INORGANIC CHEMISTRY.THE work done during the year in this department of chemistry is notdissimilar to that recorded in previous years; it reveals evidence ofactivity in a great number of directions, and from the circumstances ofthe case places on record the results of the examination of a greatvariety of materials. The revision of atomic weights and the refiningof earlier observatioiis continue still to attract workers prepared t ouse to the utmost advantage the facilities in the niariipulation of gases,liquids, and solids, which are the accumulated heritage from theresearches of the past century. The study of the rare elements, moreespecially of the rare earths, has contributed its quota t,o the year’swork, still to leave the position of some elements in doubt and dispute.The construction of fusion diagrams largely employed in the investiga-tion of alloys has been applied to the examination of the possibilitiesOF combination in the case of other solid elementary bodies.The controversy as to the cause of the rusting of iron would appear to beclosed by the investigations of Moody, who shows that carbon dioxideplays the important part in this chemical change. It is permissible toregard the discovery of a new gaseous oxide of carbon, referred to in alater section of this Report, as an addition to the facts of inorganicchemistry. The complex metalammonia compounds and the isomerismexhibited by these substances have, a t the hands of Werner and others,received further extension and elucidation, and formed the subject ofa special lecture given by Werner to the Berlin Chemical Society.Revision of Atomic Weigl~ts.Gray and Gnye2 have, from a discussion of the results of thedifferent determinations of the atomic weight of nitrogen, inde-pendently concluded that 14.010 may be accepted as the atomic weightof this element.Ann.Report, 1906, 35 ; also Trans., 1906, 89, 1173.BcT., 1906, 39, 1470, aiid A Y C ~ . Sci. ~ J L ~ J s . nut., 1905, [iv], 20, 351, aiid GupeiIIld I ) u ~ c l l a , Ci)l/Zpt. ?*end., 1905, 141, 826INORGANIC CHEMISTRT. 31Guye and Ter-Gazarian 1 have drawn attention t o a source of errorin fixing the atomic weight of silver in the potassium chloridewhich is always t o be found even in the most carefully purifiedpotassium chlorate, the amount varying from 0*022-0*029 per cent.Applying to Stas's results, a correction based upon this observationeffects a reduction in the atomic weigh€ of silver, to which the authorsassign the value 107.89 (0= 16).Baxter 2 has redetermined the atomic weight of bromine, the meanof eighteen determinations of the ratio Ag : AgBr giving a value of'79.953 for the atomic weight of bromine.From the conversion ofsilver bromide into silver chloride the value 79,952 for bromine isdeduced (C1= 35.473). The mean of all Baxter's determinations gives79.953, which is in agreement with Stas's determination, whereasScott's results are somewhat lower. I n this connexion, the determina-tion of the density of chlorine gas may be mentioned.Treadwell andChristie 3 give as the mean of two sets of determinations, representingfive observations, the value 2.4885 (air = l), a value approximating tothat found by Moissan and Binet du Jassoneix, namely, 2.490.Baxter, in association with others, has published the results of theredetermination of the atomic weights of rnnngane~e,~ ~ o b a l t , ~ andcadmium.6 The analysis of manganese bromide and chloride givesMn = 54.96, of cobalt chloride and bromide Co = 59, and of cadmiumbromide gives 112.467 for the atomic weight of cadmium. A morerecent determination of the atomic weight of cobalt from the analysisof the chloride gives a mean value of 58*995.7The analysis of Strontium bromide according to Richards8 gives anatomic weight of 87.663 for strontium, and, if C1= 35.473 be accepted,from the chloride the value 87.661 is deduced; this result is in satis-factory agreement with that derived from the bromide, whereas if theolder atomic weight of chlorine is employed the agreement is not sogood.The author concludes that the mean of these two sets ofdeterminations, namely, 87.662, may be accepted as the atomic weight(0= 16).The alternate oxidation and reduction of pure copper according toMurmann gives for the atomic weight of this element values varyingfrom 63-573-64.153 (0 = 16) ; this variation, it is suggested, arisesfrom the absorption of air by the porous copper, The atomic weightis estimated to be 63.53 with a n error of & -03.Compt. rend., 1906, 143, 411.Ibid., 1905, 47, 446.Rsxter and Coffin, ibid., 1380.Richards, ibid., 1005, 47, 145.'L Zcit.anorg. Chcm., 1906, 50, 389.4 Baxter and Hines, J. Amcr. C'ltenz. Soc., 1906, 28, 1360.7 Baxter and Coffin, Zcit. nnorg. C h e m . , 1906, 51, 171.Baxtw, Hines, and Freveit, ibid.: 770.Moaatsh., 1906, 27, 35132 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Gutbier,l in conjunction with Birckenbach and Mehler, has by threedifferent methods obtained values for the atomic weight of bismuth,the mean of which is 208.074.The determination of the electrochemical equivalent oE iodine madeby Gallo2 gives as a mean of twenty-four estimations the atomicweight 126.89 when silver is 107.93.A new determination of the atomic weight of potassium is recordedby Richards and Stahler,3 who employed for this purpose speciallypurified potassium nitrate, which they subsequently converted into thechloride.The ratios KC1 : AgCl and KC1 : Ag were deterrnined in thesame manner as the ratios NaCl : AgCl and NaCl : Ag (see Kekort, 1905,34). The results of eight experiments with three samples of potassiumchloride gave for the mtio KC1 : AgCl an average of 0.5201 18 : 1.With five different samples of potassium chloride the average of nineexperimental determinations of the ratio KC1 : Ag gave 0.691072 : 1.Taking the atomic weight of silver as 10'7.93 and t h a t of chlorine as35.473, the above results lend to 39,114 as the atomic weight ofpotassium.Brill,4 experimenting with the micro-balance, has shown thatKriiss's method of determining the atomic weights of the rare earthsdepending on the conversion of the acid sulphates into sulphates is notaccurate, the temperature employed being too low.This conversion iscomplete at 450°, and at 700" a decomposition begins resulting inthe formation of basic sulphates which is completed at 900". Thebasic sulphates are completely transformed into the oxides a t 1380".The basic sulphates, M20,*80,, of ytterbium, yttrium, erbium, lan-thanum, and samarium have been prepared by this method. Judgedby the temperature a t which these basic sulphates begin to decompose,the basicity OF the oxides of these elements increases in the followingorder : Yb, Er, Y, Sm, and La. Upon the results of this investigationBrill bases a method for the determination of the atomic weights of theelements of these rare earths.Feit and Przibylla5 show that apractical method of determining the atomic weights of these eIementsis to dissolve a known weight of the oxide in AT/2 sulphuric acidand titrate back with N/10 caust icso da solut ion,using niethyl-orangeas indicator. I n this way the following values for the atomic weightshave been obtained : La = 139.17, Pr = 140.62, Nd = 144.6, Sm = 150.56,Xu = 152.66, Gd = 157.47, Yb = 173.52, Yt = 89.40. With the exceptionof europium and yttrium, the results are in good agreement with themost trustworthy of previous determinations.Zeit. Elektrochena., 1905, 11, 831.Alti R. Aecnd. Lincei, 1906, [v], 15, i, 24.Ber., 1906, 39, 3611.Ibid., 1906, 50, 249.Zeit.nnory. Chem., 1905, 47, 464INORGANIC CHEMISTRY. 33The conversion of dysprosium sulphate, [Dy2(S0,),,8H20] into theoxide (Dy,O,) has been employed by Urbain and Demenitroux todetermine the atomic weight of the element. When the earth isprepared by the fractional crystallisation of the nitrate, the atomicweight found in this way is 162.52, whereas with the earth preparedby the fractional crystallisation of the ethyl sulphates the value162.54 was obtained.1 Hinrichsen and Sahlbom 2 have determinedthe atomic weight of tantalum by converting the metal into thepentoxide, which gives the value 181.0, somewhat lower than theatomic weight assigned t o the metal by Marignac, whose method theauthors consider to be faulty.Rare Earths.Matignon and Cazes find that samarium chloride (SmC1,) is reducedto a lower chloride SmCl,, when heated at high temperatures inhydrogen or ammonia, or with aluminium powder.This chlorideis a dark brown, crystalline powder, and is insoluble in carbondisulphide, benzene, chloroform, and the other organic solventswhich dissolve the anhydrous chlorides of the rare metals. I t isdecomposed by water, forming samarium oxide (Sm203) and an oxy-chloride (SmOC1) ; hydrogen is also liberated. The chlorides ofpraseodymium and neodymium are not reduced ; it is suggestedthat this difference may be employed i n the separation of tlieseelements. Samarium iodide4 is reduced to samarious iodide byheating in a current of hydrogen.The dehydration of the hydrated chlorides of yttrium and ytterbium,YtCl,,GH,O and YbCl,,6H20, has been studied by Matignon,5 and thephysical properties of the anhydrous clilorides so obtained described.The following compounds of neodymium have been isolated :NdC13,H20 ; NdC1,,6H20 ; NdCL3,3EtOH ; NdC13,3C,NH, ; NdOCl ;NdBr, and NdI,, also NdH,(SO,), and Nd202(S04).6 Matignon andTrannoy 7 describe a series of seven compounds, formed by the inter-action of gaseous or liquid ammonia and anhydrous neodymiumchloride ; the temperatures of dissociation and heats of formationof these compounds have been determined.The determination ofthe heats of formation of the sulphates of lanthanum, praseodymium,neodymium, and samarium, made by the dissolution of the oxidesin dilute sulphuric acid, shows the basic function of these oxidesComnpt.rend., 1906, 143, 598, and 142, 785.Ber., 1906, 39, 2600.Matigiion and Cazes, Ann. CJLim. Phyls., 1906, [viii], 8, 417.6 Ann. Chim. Phys., 1906, [viii], 8, 433 and 440.7 Compt. rend., 1906, 142, 1042 ; see also Anti. Report, 1905, 37.Compt. rem~d., 1906, 142, 183.Ibid., 243.VOL. 111. 