INORGANIC CHEMISTRY.ALTHOUGH it is not possible to report any material increase in thenumbers of papers which have been published during the year, yetthere is no doubt that much valuable work has been carried out.In general it may be said that the records have a more than com-mon interest. I n particular two investigations may be mentionedas being of especial importance. First and foremost is the dis-covery by Sir Ernest Rutherford bhat the atom of nitrogen isdisintegrated or decomposed when it is bombarded by the a-particlesfired off from radium-C. There is no doubt that the shock of thecollision is sufficient to disrupt the atom and cause it to decomposeinto two atoms of hydrogen and possibly three atoms of helium,but the latter has not yet been proved.The fundamental import-ance of this discovery must be acknowledged by everyone. Thegreat debt that chemistry owes to physicists is still further in-creased, for it cannot be denied that the influence of the sisterscience on the fundamental principles of chemistry has been pro-found. Radioactivity, on the one hand, and the energy quantumtheory on the other, exemplify the truth of this. The energy quantumtheory is not yet fully weaned, but it bids fair to have as profoundan influence on the chemistry of to-day as had John Dalt-on’s theoryof material quanta, the atoms, on the chemistry of his day.A second paper of note, is that by Dr. Maxted, on the poisoningof palladium as a catalyst by hydrogen sulphide. For the firsttime a quantitative basis has been found for the activation of agas by a catalyst.Although this may by some be thought to lieoutside the purview of pure inorganic chemistry, yet this is not afair criticism. The resolution of hydrogen sulphide into hydrogenand sulphur, and the formation of the complex Pd,S is pure inor-ganic chemistry. To mention these facts without reference to theresulting depression of the occlusive power of palladium and thequantitative relation now discovered would be an injustice to thisbranch of chemistry, which promises to become one of the mostfruitful fields of investigation of the problems of chemical reaction.2INORGANIC CHEMISTRY. 27Atomic Weights.The International Committee on Atomic Weights has issud areport on the experimental work that has been carried out onatomic weights since 1916.The report deals in particular with theatomic weights of hydrogen, carbon, bromine, boron, fluorine, lead,gallium, zirconium, tin, tellurium, yttrium, samarium, dyspro-sium, erbium, thorium, uranium, helium, and argon. Attention isdrawn to certain important determinations and it is recommendedthat new values be adopted for the atomic weights of argon, boron,gallium, yttrium, and thorium.Argon.-From the density and compressibility of this gas Leducfound the atomic weight to be 39.91.1 Since there is some uncer-tainty as to the second decimal place the new value 39.9 has beenadopted.Borort.-The atomic weight of boron has been determined by theconversion of anhydrous borax into sodium sulphate, carbonate,nitrate, chloride, and fluoride.2 Eight independent values wereobtained for boron, ranging from 10.896 to 10.905, and i t is recom-mended that the mean value of 10.90 be adopted.Gallium .-Some preliminary determinations of the atomic weightof gallium from the analysis of gallium chloride gave the values of70-09 and 70.11.3 The provisional adoption of 70.10 is recom-mended.Thom'ztm.-It is no'w recommended that the value of 232.15 beadopt'ed for the atomic weight of thorium, based on a series ofanalyses of thorium bromide .4 Two values were obtained, namely,232.152 from the silver ratio, and 232-150 from the Ag:C1 ratiowhen Br = 79.916.Yttrium.-The atomic weight of ythrium has been determinedfrom the analysis of the anhydrous chloride5 Seven specimens ofthis salt were used and the extreme values 89.30 and 89.34 wereobtained.The Committee also recommends that in place of the value of14.01 for nitrogen the more precise value 14.008 be adopted, whichis probably correct to within 1 unit in the third decimal place.Reference may be made to a determination of the atomic weight1 A.Leduc, Compt. rend., 1918, 167, 7 0 ; A., 1918, ii, 266.2 Smith and Van Haagen, Carnegie Inst. Washington, Pub.?. No. 267, 1918.3 T. W. Richards, W. M. Craig, and J. Sameshima, J . Amer. Chem. SOC.,4 0. Hsnigschmid and (Mlle.) S. Horovitz, Monatsh., 1916, 37, 105; A,,5 H. C. Kremers and B. S. Hopkins, J . Amer. Chem. SOC., 1919, 41, 718;It is recommended that the value 89.33 be adopted.1919, 41, 133 ; A., ii, 158.1916, ii, 484.A., ii, 46628 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.of scandium by analyFis of the bromide.6 Two specimens of scan-dium were employed, both of which were found to be spectro-scopically pure. From ope the value Sc=45.105 was obtained asthe mean of eight experiments, whilst the other gave as the meanof ten experiments Sc=45.093.The mean of the whole is45.099 k0.014, which differs considerably from the present inhrna-tional value 44.1. Since the last reference to the subject in theAnnual Report for 1916 a considerable amount of work has beencarried out on the atomic weights of the isotopes of lead, andseveral papers have appeared during this year.', g~ 99 10$11 It is hardlypossible as yet to draw any definite conclusions from this work asto the atomic weights of these isotopes since the values obtained bydifferent experimenters are not concordant.The problem wouldappear to be complex owing to the possibility that the isotopesthemselves are not stable.Several papers also have been published on the present-day con-ception of chemical elements, both from the point of view ofatomic structure and from the point of view of atomic stability.Amongst the former the most outstanding contribution is a newtheory of the atom developed by Langmuir.12 Although this doesnot fall within the purview of this Report, yet i t cannot be passedby without notice because the netw conception would seem to accountfor the phenomena of valency and chemical combination in a betterway than does the Bohr-Rutherford atom.Then again brief reference must be made to very striking workof Rutherford on the collision of a-particles with light atoms.13The most astonishing result was obtained with nitrogen when sub-mitted to the action of a-particles from radium-C.The resultsobtained leave little doubt that when a nitrogen molecule collideswith an a-particle the result is not nitrogen atoms but atoms ofhydrogen or atoms of mass 2. It is interesting to note that whilstthe majority of the light atoms have atomic weights representedby 4n or 4rt+3, where n is a whole number, nitrogen is the onlyone with an atomic weight of 4rt+ 2. Radioactive data would lead0. Hbnigschmid, Zeitsch. Elektrochem., 1919, 25, 91 ; A., ii, 285.T.W. Richards and N. F. Hall, J . Amer. Cham. Soc., 1917, 39, 537 ; A.,19i7, ii, 230.* A. L. Davis, J . Physical Chem., 1918, 22, 631; A . , ii, 107.lo 0. H6nigschinid, Zeitsch. Elektrochem., 1917, 23, 161 ; A., ii, 465.l1 K. Fajans, F. Richter, and (Frl.) J. Rauchenberger, Xitzurtgsber. Hcidel-l2 I. Langmuir, J . Amer. Chem. SOC., 1919, 41, 868 ; A., ii, 328.l3 Sir E. Rutherford, Phil. Mag., 1919, [vi], 37, 537, 562, 571, 581 ; A.,8. Meyer, Monatsh., 1919, 40, 1 ; A., ii, 385.berger Akad. Wiss., 1918, 28 ; A., ii, 7.ii, 256, 258, 259, 26029 INORUANIC CHEMISTRY.to the view that the nitrogen atomic nucleus consists of three heliumnuclei and either two hydrogen nuclei or one of mass 2. It is diffi-cult t o avoid the conclusion that the nitrogen atom when in colli-sion with an a-particle is disrupted into helium and hydrogen, andRutherford suggests the probability that the use of a-particles orsimilar projectiles, of still greater velocity, will result in the disinte-gration of many of the elementary atoms of small atomic weight.Although radioactive data, as Rutherford says, may have led tothe conclusion that the nucleus of the nitrogen atom is compoundedfrom three helium nuclei and two hydrogen nuclei, yet to thestudent of inorganic chemistry this observation of ita disintegrationmust form one of the most remarkable of those made in recentyears.I n his lecture before the Chemical Society Soddyl* has laidgreat stress on the far-reaching conclusions as t o atomic structurethat have been drawn from the study of radioactivity.Howeverdefinibe may be the arguments leading to a specific conclusion, i tmust be confessed that the conservatism of a chemist will onlygive way before real experimental proof. It is quite true that thewhole question of atomic structure and atomic stability was raisedby the discovery of radioactive disintegration, but to many chemiststhis phenomenon was only a troublesome property of one or twoless common elements of large atomic weight. They found securityin the statement that the radioactive disintegration of an elementis quite independent of external influence, and thus believed thelighter atoms to be perfectly stable entities. I n reality it is thisindependence that has been found to be incorrect, for it has beenfound possible to induce atomic disintegration by the use of par-ticles moving with great velocity, and the disintegration of nitrogenis the first to be observed.Molecular 'CV eigii ts.An important paper has been published on the molecular weightof