INORGANIC CHEMISTRY.ALTHOUGH perhaps the work carried out during the past year doesnot include any great and far-reaching discovery, yet it may atonce be said that a very large number of papers have beenpublished, many of which are of considerable interest and value.A great difficulty must naturally be felt in presenting a report oninorganic chemistry arising from the fundamental importance ofmany papers which perhaps may be considered more properly asbeing within the purview of physical chemistry. A few of thesewhich manifestly are of importance in relation to inorganic chem-istry have been included.I n the main the report follows in its arrangement the precedentof previous years. A section has been added on the rare earths,and as no especial reference has been made to this branch of workof late, two papers published in 1911 have been included.,4 tomic 'IVeights.During the past year a number of important papers dealing withthe determination of atomic weights have appeared.NitTogen.-Wourtzel1 has redetermined the ratio between nitro-gen and oxygen, the method involving the determination of theweight of oxygen necessary to convert a known weight of nitricoxide into nitrogen peroxide.Pive concordant experiments gave amean value for the atomic weight of nitrogen as 14.0068.Potassium and Chlorine.-Reference was made in last year'sreport to an interesting redetermination of the ratio betweenpotassium and chlorine which was based on a careful analysis ofpotassium chlorate, very great precautions being taken againstcontamination of the salt by the chloride.The final series ofresults gives the atomic weight of potassium as 39.097 and of chlorineas 35-458.2 These results have been favourably criticised by Guye,3Concpt. rend., 1912, 154, 115 ; A., ii, 248.Stahlcr mid Meyer, Zeitsch. nnorg. Chem., 1911, 71, 368 ; A., 1911, ii, 881.J. Chwn. Phys., 1912, 10, 145 ; A., ii, 552.3INORGANIC CHEMISTRY. 37who concludes that the amount of potassium chloride present inthe chlorate was a t least sufficiently eliminated to be practicallynegligible.Since the Report of the International Committee on AtoxicWeights has appeared, two papers have been published dealing withthe atomic weight of chlorine. I n one of these Wourtzel4 deter-mines the nitrogen-chlorine ratio by the formation of nitrosylchloride from nitric oxide and chlorine.As a mean of five resultsthe atomic weight of chlorine was found to be 35*460+0.003(N = 14.008).Baume and Perrot5 also determined the ration of chlorine tonitrogen from the combination of gaseous hydrogen chloride with aknown quantity of liquid ammonia, and the value found for chlorinewas 35.465 (N= 14.009).Burt and Whytlaw-Gray 6 have redetermined the density ofhydrogen chloride. The gas was absorbed by charcoal, care beingtaken to remove all mercury vapour by condensation with solidcarbon dioxide. The weight of a litre of the gas was found to be1.63915 & 0.00004, which is identical with the previously foundvalue.Bromine.-An important paper has appeared by Weber 7 describ-ing his experiments to determine the ratio between hydrogen andbromine.As a mean of ten experiments, Weber found the atomicweight of bromine to be 79.924.Fluorine.-McAdam and Smith,* as a result of experiments onthe conversion of sodium chloride into sodium fluoride, have pub-lished two preliminary determinations of the atomic weight offluorine, which are 19.0176 and 19.0133 respectively.Phosphorus.-Baxter, Moore, and Boylston 9 carried out theanalysis of phosphorus tribromide, and found as a mean of threeseries of experiments the atomic weight of phosphorus to be 31.027(Ag= 107'88).Baxter and Moore,l0 from the analysis of phosphorus trichloride,have found, as the mean of their most trustworthy experiments,the atomic weight of phosphorus to be 31.028.Mercury .-From an analysis of mercuric bromide Easley andBrannl' have confirmed the atomic weight of mercury as 200.64,This gives the atomic weight of chlorine as 35.460.C'ompt.rend., 1912, 155, 345 ; A., ii, 931.Ibid., 461 ; A,, ii, 933.Tyans. Fccia&ay SOS'~~., 1911, 7, 30 ; A., ii, 152.7 J. Amel.. Chm. SOC., 1912, 34, 1294 ; A., ii, 1162.8 ]bid., 592 ; A., ii, 549.Ibid., 259 ; A . , ii, 347.lo Ibid., 1644 ; A., 1913, ii, 43.l1 Ibid., 137 ; A., ii, 25738 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.a value agreeing with that previously found by Easley in his workon the chloride.Tellurium.-As was mentioned in last year’s report, Harcourtand Baker have thrown doubt on the work of Flint, in which thelatter claimed to have resolved tellurium into two fractions havingdifferent atomic weights.Pellini 12 has also come to a similar con-clusion. On the other hand, Flint13 has published a further paperon the subject, in which he points out that the explanation of thelower atomic weight put forward by Harcourt and Baker, namely,the presence of tellurium trioxide, is not sound. Flint gives it as hisopinion that the matter has not yet, been settled, and promises afurther paper on the subject.Dudley and Jones l4 have carried out spectrographic investiga-tions which are important in the present connexion. They findthat there is no variation in the spectrum of the different fractionsof tellurium, which indicates that no resolution of tellurium hastaken place under the conditions of fractionation.Further work ispromised in this direction.Flint’s work has been repeated by W. C. Morgan15 with largeramounts of material, and although twice as many fractionationswere carried out, no evidence of separation was obtained.Holmium.-Holmberg 16 has carried out determinations on theatomic weight of holmium by means of the sulphate method, andfinds, as a mcan of six determinations, the atomic weight to be163.45. I n the International Table of Atomic Weights for 1913this number has heen adopted.Radium-Two interesting papers on the atomic weight ofradium have appeared during the past year. I n one, Honigschmid 17found from the analysis of a large quantity of radium chloride theatomic weight of radium to be 225.95.I n the other, Whytlaw-Grayand Ramsay,18 by the conversion of small quantities of radiumbromide into chloride, found the atomic weight to be 226’36, aresult which is in agreement with the well known value of Mme.Curie and of Thorpe.Palladium.-Shinn 19 has carried out a series of determinationsof the atomic weight of palladium, the method used being thereduction of the double chloride of ammonium and palladium byP2IS141 J1617IS19Atli R. Accad. Lincei, 1912, [v], 21, i, 218 ; A,, ii, 343.J. Amer. Chem. Soc., 1912, 34, 1325 ; A., ii, 1051.Zbid., 995 ; A . , ii, 9351Ibid., 1669 ; A . , 1913, ii, 41.Arkiv Kern. Min. Geol., 1911, 4, No. 10; A., ii, 163.Momlsh., 1912, 33, 253 ; A , , ii, 523.Proc.Roy. Soc., 1912, A, 86, 270 ; A., ii, 413.J. Amer. Chm. Soc., 1912, 34, 1448; A., ii, 1178INORGANIC CHEMISTRY. 39formic acid in the presence of ammonia. As a mean of nine deter-minations the atomic weight was found to be 106.709 +_ 0.016.Uranium.-Lebeau 20 has carried out the reduction of uranylnitrate dihydrate to uranium oxide by heating the salt in a currentof hydrogen a t l l O O o , and as a mean of five experiments he obtainedthe value 238.50, thus confirming the previous value obtained.Iron.-Baxter and Hoover,21 by the reduction of ferric oxide inhydrogen, find the atomic weight of iron t o be 55.847 as a mean oftwelve analyses.Molecular Weights.RBy, Dhar, and De22 have determined the vapour density ofammonium nitrite.A weighed quantity of the salt, previouslysublimed in a vacuum, is introduced into a Hofmann tube heatedto 78O with alcohol vapour. Part of the salt decomposes intonitrogen anci water, and part into ammonium nitrate, ammonia,nitric oxide, and water, according to the equation :3NH,NO, = NH,N03 + 2N0 + 2NH3 + R20,whilst the remainder exists in the form of undecomposed vapour.The products of decomposition were estimated, and the amount ofammonium nitrite used up in their formation was calculated, Thisamount deducted from the original quantity gave the amount exist-ing in the form of vapour, and thus, as the volume of the vapour wasknown, the vapour density could be arrived at. The mean oftwenty determinations gave a vapour density of 33.5, the theoreti-cal value being 32.0.The authors point out that the presence orabsence of water made little or no difference.Ramsay23 has determined the ratio of the specific heats of neon,krypton, and xenon by comparing the wave-length of sound in thegas with that of air. He found the values as follows: Ne=1*642,Kr = 1.689, Xe= 1.666. These numbers approximate within thelimits of experimental error to the theoretical ratio 1.667, and i ttherefore follows that these three gases, like helium and argon,,must be regarded as monatomic. The molecular and atomic weightsmust be identical.Preuner and Brockmoller,2* having devised a spiral manometermade of quartz for the accurate measurement of gas pressures,determined the isothermals of sulphur, selenium, arsenic, and phos-phorus, from which they calculated the molecular weight and theCompt.Tend., 1912, 155, 163 ; A , , ii, 848.21 J. Amer. Chem. SOC., 1912, 34t, 1667; A . , 1913, ii, 55.2.2 T., 1912, 101, 1185.24 Zeitsch. physikal. Chem., 1912, 81, 129 ; A,, ii, 1145.Proc. Boy. Soc., 1912, A, 86, 100; A., ii, 25140 ANNUAL REPORTS ON THE PROURESS OF CHEMISTRY.dissociation of the vapours of these substances. They determinedthe isothermals of selenium a t 550°, 600°, 650°, 700°, 750°, and800°, and found that the equilibrium Se, 2 3Se2 takes place overthis region. A t low and high temperatures there is some evidenceof Se, and Se atoms respectively. I n the case of sulphur vapour,S,, S,, S2, and S are present in amounts depending on the conditionsof temperature and pressure.The isothermals of phosphorus andarsenic were measured a t 800°, 900°, 1000°, l l O O o , and 1200°, andfrom these curves the authors find that tetra-atomic and diatomicmolecules as well as single atoms are present.The density of phosphorus vapour has been determined withgreat accuracy by Stock, Gibson, and Stamm25 with Gibson’s mem-brane manometer.26 The investigations extended over a tempera-ture range of 500° to 1200°, and the first conclusion arrivedat is that phosphorus vapour obeys Boyle’s and Gay-Lussac’s lawsbetween 500° and 700° and 240 and 300 mm. pressure, and thati t consists entireIy of P, molecules. A t higher temperatures disso-ciation into P, sets in, especially at lower pressures. The maximumdegree of dissociation observed (twethirds of the P, having disso-ciated) was a t 1200°, and at a pressure of 175 mm.The authorsprove that the only equilibrium concerned is P, 2P2. Theyalso calculated that a t the temperature of 1300O the amount ofdissociation would be 60 per cent. a t 760 mm. and 89 per cent.at 100 mm. pressure.These results differ from those of Preuner and Brockmollermentioned above, and the authors criticise those results whichwould seem to point to there being some phosphorus atoms presenta t the higher temperatures. They point out that whereas they usedcarefully purified red phosphorus, Preuner and Brockmoller usedcommercial yellow phosphorus, which they merely washed severaltimes.Pal enc y.Some interesting experiments have been carried out by Ephraim 27on the hexammine compounds of certain metals, with the view ofobtaining a quantitative measurement of the auxiliary valencies inthese compounds.Ephraim points out that as these compoundsare formed by virtue of the secondary valencies of the principalmetallic atom in the molecule it should be possible to determinethe relative strength of these auxiliary valencies by measuring thedissociation pressure of the hexammine compounds. The author25 Ber., 1912, 45, 3527 ; A., 1913, ii, 43.2G To be described later.3 Bcr,, 1912, 45, 1322 ; A , , ii, 546INORGANIC CHEMISTRY. 