36 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.to decrease with increase in atomic weight.l For the preparation ofpure cerium compounds the following method is recommended byOrloff .2 The mixed sulphates are treated with ammonium oxalate ; theprecipitate formed consists of the oxalates of lanthanum, neodymium,praseodymium, yttrium, and some cerium, and from the filtrates con-taining the oxalates of thorium and cerium the latter slowlyprecipitates, whereas the former remains in solution.This separationis materially assisted by the addition of sodium sulphite to thesolution,A silicide of thorium, ThSi,, is formed by the interaction of thedouble fluoride of thorium and potassium and potassium silicofluorideand aluminium heated in a n electric furnace a t 1800". This samecompound is produced by heating thoria with silicon in an electricfurnace, or by heating aluminium, silicon, and thorium together a t1000°.3 The compound ThAl,4 is formed by the direct union of theelements when heated together in a vacuum, or by reducing potasFiumthorium fluoride o r thoria with aluminium in an electric furnace. I nappearance this alloy resembles aluminium; it is attacked by thehalogens and mineral acids, but not by alkali solutions, whereas it isviolently acted on by fused alkalis or carbonates of the alkali metals.preparit-tions of terbium compounds, concludes that terbium is a well-definedelement ; according to Urbain? this element forms a sulphate of thecomposition Te2(S0,),,8H20, and has an atomic weight 159.22 (0 = 16).The evidence in support of the existence of the element victorium isstill a matter of discussion, as is evidenced by the publications ofCrookes,s Urbain,g and Marc.loIn conclusion, mention should be made of Urbain's rZsum5 of hisresearches in this field of investigation made since 1S99.11Attempts to prepare metallic thorium both by the action of sodiumon the chloride and by electrolytic processes have not proved success-ful ; the product mas always found to contain some oxide (Moissan andHonigschmid).12Eberhard,5 from a spectroscopic investigation of Urbain'sThe Argon Group.The examination of the gaseous contents of forty-three thermalsprings has revealed the general presence in these gases of argon andChem Zeit., 1906, 30, 733,Hoi:igschmid, ibid., 280.Ann.h'eport, 1906, 57.Matignon, Compt. Tend., 1906, 142, 276.Xitzungsbcr. K. Acad. Jiss. Beriin, 1906, 384.3 Hijuigschmid, Comyt. rewl., 1906, 142, 157.7 Compt. rend., 1906, 142, 957.9 Cornpt. relid., 1905, 14i, 954.l1 J. Chiin. PJbys., 1906, 4, 31, 105.12 Aniz. Chim. Phys., 1906, [viii], 8, 182.Chem. News, 1906, 43, 143.lo Ber., 1906, 39, 1392INORGANIC CBEMISTRI-.35helium, the proportion of which varies with the amount of nitrogenand indirectly with the amount of carbon dioxide. I n the water ofMaziBres, the new gases formed 6.35 per cent. of the nitrogen1(Moureu).Olszewski,2 working with helium obtained from t h ~ r i a n i t e , ~ has notbeen able t o liquefy this gas, even when it is cooled by frozen hydrogenand submitted to a pressure of 180 atmospheres which was suddenlyreduced to 1 atmosphere; in this way the temperature was loweredto - 271.3'. The boiling point of helium is therefore estimated to beThe density of the vapours of zinc, cadmium, mercury, sulphur,selenium, and arsenic has been determined in atmospheres of argonand of helium, and compared with the values obtained when atmo-spheres of nitrogen and of hydrogen are employed.From the resultsof these observations, Ternent Cooke * concludes that these elementsexhibit a tendency to unite with argon and helium. Scheerer (1844),in his analysis of malacone, a silicate of zirconium, entirely over-looked the small proportion of uranium which Kitchin and Winter-son 6nd to be present in this mineral ; further, these authors show thatwhen this mineral is fused with potassium hydrogen sulphate a gas con-taining both argon and helium is given off,below - 271'.Adopting the method used in previous reports, the results of inves-tigations published during the year will be discussed under the head-ings of the several groups in the periodic system of the elements.Group I A .For the electrolytic preparation of lithium, Ruff and Johannsen 0recommend a mixture of bromide and chloride containing 13 per cent.of the latter ; this mixture melts at 520'.A carbon anode and an ironcathode are employed and an E.M.F. of 10 volts and current of 100amperes. Weyberg7 has ob-tained lithium aluminosilicates by fusing kaolin with lithium com-pounds. With the chloride and carbonate, the compound Li, AI,Si,O,,is formed, with the sulphate Li,Al,Si,O,, whilst with the bromide7Li2A12Si208, 2LiBr is produced. The hydroxides of lithium, rubidium,and cesium have been examined by Forcrand ; each forms a mono-hydrate ; the lithium compound, Li(OH)H,O, separates from its solu-tions in well-defined crystals which lose H,O when heated, formingThe metal so obtained melts at 180'.Compt.rend., 1906, 142, 1155.Ann. Beport, 1905, 38.Trans., 1906, 89, 1568.cent?.. Jfi??,., 1905, 646.Bull. Acad. Sci. Cracow, 1905, 40i.Zeit. physikal. Chem., 1906, 55, 537.Zeit. Elektrochem., 1906, 12, 186.Compf. rend., 1906, 142, 1252.0 36 ANNUAL REPORTS Oh: THE PROGRESS OF CHEMISTRY.the hydroxide melting at 445'. The hydrated rubidium and cssiumhydroxides, when dehydrated in silver crucibles, form peroxides whichattack the silver. From solutions of lithium hydroxide and chromiumtrioxide in water at 30°, the substances separating in the solid stateare [Li(OH),H,O 3, Li,Cr04,2H20, Li2Cr207,2H,0, and CrO,. Thesolubility of the chromates of ammonium, potassium, sodium, andlithium increases in the order named, whilst for the dichromatesthe order is potassium, ammonium, lithium, and s0dium.l Lebeau,2experimenting on the carbonates of the alkali metals heated in acurrent of carbon dioxide in an electric furnace, finds these carbon-ates to be non-volatile between 780' and 1 200°, save in the case of czesiumcarbonate, which dissociates somewhat at 720°, a temperature muchbelow that a t which it volatilises.are, accord-ing to Ruff and Geisel,4 mixtures of metal with solutions of the metalin ammonia.These solutions decompose on keeping, with the evolutionof hydrogen and the formation of metallic amides. Ruff and Geiselalso show that the hydrides of the alkali metals are decomposed byliquid ammonia with the liberation of hydrogen and the production ofamides. The oxidation of the solution of rubidium in ammonia yieldsa di- and tetra-oxide (Rb,O, and Rb,O,), but no sesquioxide isformed as in the case with cmium (Rer~gade),~ The peroxide ofcaesium (Cs,O,) is formed when the metal is heated a t 300--350' in astream of oxygen.6The production of caustic soda by the decomposition of sodiumsilicofluoride with lime is the subject of a patent by Reich,' who findst h i t the reaction is more complete when the amount of lime is doublethat required by the following equation :The metalammonium compounds described by JoannisNa,SiF6 + 4Ca0 = Na20 + 3CaF, + CaSiO,.The residue left after the removal of the alkali on dissolution inhydrochloric acid gives a solution of hydrogen silicofluoride which maybe recovered as potassium silicofluoride.From solutions of sodiumsulphate in sulphuric acid, two acid sulphates, namely, Na,H(SO,), andNa,H(SO,),H,O, have been obtaineda8 The investigation of the fusioncurves of mixtures of the haloids of potassium and sodium provesthese salts to form isomorphous mixture^.^ According t o van't HoffSchreineinakers, Chenz. Weekblnd.: 1905, 2, 633 ; see also Zeit, p h y s i k d Chem.,Bull, SOC. chim., 1906, [iii], 35, 5.Afin. Chim. Phys., 1906, [viii], 7, 5. Ber., 1906, 39, 828.Compt. rend., 1906, 142, 1533 ; see also Ann. Report, 1905, 43.D'Ans, Ber., 1906, 39, 1534.KurnakofT and Schemtscl~usch~y, Chem. Ccnkr., 1906, i, 526.1906, 55, 71.Kengade, ibid., 1149.' D.12.-P. 161795INORGANIC CHEMISTRY. 31and Bar~chall,~ glaserite is not a compound of the sulphates ofpotassium and sodium, but an isomorphous mixture of these sulphates.A study of the solid phases between iodine, an iodide, and polyiodideshows in the case of the alkali metals the tendency to form higherpolyiodides to increase in the following order (LiNa), K, Rb, Cs, whichis the order of decreasing solubilities of the platinichlorides (Abeggand Hamburger).2When potassium is heated in a nickel dish, a black mass results,from which nickelous nickelite, (Ni0,Ni0,,2H20), has been obtained.The corresronding cobaltous cobaltite (Co0,Cc02,2H20) is formed bythe action of potassium peroxide on cobaltous oxide.3A simple process f c z the continuous production of potassium chlorateby the electrolysis of a solution of potassium chloride containing somepotassium chromate and hydrochloric acid is described by A.Wallach.4Wohler and Kasarnowski5 consider the blue colour exhibited bysome specimens of natural rock salt may be due to organic matter, assuch specimens when heated in a current of oxygen mere found to losatheir colour and yield carbon dioxide and water. The coloratioLsproduced by heating halides with the metals probably arise frominclusion of tbe particles of the metals, or possibly from the productionof sub-halides. The former explanation is supported by the ultra-microscopic examination of coloured specimens of rock salt made bySiedentopf .6The investigations of the sulphides of rubidium and casium by Biltzand Wilke-Dorfurh show the monosulphides Rb2S,4H20 and Cs2S,4H,0to be colourless, crystalline compounds, as are also the sulphydrates.From solutions of the monosulphides and sulphur the tetrasulphidesRb,S,,2H20 and Cs,S, have been obtained, the former in yellowcrystals, the latter in reddish-yellow prisms.The pentasulphides(see Annual Report, 1905), when heated in a stream of hydrogen, losesulphur and from the residue by dissolution in water the disul-phides, Rb,S,,H,O, Cs,S2,H20, are obtained in colourless crystalsWhen the pentasulphides are heated in a current of nitrogen the tri-sulphides Rb,S,,€€,O and Cs,S,,H,O are obtained.* The sulphides ofrubidium are less hygroscopic than those of cmium, and Cs,S, is morereadily volatile than Rb,S2.Paal and Kuhng record the formation of organosols and gels ofsodium chloride and of sodium bromide by the addition of light petrol-1 Zeit.physiknl. Chem., 1906, 56, 212. Zeit. anorg. Chem., 1906, 50, 403.Hofmann and Hiendlmaier, Ber., 1906, 39, 3184.Zeit. Elektrociiew., 1906, 12, 667 ; see also Coppacloro, Gaxzetta, 1906, 36,ti Zeit. anorg. Chem., 1905, 47, 353.Chem. Cent?.., 1906, i, 388. 