molten stilphur.15 The method employed was that of surface-tension measurements made from observations of the rise in capil-lary tubes. Considerable difficulty was found a t first in obtainingpure specimens of sulphur.This was overcome by distilling thesulphur and immediately pouring the distillate into the experi-mental apparatus. The apparatus was then filled with dry nitro-gen and the sulphur boiled for 15 to 20 minutes. After coolingthe apparatus was exhausted and allowed to remain overnight.Itwas then again filled with nitrogen and the sulphur once moreboiled. This procedure was generally found sufficient to free the14 F. Soddy, T., 1919, 115, 1.A. M. Kellas, ibid., 1918, 113, 90330 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.sulphur from all impurities. It was found that a remarkablechange in the molecular complexity occurs a t about 160O. By theRamsay and Shields method of calculation the molecule of sulphurbetween 1 1 7 O and 157O is found to be S,, whilst between 160° and4 4 5 O it is S18. It wauld seem, therefore, that the suIphur moleculeundergoes an endothermic termolecular polymerisation a t about160°, 3S,=-(S6),. The author draws attention to the various criti-cisms that have been made of the Ramsay and Shields method andpoints out that all the recent methods of calculation indicate con-siderable association. Whilst the complex s, for mobile sulphuris corroborated by several methods, yet the Ramsay and Shieldsvalues seem to agree better with the experimental results.I n connexion with this work there may be mentioned an inveati-gation into the light-absorbing power of sulphur vapour betweenthe temperature limits of 400° and 1,200O.16 The remarkable fact,is recorded that the absorption of light increases to a maximum at650° and then decreases with increase of temperature.Moreover,there appears to be a definite absorption band developed, which hasa maximum intensity a t 650° and decreases rapidly in intensityas the temperature is decreased or increased.It is thus evidentthat the molecular intensity a t 650° must differ markedly from thata t 400° and 1,200°. According to Biltz the, sulphur vapour at thelower temperature consists of S, and a t the higher temperature ofS, molecules. No determinations of the vapour density of sulphurwere made by Biltz between 606O and 1,400°, but by extrapolationof his curve the density value for the temperature 650° correspondswith the formula S,. It is very noteworthy that this temperatureis the very one a t which the optical properties of the vapour aredistinct from those of the vapour composed of S, o r S, molecules.It is to be remembered that ozone has a greater absorptive powerthan oxygen.For these reasons the conclusion is drawn that at650° S, molecules form thel principal component of the equilibriumin sulphur vapour.Group 1.Investigation has been made of the normal and acid sulphates ofsodium in equilibrium with neutral and acid solutions over thetemperature range from -30° to lZOO.17 The existence of the fol-lowing salts was confirmed : Na,SO, ; Na,S0,,7H20 ; Na,SO,,lOH,O ;N%SO,,NaHSO, ; NaHSO, ; NaHSO,,R,O ; NaHSO,,H,SO, ;NaHS0,,H,S04,1*5H,0 ; Na,S04,2NaH:S04.16 Sir J. J. Dobbie and J. J. Fox, Proc. Roy. SOC., 1919, [A], 95, 481 : A . ,ii, 334.P. Pascal and Ero, Bull. SOC. chim., 1919, [iv], 25, 35; A., ii, 154INORGANIC CHEMISTRY. 31A similar investigation18 has been carried out, but under morelimited conditions as the observations were made a t only two tem-peratures, 1 4 O and 2 5 O .I n addition to N%SO,,lOH,O the follow-ing were obtained: Na&O,, NaHSO,, and NaHSO,,H,O. I n thispaper the investigation is described of the systems CuS04-H2S04-H,O, Na,SO,-CuS0,-H,O, and Na2,S0,-CuS0,-H2S04-H,0.I n the first no acid sulphates of copper were obtained, the effect ofthe sulphuric acid merely being to dehydrate the pentahydrate instages to the trihydrate, monohydrate, and the anhydrous salt.I n the second series the double salt Na,S0,,CuS04,2H,0 was ob-tained above 16*7*, whilst in the third series no salt other than thosementioned was obtained.Methods have been described for the preparation of the yellowamorphous modification of cuprous oxide.19.It is most readilyobtained by the reduction of a cupric salt by means of hydroxyla-mine in the presence of alkali. It can also be prepared electrolyti-cally, using an alkali sulphate as the electrolyte and an anode ofpure copper. The amorphous oxide when first precipitated is paleyellow and is probably a hydroxide. I n the absence of air thed o u r quickly changes to orange or brick-red, probably through lossof water. After drying, the oxide contains 2-3 per cent. of ab-sorbed water. On heating at a high temperature, above a low redheat, the water is lost and the amorphous oxide changes into thestable red crystalline modification.By the addition of a slight excess of sodium hydroxide to anaqueous solution of copper and aluminium sulphates containingapproximately 5 per cent.of CuO and 95 per cent. of A120,, a verypale blue precipitate is obtained.20 This precipitate retains itscolour after being thoroughly washed and dried a t looo. Whenground to a very fine powder and heated, first in a Bunsen flameand then in a blowpipe flame, it changes in colour to a pale greyish-blue with no signs of blackening. If the precipitate containsabout twice it9 much cuprio oxide it remains blue after heating inthe Bunsen flame, but shows signs of blackening when heated inthe blowpipe flame. It is suggested that the alumina stabilisesthe blue cupric oxide and that the change from blue to black isdue to an agglomeration of the particles. Anaiogous results wereobtained in some preliminary experiments with manganous, cobal-tous, and nickelous oxidea.l8 H.W. Foote, J . Id. Eng. Chem., 1919, 11, 629; A., ii, 361.L. Moser, Zeitsch. anorg. Chern., 1919, 105, 112; A., ii, 155.so H. E . Schenck, J. Physical Chem., 1919, 23, 283 ; A., ii, 28682 ANNUAL REPORTS O N THE PROGRESS OF CHEMISTRY.Group 11.A convenient method has been described for the extraction ofglucinum from beryl.21 The powdered mineral is treated with twoparts of sodium silicofluoride a t 850° for 30 to 40 minutes. Thesilicon fluoride which is evolved at this temperatare attacks theberyl and forms glucinum fluoride and aluminium fluoride, whichin turn combine with the sodium fluoride to give the correspondingdouble fluorides. On extracting the material with boiling water,the sodium glucinum fluoride dissolves, and the filtrate contains prac-tically the whole of the glucinum together with a little alumina andsilica.A slight excess of boiling sodium hydroxide solution isadded, and the precipitated oxides are collected and redissolved insulphuric acid. The1 solution is concentrated and the glucinum sul-phate is allowed to crystallise. By this method 90 per cent, of theglucinum present in the mineral may be readily recovered. Basedon this process a method is described for the estimation of glucinumin beryl.It has been found that alloys of magnesium and lead containingfrom 5 to 50 per cent. of magnesium are very reactive, and whenexposed t o moist air readily absorb the, whole .of the oxygenpresent.22 The two metals form the compound Mg,Pb and thisalloy, containing 90 per cent.of magnesium, is the most reactive ofthe series. During the oxidation process the alloy crumbles to ablack powder which consists of magnesium hydroxide, Mg(OH),, andhydrated lead sub-oxide, Pb,(OH),. I n dry air the mixture maybe kept unchanged, but in the presence of water the lead sub-oxideis oxidised to Pb(OH),. With the mor0 reactive alloys, the actiontakes place in the cold, but when more than 35 per cent. of mag-nesium is present heat is necessary. Alloys of magnesium andzinc are far less reactive, and indeed show greater resistance tooxidation than either magnesium or zinc.A scientific investigation of commercial superphosphates has beendescribed and although the principal subject involved is the ques-tion of the preparation and analysis, the results are of importancein connexion with monocalcium phosphate and dicalcium phos-phate.23 When increasing quantities of monocalcium phosphate aredissolved in a given weight of water at constant temperature, theproportion of free phosphoric acid continually increases and tendstowards a limit in accordance, with the equation2CaH4(P0& CaH,(PO& + CaHPO, + H,P04.21 H.Copsux, Compt. rend., 1919, 168, 610: d., ii, 192.22 E. A. Ashcroft, Trans. Faraday SOC., 1919, 14, 271 ; A., ii, 465.23 A. Aits, Anna& Chirn. Appl., 1918, 10, 45 ; A., ii, 25INORGANIC CHEMISTRY. 