41considers that if the pressure is kept constant the temperature a twhich the decomposition occurs will be a measure of the energy ofthe auxiliary valencies because the decomposition of the molecularcomplexes must be dependent on three factors, namely, the pressure,temperature, and affinities of the constituents.I n one series of experiments he finds that the temperature a twhich the dissociation pressure of the hexammine chlorides ofglucinum, nickel, cobalt, iron, copper, manganese, zinc, cadmium,and magnesium is equal to 500 mm.decreases as the atomicvolime of the metal increases. I n other words, the auxiliaryvalency of the metal is a function of its atomic volume. As Ephraimpoints out, the atomic volume is not the only factor, because copperand iron, which have the same atomic volume, have not the sameaffinity towards ammonia. The dissociation pressures of thehexammine compounds of these two metals are similar, but notidentical.The stability of the hexammine compounds alsodecreases with increase of atomic volume, and i t is an interestingfact that no single metal with an atomic volume greater than 14can form a hexammine derivative which is stable a t the ordinarytemperature. Similarly, in the case of glucinum, the atomic volumeof which is 5.26, the hexammine derivative is so stable that itdecomposes in other ways before the temperature is sufficient togive an ammonia pressure of 500 mm.He also finds that the influence of the atomic volume is less thelarger its value, for of seven hexammine chlorides examined it wasfound that the value of $F. vg was approximately constant,where T is the absolute temperature a t which the ammonia pressureis equal t o 500 mm., and v is the atomic volume.It was also found that the acid radicle has some influence on theenergy of the auxiliary valencies; f o r example, the ratio of theabsolute temperatures a t which the dissociation pressures of nickelchloride and nickel bromide is equal to 500 mm.was 1.073, and inthe case of the chloride and bromide of cadmium this ratio was1.086. This factor Ephraim calls the tension-modulus, and taking1-08 as the mean of the above two values he calculates the dissocia-tion pressure of the hexammine bromide of cadmium from theexperimentally found value for the chloride. The value obtainedis 349.3, whilst the experimental value is 347.5, which is quitewithin the limits of experimental error.I n exactly the same waythe values of the tension-moduli Br/I, Cl/I, SO,/I were measuredand found to be constant. The only exceptions were the hexamminecompounds of zinc. The constant ratio between the tension-moduliwas also found to hold good over considerable ranges of pressure;for example, the ratio for the bromide and iodide of cadmiu42 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.between the pressures of 103 and 780 mm. was found to have amean value of 1.066.This therefore appears to be a special case of Ramsay and Young'srule, and it was also found that this rule could be applied to thehexammine compounds of all the metals examined between thelimits of 200 and 700 mm. pressure. From the dissociation-pressurecurves the heat of formation, Q, was calculated from the expres-sion :T .Tz Q =L . log%.2; - Y2 p ,From the values obtained it follows that for equal dissociationpressures Q / T is very nearly constant, which corresponds withTrouton's law-.The value of the constant Q/Z' for dissociation pressures of500 mm. was found to be between 0-030 and 0.036.I n connexion with this i t is interesting to note that C'hauvenet 29found for ThC1,,18NH3 the value of &/T to be 0.032.This was previously shown by Matignon.28A llotropy.Stock, Schracler, and Stamm 30 have carried out an investigationon the transformation of yellow phosphorus into red phosphorusphotochemically. The yellow phosphorus was placed in a side-tube in an exhausted flask.By local cooling a, thin layer of yellowphosphorus was obtained on the walls of the flask, and this filmwas illuminated on the inner side by rays from a mercury lamp.The residual yellow phosphorus was removed by cooling the side-tube with liquid air for several days. It was found that the violetend of the spectrum was most active in the conversion, and thattemperature had very little influence. The density of the redphosphorus thus obtained varied from 1.95 to 2.25, the darkest incolour being tho heaviest. It is possible that some crystals werepresent.. It was also found that rapid cooling of phosphorus vapourgives red phosphorus, the amount obtained increasing with thetemperature to which the vapour was heated before being suddenlycooled.The authors conclude that the two varieties of phosphorusare chemically different, and that the red variety is formed by theunion of dissociated phosphorus particles with on0 another or withundissociated phosphorus molecules. Red phosphorus prepared inthis way is yellowish-red to blood-red in thin layers, and blackwith a violet glance in compact pieces. When ground to a powderit resembles the commercial article. I t s maximum density was2.215, and it appears to be crystalline. It is stable in air, is only2* Compt. rend., 1899, 128, 103 ; A . , 1899, ii, 273.2Q Ann. Chim. Phys., 1911, [viii], 23, 275 ; A., 1911, ii, 586.3u Ber., 1912, 45, 1514; A,, ii, 639INORGANIC CHEMISTRY. 43slowly attacked by boiling sodium hydroxide, and does not give ared solution in alcoholic potassium hydroxide solution.Hesehus21 has studied the properties of diamond, graphite, andcharcoal, and he shows that the differences between the contactelectrification of the allotropic modifications depend, not only ontheir surface density, but also on the capacity of the atoms forionic dissociation.It would seem from this result of Hesehus thatin all probability the catalytic action of the various allotropicmodifications would depend on the residual affinity.Comb us tion.Rhead and Wheeler have published papers in which they dealwith the equilibrium between carbon dioxide and carbon, and theoxidation of carbon itself. I n the first paper 32 they describe theirinvestigations into the rate of reduction of carbon dioxide bycarbon a t various temperatures.The method consisted in thecirculation for known periods of carbon dioxide over carbon heatedto different temperatures, the time taken for a complete circulationof the gas through the apparatus having been previously deter-mined. The amount of reduction of the carbon dioxide wasobtained from the increase of pressure in the apparatus. Theresults show that there is a considerable falling off in the amountof reduction taking place in a given time when the temperaturefalls below 1000°. One set of experiments was carried out with amixture containing 80 per cent. of nitrogen and 20 per cent. ofcarbon dioxide, and this has some interest from the point of viewof producer gas, because they found that owing to the short time ofcontact in the producer a temperature of more than l l O O o is necessaryfor the maximum production of carbon monoxide. The presence ofthe 80 per cent.of nitrogen causes a general retardation in the rateof reduction, and so enabled the authors to appreciate the truerelative rates a t the higher temperatures. I n the second papersthey consider the oxidation of the carbon itself. I n connexion withthe oxidation of carbon considerable evidence has been advanced,both in favour of the oxidation t o carbon monoxide and in favourof the oxidation to carbon dioxide. It is perhaps more generallyaccepted that the reaction takes place in the following way:( a ) c+o,=co,; ( b ) co,+o=2co.The work of Dixon and also that of H.B. Baker would seem tofavour the view that the carbon is oxidised first to carbon mon-oxide, as expressed by:31 J. Russ. Pliys. C h i t So,:., 1911, 43, Phys. Part, 365 ; A . , ii, 121.32 T., 1912, 101, 831.( G ) 2c 3- 0,=2CO,Ibid., 84644 ANNUAL REPOnTS ON THE PROGRESS OF CHEMISTRY.and that any carbon dioxide is due to the process:(d) 2CO + o,= 2c0,.Rhead and Wheeler point out that if the relative rates of ( a ) and(c) could be determined under conditions when the amounts of( b ) and (d) are inappreciable, the matter would be solved. Theyfind, however, that during the oxidation of the carbon at low tem-peratures neither ( a ) nor (6) takes place exclusively, and they areobliged to conclude that both carbon monoxide and carbon dioxideare simultaneouely produced by the direct oxidation of carbon.Wieland34 deals with the reaction between carbon dioxide andwater in the presence of palladium-black. He finds that the firstproduct is formic acid, which then gives carbon dioxide andhydrogen, the hydrogen being retained by the palladium. Theintermediate existence of formic acid was proved by chemicalmeans Wieland also showed that formic acid is an intermediateproduct in the hot combustion process.If a carbon monoxideflame is allowed t o play on ice, the presence of formic acid in themelted ics can readily be proved. The combustion of the hydrogenfrom the formic acid gives the water which is necessary for thecombustion t o proceed, and prevents the reversible reaction fromtaking place. Traube's view of hydrogen peroxide playing animportant part is wrong, for the hydrogen peroxide only occurs asa subsidiary product in the first stage of the combustion of thehydrogen from the formic acid.It is well known that the platinummetals catalyse the reaction :H*CO,H = E, + CO,,and in the case of palladium the catalytic influence is so greatthat the evolution of hydrogen is stronger than the absorption ofhydrogen by the palladium, with the result that hydrogen isevolved. If the palladium-black has previously been ignited, itspower as a catalyst is destroyed.F. Meyer 35 describes experiments to show that the outer zone of areversed-flame is strongly reducing, so that any substance made topass through the flame from the inside to the outside will bereduced, and remain reduced.It is necessary that the combustionproducts of the flame should be equal to one of the products formedby the reduction of the substance, for example, a chlorine-hydrogenflame must be used for the reduction of chlorides, the chlorine flameburning in a quartz jet inside hydrogen. Meyer found that thisreduces stannic chloride t o stannous chloride, arsenic chloride t oarsenious chloride, and titanium tetrachloride to titaniumtrichloride.:I4 Bcr., 1912, 45, 679; A., ii, 347.35 Ibid., 2548 ; A . , ii, 1051INORGANIC CHEMISTRY. 45rs’ulphuric Acid Munufacture.Reynolds and Taylor 315 have carried out a series of experimentswith the view of testing the theory put forward by Raschig37 of thechamber sulphuric acid process.Raschig supposes that an unstablenitrososulphonic acid is first formed by the action of sulphurdioxide on nitrous acid, as shown by the equation:and that this acid is a t once transformed by more nitrous wid intoanother unstable compound, nitrosisulphonic acid :ONS03H + €€NOz = H2NS05 + NO,and that this new acid then decomposes into sulphuric acid andnitric oxide :H,NSO,= HzS04 + NO.HNO, + SO, = ONSO,H,Raschig states that his nitrosisulphonic acid is identical with theviolet acid discovered by Sabatier.This theory was somewhat modified by Divers?* who points outthat Raschig’s nitrososulphonic acid has probably no existence, andthat the evidence is all against the existence of nityosisulphonicacid.Reynolds and Taylor confirm the conclusion of Divers thatnitrososulphonic acid has no existence, and prove by experimentthat dilute solutions of nitrous and sulphurous acids mutuallydecompose one another with evolution of nitrous oxide. They alsoconfirm Divers’ conclusion that this reaction takes place in twostages :(i) HNO,+ 2H,S0,=H2O + HON(SO,R),.