methyl alcohol is also green.It has been suggested thatthe former coloration is due to the formation of a spray of finelydivided barium hydrogen sulphate or pyrosulphate, and in thesecond case it has been shown that the colour is probably due tomethyl nitrate, as similar colorations are obtained by substitutingstrontium or ammonium nitrate for the barium salt.46For the separation of radium from barium, the fractional pre-cipitation of the chromate in 0.05-0-3N-acid solutions is as efficientas the bromide method and is advantageous when only smallquantities of the mixture are a~ailable.~'The perchlorates, as drying agents, have been further in-~ e s t i g a t e d .~ ~ Barium perchlorate trihydrate on dehydrationin a vacuum a t 100-140" gives the anhydrous salt withoutfusing, and this equals sulphuric acid in drying efficiency. Amixture of barium perchlorate with up to 35% of magnesiumperchlorate can be dehydrated at 250" and 102 mm. pressure withoutfusing ; the granules of such a mixture containing 26.5% of mag-nesium perchlorate in a drying column 15 cm.long and 2.5 cm. indiameter allow only 0.001 g. of water to remain unabsorbed whenair 60% saturated with moisture is passed through for 5.5 hoursat the rate of 53 litres per hour. The mixture is, therefore, anextremely efficient drying agent .49Group I I I .Molten mixtures of a metallic oxide, boric anhydride or borax,and a fluoride, when electrolyaed with high cathode current densities,have yielded in a molten or crystalline state most of the metals with44 W. Geilmannand E. Wiinnenberg, 8. anorg. Chem., 1927,159,271 ; A., 120.4 5 F. Krauss, Chem.-Ztg., 1926, 50, 3 3 ; A., 1926, 368.46 F. L. Hahn, ibid., p. 934; B., 1927, 106; compare Krauss, ref.45;4 7 1;. M. Henderson and F. C. Kracek, J . Amer. Chem. SOC., 1927, 49, 738;4 ~ 3 Compare Ann. Reports, 1922, 19, 45.49 G. F. Smit,h, I n d . Eng. Chem., 1927, 19, 411; A., 438.F. Krauss, ibid., 1927, 51, 38; B., 139.A., 43148 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.heats of oxidation less than that of sodium.50 A similar methodhas been applied to the production of boron, the mixture 2B,03 +MgO + MgF, being used a t 1100", in a charcoal crucible which actsas anode, the cathode being an iron rod. The cathode deposit,consisting of boron agglomerated by the solidified electrolyte, isground and extracted with hydrochloric acid, giving a 95% yield ofboron of 92 yo purity.51The constitution of the hydrides of boron, discovered by Stockand his coLworkers 52 and formulated by them on the assumptionthat the valencies involved were 3 and 4, has been reviewed byseveral authors.It is suggested that a more satisfactory explanationof the compounds may be based on the assumption that the valenciesare 3 and 5 and the co-ordination number 4. On this basis, B,H,would be formulated H3B:BH3, the ammonia compound either asthe salt (H,B:BH,--)(NH,+), or, less probably, H,B:NH,, andthe hydride B,H1, as H,B:BH,*BH,:BH,. The comparativelygreat stability of hydrides with 6 , 6, and 10 boron atoms is mostsatisfactorily explained by the adoption of ring formulae as follows :BH,=FH, BH,:BH:BH, BH,:BH-$*BH,:BH,hH ,*BH*BH, BH,:BH:hH, BH,:BH,*B*BH=BH,Formula (I) accounts for the fact that B,H, combines with 4molecules of ammonia to give B,H,(NH,),, which with hydrogenchloride evolves 4 molecules of hydrogen rapidly and a further 3molecules more slowly. The naphthalene structure of (111) is inagreement with its chemical properties and its high melting point.53Electronic interpretations of the constitution of the boron hydrideshave been advanced by Ulmann and by Main Smith.54Investigations of the freezing point of boric acid solutions, andof the solubility of the acid in water in the presence of variousproportions of hydrogen peroxide, together with the conductivityof these solutions, have thrown little or no light on the constitutionof the acid, but conductivity measurements show that the mono-borates are binary electrolytes and give no evidence of a ternarydissociation corresponding with formule of the type M,B,O,.- +(1.1 (11.) (111.)5O Andrieux, Compt.rend., 1927, 184, 9 1 ; A., 216.61 L. Andrieux, ibid., 185, 119; A,, 844.6a Compare Ann. Repork, 1926, 23, 68.64 M. Ulmann, Ber., 1927, 60, [BJ, 610; A., 399; but see A. Stock, Ber.,1927,60, [BJ, 1039 ; A., 714, who regards these explanations as unsatisfactory ;see also E. Muller, ibid., p. 1323; compare E. Miiller, 2. Elektrochem., 1925,31, 382; A., 1925, ii, 841; J. D. Main Smith, Chem. News, 1927, 135, 81;A., 813.J. A. Christiansen, 8. anorg. Chern., 1927, 160, 395; A., 399INORGANIC CHEMISTRY. 