7 Zeit. nnorg. Chem., 1906, 48, 297,Ibid., 1906, 50, 67. 9 Ber., 1906, 39, 2839 and 2863.ii, 32138 ANNUAL REI'ORTS ON THE PROGRESS OF CHEMISTRY.eum to the condensation product of ethyl chloroacetate or bromo-aoetate with ethyl sodiomalonate. When freshly prepared the precipi-tates are soluble in benzene, but become insoluble after drying in avacuum.The gsls so produced contain 58 per cent. or more of sodiumchloride.According to Mathewson1 sodium forms a series of definite com-pounds with lead, cadmium, bismuth, and antimony,Group I B.The freezing point of pure copper is 1085" ; it is depressed by cuprousoxide, the eutectic mixture solidifies at 1065' and contains, accordingt o Dejean,2 4.7 per cent. of cuprous oxide which Heyn considers toohigh, and fixes the amount a t 3.5 per cent. The maximum depressionproduced by aluminium is 1039O, corresponding to S*6 per cent, of themetal. Moissan4 has succeeded in distilling copper in an electricfurnace; the ingot left in the crucible on cooling exhibits thephenomenon of '' spitting," and is covered by a layer of graphite.Copper chloride heated in a vacuum with calcium turnings gives anorange-yellow, brittle alloy containing 18.3-1 8.8 per cent of copper.5The chief product of the action of ammonia on heated ouprous oxide isthe nitride Cu,N, ;which is converted into cuprous chloride andammonia by hydrochloric aaid ; it is attacked by nitric and sulphuricacids.6 Cuprosilicon, prepared by h a t i n g a mixture of these elements,consists in the main of the compound Cu,Si and contains some freesilicon ip a form soluble in hydrofluoric acid 7 (Annual Report, 1904,45).The influence of small quantities of phosphorus, arsenic, anti-mony, bismuth, and lead on the structure of copper has been examinedby Hiorns.8 Heyn and Bauer 9 find that copper and cuprous sulphideare not perfectly miscible in the fused state, and that selenium andtellurium produce an alteration in the microstructure of coppersimilar t o that produced by sulphur. The presence of these elementscan be readily distinguished by treating copper turnings with asolution of potassium cyanide.On the addition of cadmium acetate dis-solved in acetic acid, cuprous sulphide gives a yellow precipitate, whereasthe selenide gives an orange ; the telluride gives a coloration similar tothat of permanganate on treatment with potassium cyanide. TheseZeit. anorg. C'hem., 1906, 50, 171. RG'L'. de dlc?tnlZzwgie, 1906, 3, 233.Conzpt. wnd., 1905, 141, 853. 3 Ibid., 345.5 Hackspill, Cmtpt. rend., 1906, 142, 89.7 Lebenu, Compt. rend., 1905, 141, 889 ; 1906, 142, 154 ; and Vigouroux, ibid.Guntz and Bassett, BdZ.SOC. chim, 1906, [iii], 35, 201.1906, 142, 37.,T. SOC. Chem. Ind., 1906, 25, 616. L, MetnlZurgie, 1906, 3, 73INORGANIC CHEMISTRY. 39authors also state that sulphur dioxide does not attack copper a t900-1100', save in presence of a reducing agent ; and a t this tempera-ture the action of cuprous sulphide on cuprous oxide,cu,s + 2cu,o = so, + 3cu2,is complete.From the freezing-point curves of mixtures of cuprous sulphide andferrous sulphide Rontgenl infers the existence of the following com-pounds : 5Cu2S,2FeS, Cu,S,FeS, and 2Cu2S,FeS. The presence ofmetallic copper in copper mattes may arise from the change ofCu,SFeS into Cu and FeS,.The action of strong sulphuric acid on copper is maintained bySluiter2 to be best explained by the reduction theory, since thesesubstances heated together a t 130' with nitrobenzene yield aniline,whereas with sulphuric acid containing 12 per cent.of sulphnr trioxideno aniline is formed. On the other hand, van Deventer3 considersneither explanation satisfactory, but suggests that water plays a part inthe reaction, which he formulates as follows :CU + H,O = CUO + H, ; CUO + H2S04 = CUSO, + H,O ; H, + SO4 =2H,O + so,.The alloys of copper and phosphorus have been investigated byHeyn and Bauer ;* these alloys are harder than the corresponding tinalloye, and a compound, Cu,P, apparently exists. By employing higherpressures than the atmospheric pressure Friedrich has produced alloysof copper and arsenic containing as much a s 44 per cent.of the latterelement, The existence of several definite compounds are indicated bythe melting-point curves of mixtures of these elements. Guillet 6 haspublished an account of the examination of ' special brasses,' namely,copper and zinc alloys to which a third metal is added; whilst Pfeiffer 7concludes that copper does not alloy with iron o r with iron and carbonalloys.The production of colloidal copper oxide and of colloidal copper isdescribed by Paal and Leuze.8von Wartenberg's observations on the relative density of thevapour of silver a t 2000' show the molecule to be mono-atomic, andthat the metal volatilises a t 1950'. The reduction of silver chlorideby metallic calcium results in the formation of alloys containing from6.3 t o 16 per cent.of calcium ; these alloys are grey, brittle, crystal-Metnlliwgie, 1906, 3, 4'79.Ibid., 1906, 3, 515.MetaZ/urgic, 190t5, 2, 4 i i .7 ilfctrclluryie, 1906, 3, 281.9 Ibid., 1906, 39, 381.Chenz. Weekhlad., 1906, 3, 63.Rer. lie Xitrcllicrgic, 1906, 3, 243.Ber., 1506, 39, 1545 and 1550.Mitt. TC. Materid-prtifump Amt., 1906, 24, 9340 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.line solids, oxidised by the air and attacking water a t the ordinarytemperature.1That the electrolysis of a solution of silver nitrate gives a depositof a black powder on the anode has been observed both by Watson2and Barbieri.3 The first of these authors shows this substance to beAg,XO,,, which is decomposed by boiling water, forming silver peroxide,thus :Ag7NOI1 = AgNO, + 3Ag,02 + O,,and by ammonia with the formation of the oxide Ag,O,, from theinteraction of silver peroxide and ammonia :6Ag,O, + 2NH, = 3Ag40, + 3H,O + N,.Barbieri also records the observation that a layer of black oxide isformed on the anode when a solution of potassium hydrogen carbonateis electrolysed with silver electrodes ; this deposit gradually dissolvest o form a brown solution, which readily oxidises ferrous and cobaltoussalts.According to Lewie,4 silver suboxide is not formed by the decom-position of silver oxide Ag,O, a t 5 0 2 O , 5 2 5 O , or 445'.The investigation of the action of silver nitrate on disodium hydrogenorthophosphate shows this reaction to take place in several stages ; thesilver phosphate contains only 76 per cent.of silver, whereas Ag,PO,requires 77.32 per cent. Further, the solution always contains somephosphoric acid not precipit,ated by silver nitrate.5The freezing-point curves of mixtures of silver with sulphur,selenium, or tellurium give no evidence of the existence of compoundsother than Ag.,S, Ag,Se, or Ag2Te.6 The melting-point curves ofmixtures of silver and silver sulphide have been examined by Friedrichand Leroux,' who find that a t 906" these substances separate into twolayers, and from the eutectic mixture a t 806' almost pure silver sul-phide separates out. These authors from their investigation of thealloys of silver and arsenic fail to find any evidence in support of theexistence of the compound Ag,As ; by the reduction of silver arsenatewith potassium cyanide at low temperatures they have obtained alloyscontaining 87*3-89*5" per cent, of arsenicsSilver and zinc appear to form compounds of the formulae Ag,Zo2,AgZn, Ag,Zn,, and Ag,Zn, ; 9 with magnesium, silver forms alloyswhich are harder than either constituent ; those rich in magnesiumCompt.rend., 1906, 142, 89.Alti R. Accad. Liwei, 1906, [v], 15; i, 500.J. Anaer. Chcm. SOC., 1906, 28, 139.Laiig and Kaufmann, ibid., 1905, 27, 1515.Pdlabon, Compt. rend., 1906, 143, 294.Trans., 1906, 89, 578.Metdluryie, 1906, 3, 361. * Ibid., 192. 9 Petrenko, Zeit. anorg. Chem., 1906, 48, 347INORGANIC CHEMISTRY. 41are yellow, are readily oxidised, and decompose water more easily thanmagnes1um.lSilver forms alloys with thallium, bismuth, and antimony ; with thelast only does it form a chemical compound, namely, Ag3Sb.2I n counexion with their determination of the melting point of goldmentioned in last' year's Report, Jaquerod and Perrot3 give thefollowing values for the coefficient of expansion of gases a t tempera-tures ranging from 0" to 1067-4O: air 0.0036644, CO 0.0036638,0, 0.0036652, and CO, 0.0036756 and 04036713, that of nitrogenbeing 0 003664.Ammonia unites with aurous iodide, forming AuT,6NH3, which at28" dissociates into AuI,NH, and ammonia, and when heated formsammonia., iodine, and gold; by water it is converted into ammoniumiodide and gold.Aurous bromide forms AuBr,SNH,, readily passinginto AuBr,NH3 ; whereas aurous chloride forms AuC1,2NH3, which,losing ammonia, gives AuC1,3NH3; this compound is stable at 180",but above this temperature is resolved into ammonium chloride andgold.4Gold and zinc when alloyed appear to form the compoundsAuZn, Au3Zn,, AuZns, and with cadmium gold forms Au4Cd, andAuCd3,5 whilst with antimony it forms the compound AuSb,, but nodefinite compound is obtained with bismuth.6 For the production of goldhydrosols the use of oil of turpentine (or pinene) or oil of rosemary isrecommended by Vanino and Hart1,7 the colours produced dependingon the temperature and concentration of the solutions.Neumann 8 has examined carefully the conditions required for theelectrolytic precipitation of gold from cyanide solutions, using a lead oran iron cathode.The gold may also be deposited from these solutionsusing a carbon cathode, from which the gold may be stripped byusing the gilded carbon as anode.Group I1 A .The analysis of a sample of electrolytic calcium from the Bitterfeldworks shows it to contain some calcium carbide, iron, silica, andmanganese, the Iast-named element being possibly derived from thesteel tool used to break the calcium ( L a r ~ e n ) . ~ According to Doermer,localcium when strongly heated evolves hydrogen, which is reabsorbedSchemtschuschny, Zeit. nnorg. Chent., 1906, 49, 400.Petrenko, ibid., 50, 133.Meyer, Compt. rend., 1906, 143, 