(ii) HON(SO,H), 1- HNO,= 2H,S04 + N,O.According to Raschig, the nitrososulphonic acid being formed ina strong acid solution containing excess of sulphurous acid a t onceundergoes decomposition according to the second equation above.In his experiments potassium iodide was added to the solution toindicate whether excess of nitrous acid was present o r not.Reynolds and Taylor prove that the evolution of nitric oxide in theexperiment described by Raschig in support of his second equationis due to the action of nitrous acid on hydriodic acid.Raschigattached great importance to the violet acid discovered by Sabatier.He gives experiments to prove that although this compound isproduced in the reduction by mercury of a solution of chambercrystals in 85 to 100 per cent. acid, it was not formed in more diluteacid in this way, and he concluded that the chamber crystalscould not exist in acid containing less than 80 per cent. of sulphuricJ. SOC. Chem. €nd., 1912, 31, 367 ; A . , ii, 550.37 Ibid., 1911, 30, 166 ; A., 1911, ii, 272.33 I h X , 594; A . , 1911, ii, 59646 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.acid.Hence, as he found that the violet acid can be formed bythe action of sulphur dioxide on nitrous acid in a 70 per cent.solution of sulphuric acid, he maintains that the violet acid is thedirect product of the interaction between nitrous and sulphurousacids, and that it could not have been formed by the reduction bysulphurous acid and chamber crystals, because the latter werenon-existent.Reynolds and Taylor prove conclusively that chamber crystalscan exist even in 60 per cent. sulphuric acid, and they satisfy them-selves that Raschig’s explanation of the formation of sulphuric acidis wIong. Their evidence proves first that the only gaseous productin the interaction of sulphurous and nitrous acids is nitrous oxide,and secondly, that Sabatier’s violet acid is only formed in solutionscontaining chamber crystals ; thus they conclusively indicate thefutility of a l l attempts to account for the formation of the bulkof the sulphuric acid in the chambers by the intermediary forma-tion of nitrogen sulphonic acids.They believe that the old viewssubstantially represent the course of the process, namely :(i) SO, + H20 = H2S03.(ii) H,SO, + NO, = NO + H,SO,.(iii) 2NO + O,= 2N0,.The nitrous oxide which escapes may be a measure to some extentof the extent to which nitrogen sulphonic acids are formed,although the work of Moser 39 leaves it uncertain as to whether thismay not have been partly derived from the decomposition of thenitric oxide in contact with the water locally. It must be regardedas doubtful if nitrososulphuric acid (chamber crystals) plays animportant part in the process.The authors agree with Divers inexpressing a doubt as to whether it is ever formed except by theaction of sulphuric acid on one of the oxides of nitrogen, but assulphurous acid reacts with a solution of chamber crystals somesulphuric acid is possibly formed in this way.40In connexion with the contact process of sulphuric acid manufac-ture, Wieland 41 has investigated the reaction which takes placewhen moist sulphur dioxide is passed over palladium-black, oxygenbeing rigidly excluded. He found that considerable heat isdeveloped, and that after driving off the excess of sulphur dioxidewith carbon dioxide and extracting the palladium-black with water,considerable quantities of sulphuric acid are present in the solution.This sulphuric acid is formed according t o the equations:&SO, = SO, + H, and SO, + H,O = E,SO,.39 Zeitseh.anal. Chein., 1911, 50, 401 ; A . , 1911, ii, 598.40 For further experiments contradicting Raschig’s theories see Manchot, Zeitsch.41 Ber., 1912, 45, 685; A., ii, 343.angew. Chm., 1912, 25, 1055; A . , ii, 637INORGANIC CHEMISTRY. 47It was not found possible to detect the hydrogen which is pro-duced according to the first equation, for sulphur is deposited in itsplace, being formed by the action of palladium hydride on sulphurdioxide.SO, + ZH, = 2H20 + S,this being shown by the sulphur which is produced when a solutionof sulphurous acid is shaken with palladium hydride.Further, if apalladium-black is used which has initially been freed fromhydrogen, the yield of sulphuric acid is increased and less sulphuris formed, which suggests that in the contact process for makingsulphuric acid the above equations represent what really happens,the hydrogen formed in the first reaction being oxidised to waterby the oxygen which is mixed with the sulphur dioxide. Wielandpoints out that this view is supported by the fact that no uniontakes place between the sulphur dioxide and the oxygen if theseThe reaction took place according to the equation:gases are dry.The Rusting of Iron.An important contribution to the literature on the rusting ofiron has been made by Lambert.*2 I n spite of the fact that muchexperimental work has been done in recent years to prove thatordinary iron will always undergo corrosion when placed in contactwith water and oxygen, even in the absence of carbon dioxideand other acids, this work has been criticised by the supporters ofthe acid theory on the ground that sufficient care has not beentaken to remove all traces of carbqn dioxide from the apparatus.I n the present paper Lambert describes his experiments, in whichiron was exposed to pure water and oxygen, the most rigid precau-tions being taken t o exclude every trace of carbon dioxide and anyother acid.Both the oxygen and the water were subjected to avery careful spectroscopic test, and not the slightest trace of carbondioxide was found in either. The result of the experiments wasthat the rusting seemed to take place as quickly as in ordinaryair. It can no longer be maintained that carbon dioxide or anyother acid is the dominant factor in the atmospheric corrosion ofcommercial iron.As was pointed out in last year’s report, anelectrical theory has been put forward to explain this phenomenon,and Lambert develops this theory in his paper, and brings forwardconsiderable evidence in strong support of it. I n the first place,some absolutely pure iron was prepared, and it was found that notrace of corrosion took place when it was kept in contact witheither pure water and pure oxygen, or air and ordinary tap water,for an apparently indefinite time. The explanation of this fact is42 T., 1912, 101, 205648 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.to be sought in the greater homogeneity of the iron. I f the ironbe really homogeneous, and all parts have the same solution pres-sure, then sufficient difference of potential will not exist, and norusting will take place.Furthermore, some pieces of this pure ironwhich have been exposed to water and air for some months withoutshowing any corrosion were carefully dried; some of the pieceswere then placed in an agate mortar and pressed strongly with anagate pestle, whilst others were left untouched. All were then putin contact with the water and exposed to the air. Those pieceswhich had been pressed rapidly corroded, rust forming on thoseparts of the iron which had not been pressed, whilst the partswhich had been in actual contact with the polished agate remainedquite bright.Those pieces of iron which had not been subjected topressure showed no sign of corrosion. Evidently, therefore, theapplication of the pressure induced a change in the iron, giving ita new solution pressure, so that electrical action was set up when itwas placed in contact with water. An interesting fact was alsonoticed in that chemically pure iron can be exposed for an appa-rently indefinite time to a saturated solution of copper sulphate orcopper nitrate without the slightest trace of copper being deposited.Copper is, however, deposited on the iron if it is pressed in anagate mortar before being put in the solution, or if pressed with aquartz rod whilst under the solution.I n the case of copperchloride, however, even the purest iron reacts, and becomes coatedwith copper immediately i t is put into the solution. The authorhas no completely satisfactory answer to the question as to whythe chloride has this abnormal effect, and for the present leaves itan open question. It is interesting to note that all the samplesof pure iron may not behave exactly in the same way, even thoughprepared from the same specimen of ferric nitrate. Some will showno corrosion after contact with water in air for an indefinite time,whilst other specimens show corrosion in a few hours. The differ-ence in behaviour cannot be due to different chemical composition,and must be due to physical differences produced by variation intempering and rate of cooling in the preparation.It would seemprobable from these results that the fundamental cause of thecorrosion of iron is not carbon dioxide or any other acid, but thatthe cause is rather to be sought in the difference of solution pressureof the different parts of the iron, differences which may persisteven in the most highly purified form of the metal.Electric Discharge.During the past year Strutt has carried out further experimentsH e finds43 that the active on his active modification of nitrogen.4J Proc. Roy. SOC., 1912, A, 86, 262 ; A . , ii, 447INORGANIC CHEMISTRY. 40modification emits its energy more quickly and reverts sooner toordinary nitrogen if i t is cooled. This would appear to be the firstcase in which a chemical or physico-chemical change is increased invelocity by cooling.If the glowing gas is compressed to a smallvolume it flashes out with great brilliance, and exhausts itself in sodoing. The glow transformation seems to be a multimolecularphenomenon. The active nitrogen may revert to ordinary nitrogenin two ways. The first way consists in a volume change accom-panied by a glow, and the second in a surface action on the wallswithout glow. This is analogous to hydrogen and oxygen produc-ing water, which may be a surface or a volume effect according tothe circumstances. In a second paper Strutt 44 describes experi-ments which prove that active nitrogen is highly endothermic, butthat its energy is of the same order as that of other chemicalsubstances.I n the reversion to ordinary nitrogen, the numberof atoms ionised is a very small fraction of the total concerned.The ionisation, therefore, is a subordinate effect, and may be dueto light of very short wave-length being emitted during the change.I n this paper further proof is adduced that the change is acceler-ated by cooling, and Strutt puts forward the view that the pheno-menon is connected with a monatomic condition of the nitrogenmolecule. Strutt and Bowler *5 describe the production of variousspectra by means of the nitrogen glow, and they do not find thatthose spectra differ materially from those produced in ordinaryways. Certain minor differences are, however, to be noticed, andthe method certainly adds to our resources in producing spectra,.I n this connexion an interesting paper has been published byLowry46 dealing with the action of the spark gap and an ozonisingapparatus on the .production of nitrogen peroxide. The presenceof this gas was detected, and its amount estimated spectroscopically.A 19.5 m.[64 feet] column of the gas was used, and the light of aNernst lamp, after passage through this, was examined with asingle-prism spectroscope, and also its spectrum photographed. Theozoniser consisted of 13 Andreoli aluminium grids, 76.2 x 76.2 cm.[30 x 30 inches], separated by micanite. I n the spark gap apparatusthere were in the first experiments three spark gaps made by fixingiron studs 0.238 cm. [3/32nd inch] apart. It was found that if astream of air was passed though the ozoniser or through the sparkgaps alone, no trace of nitrogen peroxide was visible, but if the airwas passed through the spark gaps and ozoniser arranged in series,then about 1/4000th part by volume of nitrogen peroxide wasproduced.It was also noticed that if a current of air passed44 Proc. Lay. Soc., 1912, A, 87, 179 ; A , , ii, 935.45 Ibid., 1912, A, 86, 105 ; A., ii, 214. 