33u p to the saturation point at 1 5 O there thus exists a liquid phaseconsisting of water, monocalcium phosphate, and free phosphoricacid, and a solid phase of dicalcium phosphate formed by hydro-lysis of the monocalcium phosphate. It is commonly believed thatthe reaction between sulphuric acid and mineral phosphates takesplace in two stages :-3Ca3(P04), + 6H2S04 = 4H3P04 + Ca3(P04), + 6CaS04Ca,(PO,), + 4H3P04 = 3CaH4(P04),,but it would seem that the principal reaction is more correctlyrepresented by the equation:5Ca3(P04), + llH2S04 = 4CaH4(P04), + 2H3PO4 + 11CaSO4.The influence of raising the temperature on the reaction is t o in-crease the concentration of phosphoric acid in the liquid phase,whilst in the solid phase dicalcium phosphate increases in equalproportion with the free phosphoric acid.These constituentsgradually react to form monocalcium phosphates.An investigation has been made of the decomposition of bariumperoxide by heat a t atmospheric pressure, the method being thatof the observation of tahe heating curve.24 The peroxide was heatedin a carbon tube furnace and the temperature was recorded every10 seconds.Since the decomposition is an endothermic reactionits temperature range is indicated by a pronounced flattening ofthe curve towards the time axis. The temperature at which thedissociation pressure is equal to 760 mm. was found to be 795O,which is in good agreement with Le Chatelier's value1 796*.I n the presence of cupric oxide barium peroxide starts to decom-pose at 200°, the reaction becoming most vigorous at 625O to 660O.On the other hand, when the peroxide is heated with amorphoussilica the rate of rise of temperature is increased above 400O.I nall probability this is due to an exothermic reaction between thebarium oxide and silica to give barium silicate. The heating curveshows a similar indication of the formation of silicate even whenthe peroxide has been mixed with powdered quartz glass or quartz.In connexion with this it may be noted that strontium peroxidemay be prepared from strontium oxide by heating it in oxygenunder a pressure of 105 to 126 kilos. per sq. cm. a t a temperature of400° to 500O.25 The product contains more than 85 per cent. ofSrO, and resembles barium peroxide in its physical properties.The preparation of various sub-oxides of cadmium has beendescribed from time t o t-ime and the previous investigations have14 J.A. Hedvall, Zeitsch. anorg. Chem., 1918, 104, 163; A., ii, 26.' 6 J. B. Pierce, jun., Brit. Pat. 130840 ; A., ii, 413.BEP.-VOL. XVI. (34 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.now been repeated.26 The first method consisted in heating theoxalate a t 300° in a rapid stream of carbon dioxide. A small quan-tity of green material was obtained which, however, contained freecadmium. Analysis of this heterogeneous material gave valuesclosely approximating to Cdl=96.5 per cent., whilst Cd,O 27 requires96-56 per cent. By heating this substance in a vacuum it wasfound possible to distil off the free cadmium, leaving a homoge-neous green mass which on analysis was found to be Cd,O.The reduction of the brown oxide in hydrogen or carbon mono-oxide always gave an obvious mixture of the oxide and cadmium.The process described by Morse and Jones28 was repeated and smallquantities of Cd,O were obtained.Anhydrous cadmium chlorideis fused with cadmium and the product treated with wat.er. Cad-mous hydroxide is obtained from which yellow cadmium sub-oxidemay readily be prepared by dehydration. By each of these pro-m s ~ Cd20 may be prepared, but only in very small quantities.Anhydrous mercuric fluoride has been obtained by heating mer-curous fluoride in a current of dry chlorine a t 275O, or of drybromine a t 400° 29 It may also be prepared by heating mercurousfluoride at 450° under 10 mm. pressure. It forms transparent,octahedral crystals, m.F.645" and b.p. estimated a t 650O. Thevapour of mercuric fluoride is very reactive, and therefore it wasnot found possible t o measure its vapour pressure a t various tem-peratures since the containing veissels are attacked. The sub-stance reacts very readily with. moisture and becomes discoloured inthe preaence of minute t>aces of water vapur. On exposure tomoist air, hydrogen fluoride is evolved and mercuric oxyfluorideand, ultimately, mercuric oxide are formed. With small quanti-ties of water, a white, hydrated oxyfluoride, Hg3F,(OH),,3H20, isproduced. By cautious evaporation of a solution of mercuricfluoride in a 40 per cent. solution of hydrogen fluoride small colour-less crystals are deposited of the hydrated fluoride, HgF2,2H,0.Mixtures of the anhydrous fluoride with silver, copper, lead,aluminium, magnesium, zinc, tin, chromium, iron, or arsenic reactvigorously when strongly heated, yielding amalgams and metallicfluorides.The latter may readily be isolated in the pure conditionif an excess of mercuric fluoride be used. Silicon fluoride appearsto be formed when mercuric fluoride is heated with silicon, but noreaction occurs with either amorphous or graphitic carbon.26 H. G. Denham, T., 1919, 115, 556.87 8. Tanatar, Zeitsch. anorg. Chem., 1901, 27, 432 ; A,, 1901, ii, 553.28 H. N. Morse and H. C. Jones, Amer. Chem. J., 1890, 12, 488 ; A., 1890,s9 0. Ruff and G. Bahlau, Ber., 1918, 51, 1762 ; A., ii, 66.1376INORGANIC CHEMISTRY.36By heating mercurous fluoride in a current of dry chlorine at120° mercuric chlorofluoride is formed, and similarly the bromo-fluoride is obtained a t 105O. Both these substances are pale yellow.By the action of various thioamides on mercuric nitrite a complexsulphoxynitrite of mercury has been prepared.30 This compound isheavy, granular, and yellow, and can be dried in a steam-oven.By analysis the empirical formula was found to be 3(SHgNO,),HgO,but it is sugyested that the molecular formula is [3(SHgN02),HgO],,since the unimolecular formula represents an unsaturated com-pound. This substance is insoluble in water or acetone, but dis-solves in hydrochloric acid with copious evolution of nitrous fumes.When boiled with water it decomposes and is converted into blackmercuric sulphide.I n a similar manner the compound(SHgNO,),O has been obtaiiied. By tlhe action of ethyl iodide onthe sulphoxynitrite dimercuric di-iodo'disulphide, I*Hg*S-S=Hg*l,is produced. This compound is a pale yellow granular powderwhich darkens in the light, but. regains its colour when kept in thedark.The chlorine analogue1 of the sulphoxynitrite, [3 (SHgC1) ,HgO],,has also been prepared by the action of certain thioamides on mer-curic chloride.31 It forms a white amorphous precipitate which onremaining for 24 hours becomes granular.GTO'ZL~ 111.Some very interesting work has been carried out on the separationand purification of galliurn.32 I n the first place, a method has beendescribed for the recovery of gallium and also of germanium fromzinc ores.Since bot'h these metals are less volatile than zinc theyremain behind in the retorts when the zinc distils off. These resi-dues furnish a good source of gallium and germaniuq, althoughthe amounts obtainable vary very considerably. The method oftreatment, may be very briefly described. One kilo. of the oxideprepared from the zinc residues was added in small portions a t atime to 2400 C.C. of hydrochloric acid. When all had dissolved alittle potassium chlorate was added carefully until, after vigorousshaking, oxides of chlorine were evolved. The solution was thendistilled with a thermometer placed with its bulb in the liquid.Two fractions were collected, the first, up to 121°, containing verylittle germanium, and the second, up to 135-140°, containing practi-cally the whole of the germanium.The second fractions fromSQ P. C. Rky, T., 1917, 141, 101.81 Sir P. C. R&y and P. K. Sen, ibid., 1919, 115, 552.88 H. C. Fogg and C. James, J . Amer. Chern. SOC., 1919,41, 947 ; A., ii, 344.c 36 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.several quant,ities were saturated with hydrogen sulphide and thewhite germanium sulphide filtered off. The liquid left in the flaskwas diluted in water and the lead chloride allowed to settle. Theclear Equid was decanted and treated with ammonium hydroxideuntil a slight permanent precipitate was obtained, metallic zincwas added, and the whole digested at the boiling point for severalhours, after which the precipitated metals and basic salts were col-lected.Ten such precipitates were united and dissolved in hydro-chloric acid with the aid of a little potassium chlorate and the leadchloride allowed t o settle. The clear liquid was again saturatedwith hydrogen sulphide and filtered. The filtrate was boiled,neutralised with ammonium hydroxide until a permanent precipi-tate was formed, and again digested with zinc a t the boiling point.The precipitates rich in gallium were again dissolved in hydro-chloric acid, the solution nearly neutralised, saturated with hydro-gen sulphide and filtlered. The filtrate was treated with ammoniumchloride, made alkaline with ammonium hydroxide, and boiled untiljust acid. A gelatinous precipitate consisting of gallium, alumi-nium, and iron hydroxidM was filtered off and washed. The gal-lium was finally separated from the aluminium by the electrolysisof a strongly alkaline solution of the hydroxides.Independent work on tlhe preparation of pure gallium and itssalts has also been carried 3 4 35 In such work the methods oftesting the purity are of importance, especially in view of the diffi-culty of separating gallium from indium.A very delicate test hasbeen found in the spark spectrum, for by this means it is possible t odetect as little as 0.06 per cent. of indium in gallium and of 0.18per cent. of gallium in indium.It is possible to separate gallium and indium by the electrc!ysisof a dilute solution of their sulphatee, perfectly pure gallium beingobtained after 14 electrolysett.Pure gallium chloride can also beobtained from mixtures of gallium, indium, and zinc by the frac-tional distillations of the chloride in a current of chlorine. Amethod is described for the separation of gallium and indium byprecipitation of their hydroxides, Solutions containing both ele-ments are largely diluted and treated with a little hydrochloricacid, and then exactly neutralised with sodium hydroxide, anexcess of 1.5 grams of sodium hydroxide is added and the solutionboiled for several minutes. The precipitated indium hydroxide iswell washed, dissolved in hydrochloric acid, and the process r?33 L. M. Dennis and J. A. Bridgman, J. Arne?'