280.Vogel, Zeit. anorg. Chem., 1906, 48, 319 and 333.Vogel, ibid., 1906, 50, 145.7 Bcr., 1906, 39, 1696.Zeit. Ekktrochem., 1906, 12, 569.3 Arch. Xci. I'hys. Nat., 1905, [v], 20, 506.9 Chem. Cen.fr., 1905, ii, 1466.lo Ber., 1906, 39, 21142 ANNUAL REPONTS ON THE PROGRESS OF CHEMISTRY.at a lower temperature; when the metal is struck on an anvil ex-plosions not infrequently occur. The dissolution of metallic calciumin molten cast-iron is attended by considerable development of heat, andsome calcium carbide is formed. Whilst calcium reduces ferric oxide itdoes not combine with the iron produced; alloys of calcium with copper,aluminium, and magnesium are described by Stockem.l By the action ofcalcium on fused lead chloride alloys of lead arid calcium are formed.2Calcium hydride is manufactured by heating electrolytic calcium con-tained in horizontal retorts in a current of hydrogen ; the crude productcontains 90 per cent.of the hydride, and may be used as a source ofhydrogen ; when decomposed by water one kilo. yields one cubic metreof the gas.3 Hoffmeister regards the colourless gas left after treatmentof commercial acetylene with acetone and ammoniacal cuprous chlorideas a gaseous hydride of calcinm, for on burning in air it forms limeand water; with oxygen i t forms a n extremely explosive mixture.Attempts to prepare sub-salts of calcium by fusion of the metal withthe chloride, iodide, or fluoride yield products which contain lime andcalcium hydride, the latter resulting from the hydrogen produced bythe action of the metal on the traces of water present, Commercialcalcium peroxide is shown by von Foregger and Philipp 6 to consist ofcalcium peroxide and calcium hydroxide, and to contain about 60per cent.of the peroxide; when treated with water it forms calciumhydroxide and hydrogen peroxide. Calcium peroxide forms twohydrates, namely, Ca0,,8H20 and Ca0,,2H,O ; the dry peroxide can beheated a t 200’ without change and is non-explosive. With respect t oavailable oxygen it stands midway between neutral and acid calciumpermanganate. The peroxides of strontium, magnesium, and zinc aredecomposed a t a lower temperature t h m the calcium compound.Tarugi’s conception of the constitution of bleaching powder, referredto in the Report for 1905, p. 69, is refuted by Ditz,7 whilstTiesenholt 8 considers the view that, bleaching powder is a mixture ofcalcium hypochlorite and calcium chloride to be supported by the ob-servation that mixtures of calcium hypochlorite or lithium hypo-chlorite and calcium chloride are decomposed with the liberation ofchlorine by moisture or carbon dioxide.Further, when bleachingpowder is triturated with carbon tetrachloride it separates into twopowders differing in their densities and chlorine content. The evolu-tion of chlorine from aqueous solutions of bleaching powder is at-I JletnZlz6~gic, 1906, 3, 147 ; also Quasebert, ibid., 28.2 Hnckspill, Cbntpt. rend., 1906, 143, 227.3 Guntz and Bassett, jun., Bull. SOC. chi?%, 1906, [iii], 35, 404.6 J . SOC. Chem. Ind., 1906, 25, 298.8 J, Rim.Phys. Chem. Soc., 1905, 37, 834,Jaubert, ibid., 142, 788. Zeit. aizorg. Chem., 1906, 48, 137.Zeif. ni7,ptu. Chem,, 1905, 18, 1690INORGANIC CHEMISTlLY. 43tributed t o the hypochlorous acid formed by the hydrolysis ofcalcium h ypochlorite, and the fact that calcium chloride more readilythan other metallic chlorides brings about the following decomposition,Ca(OCl), + CaCI, + 2H,O 2Ca(OK), + 2C12,is considered to depend on the peculiar properties of its compounds withwater of crystallisation.Ammonium syogenite, Ca(NH,),(SO,),,H,O, formed by addingcalcium sulphate to a saturated solution of ammonium sulphate, isstable a t 25" ; observations on the solubility of gypsum in ammoniumsulphate 2 and in magnesium sulphate 3 solutions are recorded.By heating strontium hydride in a vacuum a t 1000°, Guntz andRoederer have obtained strontium as a silver-white metal, whichtarnishes in the air and melts about $00'.It is attacked by waterand alcohol; with carbon dioxide at a red heat it forms the carbonateand carbide. The authors have determined the heat of formation ofSrCl,,Aq and of SrO, and obtain results some 11 Cals. higher than thevalues given by Thomsen. The electrolysis of an aqueous solution ofstrontium chloride using a mercury cathode yields two amalgams, oneliquid and the other crystalline, of the composition SrHgll ; the latter,when heated, gives Sr,Hg, and SrHg6, and on distillation an amalgamdistils over, so that metallic strontium cannot be obtained in thisway. 5The so-called strontium-ammonium decomposes slowly in R vacuiimat 20°, leaving a residue of Sr(NH,),; by the action of carbonmonoxide strontium-ammonium is converted into Sr(C0)2, whichbecomes bright yellow when exposed to the air, and when heatedforms strontia, strontium carbonate and carbon.Oxygen convert8strontium-ammonium into strontia and the peroxide, whilst withnitric oxide i t forms the hyponitrite. Hydrogen, nitrogen, andstrontia.mide are formed by the action of ammonia on the metalat 200°, and at 800" a mixture of nitride and hydride is produced6Barium suboxide is formed by heating the oxide with magnesium ina vacuum at 1100O ; when heated together in the proportion of 3Ba0to Mg an alloy distils over, but if aluminium is substituted formagnesium the distillate consists of metallic barium.7 Bauer * hasobtained a hydrated hydroxide of the formula Ba(OH),,3K,O; strontiumand calcium hydroxides do not form similar compounds.The be-haviour of barium carbonate at high temperatures has been investi-D'Ans, Ber., 1906, 39, 3326.Bell and Taber, J. yhysiknl. C'Jcm., 2906, 10, 119.Cameron and Bell, zbid., 210.Guntz and Roederer, BzdZ. SOC. c h h . , 1906, [iii], 35, 494.Roederer, ibid., 1906, iii, 35, 715.Zeit. anorg. Chem., 1905, 47, 401.Coiiipl. r e d . , 1906, 142, 400.7 Guntz, Compt. rend., 1906, 143, 35944 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.gated independently by Finkelstein and Boeke ; both ficd LeChatelier's statement as to the fusion of barium carbonate to beerroneous ; this compound does not melt at 1350°, but fuses partlywhen heated at 1380' in a st,ream of carbon dioxide, By prolongedheating a t 1120' a basic carbonate is formed which dissolves barytaand a t higher temperatures barium carbonate. Boeke finds thatcalcite is formed from aragonite a t 470°, and that calcium carbonatedoes not melt even when heated to 1400-1500' in an atmcsphere ofcarbon dioxide under pressure.The action of the alkali bromides orbarium Carbonate is observed by Taponier3 to increase nith time,concentrat ion, and temperature, and the following represents the ordarof decreasing activity : ammonium, sodium, and potassium bromides(Amzucd Report, 1905, 45). The solubility of barium sulphate inhydrogen Feroxide solution has been established by Gawalomski.4Neuberg and Neimann record some very interesting observations onthe production of gelatinous modifications of salts of the alkalineearths ; gelatinous barium sulphate, for example, is produced bytreating the solution of barium oxide in methyl alcohol with dilutesnlphuric acid.This jelly can be dried in a vacuum without change,but when boiled with water it becomes crystalline. Gelatinous bariumcarbonate is formed by the action of carbon dioxide in the methyl-alcoholic solution of the oxide, and this, if treated with more carbondioxide, forms a white powder, BaC03,H20, which is soluble in water.A gelatinous sulphide, BaS,H,O, and gelatinous forms of other com-pounds are described. Vanino,6 continuing the study of phosphorescentsulphides referred to in last year's Report, finds that the sulphides ofcalcium and zinc do not emit Recqucrel rays, and further, that whilstcalcium sulphide loses its phosphorescence when suspended in water, itis not so affected by suspension in ether, acetone, ethyl, or amylalcohol.The phosphorescence of these sulphides is shown by Joriseenand Ringer 7 to depend on the presence of traces of salts of bismuthcadmium, and manganese, also of potassium and sodium chlorides, asthe pure sulphides themselves are non-phosphorescent.Glucinum hydroxide, heated alone or with water, or aqueous solutionsof ammonia or alkali carbonate, o r with alkali hydroxides, is graduallyconverted into a form sparingly soluble or insoluble in alkalis orBer., 1906, 39, 1585. Zeit.anorg. Chena., 1906, 50, 244.3 Bull. Soc. chim., 1906, [iii], 35, 280.r, Bioehem. Zeit., 1906, I, 166.7 Chem. Centr., 1906, i, 644.4 Chem. Centr., 1906, ii, 7.J. pr. Chcm., 1906, [ii], 73, 446 ; also Ann. Rerort, 1905, 45INORGANIC CHEmSTRY. 45acids.l The conditions determining the existence of the hexa-, tetra-and di-hydrates of glucinum sulphate, and their solubilities, have beeninvestigated by Levi-Malvano.2 Davis 3 shows that from the solutionof magnesium carbonate in carbon dioxide and water crystals of thecomposition MgCO,,SH,O? separate out when the pressure of thecarbon dioxide is reduced.OH*Mg CO,H, 2 H,O,and its production explained by the equation,This trihydrate is regarded asMg(CO,H), + 3H20 = OH*Mg*C03H,2H,0 + H2C03.When boiled with water,' the trihydrate gives a mixture ofmagnesium hydroxide and the hgdroxycarbonate, formed thus :OH*Mg*C03H,2H20 = OH*Mg*CO,H + 2H,O,OH*Mg*C03H + H,O = Mg(OH), + H,O + CO,.The addition of caustic soda to a solution of the bicarbonate doesnot produce a precipitate until boiled, because in the first instance thedouble salt, Mg(CO,Na), (Deville, 1851, and Repolds, Trans., 1898,73, 262), is formed, which on boiling is decomposed, forming a mixtureof OH*Mg*C03H,2H,0, OH*Mg*CO,H and Mg(OH),, which constitutesmagnesia alba.The first stage in the production of magnesia alba is,therefore, the formation of the double sodium magnesium carbonate,thus :2Na2C03 + MgSO, = Mg(CO,Na), + Na2S0,.The microscopic examination of magnesia alba shows it to be hetero-geneous, as the above explanation would imply.Zinc oxide heated in an electric furnace begins to volatilise a t 11 00'and doesiso rapidly at 1700", the volatilised product exhibits well-defined crystalline structure.Cadmium oxide volatilises appreciablyat 800" and readily at 1000" (Doeltz and Graumann).4Precipihated zinc carbonate precipitates iron, aluminium, anduranium completely from cold solutions of ferric chloride, aluminiumnitrate, and uranyl nitrate ; but chromium is only partially precipitatedfrom chromium nitrate solutions in the cold, and completely onboiling. Precipitated cadmium carbonate gives complete precipita-tions with solutions of ferric chloride and nitrate only?When cadmium is burnt a t a low temperature, the product containssome peroxide (CdO.JGA modification of mercurous chloride, Meyer 7 states, is producedby the reduction of mercuric chloride by lithium sulphite, whichAtti 11.Accad. Lincci, 1905, [v], 14, ii, 502. D.R.