46 T., 1912, i01, 1152.REP.-VOL. IX. 50 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.through the spark gaps first and then through the ozoniser, aboutan equal quantity of nitrogen peroxide was obtained. This resultis somewhat extraordinary, seeing that ready-made nitrogen per-oxide is bleached on being passed through an ozoniser. Lowry alsofound that if the ozoniser and spark gap were used in parallel andthe air emerging from them was mixed, nitrogen peroxide was alsoproduced. Evidently, theref ore, the nitrogen peroxide cannot bedue to a simple electrical action, but must be the result of a chemi-cal interaction between ozone and some oxidisable form of nitrogenprbduced in the spark gap.A second spark gap apparatus was setup containing seveqteen spark gaps, and the whole of the availableelectrical energy was concentrated on this. When the air issuingfrom this apparatus was rapidly passed through the absorptiontube it showed the presence of 1/7000th part of nitrogen peroxide,and when passed more slowly it showed 1/3000th part of theperoxide. No such increase of intensity of the nitrogen peroxidewas found in the gases prepared with the help of ozone. This provesthat the nitrogen peroxide formation is gradual and not instan-taneous. As nitric oxide is only slowly oxidised by oxygen andquickly by ozone, it is conceivable, of course, that ozone acceleratesthe formation of nitrogen peroxide in the above experiment.This,however, cannot be the case, since the seventeen spark gaps onlygave a yield 30 per cent. greater than the three spark gaps withozone. Lowry believes that the nitrogen is not oxidised by theozone, but that an active form of nitrogen is produced which isreadily oxidised by ozone. It is interesting that Strutt’s activenitrogen is not the same, since the latter when mixed with ozonisedoxygen a t low pressures and cooled with liquid air gives no traceof the oxides of nitrogen.It is not unreasonable to ask whetherthe production of nitrogen oxides in any form of electric dischargeis not due to the interaction of active nitrogen and ozone or atomicoxygen.Langmuir 47 claims t o have prepared a chemically active form ofhydrogen. If metallic filaments of tungsten, palladium, orplatinum are heated to a high temperature in hydrogen at verylow pressure (0.01 to 0.001 mm.) the hydrogen slowly disappears.It dissolves in the metals as hydrogen atoms, and these atomsappear to diffuse out, and then have little opportunity of joiningtogether again to form molecules, owing to the high temperatureand low pressure, and they are absorbed on the glass walls.Cool-ing the latter improves the yield. The maximum yield was7-3 cu. mm. This hydrogen appears to be extraordinarily active,giving, for example, hydrogen phosphide with phosphorus. Hydro-47 % Amcr. Chcm. Xoc., 1912, 34, 1310; A., ii, 1162INORGANIC CHEMISTRY. 51gen phosphide was placed in the apparatus, and was entirely disso-ciated, the yellow deposit of phosphorus being clearly seen. Thiscombined with the active hydrogen regenerating a certain amountof hydrogen p hosp hide.Pho t och em is t r y .Although perhaps this subject may be considered to fall moreespecially within the confines of physical chemistry, yet it is diffi-cult to omit in a report on inorganic chemistry all reference tothe extraordinarily interesting and important work which has beencarried out during the past year on photochemical reactionsbetween inorganic substances.There cannot be any doubt thatthis line of investigation promises to throw great light on themechanism of chemical reactions and on the affinity relations ofchemical compounds which are gradually being recognised as offundamental importance as the driving force of all chemical pro-cesbes. Marmier 48 has found that a dilute solution of sodiumthiosulphate (containing less than 6 grams in 1 litre) is decomposedunder the influence of light. The action of the light from a240 watt lamp placed 6-8 cm. above such a solution is to producesodium hyposulphite and some sulphur. After seventy minutes’exposure the solution was then found to contain only sodiumsulphite. The light seemed to have no action if a stronger solutionof thiosulphate was used.Reynolds and Taylor 49 have carried out an important investiga-tion on the decomposition of nitric acid by light, and they findthat the reaction takes place according to the following equation :4HN0, = 2H20 + 2N204 + 0,.The 100 per cent.acid when kept in the dark begins t o decom-pose spontaneously after about one month. When, however, theacid contains water i t is readily decomposed by light, and thereaction is reversible.Winther 50 points out that under the influence of ultraviolet lightferrous chloride is oxidised by mercuric chloride according t o thefollowing equation :2FeC1, + 2HgC1, = 2FeC1, + Hg2Cl,;this reaction is reversible, but the reverse process moves very slowlyindeed, unless it is so arranged as to give an electric current, whenit proceeds much faster.Sometimes a potential difference of asmuch as 0.1 volt has been obtained with a current of 1 milliampere.Boll and Job51 have measured the change in conductivity of the48 Compt. rend., 1912, 154, 32 ; A., ii, 112.51 Cow@. Tend., 1912, 154, 881 ; A., ii, 407.4g T., 1912, 101, 131.Zeitsch. Elektrochem., 1912, 18, 138 ; A . , ii, 318.E 52 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.hydrochIoroplatinic acids produced on exposure to light. Veryconcordant results were obtained, which showed that tbe reactionwas bimolecular. Matuschek and Nenning 52 find thah manychemical reactions are accompanied by evolution of light.Thereactions were caused t o take place in a beaker, a t the bottom ofwhich a piece of tin foil cut in the form of a star was attached,tbe whole standing on a photographic plate. The following reac-tions were found to give out light: The action between sulphuricacid, hydrochloric acid, or nitric acid and copper, tin, or lead, Alsoduring the solution of cupric oxide, cupric hydroxide, or potassiumhydroxide in acids, the slaking of lime, the setting of cement andplaster, and tho reaction between water and calcium carbide, thereis an evolution of light. This work, however, has been criticisedby Schneckenberg,53 who says that the effects obtained are probablydue to the photochemical action of the heat rays given out duringthe reaction on the photographic plate.Bennett 54 has investigatedthe photochemical reduction of copper sulphate solutions in thepresence of reducing agents, the latter being present in quantityinsufficient to cause the reduction to take place in the dark. I none set of experiments a dilute ethereal solution of phosphoruswas added to the copper sulphate solution, and it was found thatthe precipitation of copper phwphide was obtained very much morequickly under the influence of light than in the dark; in fact,when t,he relative concentrations were properly balanced, the reduc-tion would not take place in the dark at all, but took place readilyunder the influence of light.Lewis 55 has investigated the photochemical decomposition ofsodium hypochlorite solutions into sodium chloride and oxygen.The hypochlorite solution exhibits a marked absorption band inthe ultraviolet region, and i t would seem from the author’s observa-tions that this light in being absorbed produces the chemicalchange.Inasmmh as the reaction proceeds slowly in the dark,the phenomenon belongs to that class known as catalytic lightreactions. I n two papers56 a theory was put forward which wouldseem to offer an explanation of the mechanism of photochemicalreactions. Every atom in a chemical molecule possesses a certainamount of free affinity, and the force lines due t o these free affini-ties must condense together with the escape of free energy. Ittherefore follows that the reactivity of such a condensed system isless than when the system is unlocked by some suitable means.Thisunlocking may be produced by the influence of a solvent or of lightor of both together. Owing to the fact that the light is thereby6z Chem. Zeil., 1912, 36, 21 ; A., ii, 116.c4 J. Physical Cheat., 1912, 16, 782.53 Ibid., 278.65 T., 1912, 101, 2371.66 Ibid., 1469, 1475INORGANIC CHEMIS'L'RY. 53doing work against chemical forces, the light will be absorbed. 15follows naturally that the chemical reactivity of a molecular systemmay be materially increased under the influence of light of theright wavelength. This theory would also seem to explain theresults obtained by Chapman57 and his ceworkers on the photo-chemical reactions of chlorine. One of the most striking resultsof this work is the inhibiting action of nitrogen chlorides on theunion of hydrogen and chlorine.It would seem that these sub-stances exert as grsat a negative catalytic action as water exerts apositive catalytic action on other reactions. Whilst this action ofthe nitrogen chlorides might conceivably be due to their power ofabsorbing the active light rays, yet it would seem far more probablethat these substances have the power of preventing the opening ofthe closed force systems of the chlorine molecule by light. Thewriter would suggest that this may possibly be put to experimentalproof. I f a beam of light of fixed intensity be allowed to fall ona mixture of pure hydrogen and chlorine, the velocity of combina-tion can be measured, and if the light be previously passed througha tube containing chlorine the velocity of combination will bedecreased; then if a small quantity of nitrogen chloride be addedto tho chlorine in the first absorption vessel, the velocity of com-bination of the hydrogen and the chlorine in the second vesselshould tend t o increase.Colloids.Dumanski 5~4 describes the properties of colloidal arsenic tri-sulphide.It is a negatively charged colloid, which may be segre-gated by rapid centrifuging. The density of the material wasfound to be 2.938, and its specific conductivity to be 136x 10-6.The solution is coagulated by certain electrolytes, whilst othersreact with the sulphide; thus an iodine solution reacts accordingto the following equation :AS&& + 101 + 5H20 =Asz05 + lOHI + 35.SimilarIy, a solution of potassium permanganate, decolorised byhydrogen peroxide, oxidises the sulphide to arsenic oxide, no coagu-lation taking place.Solutions of alkali hydroxides and of potass-ium cyanide have no coagulating action. On the other hand, silvernitrate and copper sulphate coagulate the colloid, but the precipi-tate contains considerable quantities of silver or lead. Leadacetate coagulates the colloid, but no lead sulphide is precipitated.57 For an excellent review of the work on the influence of light on the union of55 Zeitsch. Chein, Ind. Kolloide, 1911, 9, 262 ; A., ii, 153.hydrogen and chlorine, see Chapman, Science Progress, 1912, 6, 656 ; 7, 6654 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.J. Hoffmann59 shows that the colour of ultramarine is due tocolloidal sulphur, and that a number of compounds act a6 suitablesolvents for colloidal sulphur. The silica in ultramarine can bereplaced by boron trioxide without the disappearance of the bluecolour.Furtbermore, the colour does not greatly alter when thesolvent is varied from Na,B,O,, to Na,B,,O,,. Potassium thio-cyanate when dehydrated and heated to redness becomes blue, sohere, again, a suitable solvent exists for the colloidal sulphur. Phos-phoric oxide when fused a t 900° with sodium sulphide gives a bluecolour, which, however, disappears on cooling.Hatschek 60 has carried out investigations into the production ofsubstances in the gel of silicic acid.This acid is added to a15 per cent. solution of sodium silicate, and after solidification ofthe silicic acid a hypertonic solution of a salt is added, the acid andthe positive ion of the salt