. Chem. SOC., 1918, a, 1531 ;84 T. W. Richards, W.M. Craig, and J. Sameshima, ibid., 1919, 41, 131 ;A,, 1918, ii, 456.A., ii, 157. 56 T. W. Richards and S. Boyer, ibid., 133 ; A., ii, 158INORGANIC CHEMISTRY. 37peated. Finally, it is dissolved again in hydrochloric acid, precipi-tated by ammonia, washed, dried, and ignited to onide. To thecombined filtrates and washings from all these operations sodiumsulphite is added, and the solution boiled vigorously for fourminutes, when gallium hydroxide is precipitated.It would seem, however, that the hydroxide separation methodis not very satisfactory, since the gallium thus obtained still con-tains some indium. I f , however, the gallium hydroxide is dis-solved in an acid and the slightly acid solution electrolysed thegallium is obtained pure. The melting point of the gallium pre-pared in this way is 30.8O as compared with 26*9O for the metalpurified by the hydroxide process.The compressibility of gallium has been det'ermined and for thesolid element is 2.09 x 10-6, a value which places gallium preciselyon the curve showing the periodic relation of this property to atomicweight.The compressibility of liquid gallium is 3.97 x 10-6, avalue almost identical witlh that of mercury, and nearly twice, asgreat as that of solid gallium, although its volume is less. Thedensities of solid and liquid gallium are 5.885 and 6.081 respec-tively, and thus the view that the marked expansion of galliumon solidificatlion is due to impurities is definitely negatived.I n order to obtain pure gallium chloride it is advisable to burngallium in pure dry chlorine and to distil the impure chloride inpure chlorine, in nitrogen, and in a vacuum successively.Thistreatment was found necessary to eliminate the dissolved chlorine.A preliminary determination of the atomic weight of gallium hasbeen made by the analysis of pure specimens of the chloride.Reference to this has already been made under Atomic Weights.Gallium selenate has been prepared by placing gallium hydroxidein less than the equivalent quantity of selenic acid and heating themixture nearly to the boiling point'. After several hours the solu-tion was filtered from the excess of gallium hydroxide and allowedto evaporate a t o'rdinary temperature. The crystals were collectedand dried to constant weight in the air, and on analysis were found tohave the formula Ga,(SeO,),,l6H,O.It would appear, however,that the crystals which separate from the solution have 22 mole-cules of water of crystallisation. Gallium sulphats also first cry-stallises from its solution with 22 molecules of water of crystallisa-tion.Czsium gallium selenate alum has been prepared by the evapora-tion at ordinary temperature of a solution containing msiumselenate, gallium selenate, and some free selenic acid. This salt is atypical alum, forming regular octahedra with the formula :Cs$eO,, Ga, (SeO,) 3, 24 H,O 38 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.The solubilities of ammonium gallium sulphate alum and claesiumgallium sulphate alum have been determined, in water and in 50per cent.and 70 per cent. alcohol, with the, view of testing thepossibility of using the former salt as a means of separation ofgallium. One part of the ammonium alum dissolves in 3-24 partsof water, 4600 parts of 50 per cent. alcohol, and 11,400 parts of70 per cent. alcohol. One part of the msium alum dissolves in66.2 pa-& of water, 25,800 parts of 50 per cent. alcohol, and28,000 parts of 70 per cent. alcohol.Group I V .The view has been put forward that all varieties of graphite andamorphous carbon are different physical forms of “ black carbon,’’which is to be regarded as an allotropic modification ofdiamond .367 37 The physical and chemical properties of graphitevary within such wide limits that no distinct line of demarcationcan be drawn between graphite and amorphous carbon or soot.Theproperties of different samples of graphite depend upon the condi-tions under which it has been produced, and its variable characteris to be attributed to different degrees of dispersity. The reactionsby which graphite is formed are all of the localised type which havebeen grouped together under the name of (‘ t’opochemical ” reac-ti0ns.~8 This view is borne out by the X-ray spectra of graphiteand amorphous carbon, which lead t o the conclusion that the twovarieties are identical in structure, and that amorphous carbondiffers from graphite only by its greater degree of sub-divi~ion.~~In graphite the carbon atoms are arranged hexagonally in planelayers which are superimposed on one anotqher, and as a resultgraphite, oyes its peculiar properties to its lamellar structure.In order further to elucidate the structure of graphite the, oxid*tion of graphite t o graphitic acid and the properties of the latterhave been studied.In order to guard against any possible varia-tions due t o differences in the properties of-the samples of graphite,the experiments were confined to a specially pure electrically pre-pared graphite free from ash. The oxidations were carried outwith a mixture of potassium chlorate, nitric acid, and sulphuricacid in the cold under fixed conditions. This mixture is peculiarly36 V. Kohlschutter, Zeitsch. anorg. Chem., 1919, 106, 36 ; A., ii, 151. *’ V. Kohlschutter and P. Haenni, ibid., 121 ; A., ii, 152.sB V.Kohlschutter, ibid., 1 ; A., ii, 156.89 P. Debye and P. Scherrer, Physikal. Zeitsch., 1917, 18, 291 ; A., 1917,ii, 427INORGANIC CREMISTRY. 39advantageous because it penetratea the whole mass of the graphite.Other oxidising agents which do not penetrate the graphite havelittle action or oxidise it completely t o carbon dioxide. Repeatedtreat'ment of the graphitic acid with the oxidising mixture changesits collour from green to brown or yellow, whilst the carbon contentgradually diminishes. After five oxidations the graphitic acid hasthe composition C=54*4, H=2*14, 0=43.46 per cent. Afterrepeated washings with water, t,he graphitic acid passes into solu-tion. The colloid can be precipitated by dilute acids and the gelis perfectly soluble, in water.The differently coloured graphiticacids merely differ in their degree of dispersity, the paler colouredproducts obtained by repeated oxidation being more highly dis-perse. When heated or treated with reducing agents graphiticacid is reduced to carbon. The temperature a t which the decom-position is explosive is lower the slower the heating, but if the heat-ing is very slow the decomposition may proceed quietly to completion without explosion. The black voluminous residue consists ofcarbon and has all the properties of soot, but it can be compressedinto a mass very similar to graphite. When the decomposition byheat takes place under pressure, the graphitic character of the resi-dual carbon is more marked. Treatment of graphitic acid withreducing agents, such as ferrous or stannous salts, furnishes pro-ducts with strongly marked graphitic properties giving graphiticacid again when oxidised.These results are considered t o supportthe theory that graphite and amorphous carbon are not differentallotropes, but varieties of one allotrope, black carbon.When carbonyl sulphide 40 is passed through an electrically heatedtube, packed with quartz splinters, it decomposes to give carbonmonoxide and sulphur on the one hand, and carbon dioxide andcarbon disulphide on the other. Since it has been proved that thetwo reactions are independent of one another i t is probable that thetwo reactions may be written 2COS 2C0+S2 and2COS zz CO,+ CS,. Some further experiments have shown thatthe action of heat on mixtures of carbon monotxide and carbondisulphide on the one hand and carbon monoxide and sulphur onthe other, gives the same products as when carbonyl sulphide isheated.The reactions therefore are reversible and true cases ofequilibrium exist. Similarly the combustion of carbon disulphidewith an insufficient amount of oxygen or of oxygen in carbon disul-phide vapour yields a mixture of unchanged carbon disulphide,carbon dioxide, sulphur dioxide, carbonyl sulphide, and carbonmonoxide.4 0 A. Stock and P. Seelig, Ber., 1919, 52, [B], 681 ; A., ii, 23040 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Carbonyl chloride may be prepared by the action of carbon tetra-chloride on fuming or ordinary sulphuric acid.41~ 42 With pyrosul-phuric acid the reaction is SO, + H2S04 + CCl, = COCl, + 2S0,HCI.With ordinary sulphuric acid in tlhe presence of infusorial earth asa catalyst the reaction is ZH2S04 + 3CCI4 = 3coc1, + HCI + S,O,CIand a t 150° without catalyst CCl, + H,SO, = S0,HCl + HC1+ COCI,.If slight.ly aqueous acid is used the chlorosulphonic acid gives sul-phuric acid and hydrochloric acid.The carbonyl chloride is puri-fied by solution in carbon tetrachloride and subsequent distillation.The sub-acetate and the