49Concentrated solutions of the corresponding diborates apparentlycontain B40," ions, rather than univalent diborate ions.For moredilute solutions, the experimental data are compatible with adecomposition of the diborate into monoborate and free boric acid.In concentrated solutions, the perborates behave as ternary elec-trolytes, giving the ion 2B0,,2H20~", whereas dilute solutionsapparently contain the ion B02,H202 . The fact that pentaboratesgive unusually high values for van 't Hoff's coefficient, i, is ascribedto a very considerable decomposition of the pentaborate ion, B,O,',into less complex ions and boric acid, which occurs even whenconsiderable excess of boric acid is added.55 The literature relatingto those borates of the alkali metals which can be crystallised fromaqueous solution has been critically reviewed, and X-ray diagramsand dehydration experiments have shown that the removal of thelast molecule of water from the hydrates of sodium monoborate isaccompanied by a change in the structure of the crystal.Withthe diborate it is apparently the last three water molecules whichare of structural importance, but the pentahydrate stable above 60"appears to occupy an exceptional position.56A large number of fluoroborates have been prepared, analysed, anddescribed; those of the heavy metals appear to be characterised bytheir great solubility, deliquescence, and instability at highertemp erat~res.~'A new scaly variety of aluminium hydroxide, A1,0,,4H20,d31' 1.5490, soluble in mineral acids, has been produced by thereduction of barium nitrate solution with the aluminium-mercurycouple at 0".One molecule of water is lost a t loo", and the lastmolecule, with difficulty, a t a red heat.58A vigorous discussion continues as to priority in the discoveryof element No. 61, but this can hardly be summarised here.S9 Itis more important to note that attempts to increase the amountss H. Menzel, 2. anorg. Chern., 1927, 162, 1, 22; A., 937.H. Menzel [with J. Meckwitz], ibid., 166, 63; A., 1043.67 E. Wilke-DSrfurt and G. Balz, ibid., 159, 197; A., 120; H. Funk andF. Binder, ibid., 1926, 159, 121; A., 1927, 219.HZ P. Neogi and A. K. Mitre, J., 1927, 1222; A., 741.69 W. Prandtl and A. Grimxn, 2. angew. Chem., 1926,39,1333; A., 1927, 9;L. Rolla and L. Fernandes, Gazzetta, 1926, 56, 688; A., 1927, 9; idem, 2.anorg.Chem., 1926, 157, 371; A., 1927, 31; R. Brunetti, Atti R. Accad.Lincei, 1926, [vi], 4, 518; A., 1927, 190; L. Rolla and L. Fernandes, Gaxzetta,1926, 56, 862; Atti R. Accad. Lincei, 1926, [vi], 4, 498; A., 1927, 190;R. Brunetti, ibid., p. 515; A., 1927, 190; W. A. Noyes, Nature, 1927, 119,319; A., 296; R. Brunetti, 2. anorg. Chem., 1927, 160, 237; A., 296; L.Rolla and L. Fernandes, ibid., p. 190; A., 296; W. A. Noyes, Nature, 1927,120, 14; A., 714; L. Rolla and L. Fernandes, Gazzetta, 1927, 57, 290; A.,611; B. S. Hopkins, J. Franklin Inst., 1927, 204, 1; A., 81450 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.available for X-ray exasmination have resulted in the production of asample containing about 1% of the element which gave seven linesof the L-series of illinium.6*Scandium acetylacetonate and salts of the types M3(ScF6) andK,[M(C,O4),],5H2O strongly resemble those of the tervalent elementsof the iron family, and further evidence of this analogy is afforded bythe preparation of the new compounds, which are isomorphous withthe corresponding compounds of the iron family :Samarous chloride has been prepared b y reducing the anhydroustrichloride in a mixture of ammonia and hydrogen a t 750".It formsa mass of dark reddish-brown needles, m. p. 740", readily soluble inwater to a solution which, on keeping, evolves hydrogen and depositsan insoluble oxychloride. The intense colour of the dichloride mayserve to detect samarium in the presence of large quantities of otherearths.62 A number of new compounds of europium have beendescribed, including the oxalate, nitrate, normal and hydrogentartrates, acetate, citrate, acetylacetonate, iodate, carbonate, and~yanoplatinate.~~ The ammoniates of the chlorides of lanthanum,cerium, praseodymium, and neodymium have been prepared andobservations made of their decomposition temperatures and of thecontraction in volume attending their formation.64Gallium has been used in a fused quartz thermometer which maybe employed to measure temperatures up to 1000".The removal ofgas from the liquid gallium and the effect of impurities on theamount of undercooling and on the tendency to wet quartz havebeen investigated. The ease of surface oxidation of gallium isgreater than is represented in the l i t e r a t ~ r e .