being so chosen that a salt separates ofitin the solid form within the gel. I n this way lead chloride, andcopper, calcium, and strontium phosphates were obtained in thecrystalline form. I n a second paper61 the author deals with thereduction of gold in the silicic acid gel. Both solutions and gaseswere used as reducing agents. I n the former case the reducingagent was poured over the gel which contained gold chloride, and inthe latter case the gel in a test-tube was exposed to the gas. Undersuitable conditions the gold was always deposited either in thegel or on the surface of the gel, or both a t once, depending on theosmotic relationships.Furthermore, the gold was obtained inlayers, and finally acetylene was found t o give carbon and goldlayers. The results, therefore, obtained are quik similar t o thenatural deposits which can therefore be explained in this way.A certain amount of work has been done on the preparation ofmetallic sols and metallic organosols. Pappadb 62 describes thepreparation of colloidal silver, gold, and platinum. These arenegative colloids, and their reactions towards electrolytes are similar,but the most active of the three is silver. Amberger 63 describesthe preparation of metal organosols. If lanolin, impregnated withthe aqueous solution of the salt of a heavy metal, is triturated witha solution of alkali hydroxide, the hydroxide of the heavy metalin colloidal condition is obtained.I f the oxides are readily reduced,then the colloidal metal may be obtained after the removal of waterand all electrolytes. The residue will dissolve in any lanolinsolvent, and on evaporation of this solvent solid sols with a high59 Zeitsch. Chem. Ind. Kolloide, 1912, 10, 275 ; A., ii, 752.Ibid., 77 ; A., ii, 449.61 Ibid., 265 ; A., ii, 772.62 ]bid., 1911, 9, 265, 270 ; A., ii, 157, 169.G3 Ibid., 1912, 11, 97, 100; A., ii, 1053, 1059INORGAWIC CHEMISTRY. 55percentage of the metal are obtained. The lanolin acts as a protec-tive colloid, and with the lanolin alcohols (cholesterol and isochole-sterol) sols can be obtained with an even higher metal content.Colloidal gold can be obtained in this way, hydrazine hydrate beingused as reducing agent.The final product is soluble in ether,light petroleum, fats, and liquid paraffin t o a solution having adeep blue colour, their solution in chloroform being reddish-violet.The addition of alcohol to the light petroleum solution precipitatesa pasty product containing 84 per cent. of gold. Such productswhen dried are granular and golden-brown in colour. They dissolveeasily in fats or liquid paraffin, bu6 sparingly in ether or lightpetroleum.Rare Earths.Pure metallic cerium has been prepared by Hirsch64 from thetrichloride which is dehydrated, and then electrolysed after fusionin an iron crucible. The cerium thus obtained contains about 2 percent.of impurities, f o r the most part iron and cerium oxide. I norder to obtain the pure metal the impure product is amalgamatedin boiling mercury, when the impurities may be skimmed off. Theamalgam is then distilled in a quartz vessel lined with magnesia.The pure metal has a density of 6-92 a t 2 5 O , and is almost as softand malleable as lead.Dafert and Miklauz65 have prepared the nitride and hydride ofcerium. When metallic cerium is heated in a current of hydrogenit begins to absorb the gas a t 310°, and the reaction is complete a t450°, cerium hydride being formed. This compound is dark blue orblack, and is spontaneously inflammable in air. I f moist air ispassed over the substance no ammonia is formed, but when thesubstance spontaneously burns in air some nitride is formed, whichgives ammonia with moisture.When the hydride is heated to from800° to 900° in nitrogen the pure nitride is formed. This compoundis not formed when cerium itself is heated in nitrogen; no ammoniais formed in the reaction, nor is there any ammonia formed whenthe nitride is heated in hydrogen. These authors were unable toobtain any evidence of the existence of imide compounds or of anycompound of cerium with nitrogen and hydrogen.Some doubt has been thrown by von Welsbach on the elementarycharacter of both terbium and thulium. Welsbach himself does notappear to have continued these investigations, nor do they seemto have been criticised or denied by later workers.It wouldcertainly seem that they merit further and independent investiga-64 Trans. Amer. Electrochcnt. Soc., 1911, 20, 57 ; A . , ii, 258.65 Monntsh., 1912, 33, 911 ; A., ii, 94256 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.tion. I n one paper,66 which deals with the fractional crystallisationof the double o x a l a k of the rare earths with ammonium, he saysthat the terbium is found distributed in the gadolinium anddysprosium fractions. The separation of the three elements istedious, but may be carried out satisfactorily. The first and mostinsoluble product contains the gadolinium, whilst the final onecontains the dysprosium. When optically tested these two fractionsare found to be free from terbium. The oxides are coloured deepcinnamon or ochre, and they have peroxidic properties.I n this wayit may be seen that terbium has been resolved into more than oneelement. Exner and Hnschek have shown that the spectrum ofterbium contains many lines which are common to gadolinium anddysprosium. Von Welsbach repeated these spectroscopic tests, usingmuch purer materials, and obtained the same results. The linescommon to the three spectra were found to be so strong in themiddle fraction as t o leave no doubt as to the presence of thesenew elements. He calls them TbI and TbIII, the former beingclosely related to gadolinium and the latter to dysprosium. Hepoints out that these two new elements, which are present in largeamounts in the gadolinium and dysprosium fractions of the terbiumseries, should easily be separated.Welsbach67 has drawn a somewhat similar conclusion as regardsthulium, which, from a spectroscopic point of view, he believesmust consist of a t least three elements.Meyer and Goldenbergm have carried out an investigation onthe properties of scandium oxide.They separate the thorium fromscandium by the addition of potassium iodate, which is added inslight excess so that some scandium iodate was precipitated.Different specimens of scandium oxide freed in this way fromthorium gave the atomic weight of scandium as 44-1-44.2. Thisagrees with Nilson’s value. Meyer, Winter, and Speter,69 on theother hand, using scandium which had been purified from thoriumby the sodium carbonate method, found the atomic weight to be 45.Meyer and Goldenberg investigated the arc spectrum of thescandium oxide fractions which had atomic weight 44 and 45 respec-tively, and they found that there was no difference between them.They conclude, therefore, that the method is not.delicate enoughfor the detection of 0.5 per cent. of thorium. The magnetisation-coefficient, however, affords a much more delicate test, as thescandium oxide (atomic weight = 45) gives a value of + 0.04 x 10-6,86 Chem. Zeit., 1911, 35, 658.67 Monatvh., 1911, 32, 373 ; A . , 1911, ii, 607.m Chem. News, 1912, 106, 13 ; A., ii, 768.69 Zeitoch. anorg. Chem., 1910, 67, 398 ; A., 1910, ii, 854lNORGANIC CHEMISTRY. 57whilst the scandium oxide (atomic weight = 44) gives a coefficient of-0-12 x 10-6.Scandium is diamagnetic, as also are lanthanumand yttrium, which also belong to group I11 in the periodic table.Some interesting experiments have been carried out by Joye andGarnier 70 on the hydrated oxides of neodymium. It is well knownthat’ neodymium hydroxide gives a different reflection spectrumfrom neodymium oxide. By heating the hydroxide t o various tem-peratures and observing the spectra, they have succeeded in estab-lishing the existence of two new hydrates, namely, 2Nd203,3H20and 2Nd203,2H20. The first of these is formed a t 310--325O, andthe second a t 525O.Urbain and Bourion 71 have succeeded in preparing europouschloride by reducing europic chloride in hydrogen a t 400--450°. Itis a colourless, amorphous compound, which is soluble in water to aneutral solution.A t looo the solution oxidises according to theequation :12EuC1, + 30, = 8EUC’13 + 2Eu20,.The europous chloride is more stable than the correspondingsamarium compound, which is the only other lower chloride knownin the group.Two papers have been published dealing with isomorphic series ofsalts of the rare earths. In the first of these, Billows 72 describes thepreparation of a complex molybdate having the composition(NH,)6R,Mo,,02,,24H,0, where R = Ce, La (Ce La), Nd, Pr, Sm.He found that all these crystallise in the triclinic system, and areperfectly isomorphous. I n the second paper, Jantsch 73 hasdescribed the double nitrates of formula M”’(NO3),M,”,24(H2O),where M’I’=La, Ce, Pr, Nd, Sa, Gd, and M/’=Mg, Zn, N, Co, Mn.The only exception is the double nitrate of gadolinium and man-ganese, which is too soluble t o crystallise well.All the other saltsare isomorphous, and crystallise well; all melt in their water ofcrystallisation a t definite temperatures, and these temperaturesfall as the elements are substituted in both series in the order inwhich they are given above. The solubilities follow the same order,the magnesium salts being the least soluble. The cerium doublesalts, however, are less soluble than those of 1ant.hanum. Themolecular volume curves show close parallelism in the case of themagnesium, manganese, nickel, cobalt, and zinc salts. The molecularvolume of the cobalt salts is greater than that of the nickel salts,and that of the neodymium salts greater than that of the praseo-7O Compt.rend., 1912, 154, 510 ; A., ii, 352.72 Zeitseh. Kryst. Min., 1912, 50, 500; A . , ii, 560.l3 Zcitsch. anorg. Chem., 1912, 76, 303 ; A., ii, 767.Ibid., 1911, 153, 1155 ; A., ii, 16258 ANNUAL REPOnTS ON THE PROGRESS OF CHEMISTRY.dymium salts, so that in both cases the order is the reverse of thatwhich would be expected from the periodic law.A considerable amount of work has been carried out on the rareearths by James and his c0lleagues.7~ I n one paper are describedthe preparation and properties of the salts of samarium and neo-dymium of a very large number of organic acids, mainly sulphonicacids of the aliphatic series. This formed part of a search forsatisfactory salts for the purposes of fractionation.In two otherpapers,75~ 76 the application of sebacic acid is described for the quan-titative precipitation of thorium and yttrium. In the case ofthorium an aqueous solution of the acid is added to a neutral solu-tion of the thorium salt., when thorium sebacate is precipitated.This method can be used for the quantitative separation of thoriumin the presence of cerium, lanthanum, praseodymium, neodymium,samarium, and gadolinium. It is also of value for the separationand purification of thorium. Pyrotartaric acid may also be used,and when added to a cold neutral solution of the thorium salt noprecipitate is obtained, but, on boiling, the thorium pyrotartrate isquantitatively precipitated. In the case of yttrium the elementmay be quantitatively separated from certain other elements asfollows: A single precipitation with ammonium sebacate gives aquantitative separation of yttrium from sodium, whilst a doubleprecipitation gives a precipitate completely free from potassium.On the other hand, oxalic acid gives the most satisfactory separa-tion of yttrium from iron, aluminium, lithium, and magnesium.Finally, James77 has described a complete scheme for the separa-tion of the rare earths from one another.This scheme is, of course,far to9 long to be detailed in this report, but no account of theprogress of work done in this field of research would be completewithout including, a t any rate, this brief mention.Group I .Ermen78 has investigated certain basic salts of copper, and hefinds that, with N / 10-solutions, the addition of sodium hydroxideto copper sulphab until all the copper is precipitated, or the addi-tion of copper sulphate to sodium hydroxide until the copper justappears in the solution, two molecules of copper sulphate and threemolecules of sodium hydroxide always react together.The bluebasic sulphate, CuS04,3Cu(OH),, is precipitated, the same com-74 James, Hoben and Robinson, J. Amr. Chem. SOC., 1912, 34, 276, 281 ; A.,i, 233.75 Smith and James, ibid., 281 ; A., ii, 390.78 Whitternore and James, ibid., 772 ; A., ii, 690.77 Ib'bid., 757 ; A . , ii, 690.76 J. SOC. Chem. I n d . , 1912, 31, 312 ; A . , ii, 453INORGANIC CHEMISTRY. 