-P. 165488.Iiohn, Zeit. unorg. Chew, 1906, 50, 315.3 J. Xoc. Chem. I7ad., 1906, 25, 788.ti Manchot, Ber., 1906, 39, 170.Metallwgie, 1906, 3, 212 and 372.Zeit. niiorg. Chem., 1905, 47, 39946 ANHUAL REPORTS ON THE PROGRESS OF CHEMISTRY.gives the ordinary calomel, in the filtrate from it the new formseparates out in lustrous scales. From the study of the solutions ofmercuric iodide in nitrobenzene, m- and p-nitrotoluene, and cc-nitro-naphthalene, it appears that the yellow modification only exists inthese solutions.1By the interaction of solutions of mercuric chloride and borax twooxychlorides of mercury are formed, one crystallising in brown needlesand the second in lustrous, golden scales.2 By the action of iodine onmercurous sulphate, mercurous iodide, mercuric sulphat'e, and freesulphuric acid are formed.The action with mercuric sulphate doesnot take place so readily; in presence of water in excess the productconsists of mercuric iodide and iodate ; in presence of alcohol, sulphurtrioxide is formed in addition and the alcohol is oxidised toacetaldehyde.3Foote and Levy have investigated the double salts which thechlorides of potassium, rubidium, and sodium form with mercuric~ h l o r i d e .~ Duboin has extended-his investigation j of the compoundsformed by mercuric iodide and the iodides of other metals, and hasdescribed the compounds containing the iodides of calcium, strontium,barium, zinc, cadmium, magnesium, and manganese.6A number of investigations have been published during the yeardealing with ;the alloys containing metals of this group. Sodiumdissolves in magnesium to the extent of 2 per cent., forming an alloymelting a t 638--650°, and magnesium dissolves in sodium to theextent of 1.6 per cent., forming a mixture which has the melting point ofsodium. Zinc and sodium are only partially miscible ; they form a com-pound which is either NaZnll or NaZn12.7 Magnesium appears to formdefinite crystalline compounds of the formulze CdMg, Zn2Alg, Bi,Mg,,and Sb,Mg,.8 The statement of Monkemeyer that the compound ZnSbmelts at 561" is not confirmed by Schemtschuschny,1° who finds thatthe compound breaks up a t 5 3 7 O into Zn,Sb, and Sb, which, on cooling,form a stable system of ZnSb and Sb.Cadmium and copper formthe compounds Cu,Cd and Cu2Cd3.11 The freezing-point curves ofzinc and arsenic do not afford definite information as to whether theseelements f orm compounds or solid solutions.12. The alkali andalkali earth metals, according to McPhail Smith,l3 dissolve in mercuryAtti R. Accad. Jineei, 1906, [v], 15, ii, 192.Zeit. anorg. Chem., 1906, 49, 336.Amer. Chenz. J., 1906, 35, 236.Comnpt. rend., 1906, 142, 40, 395, 573, 887, 1338, and 143, 313.!I Abstr., 1905, ii, 171.3 Briickner, Monatsh., 1906, 27, 341.Ann.Report, 1905, 46.7 Mathewson, Zeit. anorg. Chem., 1906, 48, 191.8 Grube, ibid., 49, 72.lo Chem. Centr., 1906, i, 536.l2 Friedrich aiid Leroux, Metallz~rgie, 1906, 3, 472.l1 SalIinien, Zeit. anorg. Chem., 1906, 49, 301.Amer. Chenz. J., 1906, 36, 124lNORGANIC CHEMISTRY. 41in the form of solutions of compounds of tho formula MHg,,, whilstzinc, cadmium, bismuth, lead, and tin dissolve in mercury and do notform amalgams. The influence of small amounts of lead and ofcadmium on the properties of zinc has been studied by N0vak.lGroup I11 A .The mono-, di-, and penta-borates of potassium, Sodium, andammonium are described by Atterberg,2 also barium monoborate andsesquiborate, 2Ba0,3B20,,7H,0 ; and Dukelski,3 from an investiga-tion of the equilibrium at 30° in the systems (1) caustic potash, boricacid, and water, (2) caustic soda, boric acid, and water, concludes thatsolid borates OF the following composition exist : K20,B20,,2QH,0 ;K,0,2B20,,4H20, K,0,5B,03,8H20, Na20,B203,4H20,Na209B203,8H207Na20,2B20, 10H,O, and Na,0,5B20,, 1 OH,O.Zinc and magnesiumperborates are produced by the action of boric acid and sodiumperoxide or of sodium perborate on zinc or magnesium salts; the zinccompound contains 9.5 per cent. of active oxygen, whilst the mag-nesium compound contains 11.9 per cent.4 Boron sulphide, B,S,, isformed by passing hydrogen sulphide over powdered f erroboron heatedat 300-400°, and condensing the product in a U-tube surrounded byiceas Ouvrard has prepared the chloroborates and bromoborates ofstrontium and barium similar to the calcium compounds mentioned inlast year's Beport (p.46).Silver ultramarine is formed by the action of silver nitrate andwater on ultramarine at 1 10-180° ; attempts t o prepare ultramarinecontaining organic radicles such as ethylene, napht hyl, and triphenyl-methyl have been unsuccessful.7 The blue cdour observed by Knapp inhis investigations of boron glass is due, as pointed out by Hoffmann,8 tothe production of a boron-ultramarine by the action of sulphur on thesulphides present in the materials employed. Blue compounds areformed by heating mixtures of boron trioxide and borax in hydrogensulphide or carbon disulphide.Group 111 B.A crystalline basic aluminium sulphate of the composition,4 0 3 , 2 so,,is manufactured by heating an excess of alumina with sulphuric acidZeit.anorg, Chmz., 1905, 47, 421.Ibicl., 1906,50, 38.Hoffmann, Zeit. nngew. @hem., 1906, 19, 1362.G'ompt. rend., 1906, 142, 281.ChabrG and Levallois, ibid., 1906, 143, 222,Zeit. a?zqew. Chem., 1906, 19, 1089.l b i d . , 1906, 48, 367.D.R.-P. 165278 a i d 16527948 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.(s. g. 1.475) ; the solution after treatment with calcium carbonate orlime is evaporated under reduced pressure.1 Although silicon andaluminium do not combine when heated together, yet in presence ofthe oxide or salt of another metal, compounds of silicon, aluminium, andthe metal, so-called silico-aluminides, are formed. Several such silico-aluminides are described by Vigouroux ; 2 they are metal-like, crystal-line, hard and brittle substances; the majority resist the action ofacids, except hydrofluoric acid, and are not attacked by solutions of thealkali hydroxides.J.Meyer draws attention t o sources of inaccuracy in the determina-tion of the atomic weight of indium in the volatility of the oxide athigh temperatures, and the incomplete decomposition of the nitrateeven a t llOOO.By treatment of an alkaline solution of n thallous salt withhydrogen peroxide a t low temperatures a dark brown, flocculentthallic oxide is formed, and at 80-100' it is produced as a black,heavy, sandy powder.4 The increase in weight observed when thallicoxide is heated over a bunsen burner arises from the formation of amixture of normal and thallous sulphates, produced by the action ofsulphur dioxide derived from the burning gas.5 Thomas6 has de-termined the heats of solution of the tetrahydrates of thallic chlorideand t h rllic bromide, also of the chlorobromides, T1C12Br,4H20 andT1ClBr2,4H20.Thallous chloride unites with chlorine a t the tempera-ture of liquid chlorine to form T12Cl,, and a t the ordinary temperatureTl,CI, is produced.The aluminium alloys which Pkcheux' has shown t o decomposewaters also precipitate cjpper from solutions of copper sulphate.Thallium is only partially miscible with copper at 954"; these metalsform neither chemical compounds nor mixed crystals.Thalliumbehaves in a somewhat similar manner with aluminium ; withantimony it is miscible in all proportions.1° Aluminium and tin donot combine chemically with one another; neither do aluminium andbismuth ; further, the two latter metals are only slightly soluble in oneano t her.llGroup I V A .A discovery of considerable interest and importance is that of anew gaseous oxide of carbon, which Diels and Wolf have found1 Spence, D.R.-P. 167.119.3 &it. anorg. Chem., 1905, 47, 281.7 Ibid., 575.9 Doerinckel, Zeit. anorg. Chew., 1908, 48, 185,Cowapt. rend., 1905, 141, 951.Rabe, ibid., 1906, 48, 427.Compt. rend., 1906, 142, 838.Ann. Report, 1905, 46.11 Gwyer, ibid., 49, 311.l b i d . , 50, 158.Williams, ibid., 50, 127INORGANIC CHEMISTRY.49amongst the products of the decomposition of ethyl mslonate byphosphorus pentoxide at 3 O O O . l Its composition is represented by theformula C,O,; i t is combustible, burning with a blue, smoky flame toform carbon dioxide ; by its reaction with water, ammonia, and hydro-chloric acid, malonic acid, malonamide, and malonyl chloride areformed. To the name, carbon suboxide, proposed by its discoverers,lBerthelot 2 objects, as that name has been already appropriated forcompounds discovered by Brodie in 1873, and examined by Berthelothimself. The constitutional formula 0:C:C:C:O has been assignedto this substance, but Michael3 considers it t o be the lactone ofC /Y P-hydroxypropiolic acid, for which the formula C 0 is suggested.\/The production of lamp black and graphite by the explosive decom-position of acetylene under a pressure of six atmospheres is describedby Frank,4 who points out that the carbon resulting from the explosionof a mixture of acetylene and carbon monoxide or dioxide is entirelyfree from oily matters associated with the carbon obtained fromacetylene alone.The carbon so produced is of higher electrical con-ductivity than ordinary charcoal, and forms a dense black pigment ofhigh covering power. Calcium carbide heated in a current of carbonmonoxide or dioxide yields graphite and lime ; strontium and bariumcarbides behave similarly. Commercial calcium carbide containsgraphite, representing carbon which has been dissolved by the fusedcarbide and separates out as graphite on cooling ; the proportion of dis-solved carbon depends on the intensity of the current used in the manu-facture of the ~ a r b i d e .