~ ~Group I V .The vapour pressure of carbon suboxide (C30,) has been measuredbetween -62" and 4", giving a calculated b. p. a t atmosphericpressure of 6.8" and a heat of vaporisation of approximately 6 kg.-cal. The pure gas is very stable in contact with dry, clean glass,but decomposes in contact with mercury or glass contaminated withthe polymerised form.6660 J. M. Cork, C. James, and H. C. Fogg, Proc. Nut. Acad. Sci., 1926, 12,61 G. Urbain and P. B. Sarkar, Compt. rend., 1927, 185, 593; A., 1010.62 G. Jantsch, H. Riiping, and W. Kunze, 2. unorg. Chem., 1927, 161, 210;63 M. P. B. Sarkar, Bull. SOC. chim., 1927, [iv], 41, 185; A., 326.64 F. Ephraim and R. Block, Ber., 1926, 59, [B], 2692; A., 1927, 121.65 S.Boyer, J . Opt. Xoc. Amer., 1926, 13, 117; A, 1927, 100.66 M. J. Edwards and J. M. Williams, J . , 1927, 865; A,, 506.696; A., 1927, 190.A., 530INORGANIC CHEMISTRY. 51A brown solid, having the composition C,O,,xH,O, has beenproduced by circulating carbon monoxide, a t 20-69 cm. pressureand continuously freed from carbon dioxide, through a Siemensozoniser actuated by alternating electric current of the order of20,000 volts per cm. a t 250 cycles per second. This substance wasvery hygroscopic, reacted with water, giving approximately 1 mol.of CO, per mol. of C,O,, and formed a solution containing oxalicacid (1 mol. per 3 mols. C,O,), colloidal particles, and a dark; in-soluble residue.67A long series of experiments on the catalytic synthesis of hydro-cyanic acid from nitric oxide and hydrocarbons a t high temper-atures has led to the general conclusion that the first stage in thereaction is the reduction of nitric oxide to ammonia, part of whichreacts to form hydrocyanic acid, whilst part dissociates into hydrogenand nitrogen.Ethylene decomposes gradually, giving methane,hydrogen, and carbon, but acetylene and highly unsaturatedhydrocarbon residues (CH,= and CHZ) are probably intermediateproducts. The formation of hydrocyanic acid is then completed bythe action of ammonia either on ethylene or on one of these inter-mediate products. It is unlikely that solid carbon plays anyappreciable part in the reaction. Carbon monoxide is probablyproduced by the interaction of ethylene with the water vapourformed in the reduction of nitric oxide.6sThe reactions of alkali thiocyanates which have been stronglyacidified with sulphuric acid differ greatly from those of the neutralsalt.Thus ferric salts give a red colour which rapidly disappears,wliilst the white precipitates obtained with silver, lead, and mercurysalts slowly become yellow, and cobalt nitrate yields a green precip-itate of perthiocyanogen, (HC,N,S,), with traces of cobalt saltsand of perthiocyanic acid, (H,C,N,S,). When acids (preferablynitric acid) are added to a solution containing two or more thio-cyanates, precipitation of a complex thiocyanate of definite com-position may result. The precipitate must be collected immediately ;otherwise a violent reaction may take place after a few minutes,with the decomposition of the precipitate and evolution of oxides ofnitrogen, sulphur dioxide, hydrogen cyanide, etc.The compoundsHgCo(CNS) (blue) (which may be used as a fairly delicate qualitativetest for mercuric ions), PbBi(CNS), (red), and possibly CdHgBi( CNS)(red) have been obtained by this meth0d.6~Pure silicon has been prepared by interaction of pure silicontetrachloride vapour and hydrogen at the surface of a glowing67 R. W. Lunt and R. Venkateswaran, J., 1927, 857; A., 531.6 8 E. Elod and H. Nedelmann, 2. Elektrochem., 1927, 33, 217; A., 838.B. Ormont, 2. anorg. Chem., 1927, 161, 337; A., 53152 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.carbon filament 0.3 mm. thick, maintained at 1000" by an electriccurrent.The product has a metallic appearance, is very stable inair, and appears to become more brittle as the temperature rises.The coefficient of linear expansion between 18" and 950" is 3.55 x; the electrical conductivity increases with rise in temperature.?OThe absolute density and coefficient of thermal expansion ofsilicon tetrachloride have been found to be 1.481475 & 0*0,68/20"and 0*0014044 5 0*0687/20", respectively.71Pure