59pound being obtained if the sodium hydroxide solution is added tothe solution of copper sulphate when boiling. On the other hand,the addition of copper sulphate t o a boiling hydroxide solutionprecipitates the black oxide of copper, which is not completely con-verted into the basic salt even on prolonged boiling.Evidencewas also obtained of the existence of a green basic salt containing3.5 of copper and 1 of SO,. On mixing cold dilute solutions ofcopper sulphate and sodium carbonate, one molecule of each saltreacts together, and the blue basic carbonate, CuC'O,,Cu(OH),,H,O,is precipitated. On keeping, this compound changes into malachite,CUCO,,CU(OH)~. From this substance sodium hydroxide removesall the carbon dioxide. When sodium carbonate solution is runinto a boiling solution of copper sulphate, the blue compound,CUSO,,~CU(OH)~, is also obtained. When copper sulphate isadded t o aqueous ammonia until a permanent precipitate is pro-duced, one molecule of copper sulphate is found t o have reactedwith .four molecules of ammonia.The addition of more coppersulphate until the solution becomes colourless gives the compoundCuS04,2Cu(OH),. If an aqueous solution of copper sulphate issaturated with ammonia gas a precipitate is obtained having thecomposition CuS0,,4NH3,H,0. This substance is also formed bytriturating copper sulphate crystals with strong aqueous ammonia,or by adding anhydrous copper sulphate to strong ammonia SO~U-tion. It is also obtained by exposing the compound CuS04,5NH,to moist air.Gumming and Gemme1179 have investigated the trihydrate ofcopper nitrate.It is best prepared by dissolving cupric oxide inconcentrated nitric acid. The salt may be recrystallised from hotconcentrated nitric acid, and may be dried on a porous tile overpotassium hydroxide. If the nitric acid is obtained by distillationfrom sulphuric acid in a current of carbon dioxide, the cupric oxidereacts with i t a t a temperature of 70-75O. Nitrogen peroxide andoxygen are evolved, the process taking place according to theequation :CUO + 6HN03= C'U(NO~)~,~H,O + 4N02 + 0,.The hexahydrate of copper nitrate when dried in a vacuum overphosphoric oxide readily loses three molecules of water, giving thetrihydrate. It would appear that the only basic nitrate of copperis Cu(N03),,3Cu(OH),, and this may readily be prepared by heatingthe trihydrate to looo or by boiling a concentrated aqueous solutionof copper nitrate with cupric oxide. The transition temperaturebetween the two hydrates of copper nitrate has been found to be24'66O with the dilatometer and 24-65O thermometrically.59 PTOC.Eoy. Sac. Edin., 1912, 32, 4 ; A., ii, 55660 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY,BaubignygO finds that on heating copper sulphate solution withan excess of sodium sulphite solution so that the precipitate firstformed is redissolved, and if the resulting colourless solution isheated on a water-bath, a red precipitate is formed, which is adouble compound of sodium sulphite and cuprous sulphite, whilstthe solution contains sodium dithionate. The process takes placeaccording to the equation :2CuS0, + 4Na2S03 = 2Na,SO, + CI.$~O,,N~&~O, + Na.$3,06.If aqueous solutions of copper sulphate and sodium sulphite aremixed in a sealed tube a t the ordinary temperature, then after somehours colourless crystals are formed having the constitutionCu,S03,Na2S0,,1 2H,O.Some interesting work has been carried out by D’Ans andFriederich 81 on the preparation of derivatives of hydrogen per-oxide, in which the hydrogen atoms have been partly replaced bysodium or potassium. An ethereal solution of pure hydrogen per-oxide reacts with metallic sodium to give a white solid havifig theformula 2NaH0,,H,02.This compound is identical with that p r epared from the action of hydrogen peroxide on sodium ethoxide.82Metallic potassium acts more vigorously than sodium, and gives asubstance having the formula 2KH0,,3H20,.When an absolutealcoholic solution of potassium hydroxide and hydrogen peroxideare mixed they then give 2KH0,,H20,, a compound which haspreviously been obtained by Schone.83The phosphides of czsium, rubidium, potassium, and sodiumhaving the general formula M,P, have been studied by Hackspilland Bos~uet.*~ Two or three grams of the metal and a largeglobule of phosphorus are distilled successively into an exhaustedtube, and the mixture heated to 400--5OOO. The black mass thusobtained still contains free metal, but after continuing the heatingfor about one hundred and fifty hours this is volatilised, and thephosphide appears reddish-brown. A t temperatures between 00and looo the four compounds resemble cadmium sulphide in colour,they become darker at higher temperatures, and are almost colour-less a t the temperature of boiling nitrogen.They melt at 650°with loss of phosphorus. They decompose in air, and when treatedwith water give a solid hydride of phosphorus with a small amountof hydrogen phosphide and hydrogen.Barhieri s5 claims to have prepared a compound containinggo Cowpt. rend., 1912, 154, 434, 701 ; A., ii, 351, 447.g1 Zeitsch. unorg. Chem., 1912, 73, 325 ; A., ii, 151.d2 \Volffenstein and Peltner, Ber., 1908, 41, 280 ; A., 1908, ii, 180.83 Airnalen, 1878, 193, 276, 289 ; A., 1878, 931.a1 Compt. rend., 1912, 154, 209 ; A., ii, 252.Atti R.Acead. Lincei, 1912, [v], 21, i, 560; A., ii, 763INORGANIC CHEMISTRY. 61bivalent silver. If a solution of silver nitrate in pyridine is addedt o a cold solution of potassium persulphate, a yellow precipitate isobtained, with tliu formula AgS20,,4C,N’H,, and the substanceresembles the corresponding cupric compound. I f very dilutesolutions are used the silver and copper compounds can be precipi-tated simultaneously, and, by varying the proportions of the silverand copper, mixed crystals are formed, showing complete grada-tions of colour between the orange silver compound and the violetcupric compound, thus proving the complete isomorphism of thetwo.Group ZZ.Barre 86 has prepared the double carbonates of calcium withsodium and potassium.If precipitated calcium carbonate is boiledwith a concentrated solution of sodium carbonate, orthorhombiccrystals are formed of Na2C0,,CaC0,,2H,0. A t 9 8 O the compoundis not formed unless the solution contains a t least 21.06 per cent.of sodium carbonate. Under the same conditions potassiumcarbonate gives prismatic needles of I.,CO,,CaCO,, which arereadily hydrolysed by water, and are only stable in a solutioncontaining a t least 59-25 per cent. of potassium carbonate.Niederstadt87 has shown that aragonite and calcite may bedistinguished in that thc former is readily coloured lilac by cobaltsalts, whilst the latter only assumes a bright blue colour on pro-longed boiling. Aragonite is the more active towards manganese,zinc, and iron salts, whilst calcite more readily precipitates copper,lead, and silver.Group ZZI.An important paper has been published by Travers and Rayeson borohydrates.Magnesium powder is heated with boron trioxideto a bright red-heat in a current of hydrogen. When the resultingpowder is treated with water a yellow solution is obtained, whichhas powerful reducing properties; it is slightly alkaline to litmus,and on boiling it gives off hydrogen. The addition of dilute acidcauses an evolution of hydropn, together with a trace of someboron compound. The original solution both before and after theaddition of acid is strongly reducing. It takes up iodine, andprecipitates the heavy metals from solutions of their salts. Theauthors find that the amount of hydrogen evolved on the additionof acid and the iodine equivalent are in the ratio of 1 to 1 o r 1 to 2,86 Conzpt.rend., 1912, 154, 279 ; A . , ii, 254.137 Zeeitsch. angew. Chem., 1912, 25, 1219 ; A., ii, 760.8s Pt-oc. Roy. .voc., 1912, A, 87, 163 ; A . , ii, 93862 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.and analyses show figures that are very near to one or the other ofthese ratios. Furthermore, the solutions contain one atom of boronfor every atom of hydrogen that is evolved when the solution istreated with acid. When the aqueous solutions are evaporated ina vacuum a t low temperatures with the help of a condenser cooledin liquid air, little or no gas is evolved until all the water has beendistilled off. A viscous residue is left, which now proceeds to giveoff hydrogen rapidly, and then sets t o a semi-crystalline or glassysolid.The amount of hydrogen evolved a t this stage is equal tohalf the volume of hydrogen which is given off when the originalsolution is acidified. If the bulb containing the glassy solid is leftconnected to the pump, a further quantity of hydrogen is slowlyevolved, and the amount of this second quantity of hydrogen isequal to the first quantity. On gently heating the substance athird quantity of hydrogen is evolved, which again is about equalto the two amounts previously evolved. It is found that half theboron distils over during the evaporation, and as the distillate doesnot take up much iodine the authors conclude that this volatileboron compound is oxidised.They carried out certain experimentswhich proved that this volatile boron compound is not boric acid,and put forward the view that a polymeride of boron trioxideexists which is not acid, and is volatile; they suggest that theoriginal substances are hydrated derivatives of B402 and B,O,.A second important paper has appeared during the year, namely,one dealing with the hydrides of boron, by Stock and Massenez.89These authors prepared the hydrides from magnesium boride, andsome difficulty was experienced in the preparation of this compoundso as to obtain it good yield of the boron hydrides. The best resultis obtained as follows: one part of very finely powdered borontrioside is intimately mixed with three parts of magnesium powder,and 10 grams of the mixture are rapidly heated in a thin ironcrucible in a stream of hydrogen.The powdered product is slowlydropped into not too dilute hydrochloric acid a t 50-80". The gasevolved is fractionated by means of liquid air. Two hydrides wereisolated, namely, B4HIo and B6H12. The former was analysed bytreatment with water, when the following reaction takes place :B4HI0+ 12H20=4B(OH),+ 11H2.The results of the analysis and the density determinations agreedwith the formula B4H,,. The compound B6H1, was analysed in asomewhat similar way, and its density was also determined, butowing to the small quantities available there is still some doubtas to the number of hydrogen atoms in the molecule.The hydride B4H10 melts a t - 1 1 2 O , and boils between 1 6 O andx9 Ber., 1912, 45, 3539; A ., 1913, ii, 44INORGANIC CHEMISTRY. 