sub-sulphate of lead43 have been pre-pared by methods analogous to t'hose used for the sub-haloid salts ofthis meta1.44 The sub-acetate was obtained by the action of thevapour of acetic anhydride on the sub-oxide a t 195O, whilst thesub-sulphate was formed by the action of methyl sulphate vapouron the sub-oxide at 2 8 0 O .The latter salt decomposes on solutionin water, but is stable to the action of heat, for no sign of anychange could be observed on heating it a t 440O.A double nitrate and hypophosphite of lead has been described.45It is obtaineld by adding, with stirring, pure crystalline lead hypo-phosphite (250 grams) to a boiling solution of lead nitrate (500grams) in water (1.5 litres). The mixture is then rapidly cooledwhen the double salt separates out. It has the formulaPb(NO,),,Pb(H,PO,), and is explmive. Its use in percussionfuses is recommended.Some further work 011 zirconium compounds may be reported.46The basic compounds formed by the hydrolysis of zirconium sul-phate in aqueous solution are much more complex than has beensupposed and three basic sulphates have been isolated in the crys-talline form.I n spite of the Crystalline character of these com-pounds their solutions are essentially colloidal. The following basicsulphates have been prepared : Zr,( SO,),(OH),,,lOH,O,Zr8(S0,),(OH),,,8H,0, and [Zr,(S0,),(OH)8]H,,4H,0. The lastnamed had previously been described and given the formula2Zr0,,3S0,,5H20, but the new formula takes into account its acidproperties and explains its ready conversion into either of the firsttwo compounds. The usually accepted formula for potassiumzirconyl sulphate, Zr,O,(SO,),K,, is now shown to be incorrect and41 V. CYTrignard and E. Urbain, Compt.rend., 1919,169, 17 ; A., ii, 340.42 C. Mauguin and L. J. Simon, ibid., 34 ; A., ii, 341.44 Ibid., 1917, Ill, 29 ; 1918,113, 249.45 E. von Hem, Zeitsch. ges. Schiess. u. Sprengstoflw., 1916, 11, 365, 388 ;46 0. Hauser and H. Herzfeld, Zeitsch. anorg. Chem., 1919, 106, 1 ; A.,H. G. Denham, T., 1919,115, 109.p.. ii, 284.ii, 290INORGANIC CHEMISTRY. 41a definite salt, Zr4(S04)5(OH)8K,, has been prepared. Theammonium salt of the above zirconylsulphuric acid,(NH4),Zr4(OH),( S0,),,4H20, has been obtained, together with aless basic salt, (NH4)4Zr(S04)4,5H,0. A basic salt,K,[Zr,( OH),( S04),]8Hz0, has also been prepared.Group V .Reference wits made in the Annual Reports for 1912 and 1915to Franklin's work on the ammonia system of acids, bases, andsalts.' A further contribution has been made during this year andcertain new compounds have been described .47 Dipotassiumammonosodiate, [Na(NH,)3]K,, is obtained by the act,ion of potass-amide on sodamide in liquid ammonia, by the action 'of sodiumiodide on an excess of potassamide in liquid ammonia, or by theaction of sodium on potassamide in liquid ammonia in the presenceof a small quantity of platinum black. This compound crystal-lises well, and does not lose ammonia a t looo in a vacuum. Mono-rubidium ammonosodiate, [Na(NH,),]Rb, is formed by the actionof sodium and rubidium simultaneously on liquid ammonia. Thiscompound is readily soluble in liquid ammonia, and is violentlydecomposed by water with the formation of the hydroxides of themetals.Dirubidium ammonosodiate, [Na(NH,),]Rb,, is formedfrom the mother liquors of the previous compound by the additionof a large excess of rubidamide. Dipotassium ammonolithiate,[Li(NH,),]K,, is prepared by the action of potassamide on lithiumiodide in liquid ammonia solution, and also by the action of lithiumand potassium simultaneously on liquid ammonia in the presence ofplatinum black. Monorubidium ammonolithiate, [Li(NH,),]Rb,is prepared by the action of an excess of a solution of rubidamidein liquid ammonia on metallic lithium in the presence of platinumblack.An important investigation has been carried out on the densitiesof liquid nitrogen peroxide and mixtures of nitrogen peroxide andnitric acid.48 The specific volume of nitrogen peroxide between4 O and ' 180 is expreesed by ZJ = 0.67027 + 0.0010075t + 0*000003t2.The boiling point of the liquid is 21.9+0.1°.Nitrogen peroxide issoluble in nitric acid up to about 54 per cent. and nitric acid ismuch less soluble in nitrogen peroxide, about 7 per cent. Withinthese limits the densities of mixtures of the two substances havebeen determined at 4O, 1l0, and 1 8 O . A maximum density is ob-tained with the mixture cont,aining about 44 per cent. of nitrogen47 E. C. Franklin, J . Physical Chem., 1919, 23, 36; A., ii, 191.W. R. Bornfield, T., 1919, 115, 45.042 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.peroxide. On the other hand, the maximum contraction on mixingis found with mixtures containing 49.3 per cent.of nitrogenperoxide. There is considerable heat evolved on mixing which indi-cates a powerful combination between the two substances. Thecompound 3HN03,2N,0, corresponds with 49.33 per cent. ofN,O,. The maximum value of the density does not indicate theexact composition of the compound since it is so largely deter-mined by the mere density differences of the two components.When the temperature coefficients of expansion are examined theseare found to show a definite minimum for the mixture containing26.7 per cent. of N,O,, which corresponds with the definite compo-sition 4RNo3,N@,. It may therefore be concluded that two specificcompounds exist, namely, 3HN03,2N,04, and 4HN03,N,04.The maximum density of the solution containing about 43 percent.of N,O, has been independently observed,49 and the authorsconsider that this gives evidence of the existence of the compound2HN0,,N,04 or N,05,N204,H,0. They state that the existence ofthis compound is confirmed by a thermal study of the reciprocalsolubilities of nitric acid and nitrogen peroxide. The compound isstable below - 48*5O, and a t tlhis temperature it dissociates, liberat-ing nitrogen peroxide. The density of nitrogen peroxide is ex-pressed by D: = 1.490 - 0.00215t.It has generally been believed that the combustion of ammoniain a deficiency of oxygen proceeds according to the equation4NH3 + 30,= 2N, + 6H,O. It has, however, been shown that underthese conditions the resulting gas consists of about 59 per cent. ofnitrogen and 41 per cent.of hydrogen,50 the explanation beingoffered that it portion of the ammonia is dissociated into nitrogenand hydrogen a t the high temperature of the flame. If this werethe correct explanation considerable quantities of nitrogen shouldbe formed when ammonia is burnt in excess of oxygen, which is notthe case. A possible explanation may be found in the formation ofthe hypothetical di-imide, when insufficient oxygen is present,which would decompose into nitrogen and hydrogen.51 The reactionmay be expressed therefore by the equation 2NH, + 0, = N,H,+ 2H20. When potassium hydra,zinesulphonate was heated withpotassium hydroxide in the expectation that thel hydroxyhydrazineprimarily formed would immediately give di-imide by loss of water,the calculated amount of potassium sulphite was obtained and amixture of equal volumes of hydrogen and nitrogen.A second4 0 P. Pascal and M . Gamier, Bull. So'c. chim., 1919, [iv], 25, 309 ; A.,ii, 339.Muller, Zeitsch. physikal. Chem. Unterr., 1913, 169.ti1 F. Raschig, ibid., 1918, 31, 138 ; A., ii, 149INORGANIC CHEMISTRY. 43alternative is that the combustion proceeds with the intermediateformation of hydrazine, 4NH, + 0, = 2N,H4 + 2H20. The form*tion of hydrazine when oxygen burns in ammonia has actually beenproved by the formation of benzylideneazins. It is concluded thatthe combustion gives mainly di-imide and also hydrazine to a lessextent, and that the direct formation of the nitrogen according tothe commonly accepted reaction does not take place at all.The, reduction of arsenious acid by means of stannous chloridehas been the subject of systematic study.52 It was recorded byBettendorf 53 that a voluminous, brown precipitate of arsenic isobtained, accompanied by traces of non-removable tin.There seemslittle doubt that as the first product of this reaction arsenic is ob-tained as the yellow allotropic modification. A portion of it issoluble in carbon disulphide and the yield of yellow arsenic is in-creased if the reaction mixture is shaken with carbon disulphideduring the process of reduction. The results indicate that the veryearliest deposit of arsenic is of the yellow type, but that unless cer-tain unascertained conditions obtain, the yellow variety spontane-ously becomes brown or grey.The reaction does not take place if the two chlorides are anhy-drous, but readily proceeds if the mixture of anhydrous chloridesis moistened with water.I n general it is found that the reactionis accelerated chiefly by increase in the concentration of hydro-chloric acid, next by that of arsenic chloride, and least by that ofstannous chloride. The results obtained in velocity measurementslead to the view that the reaction takes place between arseniousand chloride ions and the complex H2SnC1,. Dilution with waterrapidly decreases the. velocity, owing no doubt to the hydrolysis ofthe amenious chloride.A number of double compounds have been described of arseniousoxide and the iodides of bivalent light and heavy metals.64 Themoderately concentrated solution of the iodide is saturated in thehot with arsenious oxide and