germanium, prepared by reducing specially purifiedgermanium dioxide with hydrogen and graphite, has m. p. 959" inhydrogen, 958" in carbon dioxide, and 975" in a vacuum; it isvolatile in hydrogen at atmospheric pressure below SOO", and in avacuum below 760". Molten germanium (1 g.) absorbs hydrogen(0.1839 g.) on cooling and reduces germanium dioxide to the mon-0xide.7~ The metal has also been prepared by reduction of thedioxide with carbon under a flux of sodium chloride in a graphitecrucible heated in an induction furnace.73The formation of complex compounds of stannic and stannousiodides with other metallic and organometallic halides has beenextensively investigated, and a number of stable compounds havebeen reported.74The specific heats at 8-13" of chemically and physically purewhite and grey tin are reported as 0.0537 r f 0.0003 and 0.0493 &-0.0002, respectively,75 and the transition temperature of the twoforms is found to be 13".76 It has also been shown that the metalheated at 350" for some minutes gives a calorimetric curve withoutdiscontinuities, whilst if heated at 500" for 2 hours, it gives a curveshowing a sharp break a t 170-171", which is accentuated by in-crease in time or temperature.The dependence of such results onthe previous thermal treatment of the metal, irrespective of itspurity, explains the anomalous results previously obtained. 77Croup v.The view that active nitrogen is a metastable molecular formwith an energy of about 42,500 cal. per g.-mol. (Le., approximately7O R. Holbling, 2. angew. Chem., 1927, 40, 655; A., 844.7 1 P. L. Robinson and H. C. Smith, J., 1926, 3152; A., 1927, 102.72 J. H. Muller, E. F. Pike, and A. K. Graham, Proc. Amer. Phil. SOC.,73 (Miss) K. M. Tressler and L. AT. Dennis, J . Physical Chem., 1927, 31,74 T.Karantassis, Ann. Chim., 1927, Ex], 8, 71; A., 950.75 E. Cohen and K. D. Dekker, 2. physikal. Chem., 1927,127, 183; A,, 818.7 6 Idem, ibid., p. 178; A,, 818.7 ' A. Travers and Huot, C m p t . rend., 1927, 184, 162; A., 194.1926, 65, 15; A., 1927, 121.1429; A., 1046INORGANIC CHEMISTRY. 532.0 volts) '* is said to be difficult to reconcile with the spectroscopicA study of the reactions of active nitrogen with othergases of varying critical increments, confirmed the energy givenabove as a mean value, but was held to show that the gas containsthree species of active molecules, some having energies just aboveand others energies just below the mean value.8O From laterexperiments, the conclusion is drawn that the glowing and chemicallyactive forms of nitrogen are distinct, the former being due to re-combination of atoms with a heat of formation of 250,000 cal./g.-mol., and the latter, which possesses an energy of about 45,000cal./g.-mol., may be metastable molecular nitrogen or a more complexsubstance such as N3.81 The decay process, presumably of thisform, in contact with metallic filaments is bimolecular and gives anenergy value of 46,000 cal./g.-mo1.82 On the other hand, a study ofthe electrical behaviour of glowing active nitrogen indicates that itis molecular nitrogen in a metastable state with an energy between9.4 and 10.4Sodium azide when rendered unstable thermally in an atmosphereof oxygen is converted almost quantitatively into sodium nitriteand nitrogen.Catalysts or carriers are unnecessary, but the presenceof free alkali is essential to delay the transformation 3NaN-Na3N + N, until the oxygen molecule can penetrate the nitrogenzone.Oxidising agents such as copper oxide, manganese dioxide,and lead dioxide merely facilitate the remoyal of nitrogen byoxidising the sodium, whereas peroxides such as barium peroxidehave the same effect as gaseous oxygen, since they furnish molecularoxygen. It therefore appears probable that the direct combustionof ammonia to nitrite and subsequently to nitrate at a basic contactdepends on an unstabilising of the molecule, prior to thermal dis-sociation, which may be represented by the scheme HN-H,;molecular oxygen then penetrates the molecule and displaceshydrogen, which is subsequently oxidised to water.In a similarmanner, hydrogen converts sodium azide into nitrogen and sodamide,the greater velocity of diffusion allowing it to penetrate the nitrogenzone so rapidly that the presence of alkali is not necessary to preventthe transformation 3NaNzzr+ Na,N + N,. If sodium azide isdecomposed in an atmosphere of carbon dioxide, the NaN:::residues unite with loss of nitrogen and production of sodiumnitride, which is very readily hydrolysed to sodium hydroxide78 Ann. Reports, 1926, 23, 65.79 R. C. Johnson, Nature, 1927, 119, 9; A., 85.