~ Manville 6 has shown that the temperatures offorniatibn of carbon monoxide and dioxide by the direct union of theelements are influenced by the form of the carbon and the velocity ofthe stream of oxygen. When a given form of carbon has beenrepeatedly heated in a vacuum and cooled, the temperature at which itreacts with oxygen is materially changed.Pring and H ~ t t o n , ~ investigating the action of carbon and hydrogenon one another at temperatures ranging from 10003 to 2800", findthat methane is produced at all temperatures, and that the pro-duction of acetylene begins at 1870' and continues up to 2800'.I nthese experiments the carbon rods were heated electrically and thetemperatures estimated by the Wanner pyrometer. The reduction ofBer., 1906, 39, 689.Ber., 1906, 39, 1915.U a h , Conzpt. mzd., 1906, 143, 49.l'rans., 1906, 89, 1591.Compt. rend., 1906, 142, 533.Zeit. anyew. Chem., 1905, 18, 1733.IDitl., 1906, 142, 1190 xiid 1523.VOL. I". 50 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.water vapour by carbon monoxide, according to Gautier,l takes placea t 1200-1250", and equilibrium is established when the volume ofhydrogen is double that of the carbon monoxide, thus :3CO + 2H,O = 2C0, + 2H, + CO.The reductionof carbon dioxide proceeds a t 1300°, and the conditions of equilibriumare represented by the following equation :I n this reaction traces of formic acid are produced.CO, -I- 3H, = CO + H,O + 2H,.These observations explain the presence in volcdnic gases of carbonmonoxide and dioxide, of water and hydrogen, also of formic acid. Ithas been observed by Farup2 that carbon dioxide and water vapourreact with equal velocity 011 carbon at 850°, whereas oxygen attains thesame rate of reaction at 450'.When the diamond is heated in aporcelain tube in a current of carbon monoxide there is no reaction a t1000°, but a t 1250' carbon is deposited on the porcelain but not on thediamond ; thus the catalytic action of the porcelain is greater than thatof the diamond.A carbide of boron, BGC, is formed when petroleum coke and borontrioxide are heated together in an electric furnace; the carbide, BC,previously described by Muhlhiiuser, is probably a mixture of thecarbide B,C and grayhite.3By heating titanium containing some 2 per cent.of carbon in anelectric furnace with a current of 1000 amperes at 55 volts, Moissan4succeeded in distilling the metal. The distillate consists of minutecrystals and contains some nitride. Iron, chromium, manganese, andtungsten can be distilled in the electric furnace, and from theseexperiments Moissan estimates the temperature oE the sun to benearer 2000-3000° than 6950°, as estimated by W i l ~ o n . ~ Chloroformvapour passed over titanium dioxide heated in a hard glass tube givesthe tetrachloride ; the action is a complex one ; titanium, titaniumdichloride and tricbloride are formed, also a colourless, gaseoushydride, TiH,.Tin tetrachloride but not silicon chloride can beobtained in a similar way.6 Titanium silicide, TiSi,, and the corre-sponding zirconium compound are formed by heating together powderedaluminium, sulphur, fine sand, and either titanic acid or potassiumtitanofluoride, or the corresponding zirconium compounds ; the mixtureis covered with a layer of magnesium powder, and the action startedby a Goldschmidt pastile.7 Valuable observations on the autoxidationConzpt. rend., 1906, 142, 1382.J. Arner. Chetn. XOC., 1906, 28, 605.P ~ o c . Boy. SOC., 1902, 69, 312.Hb'nigschmid, Compt. rend., 1906, 143, 224.Zeit. anorg. Chem., 1906, 50, 276.Compt. rend., 1906, 142, 673.Renz, Ber., 1906, 39, 249INORGANIC CHEMISTRY.51of tervalent titanium compounds have been made by Manchot andKichter.l Titznium trioxide shaken up with caustic potash andoxygen absorbs more than an equivalent of oxygen ; hydrogen peroxideis formed first, and this converts the trioxide into titanium dioxide, orpossibly into pertitanic acid. When baryta is substituted for causticpotash the hydrogen peroxide formed is retained as such.Group I V B.The temperature of formation of crystalline carborundum appears toto be 1950°, and a t 2220' it is decomposed into carbon and silicon.2Amongst the products formed by heating lime and coke together inan electric furnace a silicide of iron, FeSi, and also one of the composi-tion FeSi,, have been obtained ; the iron was derived from the coke.3When solutions of sodium disilicate (Nn,Si,O,) are decomposed byhydrochloric acid so as to form solutions containing 1 per cent.ofsilica, the silicic acid so produced exists in two modifications, an U-and p-form; the former is not precipitated by egg-albumin, and isconverted into the latter by heat. Solutions of silicates of the typesR,SiO,, R,SiO,, and R2Si205 all yield solutions containing the a-silicicacid; solutions of water-glass yield the p-form in addition. Themolecular weight of the a-modification is estimated at 155, whilstaccording to Sabandeff the p-modification has a molecular weight of49000.4 Observations on the melting points of silicates and of the rateof reaction in fused silicates have been published by Doelter ; 5 fusedcomplex silicates form viscous masses from which crystallisation takesplace very slowly; consequently the first to separate out from such afused mass are the simpler silicates.The electrolysis of an aqueous solution of stannous chloride con-taining some hydrochloric acid with two tin anodes on either side of arevolving copper cathode gives a deposit of spongy tin.6 The freezing-point curves of mixtures of t i n and sulphur and of tin and seleniumshow that sulphur and selenium behave alike towards tin, which appearsto form with sulphur SnS, Sn2S3, and SnS,, whilst in the case of telluriumthere is only one break in the curve, which corresponds to the pro-duction of a compound, SnTeS7 The solutions of orthostannic acid insulphuric acid give with like solutions of calcium sulphate cubicalcrystals of the composition Sn(S0,),CaS0,,3H,0.S Similar compoundsRer., 1906, 39, 320 and 488,Tucker and Lampen, J. Amer.Chem, SOC., 1906, 28, 853,Vanzetti, Gazzetla, 1906, 36, i, 498.Mylius and Groschuff, Ber., 1906, 39, 116.Zeit. Elektrochem., 1906, 12, 413, and Moncctsh., 1906, 27, 4%.Tommasi, Compt. Tend., 1906, 142, 86. Pklabon, ibid,, 1906, 142, 1147,* Weinland and Kuhl, Ber., 1906, 39, 2951.E 52 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.are formed with the sulphates of strontium, barium, and lead ; they areregarded as derivatives of a compound, Sn[(SO,H),(OH),]M”, corre -sponding to the orthostannates of Bellucci and Parravano (see AnnualReporis, 1904 and 1905).By using a revolving cathode and low current density, lead can bedeposited as an adherent metallic film from solutions of the acetate.lA yellow and red modification of lead monoxide are described byLead peroxide is formed by the action of chlorine on amixture of magnesia and lead sulphate suspended in water ; the crudeproduct is treated with caustic soda and then with nitric acid.3 Crystal-line lead chromate (crocoite) is gradually deposited on exposing to air asolution of lead chromate in caustic soda ; an alkaline solution of leadmolybdate gives under similar conditions crystals of ~ u l f e n i t e .~ I nthe investigation of the freezing-point curves of mixtures of leadoxide and lead chloride the formation of PbCl,,PbO (matlockite),PbCl2,2Pb0 (mendipite), and of PbC12,4Pb0 is observed ; it has alsobeen noted that mixtures containing less than 88 per cent. of leadoxide may be fused in platinum vessels without attacking the p l a t i n ~ m .~The freezing-point curves of mixtures of galena and lead and themicrographical examination of the fused mixtures afford no evidenceof the existence of Pb,S or Pb,S;G further, in a similar way, leadsulphide and iron sulphide when fused together are shown not tocombine with one another.7 Alloys of lead and arsenic containing upt o 34.4 per cent, of arsenic have been prepared by Friedrich ; s theinvestigation of these alloys affords no evidence of the existence ofthe compounds described by Descamps 9 and Spring.Io Bock l1 has sug-gested that in the Pattinson process for the concentration of silver inlead, the iron of the vessels in which the metals are melted plays animportant part in promoting a separation of the metals.This sug-gestion is shown by Friedrich 12to be unfounded, as the separationis as perfect in porcelain as in iron vessels.Group V A .Erdmann l3 describes an apparatus for the condensation of largequantities of nitrogen, which can now be obtained comaercially pure ;Snowdon, J. Plqsical Chenz., 1906, 10, 500.Zeit. anorg. Chem., 1906, 50, 265.Cesliro, Bull. Acad. Roy. Belg., 1935, 327.Zeit. anorg. Chein., 1906, 49, 365.Friedrich and Leroux, Metallzwgie, 1905, 2, 536.Abstr., 1878, 705.l1 Chem. Zeit., 1905, 29, 1149.l 3 Ber., 1906, 39, 1207.3 Cheni.Centr., 1906, ii, 465.7 Weidmann, ibid., 1906, 3, 660. 8 Ibid., 41.10 I b i d , , 1883, 650.12 JfetnllrrTgic 1906, 3, 396INORGANIC CHEMISTRY. 