6317O/760 mm. The hydride B,H,, should boil a t about looo, butboth compounds are unstable, and tend to decompose a t theordinary temperature, and give a series of further hydrides yetto be investigated. The hydride B,HI, is rapidly absorbed byalkali, and the solution slowly evolves hydrogen, the completereaction being similar to that with water mentioned above. Thehydride B6HI2, on the other hand, gives up hydrogen immediatelywhen treated with alkali.JohnsonQO has found that the alumina formed by dehydratingthe hydroxide a t low temperatures is an excellent drying agent forgases such as hydrogen phosphide, hydrogen bromide, and hydrogeniodide, where phosphoric oxide is useless.It is better than calciumbromide, zinc bromide, or zinc chloride, and is more efficient thansulphuric asid. A tube filled with this alumina can be usedindefinitely, as the substance can readily be dried again by heatingwith a smoky flama in a stream of air dried by sulphuric acid.Terni91 has prepared a peroxide of aluminium by adding anexcess of 30 per cent. solution of hydrogen peroxide to aluminiumhydroxide dissolved in the minimum quantity of 50 per cent.solution of potassium hydroxide. An amorphous, white powder isformed, having a constitution AI,O,,Al,O,,lOH,O. It is a trueperoxide, and Terni believes that the first product of the reaction isAIBO,, which is changed by hydrolysis t o the above substance.Bosshard and Zwicky 92 have investigated the perborates.Twentyof these compounds were heated t o a temperature of 55--65O, eitherunder diminished pressure or in a current of air free from carbondioxide. In only one case was a trace of hydrogen peroxide foundin the distillate. Usually the water of crystallisation distilled over,and the percentage of active oxygen in the salt was increased.Potassium peroxide perborate when dissolved in water evolves onlytraces of oxygen, and when dried in a vacuum over phosphoricoxide four molecules of the salt lose one molecule of water, so it isimprobable that the salt contains hydrogen peroxide of crystallisa-tion. Further, it is found that sodium perborate evolves oxygenwhen treated with colloidal platinum solution more slowly thanhydrogen peroxide solution under the same conditions.Theauthors conclude, therefore, that the perborates do not containhydrogen peroxide of crystallisation, and that they have theconstitution :0M*O*B<A and not M*O*O*B:O.J. Amcr. Chem. Soc., 1912, 34, 911 ; A . , ii, 847.y1 Atti It. Accad. Lincei, 1912, [vJ, 21, ii, 104 ; A., ii, 944.92 Zeilsch. nngew. Chem., 1912, 25, 993; A . , ii, 64064 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.According t o this the potassium salt mentioned above would have00the coiistitutiori K*O*O*B<I, and not K*O.O*O*B:O, as, owing tothe three oxygen atoms existing in the chain, a compound of thelatter constitution would probably be very unstable. They pointoat that Riesenfeld’s test (see under Group IV) fails with per-borates.Group IV.A certain amount of work on silicon hydride has to be reported.Adwentowski and Drozdowski 93 state that pure silicon hydride,SiH,, can only be prepared by repeated fractionation of the gasusually obtained.By the action of hydrogen chloride on mag-nesium silicide a gas is produced from which liquid fractions areobtained, which are mixtures of several silicon hydrides. The pureSiH, is spontaneously inflammable in plenty of air. The weightof one litre is 1.4538 grams. It boils a t -116°/740 mm. Itscritical temperature is - 3 * 5 O , and its critical pressure is 47-8atmospheres.Besson 94 describes certain reactions of silicon tetrahydride.Powdered magnesium with half its weight of powdered quartz isstamped into a crucible and fired by the ignition of some powderedmagnesium placed on the top of the mixture., The magnesiumsilicide obtained when treated with hydrogen chloride gives a gascontaining 6-7 per cent.of SiH,. A gas containing less than0.5 per cent. is not spontaneously inflammable, but fumes in moistair, giving a colourless solid, which varies in composition betweenH,Si,O, and H,Si,O,. In his second paper Besson points out thatthe gas from the magnesium silicide contains hydrogen phosphideand hydrogen sulphide as impurities. These impurities are readilyremoved by passing the gas through a dilute solution of iodine inpotassium iodide or hydriodic acid, and the gas may then be driedwith phosphoric oxide.During the preparation silico-oxalic acid,HzSi20,, separates as a white solid in the reaction mixture, theamount of this being from 25 to 30 per cent, of the weight of themagnesium used. The dry gas is not acted on by light, but underthe influence of the electric discharge it gives a yellow to brownsolid deposit of uncertain composition, which is spontaneouslyinflammable in air. The gas has no action on concentratedammonia solution or liquid ammonia when dry. In the presenceof moisture it gives with liquid ammonia a white solid, which whendried a t looo has the composition H,Si,O,. I f the white solid93 Bt611. Acad. Ski. Cracow, 1911, A, 330 ; A., ii, 44.94 Compt. .rend., 1912, 154, 116, 1603; A., ii, 256, 641INORGANIC CHEMISTRY.65obtained when the gas fumes in moist air is dried a t looo in avacuum for several hours, it has the composition H3Si305. Whencarefully heated in a vacuum it loses hydrogen, and the compositionof the residue approximates t o Si30,.Riesenfeld and Mau 35 discuss the constitution of the percarbon-ates, and they differentiate between t.he true percarbonate and theordinary carbonate containing hydrogen peroxide of crystallisation.Whilst the former give a quantitative separation of iodine fromneutral potassium iodide solution, the latter liberate practically noiodine.96 Even the addition of hydrogen peroxide to a solution ofpotassium percarbonate in the proportion of 2 molecules to 1 causesvery little change in the quantity of iodine liberated.A solutionof potassium percarbonate a t 15O, after keeping for an hour, stillliberates iodine quantitatively from potassium iodide solution,whilst the so-called sodium percarbonate does not liberate iodineeither a t 1 5 O or Oo or even when added as a solid. The authorsconclude that the latter possesses the constitutionSimilarly, Peltner's 97 rubidium percarbonate must have the formulaRb2C0,,3H,02. They also point out that persulphates andsulphates with hydrogen peroxide of crystallisation 96 react towardspotassium iodide in a similar way to the percarbonates. In theiropinion, therefore, this reaction may be considered as a generalone for the differentiation between true per-salts and salts contain-ing hydrogen peroxide of crystallisation.It may be pointed out,however, that the test fails in the case of the perborates (see underGroup 111). Based on the experience gained from these results,Riesenfeld and Mau have investiga€ed the salts formed by theaction of carbon dioxide on sodium peroxide. The four saltsprepared by Wolffenstein and Peltner 99 are found by Riesenfeldand Mau to be really salts of two different acids, namely, monoper-oxycarbonic, H,CO,, and monoperoxydicarbonic acid, H2C,0,, andto have the constitution Na2G04,Na2C04,H202,Na&206 andThey prove this conclusion by comparing the reactions of the foursalts with potassium iodide. They also found that the two pairsof salts were mutually convertible by the addition or withdrawal ofhydrogen peroxide. It was observed that the potassium percarbon-ate, &C206, prepared by electrolytic methods, behaves differently96 Ber., 1911, M, 3589, 3595 ; A., ii, 156.96 Compare Reisenfeld and Reinhold, ibid., 1909, 42, 4377 ; Tmatar, ibid., 1910,97 Ibid., 1909, 42, 1777; A., 1909, ii, 574.ys Willstatter, ibid., 1903, 36, 1828 ; A., 1903, ii, 537.Ber., 1908, 41, 280; A., 1908, ii, 180.Na2C03>q02,&H20 *NazC2Ot3,H%O2.43, 127, 2149 ; Reisenfeld, ibid., 566, 2594 ; A., 1910, ii, 33, 203, 290, 774, 952.REP.-VOL.IX. 66 ANNUAL REPOlllv UN THE PROGRESS OF CHEMISTRY.towards potassium iodide from the corresponding sodium com-pounds, for the potassium a salt gives a quantitative liberation ofiodine, whereas in the case of the sodium salt only about 50 percent, of its active oxygen liberates iodine.By the action of carbondioxide on potassium peroxide a compound, K2C206, was formedwhich behaves towards potassium iodide solution similarly to thesodium compound. This compound, therefore, must be isomericwith ordinary potassium percarbonate, and they suggest that itsconstitution may be represented by K*O*O*CO*O*CO*O~K, asdistinct from KO*CO*O*O*CO*OK.Stock and Praetorius 1 have contributed considerably t o our know-ledge of carbon subsulphide, C,%, first prepared by von LengyeI.2A modificatiori of von Lengyel's method was used in its preparation.An electric arc is maintained with 10 to 15 amperes a t 110 voltsunder pure carbon disulphide between a graphite cylinder ascathode and an anode of graphite and antimony.Finely powcteredantimony and graphite were mixed with sugar solution, and pressedinto rods, which after being dried were heated to carbonise thesugar, A slow stream of carbon dioxide was passed thrpugh theapparatus, and the resulting fluid was shaken with mercury andsome phosphoric oxide to remove the sulphur. When the carbondisulphide had been distilled off, the subsulphide was sublimed in ahigh vacuum using a condenser cooled with liquid ammonia. A tthe ordinary temperature C3S2 forms a clear, red, strongly refractiveliquid, and in dilute carbon disulphide solution it exhibits anabsorption band between 530 and 455 p p . Under the influence oflight it polymerises finally into a black substance, which is unaf-fected by water, hydrochloric acid, sodium hydroxide solution, orchlorine water.The subsulphide reacts a t once with aniline to givethiomalonamide :C3S2 + 2C6H5*NH2 = C'H2(C'S*NH*C6H5),.It is clearly the analogue of carbou suboxide, C,O,.Group V .During the year Franklin has published four papers describingthe preparation and properties of certain amine compounds, thuscontinuing his work based on his ammonia system of acids, bases,and salts. I n a series of papers3 in previous years he has pointedout the analogy between liquid ammonia and water as electrolyticBET., 1912, 45, 3568; A, 1913, ii, 46.Ibid., 1893, 26, 2690 ; A., 1894. ii, 90.3 Frsnkliii and Kraus, Arne?.. Chem. J., 1899, 21, 1, 8; 1900, 23, 277 ;Franklin and Stafford, ibid., 1902, 28, 83 ; Franklin, J.Arne?.. CJienz. Soc., 1905,27, 820 ; A., 1899, ii, 284; 1900, ii, 382 ; 1902, i, 748 ; 1905, ii, 581INORGANIC CHEMISTRY. 67solvents, and the analogy between the reactions which take placein the two solvents. He formulated a system of acids, bases, andsalts on the basis of ammonia as the typical substance, analogous tothe ordinary system of oxygen acids, bases, and salts as derivativesof water. During the present year the system has been furtherdeveloped, and the relationships have been shown to be fully justi-fied. I n the first paper 4 he describes the preparztion of potassiumammoniocadmiate, Cd(NHK),,2NH3, which is obtained by addingpotassamide to cadmium iodide or nitrate dissolved in liquidammonia. This substance corresponds with the zinc compounddescribed by Bitzgerald.6I n a second paper6 Franklin and Hine describe the reactionbetween titanium nitride bromide and potassamide in liquidammonia, which, although the analyses show that absolutely pureproducts were not obtained, would certainly seem to take placeaccording to the equation :NiTi*Br + 2KNH2=NiTi*NHK +I(&+ NH,.In a third paper 7 Franklin describes the action of potassamideon a liquid ammonia solution of Cu(NO3),,4NH,. These compoundsreact according to the equation:3Cu(N03),+ 6KNH,= @u3N,nNH3+ 6KN0, + (4 - n)NH, + N.The copper compound is precipitated, and when warmed to thelaboratory temperature in a vacuum it is converted into cuprousimide, @u2NH, and when heated to 160°, cuprous nitride, Cu,N, isformed.The compounds with the general formula Cu3N,nNH3readily dissolve in liquid ammonia solution of potassamide, and inthis way well-crystallised specimens of a colourless compound havingthe formula CuNK2,3NH3 were obtained. This potassium ammonio-cuprite with three molecules of ammonia of crystallisation loses onemolecule of ammonia in a vacuum at the ordinary temperature, andat higher temperatures it loses a further molecule, formingG'uNK,,NH,.In a fourth paper8 Franklin describes the