*O E. J. B. Willey and E. K. Rideal, J., 1927, 669; A., 431.s1 E. J. B. Willey, Natwe, 1927, 119, 924; A., 635.88 E. J. B. Willey, J., 1927, 2188; A., 1038.83 P. A. Constantinides, Phy&al Rev., 1927, [ii], 30, 95; A., 91654 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.and ammonia, even when the concentration of aqueous vapour isvery l0w.84The behaviour of nitrogen tetroxide and trioxide in a number ofadditive compounds with tin and titanium tetrachloride, chosenbecause the co-ordination number is invariably six, has led to thediscovery of certain new compounds stable only at low temperaturesand to the conclusion that the oxides may best be representedrespectively as (O:),N*O*N:O, and O:hT*O*N:0.85 A study of thethermal decomposition of nitrogen pentoxide in the presence offoreign gases has shown that hydrogen, carbon monoxide, andbromine are without influence, but that certain organic vapoursbring about rapid, almost explosive, reaction ; whilst nitric oxide isoxidised immediately.Hydrogen and air have no 'effect on thephotochemical decomposition, but bromine, possibly by opticalscreening, retards it. The mechanism of the decomposition is con-sidered to be N,05--+N0 + NO, + 0, (slow); NO + N205+3N0, (rapid).86Since the reaction 4NO + 30, + 2H,O + 4HN0, involves adiminution in volume, attempts have been made to use nitric acid asa carrier of oxygen present a t a high pressure. In this way, withoxygen at 20 atm., sodium nitrite has been oxidised to nitrate andarsenious oxide has yielded arsenic acid; but no change took placein the absence of nitric acid. The method has been applied withsuccess to arsenic sulphides, which are converted into arsenic acidand sulphuric acid.87Nitrogen sulphide has been prepared by passing ammonia mixedwith air into an anhydrous benzene solution of sulphur dichloride.It had m.p. 179", d 2-2, and M 184-3, corresponding to the acceptedformula N4S4. It sublimes near the melting point and explodesa t higher temperatures. Its solubility in various solvents and itsreactions with water, acid, and alkali, respectively, have been fullyinvestigated. The structural formula suggested as according bestwith its properties is S:S(:N*SiN),.s8It is only possible here to refer to papers on the chemiluminescenceof phosphorus vapour 89 and on the nature of the ions produced byglowing phosphor~s.~~ Pure crystalline phosphorus tri-iodide,84 K. A. Hofmann and U. Hofmann, Ber., 1926,59, [B], 2574; A,, 1927, 31.8 6 W.F. Busse and F. Daniels, J. Amer. Chern. SOC., 1927, 49, 1257; A,,8 7 P. Askenasy, E. Elod, and H. Zieler, 2. anorg. Chem., 1927, 182, 161;68 S. A. Vosnessenski, J . Rum. Phys. Chern. SOC., 1927, 59, 221 ; A,, 741.H. Reihlen and A. Hake, Annalen, 1927, 452, 47; A., 219.635; compare R. G . W. Norrish, Nature, 1927, 119, 123; A,, 119.A,, 635.E. J. Bowen and E. G. Pells, J., 1927, 1096; A., 633.W. F. Busse, Ann. Physik, 1927, [iv], 82, 873; A., 633INORGANIC CHEMISTRY. 55m. p. 61.0", and di-iodide, m. p. 124.5", have been prepared.g1Besson's supposed suboxide, P20, has been shown to be a mixture offinely divided amorphous phosphorus with adsorbed phosphorousa ~ i d . ~ 2 It has been shown that a variable mixture of phosphorussulphides is formed when phosphine is heated with sulphur or withhydrogen s ~ l p h i d e .~ ~The glow of arsenic in air or oxygen, which appears suddenly attemperatures between 250" and 310" when the pressure is reducedbelow a critical value, differs from that of phosphorus, since it is notaffected by small quantities of carbon tetrachloride, nitrobenzene,or sulphur dioxide. The appearance of the glow is favoured by arapid removal of the arsenic trioxide formed in the reaction, and itis suggested that the velocity of removal of the product by evapor-ation and condensation is the determining factor in the formationof theAn extensive research is reported on the oxides of antimony,and a new oxide, Sb,O,, is described which is very stable and maybe heated for a long period a t 800" without decomposition; oncedecomposition has begun, however, it proceeds rapidly a t lowertemperatures.The paper contains much thermal and vapour-pressure data which cannot be summarised here.95Investigations of tervalent vanadium 96 and of the action ofhydrogen peroxide on acidified solutions of vanadic acid have beenmade.97croup V I .The fact that the density