53dry oxygen is liquefied by cooling with Iiqiiid nitrogen, but not whencooled with liquid air. Ozone dissolves in liquid nitrogen, forming ablue solution. The production of oxidised products of nitrogen hasengaged the attention of many workers. Schmidt and Bockerl findthat when ammonia and air or oxygen are passed over platinum orplatinised asbestos heated to redness, some 75 per cent. of theammonia is oxidised to nitric oxide, which is subsequently convertedinto nitrogen trioxide. The products OF the electrolysis of ammoniain presence of caustic soda are influenced by the nature of the anodeemployed : with platinum electrodes, sodium nitrate, a little nitrite,nitrogen and oxygen are formed; when the anode is either of copper,nickel, iron, or cobalt, sodium nitrite, nitrogen, and oxygen are aloneproduced.2 The formation of nitrite has also been observed to takeplace when a solution of ammonia is electrolysed in presence of acopper salt and an alkali.3'I'he production of alkali nitrites by passing a mixture of air oroxygen and ammonia over heated metallic oxides is the subject of apatented process.With ferric oxide at 700' a continuous current ofnitrous gas is obtained, which is converted into sodium nitrite byabsorption in caustic soda.4 I n discussing the fixation of atmosphericnitrogen, Guye5 points out that, whilst the yield of nitric oxide israised by employing high temperatures, difficulties arise from thedecomposition of this gas at such temperatures.The best effect isobtained when the gases are rapidly swept out of the region of the arc,or where a device is employed by which the arcs are rapidly lighted andinterrupted several times a second, or again by causing the arc to playin different regions of the gases. The gases as they pass out oE thevessels in which they have been submitted to the action of the electricarc contain from 1 to 2 per cent. by volume of nitric oxide, and oncooling to 500' or 600°, nitrogen peroxide and trioxide are formedwhich are absorbed by water. Under the influence of the silentelectric discharge the nitrogen and oxygen of the air in presence ofwater or an alkali combine, forming nitric acid or nitrates, thus :2N, + 50, + 2H20 + Aq = 4HN0, (dil.).Neither nitrous acid nor ammonia is formed; the reaction isindependent of the relative proportions of the two gases, and proceedsuntil the whole of the oxygen is used up (Berthelot).6 The oxidationof nitrogen by the passage of silent electric discharge through air hasalso been investigated by Warburg and Leithauser.7Bcr., 1906, 39, 1366.Miiller and Spitzer, Zeit.Elektrochem., 2905, 11, 917.J. SOC. Chenz. Ind., 1906, 25, 567.Ann. Phpsik., 1906, [iv], 20, 743.Tranbc and Biltz, Ber., 1906, 39, 166. D.R.-P. 168272.Conzpt. rmd., 1906, 142,136754 ANNUAL REPORYS ON THE PROGRESS OP CHEMISTRY.Kuriloff 1 concludes that the compounds of ammonia with metalliccompounds are of three classes : (i) those of constant compositioa, inwhich the proportion of ammonia to the nietallic compound is simple;(ii) those in which the composition is not constant, containing avariable amount of ammonia to one molecule of the metalliccompound; (iii) colloidal class, in which thore is no relationshipbetween the ammonia and the metallic salts.The second class areanalogous to hydrates, and the third class to hydrogels.Raschig 2 describes tt potassium hydroxylamineisodisulphonate(SO,K*NH*O*SO,K), obtained from acid solutions of potassium hydr-oxylaminetrisulphonate, this compound is regarded as a derivative ofthe amide of Caro’s acid, namely, H,N*O*SO,H. Ruff and Stiduber 3 givea full account of their investigation of nitrosyl fluoride referred to inlast year’s Repovt.The chlorides of tin, antimony, and molybdenumreact with nitrogen sulphide dissolved in chloroform, the additivecompounds, SnC1,,2N4S,, SbC15,N4S4, and MoC15,N4S4, being formed ;whereas the chlorides of titanium and tungsten are first reduced andthen the reduced chlorides combine with the sulphide to formTi,CI,,N,S, and WC14,N,S,.4 The action of liquid ammonia on solidnitrogen peroxide is one of great violence, but when moderated bypassing ammonia gas over nitrogen peroxide a t - 20°, the products ofthe interaction are nitrogen, nitric oxide, water, ammonium nitrate, anda trace of nitrite, Nitrogen peroxide reacts slowly with ammoniumchloride in the cold, but a t lOOOin sealed tubes the action is complete andresults in the production of chlorine, nitrogen, nitrogen monoxide andtrioxide, nitrosyl chloride, water, and nitric acid ; ammonium nitrateand sulphate are also attacked by nitrogen p e r ~ x i d e . ~ I n connexionwith the removal of nitrous acid from concentrated nitric orsulphuric acids, Silberrad and Smart find that carbamide, leadperoxide, oxamide, methylamine nitrate, or aminoguanidine nitratereact but slowly with nitrous acid i n presence of these acids, whereasthe action of hydrazine sulphate is very violent.DelBpine7 drawsattention to the possible error involved in the use of platinum wireor foil in the Kjeldahl process arising from the decomposition ofammonium sulphate, due to the following reactions :4H,SO, + Pt = Pt(S04), + 2S0, + 4H20SPt(S0,) + 2(NH,),S04 --- 2N, + 3Pt + 8H2S0,.The formation of nitric acid by the interaction of silver sulphate,ammonium persulpbate, and dilute sulphuric acid is recorded by Kempf .*Ann.CJ~ivz. Phys., 1906, [viii], 7, 568.Zeit. anorg. Chem., 1905, 47, 190, 192.J. SOC. Chem. Ind., 1906, 25, 156.Ber., 1906, 38, 3963.Bey., 1906, 39, 245.Trans., 1906, 89, 1575.Conipt, rend., 1905, 141, 886.rj Besson and Rosset, Comnpt. repid., 1906, 142, 653INORGANIC CHEMISTRY. 55Group V R .A convenient method of preparing phosphoric di-iodide is describedby Doughty.1 The conclusions drawn by Giran2 as to the existenceof various sulphides of phosphorus, based upon the behaviour oncooling of mixtures of these elements, are adversely criticised byBouIouch,S who maintains that the trisulphide does not exist. Stock,*in conjunction with others, has made a study of the action of liquidammonia on phosphorus pentasulphide (see Report, 1905), aninvestigation which has resulted jn the isolation of a numberof phosphorus compounds.The additive product, P2S,,7NH,, is amixture of ammonium iminotrithiophosphate, P(SNH,),:NH, andammonium nitrilodithiophosphabe, P( SNH,),iN. From the former,other salts of iminothiophosphoric acid, for example, SH*P(SNH,),:NH,have been prepared, and by the action of water on it ammonium thio-phosphate, PO(SNH4),,K20, is obtained. When heated at 180° in avacuum, the ammonium iminothiophosphate gives thiophosphoricnitrile, NiP:S, which has also been produced by heating ammouiumnitrilodithiophosphate at 300' in a vacuum.Many other derivativesof these compounds are described in the papers referred to.Phosphorus chloronitride, PCI,N, is the product of the interactionof phosphorus pentachloride and ammonium chloride ; i t is insolublein water, but soluble in light petroleum and several organic solvents,and is dissolved by phosphorus oxychloride, sulphur dioxide, ornitrogen peroxide. A cryoscopic determination gives a molecularweight corresponding to the formula (PCI,N),.5 From the measure-ments of the electrical conductivity, Parravano and Alarini 6 attributeto sodium hypophosphate the formula Na,H,P,O,, whilst Rosenheim,Stadler, and Jacobsohn 7 from their determinations deduced the mole-cular formula NaHPO,.From a mixture of a phosphate and sugaron treatment with siilphuric acid, the phosphorus can be completelyvaporised by heat.8 Hydrogen phosphide is evolved from electrolyticferrosilicon by treatment with water; the fatal cases of poisoningwhich have occurred on boats carrying ferrosilicon and the explosionsin cargoes of this material in Liverpool in 1904 are to be attributedto the hydrogen phosphide formed in this way from ferrosilicon.9The examination of the different methods for producing arsenic andsulphur compounds shows that only realgar and orpiment are produced,and that the pentasulphidc, As2SS, is not forrned.l0 Arsenic penta-J. A m e y . Cham. Soc., 1905, 21, 1444.Ibid., 1045, n~id 143, 41.C ' o ? ~ p f .r e d . , 1906, 142, 398..I &r., 1906, 39, 1967.Bey., 1906, 39, 2857.r, Bessoii a i d Rosset, Cmnpt. md., 1906, 143, 37.fi Atti X. Accad. Lincei, [v], 15, ii, 203.lo Borodowski, Chenz. Cclttr., 1906, ii, 297.Ibztl., 1906, 39, 2625. Zcit. ATnJw. Genzcsmn., 1906, 12, 13256 ANNUAL KEPOKTS ON THE PROGRESS OF CHEMISTRY.fluoride (AsFJ is formed by the action of bromine and antimonypentafluoride on arsenic trifluoride. It is a colourless gas forminga yellow liquid a t - 50' and a white solid at - 80'; when dry, it doesnot attack glass, and when heated with silicon, arsenic and siliconfluoride 1 are produced.A black and a yellow modification of antimony are produced by theaction of oxygen on liquid antimony hydride a t -40' and -90'respectively.2 The selenide (Sb,Se,) and telluride (Sb,Te,) areformed by fusion of antimony with selenium or tellurium ; thesecompounds form liquids when fused with either constituent, a propertyby which cryoscopic constant for antimony has been determined.3ChrBtien 4 maintains that other selenides are formed,and gives 628' asthe melting point of antimony.Antimony sulphate, Sb,(SO,),, with water yields the basic salt,(SbO),SO, ; with alcohol, the compound Sb2O3,2S0, is formed.Withalkali sulphates, double salts of the formula Sb,(SO,),,M',SO, areformed ; the dissolution of antimony or antimony sulphide in sulphuricacid is greatly facilitated by the addition of an alkali sulphate.5 ThisIsst observation is applied in a patented process for the preparation ofantimony oxide from antimony sulphide.6The double compounds BiCl,,BKCI and BiBra,2KBr, also.BiBr,, 2 H Br, 4H,O,are described by hloy and F r e b a ~ l t .~ Gutbier and Bunz,8 investigat-ing the so-called peroxides of bismuth, show that the product formedby the oxidation of bismuth oxide by chlorine in presence of analkali is not a uniform material, nor is that which is formed by theoxidation with alkaline potassium ferricyanide as described by Hauserand Vanino (see Report, 1904) ; further, there is no evidence that theseproducts possess acidic properties.By the electrolytic reduction of vanadyl sulphate, vanadium hydrogensulphate, Vd,(SO,),,H2SO,,12H,O, is formed ; it gives double sulphateswith those of rubidium and amm~nium.~ Rutter lo finds that vanadicacid cnn be reduced electrolytically to vanadous sulphate by using amercury cathode, and when a platinised platinum cathode is used thevanadic acid is reduced to a vanadic salt.Vanadous salts reducesilver bromide to metallic silver, and vanadic salts are oxidised bysilver sulphate, a reaction assisted by copper sulphate, which is, how-l Ruff, Graf, and Heller, Ber., 1906, 39, 67.Stock and Siebert, ibid., 1905, 38, 3837.Pdabon, Conapt. Tend., 1906, 142, 207. Ibid., 1906, 142, 1339 and 1412.Metzl, Zeit. ~tnorg. Chenz., 1906, 48, 140. D.E.-P., 161776.Bdl. SOC. cJiinL., 1906, iii, 35, 346.Zeit. n n o ~ g . Chewz., 1906, 45, 162, 294 ; 49, 432 ; a i d 50, 210.Stihler and VTirthwein, Ber., 1905, 38, 3978.lo .%it.h'lektrochena., 1906, 12, 230lNORGANIC CHEMISTRY. 