preparation of thecorresponding thallium compounds. When a liquid ammoniasolution of potassamide is added to a liquid ammonia solution of athallous salt, a black precipitate of thallium nitride is formed ;thus, with thallous nitrate the reaction takes place according tothe equation:3TlN0, + 3KNH, = T13N + 3KN03 + 2NH3.J.Amer. Chem. Soc., 1907, 29, 656; A,, 1907, ii, 546.4 Amer. Chem. J., 1912, 47, 285 ; A., ii, 451.6 IW., 1912, 341, 1497; A., ii, 1168.7 &id., 1501 ; A, ii, 1174.ti J. Physical Chew., 1912, 16, 682 ; A . , 1913, ii, 52.F 68 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.This precipitate readily dissolves in a liquid ammonia solution ofaiiimonium nitrate or potassamide, and so must be assumed to beamphoteric in character. When it is dissolved in a liquid ammoniasolution of potassamide, a yellow solution is formed, from whichyellow crystals of the compound, T1NK2,4NH,, may be obtained.This potassium ammoniothallate readily loses portions of itsammonia of crystallisation, giving the compounds TlNK2,2NH, andTlNK,,l&NH,.When excess of potassamide is used, well-crystallisedproducts are obtained, which are impossible of formulation asdefinite chemical compounds. Franklin looks upon these as iso-morphous mixtures of the compound TlNK2,4NH, with potass-amide, or perhaps rather an isomorphous mixture of the unknownthallous amide and potassamide. Considerable difficulty was foundin the work owing to the great instability of these compounds.They explode violently when heated or subjected to shock, andtheir satisfactory analysis was a matter of considerable difficulty.Dafert and Miklauz9 have made investigations on the amide,imide, and nitride of lithium. Under the influence of sunlightlithiuinimide decomposes with the production of a dark red colour,the reaction being represented by the equation:2Li2NH = Li3N + LiNH,.They find that by the action of nitrogen on lithium hydride,hydrogen on lithium nitride, a mixture of nitrogen and hydrogenor of ammonia on lithium hydride or nitride, lithium-imide, -amide,or trilithiumamide, or mixtures of these compounds are formedaccording to the temperature and other conditions. Lithiumamideis formed by the action of ammonia on amorphous lithium nitriciea t 130-350°, or on crystalline lithium nitride a t 410430°, or onlithium hydride a t 440-460°, thus:Li,N + 2NH, = 3LiNH, or LiH + NH,= LiNH, + El,.At 450° hydrogen reacts with lithiumimide according to theequation :2H2 + 3Li2NH= 2Li,NH, + NH,,The trilithiumamide when heated to 600° in a stream of nitrogenreacts according to the equation:4Li,NH2 + N2 = 6Li,NH + H2.Ruff and Martin10 have prepared metallic vanadium in a highstate of purity. A mixture of vanadium trioxide and carbon aremoulded with starch into rods, and these are sintered in an electricfurnace a t 1750°, and finally fused in the arc.The product contains95 to 97 per cent. of vanadium. Vanadium carbide can be preparedHonatsh., 1912, 33, 63 ; A., ii, 253.lo Zeitsclb. aizoyy. Chm., 1912, 25, 49 ; A., ii, 16G1 NOEGAN IC CHERIISTKY. 69in the resistance fnrnace. It is silvery-white, highly crystallino,and very hard. It melts a t 2750O. This carbide if powdered andmixed with the trioxide and a little starch solution, and thenheated to 1950O in a zirconia crucible, gives the fused metal, whichon analysis is found to contain 98 per cent.of vanadium and1-91 per cent. of carbon. The authors find that the pure metalmelts at 1 7 1 5 O .Renschler has shown that ammonium metavanadate can readilybe reduced electrolytically to tervalent vanadium salts. Thecathode is placed in a porous cell; both the anode and cathode arelead cylinders, and dip in 50 per cent. sulphuric acid. Ammoniumvanadate is added to the cathode cell, and the liquid well stirred,and a current of 4 amperes passed through until the solutionbecomes green. It is then removed and kept, when ammoniumvanadium alum separates in good yield.Group VI.Von Pirani and Meyer12 have determined tlie melting points oftungsten and molybdenum, finding that tungsten melts a t3100O +_ 60° and molybdenum a t 2450O +_ 30°.Ruder13 has studied the action of acids and alkalis on wrongbttungsten and molybdenum.He finds that both metals are some-what resistant t o acids owing to the formation of a coating of oxide.Tungsten is most rapidly attacked by fuming sulphuric acid, butonly 1.2 per cent. dissolves in eight hours. Molybdenum is moreeasily attacked than tungsten, but it is fairly resistive to concen-trated hydrochloric acid below 125O, and is not affected by hydro-fluoric acid. The metals are not attacked by aqueous potassiumhydroxide, but are slowly dissolved by fused alkali.Sanger and Riegel14 have succeeded in preparing pyrosulphurylchloride and chlorosulphonic acid in the pure state.I n the pre-paration of the former a mixture of fuming sulphuric acid andsulphur trioxide is added gradually t o an excess of carbon tetra-chloride in a heated flask, when the following reaction takes place :but some chlorosulphonic acid is also formed a t the same time.The liquid is purified by repeated fractional distillation off sodiumchloride.The pure pyrosulphuryl chloride boils a t 152.5-153O/760 mm.2s0, + CCl, = coc1, + S,O,Cl,,2 c i M i . Elcktrochem., 1912, 18, 137 ; A!., ii, 356.Her., 1912, 14, 426 ; d., ii, 560.l3 J. A?ner. Chem. Xoc., 1912, 34, 387 ; A., ii, 454.l.4 Proc. dmer. Acad., 1912, 47, 673 ; Zcitsch. nnorg. CJLCIL, 1912, 76, 79 ; / I . ,ii, 75270 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Chlorosulphonic acid is obtained by passing dry hydrogen chlorideinto a mixture of sulphuric acid and sulphur trioxide, the mixtureis distilled in a stream of hydrogen chloride, and the fractionpassing over between 145O and 160° is collected.It is purified byfractional distillation a t low temperatures. It boils a t 151--152O/765 mm. and a t 74--75O/19 mm. The two compounds may readilybe distinguished, as powdered selenium or tellurium give colourswith chlorosulphonic acid and not with pyrosulphuryl chloride.Group VZZ.A further investigation into the reactions of bleaching powderhas been carried out by Taylor and Bostock.16 When bleachingpowder and thirty times its weight of water is distilled withsulphuric acid, hydrochloric acid, or nitric acid in quantity slightlygreater than is necessary t o neutralise the free lime present, hypo-chlorous acid, and a small quantity of chlorine are evolved.Whenthe quantity of acid is sufficient t o neutralise all the free lime andto decompose all the hypochlorite theoretically present, the propor-tion of chlorine evolved is considerably greater. With more acidthe quantity of chlorine is still further increased until finallynothing but chlorine is evolved. There is no marked differencebetween the results obtained with each of the three acids present.Acetic and phosphoric acids are very much alike in their action,but they differ considerably from the mineral acids in theirbehaviour to bleaching powder.Even with large amounts of theseacids the proportion of hypochlorous acid does not fall much below50 per cent. When bleaching powder is distilled with boric acidand sufficient water, almost pure hypochlorous acid is produced.There is very little difference, even if three parts of boric acid areemployed to one of bleaching powder. The aut.hors suggest thisas a very convenient method for the preparation of aqueous hypo-chlorous acid.If carbon dioxide is bubbled through a mixture of bleachingpowder and water at the ordinary temperature nothing but chlorineis evolved; when, however, the solution is warm, hypochlorous acidcomes off in an amount increasing with the temperature, untilwhen the mixture is boiled hypochlorous acid is evolved practicallyfree from chlorine.Higgins16 has contributed a further important paper on theprocess of bleaching, and he finds that the action is due to bothhypochlorous acid and hypochlorites. Since the former is lessstable than the latter i t possesses a greater bleaching effect.Henceany agent which by hydrolysis increases the amount of the freeacid in solution tends to increase the efficiency of that solution.l5 T., 1912, 101, 444. Ibid., 222INORGANIC CIIEMISTKY. 7 1Willard 17 has prepared pure perchloric acid dihydrate,HC10,,2H2O, from ammonium perchlorate by oxidation of theammonia with aqua regia. The acid has the advantage of beingvery cheaply prepared in the pure state. It is not poisonous orexplosive. As a rule it is non-oxidising except a t the boiling point(203O). It will completely displace hydrofluoric, nitric, hydro-chloric, or any other volatile acid, owing t o its high boiling point.I t s salts are soluble in water, alcohol, or acetone, and are verysuitable for electrochemical work as they are not reduced. Itsprincipal use is for the separation of sodium and potassium.Furthermore, it serves as an excellent standard solution in acidi-metry, and it may be used in place of sulphuric acid in potassiumpermanganate titratioas.Group VIZI.Hantzsch 18 has put forward a suggestion regarding the constitu-tion of the blue and red cobalt hydroxides. It is generally thoughtthat the blu0 compound is a basic salt, for when it is precipitatedfrom the sulphate by alkali and washed with cold water until nomore sulphuric acid comes out it still contains SO,.Hantzsch states that this may be removed by repeated boilingwith air-free water, without the d o u r being impaired. The redhydroxide is obtained by precipitating with excess of alkali, andwashing the precipitate with hot water in an atmosphere ofhydrogen, and giving it a final washing with alcohol and ether.It still retains some water after prolonged heating in an atmo-sphere of nitrogen a t 300°, whilst the blue compound is completelydehydrated a t 170O. Acetyl chloride and benzoyl chloride reactmore vigorously with the red compound than with the blue.Hantzsch calls this a case of chromoisomerism, and he suggeststhe formula Co(OH), and COO . . . H,O for the red and bluecompounds respectively.Recoura l9 discusses the constitution of ferric sulphate trihydrateand ferric fluoride hexahydrate. I f ferric sulphate hexahydrateis dehydrated a t 108O i t loses three molecules of water, and isconverted into the yellow Fe2(S0,),,3H,0. This substance whendissolved in alcohol gives no precipitate with barium chloride.The alcoholate, Fe,(S04)3,2H,0,2C,H,0, previously described,20I7 J. Aqncr. Chem. Xoc., 1912, 34, 1480 ; A., ii, 1163.Zcitsch.. nnorg. Chcnt., 1912, 73, 304 ; A., ii, 166.l9 Compt. rend., 1911, 153, 1223 ; 1912, 154, 655 ; A , , ii, 165, 353 ; BuU. SOC.20 Ann. Chim. Phys., 1907, [viii], 9, 263 ; Compt. rend., 1907, 144, 1427 ; A . ,chim., 1912, [iv], 11, 370, 941.1907, ii, 552, 69372 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.loses half of its water a t 1 0 5 O without changing its appearance,whilst the remainder is lost together with the alcohol at 115O, thematerial becoming txmporarily black. The alcohol molecules do notappear to be attached to the sulphuric acid as they are removedon solution in water, unlike the case of ethyl ferrisulphate. Recourasuggests for the trihydrate the formula Fe,( S0,)3(0H),, similar tothat of the green pentahydrate of chromic sulphate. He considersthat the formula of ferric fluoride hexahydrate should be writtenFe2F,(OH),,(HF),,4H20, and for this he advances the followingreasons. Only a third of the iron reacts with barium fluoride,although the phenomenon is complicated by the precipitation ofthe double salt, Fe,F6,3BaF,. When the solid salt is boiled withalcohol only one-third of the fluorine is rapidly eliminated ashydrogen fluoride, the remainder being lo& very slowly. Whenheated to 9 5 O the salt loses water rapidly and proportionally to theloss of fluorine, until one-third of the latter has been eliminated,after which no more water is lost.E. C. C . BALY