of solid oxygen a t -252" is 0.034 unitgreater than the accepted value (wix., 1.46) affords additional evidenceof the existence of a second, denser form of solid oxygen stable a ttemperatures much lower than the melting point. The existence oftwo forms of oxygen accords with the properties of other membersof the same group.gsWhen mixtures of sulphur and chlorine are heated at 100" insealed tubes, or with the addition of a trace of iodine as catalyst, theproduct gives a freezing-point curve showing, in addition to thefamiliar maxima due to the crystallisation of S2C12 and SCl,, twowell-defined breaks which are attributed to the crystallisation ofF. E.E. Gemann and R. N. Traxler, J . Arner. Chem. SOC., 1927, 49, 307 ;A., 328.92 L. J. Chalk and J. R. Partington, J., 1927, 1930; A., 950.VS L. Delachaux, Helv. Chirn. Acta, 1927, 10, 195; A., 326.94 H. J. Ernelkus, J., 1927, 783; A., 497.O 5 A. Simon and E. Thaler, 2. anorg. Chem., 1927,162, 253; A., 730.g6 J. Meyer and E. Markowicz, ibid., 1926, 157, 211 ; A., 1927, 32.97 J. Meyer and A.Pawleth, 2. physikal. Chem., 1927, 125, 49; A., 326.g * J. C. McLennan and J. 0. Wilhelm, Phil. Mug., 1927, [vii], 3, 383; A,,29756 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.SCl, and S3C14. Freshly prepared mixtures of sulphur mono-chloride with an equilibrium mixture containing about 70% ofchlorine exhibit a maximum freezing point at the compositionSCI,, and a solid of this composition is obtained on freezing.99It has been shown that the burning of sulphur vapour in oxygengives rise to sulphur trioxide; this is formed by direct union andnot by a secondary oxidation of sulphur di0xide.lThe molecular weights of the various forms of sulphur trioxidehave been determined in various solvents and by measurementsof vapour density.2 Because negative catalysts such as sulphur,tellurium, carbon tetrachloride, and phosphorus oxychloridestabilise the a-form of sulphur trioxide, it is suggested that it hasthe structure S( :0)3, whilst the p-form, disulphuric anhydride, hasthe constitution 0 0>S<g>S<g.3A long and very important paper has appeared on the inter-relationships of the sulphur acids; it contains so much valuabledetail that it cannot usefully be summarised here, and the reader isreferred to the original communication.A new anhydro-acid,H,S,O,, is described which is oxidised by methylene-blue to tetra-thionic acid. The acid results from one of the three possible modesof decomposition of thiosulphuric acid : H,S,03 e H , S 0 3 + S ;2H,S,03 H,S +- H2S306 ; 2H2S,03 H,O + H2S405.4Sulphur trioxide readily reacts with nitric oxide a t 60", yieldingthe product 2S03,N0, m.p. 215-220' after darkening and softeninga t 180", b. p. 275"/715 mm. The substance is readily decomposedby water into sulphuric acid and nitric oxide, but does not react withferrous sulphate or cupric sulphate dissolved in concentratedsulphuric acid. When heated, it decomposes into sulphur dioxideand nitrogen peroxide, from which it may be prepared a t 200-300".If, however, the gases are moist, nitrosylsulphuric acid is formed inwhich drops of " blue acid " appear after some time. The " acid "is regarded as an oxide of nitrogen intermediate between NO,.,and NO. Raschig's conception that " blue acid " has the composi-tion H,SNO, is rendered improbable by the observation that itsabsorption spectrum does not resemble those of the compoundsCuSO,,NO and FeS04,N0.599 T.M. Lowry, L. P. McHatton, and G. G. Jones, J., 1927, 746; A., 505.1 J. Cornog, W. Dargan, and P. Bender, J. Amer. Chem. SOC., 1926, 48,G. Odd0 and A. Casalino, Gazzetta, 1927, 57, 60, 76; G. Oddo, ibid.,G. Odd0 and A. Sconzo, ibid., p. 83; A., 432.4 H. Bassett and R. G. Durrant, J., 1927, 1401; A,, 843.W. Manchot [with J. Konig and S. Reimlinger], Ber., 1926, 59, [B], 2672;2757; A., 1927, 32.pp. 29, 104; A,, 312, 300, 432.A., 1927, 32INORGANIC CHEMISTRY. 57If the solution, obtained by the action of carbon monoxide onmagnesium ethyl in the presence of chromic chloride is hydrolysedwith sulphuric acid at 0", and the ethereal layer is separated,neutralised, dried: and evaporated, crystals of chromium carbonyl,Cr(CO),, separate.The pure carbonyl has dlS* 1.77, is slightlysoluble in benzene and ether, but more soluble (2%) in chloroformand carbon tetrachloride, decomposes rapidly into chromic oxidea t 210", melts in a sealed tube at 149-150", and irreversibly depositsa chromium mirror a t 230". The carbonyl is unattacked by diluteacids, bromine, and iodine; fuming nitric acid converts it intochromic nitrate and free carbon.6Molybdenyl monochloride, MoOC1,4H20, has been shown to existin two stereoisomeric forms,' and the behaviour of the othermolybdenyl salts has suggested that in certain circumstancesoxygen may occupy two positions in the co-ordination sphere.8 Alarge number of hydrated molybdotungstates have been preparedand analy~ed.