allowed to cool, when the compoundseparates in crystals, sometimes with a little free arsenious oxide.The following have been prepared : G11,,3As,0,,8H20 ; Mg(or Ca orSr)I,,3A~03,1ZH,0 ; Ba12,3As,03,9H,0 ; Zn12,3A~0,,10H20 ;Mn(or Fe or Co)I,,4As20,,12H20 ; NiI,,4As203,10H20 ;A113,6A~20,,18( 1 )H,O ; LiI,2As20,,3H,O.In properties these com-pounds resemble the arsenites of the metals, the characters of theiodides being suppressed. They are moderately stable in dry air,but tend to become oxidised on keeping. With the exception ofsB R. G. Durrant, T., 1919,115, 134.53 Sitzungsber. Miederrhein. Ges. Bonn, 1869, 128.R. F. Weinlaad and P.Gruhl, AT&. P h m . , 1917,255, 467 ; A., ii, 411.a* 44 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.the magnssium compound these salts are sparingly soluble in water,and when heated with water they tend to undergo partial dissocia-tion into the iodide and arsenious oxide. It is believed that thesecompounds approximate in constitution t o a complex salt with sim-ple metallic cation.Investigation has shown that antimony pentoxide does not formthe definite hydrates, ortho-, pyro-, and meta-antimonio acids, ashas hitherto been supposed.65 Analyses of the various hydrates havebeen made, and their dehydration curves obtained by keeping themover sulphuric acid in a desiccator. It is found that the propertiesof the antimonic acids vary with the method of preparation andwith previous treatment.Thus the modifications prepared eitherby the action of water on antimony pentachloride or by the actionof acids on antimonates, or by the hydrolysis of antimony tri-chloride in the presence of nitric acid differ in their water content,stability, solubility, and their action towards acids and alkalis.The differences observed are in all probability due to variations inthe size of the particles. It appears that the soluble antimonicacids are hydrosols with small stability and that the definitehydrates, H,SbO,, H,Sb,07, and HSbO,, can have no free existence.The hydrates exhibit marked selective absorption towards dilutealkalis t o give amorphous substances which apparently are alkaliantimonates.The hydrates dissolve in concentrated alkali solutionsand from these solutions by cautious evaporation a t low tempera-ture various alkali antimonates may be crystallised. The nature ofthese salts depends, however, on the concentration of the motherliquor.When magnesium which has been covered with thorium4 (anisotope of bismuth) is dissolved in hydrochloric acid, a small quan-tity of a radioactive hydride is obtained. This observation leads tothe belief that bismuth forms a volatile hydride, and the existenceof this compound has been definitely proved both by radioactivemethods and also by its preparation from non-radioactive mate-rials.56J57 The second method has a greater importance since allcriticism arising from possible misconception of radioactive pheno-mena is eliminated.A bismuth magnesium alloy is prepared bymelting together equal weights of powdered bismuth and magnesium(as free from silicon as possible) in an iron crucible in a rapid cur-rent of hydrogen. When this alloy is dissolved in approximately4N-hydrochloric or sulphuric acid sufficient bismuth hydride is ob-S5 G. Jander, Kolloid Zeitsch., 1918, 23, 122 ; A., ii, 29.5~ F. Paneth, Zeitsch. Elektrochem., 1918, %, 298 ; A., ii, 30 ; Ber., 1918,I 7 F. Paneth and E. Winternitz, Ber., 1918, 51, 1728 ; A., 3, 68.51, 1704 ; A., ii, 67INORGANIC CHEMISTRY. 45tained to permit its detection either by the bismuth mirror test orby the luminescence test. The bismuth mirror as obtained in theusual Marsh's apparatus very closely resembles the antimony mir-ror, and consists of a strong brown ring in front of the heated spotand a fainter ring behind it.Only about 5 x 10-5 of the bismuthused is converted into the hydride but the optimum conditions havenot yet been determined. The luminescence test gives a very satis-factory method of detection of the bismuth hydride. The gasesissuing from the Marsh's apparatus are ignited, and a piece of purecalcium carbonate is held on a platinum loop in the flame. Thehydride is decomposed and a portion of the bismuth is depositedon the lime. The lime is allowed to cool and is then placed a t theedge of the hydrogen flame, when the presence of bismuth is shownby a cornflower-blue phbsphorescence. A sky-blue phosphorescenceis shown by antimony.Bismuth hydride is absorbed to some extent by water and4N-sulphuric acid.The gas is absorbed more readily byO.5N-sodium carbonate and N-potassium hydroxide, and also bycalcium chloride or soda-lime. It is completely decomposed by con-centrated sulphuric acid.Group V I .The boiling point of sulphur as a standard temperature hasbeen discussed together with the conditions which must be ob-served.58 The vapour pressure over the range 700-800 mm. hasbeen measured and within these limits the relation between tem-perature and pressure is given by:t =444*60 + 0*0910(p - 760) - 0.000049(p - 760)2.By the action of liquid sulphur dioxide on the iodides of sodium,rubidium, and caesium, sulphones of the type MI,,3soz have beenprepared.69 The sodium compound is amorphous, but the other twocan be crystallised from their solution in liquid sulphur dioxide.The dissociation pressures of these compounds have been measuredbetween - 22.5O and 20.9O, and they show increasing stability.withrise in atomia weight of the metal.The molecular weight of sulphur monochloride in solution inbromoform has been found to correspond with the formula SzC1,.The freezing point of bromoform is depressed by the addition ofsulphur and sulphur monochloride less than the sum of the depres-68 E.F. Mueller and H. A. Burgess, J . Amer. Chern. Soc., 1919, 41, 745 ;59 R. de Forcrand and F. Taboury, Compt. rend., 1919, 168, 1253 ; 169,A,, ii, 446.162 ; A., ii, 341, 36646 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.sions of these two taken separately.This points to the existenceof polythionic chloridee and the highest of these detected in bromo-form solution is S4Cl2. It is probable that chlorides richer in sul-phur are formed when sulphur is dissolved in the monochloride.It is suggested that the existence of these polythionic chloridesaffords an explanation of the formation of polythio-derivatives whenorganic substances are treated with sulphur monochloride.60In the Annual Report for 1913 the preparation was described ofthe fluorosulphonates and of sulphuryl fluoride. The ammoniumsalt, NH4S0,F, is most conveniently prepared by the gradual addi-tion of dry ammonium fluoride to sulphuric acid containing about70 per cent.of sulphur trioxide, and treatment of the product witha small excess of ammonia dissolved in methyl alcohol.61 The saltmelts at 245O and readily reacts with ammonia, especially at lowtemperatures, to form liquid ammines. The alkali metal salts areprepared by the action of the alkali hydroxide on the ammoniumsalt in aqueous solution. The following have been obtained:KSO,F, RbSOP, LiSO,F, and LiSO,F,SH,O. These salts can becrystallised from water if the operation is rapidly performed. Thealkali fluorosulphonates are very stable towards heat, thus thepotassium salt, when heated for some time, to bright redness, suffersonly slight decomposition, with the evolution of sulphur dioxide,sulphur trioxide, hydrogen fluoride, and oxygen. The barium salthas not been obtained in the pure condition and the crude salt isdecomposed at a red heat into barium sulphate and sulphurylfluoride.This affords the most convenient method for the prepara-tion of sulphuryl fluoride.The alkali fluorosulphonates possess the remarkable property ofexchanging the fluorine atom for an amino-group when treatedwith an aqueous solution of ammonia or substituted amine.62 Aportion of the fluorosulphonic acid depending on the strength ofthe base is hydrolysed to hydrogen fluoride and sulphuric acid,which may readily be removed by chalk or barium hydroxide. Inaddition to many substituted aminosulphonates the following havebeen prepared : aminosulphonic acid, barium hydrazinosulphonate,and potassium hydrazinosulphonate.This method of preparationhas the great advantage of not requiring the isolation of the basein the anhydrous condition.An important paper has appeared on the preparation and purifi-60 G. Bruni and M. Amadori, Atti R. Accad. Lincei, 1919, [v], 28, i, 217 ;61 W. Traube, J. Hoerenz, and F. Wunderlich, Ber., 1919, 52, [B], 1872 :68 W. Traube and E. Brehmer, ibid., 1284 ; R., i, 434.A., ii, 281.A., ii, 364INORGANIU UHEMISTRY. 