~Selenium oxyfluoride and tetrafluoride, both colourless, fumingliquids, have been prepared, and there is some evidence that alower fluoride, Se2F,, may exist.1° Attempts to repeat theproduction of selenium trioxide previously reported l1 have beenunsuccessful.12Group V I I .The oxidising action of fluorine on water, alkali hydroxide,sulphuric acid, phosphoric acid, phosphates, carbonates, andborates gives highly oxygenated compounds of the peroxide, peracid,or ozonide type in most cases.13 Passage of fluorine into a fairlyconcentrated solution of cobaltous sulphate in sulphuric acid yieldscobaltic sulphate, which is extremely unstable : the cobaltic salt isformed only with difficulty in dilute s01utions.l~A gaseous, oxygen compound of fluorine is produced when wateris present in the electrolysis of molten acid potassium fluoride below100".The gas has not been prepared in the pure state, but studiesof its mixtures with oxygen point to the formula F,O. Such mixtures6 A. Job and A. Cassal, Bull. SOC. chim., 1927, [iv], 41, 1041; A., 1044.7 W. Wardlaw and R. L. Wormell, J., 1927, 130; A., 296.8 Idem, ibid., p. 1087; A., 636.9 L. Fernandes, Bazzetta, 1926, 56, 655; A., 1927, 33.10 E. B. R. Prideaux and C. B. Cox, J., 1927, 928; A., 532.11 Ann. Reports, 1923, 20, 52.12 J. Meyer and A. Pawletta, Bey., 1927, 60, [B], 985; A., 532.13 F. Fichter and W. Bladergroen, Helv. Chirn. Acta, 1927, 10, 549, 653,1 4 F. Fichter and H.Wolfmann, Helv. C h h . Acta, 1926,9, 1093; A., 1927,559, 566 ; A., 741.12358 ANNUAL REPORTS ON TEE PROGRESS OF CHEMISTRY.have the odour of fluorine, do not act on glass even at high temper-atures, are stable in the presence of water, but liberate iodine frompotassium iodide.15A compound, K,Mn(CN)3, containing univalerit manganese, hasbeen produced by the reduction of potassium manganocyanide byaluminium filings or Devarda’s alloy; it has very strong reducingproperties.16Group V I I I .A number of salts have been described in which cobalt and nickelare believed to be univalent,17 and a description has been given ofthe preparation and properties of certain perferrates. Potassiumperferrate is more stable than potassium ferrate and can be purifiedby cautious sublirnation.l*The passivity of iron has been shown to be due to the existence ofa protective film which is too thin to give interference tints.Thisfilm, which consists of ferric oxide or hydroxide, has been obtained asa transparent envelope by dissolving the metal beneath by anodictreatment in sodium chloride solutions. The transparent skin canalso be removed by treatment with iodine. Chlorides favour theactivation, since the film is permeable Do chloride ions, and underanodic conditions the metal is dissolved away beneath the skin;mere immersion in a chloride solution, under conditions precludingthe flow of local currents, does not cause activation. The attack onpassive iron is often localised a t the water surface, the protectivefilm tending to cling to the gas-liquid rather than to the metal-liquid interface, thus initiating a breakdown.Nitric acid is a ratheruntrustworthy reagent for producing passivity. Transparentfilms have also been isolated from the surface of passive copper andaluminium. l9Differences of density and viscosity in a series of aqueous solutionsof cobalt chloride and hydrochloric acid show when plotted awell-marked inflexion and maximum respectively, which are con-sidered to mark the concentration of acid a t which equal numbers ofthe blue and the red ions are present. This point does not correspondt o the maximum colour change because the blue of CoCl4” is moreintense than the red of C O ( H ~ O ) ~ * * .~ ~ Other workers prefer toP. Lebeau and A. Damiens, Compt. rend., 1927, 185, 652; A., 1044.W. Manchot, ibid., 1926, 59, [Bj, 2445; A., 1927, 33; G. Grube [withH. Lieder and P. Schiichterle], 2. Elektrochem., 1926, 32, 561 ; A., 1927, 119.D. K. Goralevitch, J . Russ. Phys. Chein. SOC., 1926, 58, 1129; A,, 1927,433.l6 W. Wanchot and H. Gall, Ber., 1927, 60, [B], 191; A., 220.Is U. R. Evans, J., 1927, 1020; A., 619.0. R. I-Iowell, J., 1927, 158; A., 205INORGANIC CHEMISTRY. 