47cation of selenious and selenic acids, and certain new selenium com-pounds are de~cribed.6~ Two sources of selenium were employed,namely, smelter flue-dust and anode slimes from a copper refiningworks. The flue-dust contained about 22 per cent. of selenium,small amounts of silica, iron, and aluminium, and a trace of tellu-rium.The flue-dust was finely ground and fused with sodiumcarbonate and sodium peroxide. The product was treated withwater and the insoluble material was filtered off. The filtrate wasthen nearly neutralised with concentrated hydrochloric acid whichprecipitated the greater part of the aluminium and zinc as hydr-oxides. The solution was filtered and the liquid diluted with threetimes its volume of concentrated hydrochloric acid and boiled for30 minutes to reduce the selenic acid to selenious acid. Any silicaprecipitated at this stage was removed by filtration. The filtratewas heated to 80° and treated with sodium sulphite and digestedat 80° for several hours to convert the selenium into the greymodification.The anode slimes contained 96 per cent.of selenium and a con-siderable amount of tellurium. The dry, finely powdered slime wasadded to concentrated nitric acid diluted with onefifth its volumeof water. After the vigorous reaction had moderated the mixturewas heated to complete the oxidation. The filtrate was evaporatedto drynem, the residue dissolved in hydrochloric acid (3 : 1) and theselenium precipitated by sulphur dioxide or sodium sulphite.Pure selenium dioxide was obtained by oxidation of the seleniumobtained as above with nitric acid and evaporation of the solutionto dryness. The, selenium dioxide thus obtained was sublimed ina glass tube, the vapours passing through a 2 cm. plug of glasswool.Pure selenious acid was prepared by the evaporation of the solu-tion obtained by oxidation of the anode slimes by nitric acid untilit had a syrupy consistency.The solution on being allowed to re-main deposited crystals of selenious acid, and these after fourrecrystallisations from water were found to be quite free fromtellurium .For the preparation of pure' selenic acid three separate methodswere employed. In the first silver selenite was oxidised by bromineaccording to the reaction Ag,SeO, + 2Br + H20 = H2Se0, + 2AgBr.The silver salt was suspended in the necwsary volume of water toyield a 3 per cent. solution of selenic acid, and bromine water witsadded until the, solution assumed an orange colour. After twohours the precipitated silver bromide was removed by filtration and63 L. M. Dennis and J.P. Roller, J . Amer. CTsena. Soc., 1919, 41, 949; A.,ii, 33648 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.the excess of bromine removed by a current of air. Silver nitratesolution was then added to precipitate the hydrobromic acid andthe solution again filtered. The filtrate was then evaporated on asteam-bath to half its volume and then further concentrated undera pressure of 25 mm. at 120° to remove nitric acid. This processwas rendered necessary because it was noted that bromine waterwhen exposed to light gives considerable quantities of hydrobromicacid.The second method was the oxidation of selenious acid by chlorinein presence of copper carbonate, which dissolves to form cupricchloride and cupric selenate. After the reaction w u complete thesolution was filtered and evaporated t o a small volume at 65O.Oncooling cupric sslenate separated out and after recrystallisation fromwater the salt contained only 1-2 per cent. of cupric chloride. Thecrystals w0re dried in the, air and extracted with acetone in aSoxhlet apparatus to remove the cupric chloride. After one furtherrecrystallisation from water the cupric selenate was found to bepure. Selenic acid was obtained by cautious electrolytic deposi-tion of the copper.I n the third method the selenious acid was converted to selenioacid by anodic oxidation.It was proved that the electrolysis of selenic acid and its saltsunder the same conditions as are favourable for the formation ofpersulphates does not yield either perselenic acid or perselenates.Methods are described for the detection of small amounts of seleni-ous acid in selenic acid, as well as small amounts of sulphuric acidin selenic acid.Very delicate tests are given for the detection oftellurium in the presence of selenium. A method is also describedfor the estimation of selenium in selenates by the use of hydrazinehydrate.Copper selenate crystallises from its aqueous solutions with fivemolecules of water of crystallisation. This pentahydrate on heat-ing at 104O loses 4H,O and yields the monohydrate. Completedehydration is effected a t 230-235O, and the resulting anhydroussalt is stable up to 280O. When the pentahydrate is treated withacetone it is converted into the trihydrate.Certain ammonia derivatives of copper selenate are described,namely, CuSe0,,4NH,,Hz0, CuSe0,,3NH3,H,0, and CuSe0,,4NH3.A critical examination has been made of the precipitation oftellurium as the sulphide.64 When an aqueous solution of tellurousacid is treated with hydrogen sulphide the tellurium is precipitatedas TeS2, but from this substance after drying the greater part ofthe sulphur can be extracted by carbon disulphide.The fact that^64 A. M. Hageman, J . Amer. Chem. SOC., 1919,431, 329; A., ii, 190INORUANIC CHEMISTRY. 49about one per cent. of sulphur is retained has supported the con-tention that ordinary tellurium is not a pure chemical element, the‘suggestion being made that the retention of the sulphur is due tothe presence of a more basic element forming a more stable sulphide.There is no doubt that TeS, is initially precipitated, but that it is anunstable substance a t the ordinary temperature, dissociating intotellurium and sulphur.Below -20° TeS, is stable and the velocityof dissociation above that temperature increases as the temperatureis raised.The previous observations as to the impossibility of extractionof the total quantity of sulphur from the compound have been con-firmed. After extraction with carbon disulphide for nine days,followed by treatment with boiling alcohol for 30 days, the tellu-rium still retains a t least 0.95 per cent. of sulphur. This residualsulphur does not exist as a sulphide that can be decomposed byhydrochloric or hydrobromic acid, or as an allotropic modificationof sulphur insoluble in carbon disulphide.The question as to thecondition in which it exists still remains unanswered.It is shown that the monosulphide, TeS, which has often beendescribed, has no existence.Group VZI.For the preparation of fluorine the electrolysis of molten potass-ium hydrogen fluoride or sodium hydrogen fluoride is recorn-mended.65 The electrolysis is carried out in an electrically heatedcopper vessel which serves as the cathode. The anode is made ofgraphite and is enclosed in a permeable diaphragm, which preventsthe hydrogen from mixing with the fluorine. The most efficientconditions are obtained with a current of 10 amperes a t 15 voltsand a temperature of 240-250°, when t h e , current efficiency is+out 70 per cent.As the electrolysis proceeds, the alkali fluorideand copper fluoride are deposited, and after a time it becomes neces-sary t o regenerate the electrolyte. It is of course necessary thatthe alkali hydrogen fluoride be absolutely dry, and this is moreeasily realised with the sodium salt since the potassium salt ishygroscopic. The sodium salt has also the advantage in being lessexpensive. Moreover, it cont?ains a relatively larger proportion ofavailable hydrogen fluoride and it decomposes below its meltingpoint.The original investigations 66167 of crystallised sodium hypo-66 W. L. Argo, F. C. Mathers, P. Humiston, and C. 0. Anderson, J . Physicui66 M. Muspratt end E. Shrapnel-Smith, J . SOC. Ohm. Ind., 1898,17, 1096,67 M.Muspratt, ibid., 1903, 22, 691.Uhern., 1919,23, 348 ; A., ii, 332.1899 ; 18, 210 ; A., 1899, ii, 281, 63360 ANNU& REPORTS ON TEE PROGRESS OF CHEMISTRY.chlorite have been repeated.6* The salt was originally found tohave a composition corresponding. approximately with a hexa-hydrate NaOCl,GH,O. In the present investigation the hypo-chlorite solutions were prepared by treating 35 per cent. sodiumhydroxide solution, cooled in ice-water, with chlorine, removing theprecipitated sodium chloride, adding sodium hydroxide equivalentto the sodium chloride precipitated, and repeating the treatmentwith chlorine until the solution was about 5 N . The solution, whichhas been freed from precipitated sodium chloride, iS cooled to-loo and induced to crystallise by shaking. The sodium hypo-chlorite separata as a mass of very fine, hair-like crystals fillingthe whole liquid, whilst the temperature rises considerably.Whenthe whole has again been cooled to - loo the crystals are removed bysuction. Considerable difficulty was met with in the analysis ofthese crystals owing to the fact that they are +cry deliquescent andalso to the fact that they melt between 1 8 O and 19O. The analysesseemed to show that the salt approximates more nearly in composi-tion to a heptahydrate than to a hexahydrate, but further investi-gation may show that more than one hydrate is present.The heptahydrate melts to a cloudy liquid and if this liquid iscooled t o ordinary temperature large and well-f ormed crystals of a, newhydrate, NaOC1,5H20, are obtained, This pentahydrate, meltingat 27O, is also very deliquescent, but may be kept unaltered in awell-stoppered bottle.Aqueous solutions of hypochlorous acid containing 25 per cent.ofthe acid are readily obtained by distilling a mixture of chlorinehydrate and yellow mercuric oxide in a good vacuum.69 I n attempt-ing to prepare, a more concentrated solution or the anhydrous acidby distillation of this solution and condensation of the distillate inreceivers maintained a t Oo, -20°, and -80°, it was found that inthe first two flasks 25 per cent. hypochlorous acid was collected,whilst in the third pure chlorine monoxide condensed. It is evi-dent, therefore, that in the aqueous solution there exists the equili-brium, 2HC10 -7-t H20 + C1,O. This equilibrium has been investi-gated by agitating aqueous solutions of hypochlorous acid with car-bon tetrachloride at Oo.The equilihrium lies greatly in favour ofthe hypochlorous acid, for an approximately N / 5-solution contains0-2 per cent. of chlorine monoxide. It is probable that the greateroxidising properties of hypochlorites in acid solution are due to thepresence of chlorine monoxide,.The absorption spectra of hypochIorous acid, its ethyl ester and68 M. I?. Applebey, T., 1919,115, 1106.09 S. Goldschmidt, Ber., 1919, 52, [B], 763 ; A., ii, 227INORGANIC CHEMISTRY. 51metallic salts have been observed.70 Whilst the acid and ester havethe same absorptive power, the salts differ very materially and showwell marked absorption bands.This is interpreted to mean thatthe constitution of the salts and the free acid is different. Theauthors put forwardC1-0-H and thatequilibrium :the view that the acid has the constitutionthe sodium salt is to be represented by theSimilar differences are found in the case of the chlorites which givethe authors another opportunity of making the same suggestion,namely, that the free acid has the constitution 0 = C1- OH and thesalts the constitution Cl<&::*M. It is hardly necessary to mentionthat there is not the slightest evidence in favour of this fancifulsuggestion, and that the great mass of experimental evidence isagainst any such explanation. These substances form typicalexamples of the same nucleus, C10 - or C10,-, having differentenergy contents when in combination with hydrogen or an alkylgroup and with a metal. Another example is afforded by thenitrates.A convenient method has been described for the, preparation ofhydrobromic acid solutions in the laboratory.71 To 25 C.C.of potass-ium bromide (oontaining 15 grams of the salt) are added 0.2 gramof stannous chloride and 3.4 C.C. of concentrated sulphuric acid.By distillation a t 120-127O a solution of hydrobromic acid is ob-tained, free from tin and almost free from hydrochloric acid, theyield being 95 per cent.The red compound, CaOBr,,H,O, formed by acting on quicklimewith bromine and water in the proportion of 100 grams of lime,41 c.c of bromine, and 36 C.C. of water, on heating at looo losesbromine and water, and yields a new basic hypobromite,CaO,CaOBr,,H,0.72 This compound is a pale yellow powder andcontains about 33 per' cent.of available bromine.0.Growp V l l l .The solubility of the ammoaium salts of chloroplatinic, bromo-platinic, and chloroiridic acids in water has been determined a t a70 K. Schaefer and W . K6hler, Zeitsch. physikal. Chem., 1919,93, 312 ;A., ii, 207.71 A. Pickles, Chem. News, 1919, 119, 89 ; A., ii, 411.7' J. S. Arthur and L. G. Killby, Brit. Pat. 131750 ; A,, ii, 46552 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.number of temperatures.73 The platinibromide is somewhat morereadily soluble than the platinichloride, whilst the solubility of theiridichloride is nearly twice as great as that of the platinichloride.In the presence of ammonium chloride the solubility of ammoniumplatinichloride and iridichloride is much reduced, but that of theiridichloride is several times as large as that of the platinichloride.Similarly, ammonium bromide reduces the solubility of ammoniumplatinibromide.I n all three cases the reduction in solubility isproportional to the concentration of the ammonium haloid. Thedifference in the solubility of ammonium platinichloride and iridi-chloride furnishes a good method for the complete separation ofplatinum and iridium.The inhibiting influence of various substances on the absorptivepower of palladium for hydrogen has long been known, but onlyin a qualitative sense. The influence of hydrogen sulphide has nowbeen quantitatively determined and the observations have led to amost interesting and important result.74 I n the first place it wasnecessary to fix the conditions of experiment.Since the bulk ofthe absorbed hydrogen is evolved at looo in a vacuum, and sincealso it was advisable t o avoid the danger of changing the activityof palladium by heating, it was decided to fix looo the maximumtemperature a t which the metal should be heated.I n the second place it was necessary to determine the amount ofhydrogen evolved by a given quantity of palladium at looo, andalso the amount absorbed by this dehydrogenated palladium a tordinary temperature. This volume of hydrogen was found to be68-5 C.C. for one gram of palladium. On treating the dehydro-genated palladium with hydrogen sulphide it was found that in afew minutes the gas was rapidly absorbed up to about 13.5 C.C.pergram of palladium. This was followed by a slow and continuousabsorption of a secondary nature, the total volume of hydrogensulphide absorbed in 40 hours being 22.5 C.C. per gram of palladium.The hydrogen sulphide thus absorbed was not removed to any greatextent by exhaustion at ordinary temperatlure, this being especi-ally the case when the hydrogen sulphide content of the palladiumwas comparatively low, and on treatment of the palladium withhydrogen occlusion no longer took place.On heating the palladium containing hydrogen sulphide in avacuum at looo a volume of gas, approximately equal to that ofthe hydrogen sulphide contained in the palladium, was evolved.This gas, however, was found to consist almost entirely of hydrogen,78 E.H. Archibdd and J . W. Kern, Trans. Roy. SOC, Canada, 1917-1918,[iii], 11, 7 ; A., ii, 70.74 E. B. Maxted, T., 1919,115, 1050INORGANIC CHEMISTRY. 53the sulphur being retained by the palladium. An interestingobservation was made with respect to the specific influence of thesulphur absorption compound on the occlusive power of the palla-dium for hydrogen, in that, whilst about 13.3 C.C. of hydrogen sul-phide are sufficient completely to inhibit the occlusive power forhydrogen of one gram of palladium, the equivalent quantity ofsulphur, which remains behind after exhaustion at looo, is by nomeans sufficient completely to prevent the occlusion of hydrogen.The influence of the sulphur retained by the palladium after ex-haustion a t looo on the occlusive power for hydrogen wits quanti-tatively determined. The mean occlusive power is approximatelya linear function of the sulphur cont<ent, and each atom of sulphurrenders1 almost exactly four palladium atoms incapable of occlud-ing hydrogen, the remainder of the palladium being capable ofoccluding normally. This obviously raises the question of theformation of a definite sulphide of palladium, but, as the authorpoints out, there is insufficient evidence to justify such an assump-tion, and he mentions the fact that palladium foil remains untar-nished in pure hydrogen sulphide both a t the ordinary temperatureand at looo.There is no doubt that this result is one of great importance forit may be discussed from an aspect not mentioned by the author,namely, the catalytic activity of the hydrogen occluded by metals.There is little doubt that this activity is due to the supply by themetal of energy to the hydrogen molecules sufficient to dissociatethem into atoms. No quantitative data are to hand as regards thenumber of molecules of hydrogen activated by a given number ofmetallic molecules. The results described above would seem toafford the first instance of a definite quantitative relation, sincefour palladium atoms can supply sufficient energy to dissociateone molecule of hydrogen sulphide into a molecule, of hydrogenand an atom of sulphur. It is obvious, therefore, how it comesabout that more palladium is poisoned by a given volume of hydro-gen sulphide than is accounted for by the formation of the Pd,Scomplex. I n forming this complex one gram of palladium requiresabout 52.5 C.C. of hydrogen sulphide, and during the formation ofthis complex 52.5 C.C. of hydrogen are set free. This volume ofhydrogen will be absorbed by a further quantity of palladium whichwill then no longer have any power of absorbing hydrogen. It isnot possible from the evidence a t hand to calculate, the, equilibriumconditions which exist.If, however, Maxted’s figure for the poisoning of palladium byhydrogen sulphide is correct, namely, that one gram of palladiumis completely poisoned by 13.5 C.C. of hydrogen sulphide, the rela54 ANNUAL REPORTS ON TEE PROGRESS OF CHEMISTRY.tive amounts of energy required to dissociate the molecule of hydro-gen into atoms and the molecule of hydrogen sulphide into hydro-gen and sulphur may be roughly approximated: 13.5 C.C. ofhydrogen sulphide in forming the complex Pd,S account for 0.257gram of palladium, and therefore, 0.743 gram of palladium isrequired t o activate 13-5 C.C. of hydrogen. Thus about 12 gram-atoms of palladium are required completely to activate one gram-molecule of hydrogen. Assuming that in each case the amount ofenergy available from each atom of palladium is the same, it followsthat three times as much energy is required to dissociate one mole-cule of hydrogen into atoms as is required to dissociate one mole-cule of hydrogen sulphide into a molecule of hydrogen and an ;tornof sulphur. On the energy quantum theory this would lead tothe conclusion that the frequency of the ultra-violet absorptionband of hydrogen must be about three times that of the absorp-tion band of hydrogen sulphide. E. C. C. BUY