59explain the change in terms of complex formation, and compoundsof the type CoCl,,H,O (blue or violet), CoC1,,4H20 (red), andCoCl,,GH,O (rose) are postulated.21Ruthenium dichloride and dibromide have been produced byreducing the tri-halogen salts with hydrogen in suitable organicsolvents in the presence of catalysts, and their properties have beenfurther described.22 The structure of Wilm’s rhodium chloronitratehas been investigated and the following formula has been suggested:[ ~ ~ [ R h & 6 ] ~ ~ ~ ] SH4.23Systems.A great deal of work has been done on various types of, systemswhich it is impossible to describe in this Report.It may be useful,however, to give here the titles, in the order in which they appear inthe Abstracts.Potassium zincicyanide-potassium mercuricyanide-potassiumnickelocyanide-potassium cadmicyanide 24 ; mercuric chloride-silver iodide 25 ; lanthanum sulphate-ammonium sulphate-water 26 ;beryllium chloride-chloride of lead or silver or cadmium 27 ; sodiumnitrate-sodium chloride-water ; sodium oxide-nitrogen pentoxide-water 29; sodium, or potassium, or rubidium or cEsium chloride-cobaltous chloride-water 30 ; magnesium sulphatezinc sulphate ;lithium chloride or bromide-water 32 ; zinc hydroxide-zinc oxide-sodium zincate-sodium hydroxide 33 ; lithium chlorate-water 34 ;2 1 A.Hantzsch, 2. mzorg. Chem., 1927, 169, 273; A., 205; compare J.Gr6h and R. Schmid, ibid., 162, 321; A., 728; A. Hantzsch, ibid., 166, 237;A., 1023.22 H. Gall and G. Lehmann, Ber., 1926, 59, [B], 2856; A., 1927, 123.23 0. E. Zwiagincov, J. Russ. Plqs. Chem. SOC., 1926, 58, 170; A., 1927,24 A. S. Corbet, J., 1926, 3190; A., 1927, 112.25 A. G. Bergman and T. A. Henke, J . Russ. Phys. Chem. SOC., 1926, 58,z 6 F. Zambonini and (Miss) A. Stolfi, Atti R. Accad. Lincei, 1926, [vi], 4,2 7 J. M. Schmidt, Bull. SOC. chim., 1926, [iv], 39, 1686; A., 1927, 112.28 F. Hold a2nd H. Crotogino, 2. anorg. Chem., 1926, 159, 78; A., 1927, 207.e9 W. T. Nikolaiev, J. Russ. Phys. Chem. SOC., 1926, 58, 557; A., 1927, 3138so H. W. Foote, Amer. J . Sci., 1927, 13, 158; A., 313.31 H. G. K. Westenbrink, Proc. K. Akad. Wetensch. Amsterdum, 1926, 29,3a.G. F. Huttig and W. Steudemann, 2. physikul. Chem., 1927, 126, 105;a3 E. Muller, J. Muller, and A. Fauvel, 2. Elektrochem., 1927, 33, 134; A.,34 C. A. Kraus and W. M. Burgess, J. Amer. Chem. Soc., 1927, 49, 1226;123.80; A., 1927, 112.424; A., 1927, 112.1374; A., 1927, 417.A,, 617.518.A., 62760 ANNUAL REPORTS ON THE PROGRESS O F CHEMISTRY.carbon disulphide-hydrogen sulphide 35 ; manganese-zinc 36 ;calcium f errocyanide-sodium f errocyanide-water 37 ; barium iodide-iodine-water 38 ; sodium chloride-potassium chlorate-sodiumchlorate-potassium chloride 39 ; potassium oxide-ammonia-phos-phorus pentoxide-water 40 ; cobalt chloride-rubidium or lithium orzinc or cadmium chloride-water 41 ; uranyl sulphate-ammoniumor potassium or sodium sulphate-water 42 ; copper-aluminium-manganese 43 ; iron-silicon 44 ; iron-cobalt-nickel45 ; neodymiumsulphate-rubidium sulphate-water 46 ; silver nitrate-lithium orrubidium nitrate 47 ; neodymium sulphate-ammonium sulphate 48 ;cerium sulphate-sodium sulphate 49 ; aluminium nitrate-water 50 ;sodium or lit,hium chloride-lead chloride-water 51 ; boron trioxide-sulphur trioxide or phosphorus pentoxide-water .52H. V. A. BRISCOE.P. L. ROBINSON.36 W. Biltz and M. Briiutigam, 2. anorg. Chem., 1927, 162, 49; A., 627.36 C. L. Ackermann, 2. MetaZZk., 1927, 19, 200; A, 627.38 A. C. D. Rivett and J. Packer, ibid., p. 1342; A., 731.39 C. Di Capua and U. Scaletti, Cazzetta, 1927, 57, 391 ; A., 731.40 E. Jiinecke, 2. physikal. Chem., 1927,127, 71; A., 731.41 A. Benrath, 2. anorg. Chem., 1927,163, 396; A., 829.42 A. Colani, Compt. rend., 1927, 185, 273; A., 830.43 W. Krings and W. Ostmann, 2. anwg. Chem., 1927, 163, 145; A., 830.44 T. Murakami, Sci. Rep. Tdhoku Imp. Univ., 1927, 16, 475; A,, 830.4 5 T. Has& ibid., p. 491; A., 830.d6 F. Zambonini and V. Caglioti, Atti R. Accad. Lincei, 1927, [vi], 5, 630;47 A. P. Palkin, J . Russ. Phys. Chem. SOC., 1926, 58, 1334; A., 1927, 939.48 F. Zambonini and A. Stolfi, Atti R. Accad. Lincei, 1927, [vi], 5, 832;** F. Zambonini and S. Restaino, ibid., p. 828 ; A., 940.6O G. Malquori, ibid., p. 892; A., 949.51 G. E. R. Deacon, J., 1927, 2063; A., 1030.62 31. Levi and L. F. Gilbert, ihid., p. 2117; A., 1030.(Miss) M. Farrow, J., 1927, 1163; A., 628.A., 842.A., 949