57ever, not reduced to metallic copper. The production of ammoniumvanadate and sodium uranate from Utah sand, containing carnotite,is described by Ohly.1Double salts of columbium oxychloride and oxybromide of the typesCbOCl,,RCl and CbOC13,2RC1 are formed either by adding the halidesto solutions of columbic acid in hydrochloric or hydrobromic acids, orby adding the halides dissolved in alcohol with hydrogen chloride orbromide t o alcoholic soiutions of the oxychloride o r oxybromide ofcolumbium. Caesium, rubidium, quinoline, and pyridine compoundshave been prepared.?G~ozcp V I A .Many workers have busied themselves with the study of theproblems connected with the conversion of oxygen into ozone, Thus,Fischer and Braehmar have succeeded in demonstrating the formationof ozone and of nitrogen trioxide in the burning of hydrogen, carbonmonoxide, acetylene, sulphur, and charcoal below or a t the surface ofliquid air.Ozone is also formed when platinum heated electrically toa white heat is submerged under liquid air or oxygen. Further, i t hasbeen observed that the production of ozone is favoured by rapid heat-ing followed by rapid cooling, whereas the formation of nitric oxidetakes place more readily when these conditions are reversed.* As to theinfluence of the form of the electrode on the yield of ozone, Warburgand Leithiuser find that for small concentrations (4 grams of ozoneper cubic metre) a highly positively charged sphere is the best, and forhigh concentrations (9 grams per cubic metre) a negatively chargedsphere.Moisture reduces the yield of ozone, whereas a rise intemperature up to 80’ has but little effect on the yield from oxygen,but produces a considerable reduction in the yield of ozone fromair. Chassy 6 has observed that with a given discharge reduction ofpressure lowers the yield of ozone; observations on this subject arealso recorded by Cermak.7 The formation of ozone in the electrolysisof solutions of hydrofluoric acid and of potassium fluoride describedby PrideauxS is easily explained in the light of the properties offluorine.The amount of ozone found by Lespieau9 in the air above theglaciers of Mont Blanc is 4.5 milligrams per 100 kilograms, and isshown to be independent of the altitude.A redetermination of the density of ice made from water speciallyWeiiiland and Storz, Bw., 1906, 39, 3056.Ibid., 2857.C’lzem.Cknlr., 1906, ii, 166.Bw., 1906, 39, 940.F, Ann. Physik., iv, 1906, 20, 734 and 751.Chem. Ccntr., 1906, ii, 585.BulZ. SOC. chim., 1906, [iii], 35, 616.ti Compt. mizd. , 1906, 143, 220.Trans. F’nradny S’oc., 1906, 2, 3458 ANNUAL REPORTS ON TlIE PROGRESS OF CHEMISTRY.freed from gases by boiling gives the value 0.91752, whereas Bunsenfound it t o be 0.9176.1Doring 2 finds that chromium, prepared by the alumino-thermalmethod, containing some 97 per cent. of this element is less readilyattacked by the halogen hydracids than those preparations containinga smaller proportion.Chromium can be electrolytically depositedfrom violet solutions of chroma alum, but not from the green solu-tions.s The deposition of the metal from chromic salts is not neces-sarily preceded by the reduction to a chromous salt. hasextended his investigations of the sulphates of chromium referred toin last year’s Report, and describes a green compound,which is formed by reducing with sulphur dioxide n mixture of 2molecular quantities of chromic acid and 3 molecular quantities ofsulphuric acid dissolved in its own weight of water. A green chloro-sulphate, CrCISO4(5H20),3H,O, and a violet compound,are described by Weinland and Krebs; the latter compound reactswith silver nitrate solutions, whilst the former does not.Bjerrum,Gdoes not agree with the conclusions of Weinland and Krebsas to the constitution of the chlorosulphates of chromium, forwhich in a later communication these authors have adopted differ-ent f o r m u l ~ . ~ It is only possible to refer to the researches ofWerner and his pupils on the complex chromium compounds; thespace at disposal in a Report of this kind does not suffice to dojustice t o the results of these important investigations,* nor t oPfeiffer’s 9 discussion of the isomerism exhibited by complexchromium compounds. That solutions of chromium trioxide containH2Cr,07 and H2Cr0, is supported by ebullioscopic and cryoscopicmeasurements, but no such evidence is forthcoming from electricalconductivity deterrninations.1° From experiments on the oxidisingaction of chromic acid and tho results of the study of the action ofdilute solutions of this substance on potassium iodide and hydro-chloric acid, Seubert and Carstens11 conclude that, as solutions ofchromic acid contain Cr,O,” and CrO,” ions, also undissociated CrO,, itmay be assumed that the reaction is represented as follows : CrO,+H’ + I’ = CrO,I*OH, the complex CrO,I*OH reacting with hydriodicacid to form iodine, water, and a chromic salt.ColsonCr,(OH)2(SO,H),,1 OH@,[CrSO,(6H,0),2H,O]Cl,Lednc, Compt.wmL, 1906, 142, 149.:: Zeit. Elektrochem., 1906, 12, 329.Zeit. anorg. Chem, 1906, 48, 251.7 Zeit. anorg. Chenz.,~,1906, 49, 157.Aznalen, 1906, 346, 28.J. pr. Chcm., 1906, [ii], 73, 393.Compt.?end., 1906, 142, 402.Bcr., 1906, 39, 1547.Ber., 1906, 39, 329, 1823, 2656.lo Costa, Gaxzetta, 1906, 36, i, 535.l1 Zeit. anorg. Chern., 1906, 50, 53INORGANIC CHEMlSTRY. 59Alloys of molybdenum and boron are formed by the reduction ofmolybdenum dioxide heated with boron in magnesia crucibles. Thealloys are non-crystalline, the hardness increasing with the percentageof boron ; the highest proportion of boron found in these alloys being 46per cent. These substances are not attacked by hydrochloric or hydro-fluoric acids or alkalis, but are dissolved by dilute sulphuric andnitric acids, and are acted on by fluorine and chlorine at a red heat.1By heating together molybdenum and manganese at 1500O alloys areformed, and alloys richer i n molybdenum are produced by reducingthe mixed oxides with aluminium pomder.2 These alloys are silverwhite, non-magnetic, hard, brittle solids, consisting of mixtures ofmanganese with one or other of the compounds MnzMo, MnMo, orMnMo2.Sodium silicomolybdate is formed by heating a mixture ofsodium silicate and molybdic acid in the proportion1.2 Moo, : SiO, : 2 Na20,with a little water in sealed tubes a t 150'. From the sodium salt nseries of metallic silicomolybdates have been prepared, which areusually isomorphous with the corresponding silicotungstntes. Thecrystals of potassium silicomolybdates and silicotungstates are optic-ally active; dextro- and hvo-crystals have been obtained; the solu-tions are optically inactive.3It has also been observed by Copaux that the acidSi0,,12Rfo03,2H,0,31H20,and the lithium salt,, SiO,,l2Mo0,,2Li,0,29H,O, form mixed crystals,and, further, that the acid, after the loss of 7H,O, forms mixedcrystals with the barium salt, SiO,,l 2Mo03,2Ba0,22H,0.Theauthor concludes that isomorphism is dependent upon crystalline formrather than chemical composition.The following compounds of iron and molybdenum, Fe,Mo, Fe3M04,FeMo, and FeMo,, have been isolated by Vigouroux from the productof the reduction of the mixed oxides by the Goldschmidt method, alsoby heating a mixture of the finely-divided metals in B stream ofhydrogen.Group VI B.I n the formation of polythionic acids by the interaction of sulphurdioxide and hydrogen sulphide, a hydrate of sulphur, S,,H,O, is formed,and not a new allotropic form of sulphur, as suggested by Debus in 1888(Spring).6 Sulphur heated in sealed tubes at 150-180' with aqueoussolutions of cupric chloride reduces i t to cuprous chloride, and underAi*i*ivant, Coinpt.T e n d . , 1906, 143, 285.Chem. Centr., 1906, i, 1673.Ibitl., 464.Cowpt. mid., 1906, 142, 889 and 928.3 Copaiix, Anu. chinz. phys., 190G, [viii], 7 , 118.fj Xec. t m v . chim., 1906, 25, 25360 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRYsimilar conditions reduces potassium dichromate to chromium sesqui-0xide.lThe variations in the vapour pressures of mixtures of sulphurchloride (S,CI2) and of chlorine, and of the boiling points also,confirm the existence of the dichloride SCl,, a conclusion supportedalso by dilatometric measurements., Briickner 3 finds that whenanhydrous metallic sulphates are heated with sulphur in glass tubesthrough which the vapour of sulphur is passed, complex reactions takeplace, which in the cam of the sulphates of the alkali metals and ofthe alkaline earth metals result in the production of sulphur dioxide,and a mixture of sulphide and thiosulphate of the metal.Magnesium,aluminium, and glucinum sulphate do not react with sulphur at a redheat; chromium sulphate gives a black Cr,S, not attacked by hydro-chloric acid ; the sulphates of other metals yield sulphides and sulphurdioxide.Lunge 4 defends his theory of the lead-chamber process for the manu-facture of sulphuric acid against the criticisms of Raschig, and main-tains that the production of nitrous oxide and nitrogen by the actionof sulphuric acid on nitrogen peroxide does not take place; further,that nitrogen trioxide is not an intermediate step in the conversion ofnitric oxide to nitrogen peroxide.The following equations representthe changes taking place in the production of sulphuric acid.I. SO, + NO, + H20 = SO,NH, (sulphonitronic acid).The sulphonitronic wid is decomposed by oxygen or nitrogen per-oxide, forming nitrosylsulphuric acid (S0,NH).11. BSO,NH, + 0 = H,O + 2S0,NH.111. 2S05NH2 + NO, = H20 + 2S0,NH + NO.The nitrosylsulphuric acid is then decomposed, thus :IV. 2S0,NH + H20 = H,SO, + NO + NO,.V. ZS05NH + SO, + 2H,O = H,SO, + ZSO,NH,.The nitric oxide is finally oxidised to nitrogen tetroxide.Examining the question of the loss of nitrogen in the lead-chamberprocess, Inglis5 finds that the reduction to nitrous oxide is veryslight, and the chief source of loss of nitrogen compounds is theinsufficient absorption in the Gay-Lussac tower.Selenium is found inthe manufacture of sulphuric acid from pyrites containing this elementin the elementary form, as selenious acid and as the compound SeSO,,which dissolving in sulphuric acid imparts a green colour to the acid.Zeit. Yhysik., 1905, 54, 55,Zed. anorg. Clwn., 1906, 19, 807, 857, and 881.AZti 22. Accad. Limei, 1906, [v], 15, i, 703.Alo.i.zntsh., 1006, 27, 49.J. Xoc. Chem. Ind., 1906, 25, 149INORGANIC CHEMISTRY. 61Ten grams of selenium per ton of pyrites is suficient to show itself inthe process; the red sludge collecting in the Glover tower contains2-4 per cent.of selenium. Littmannl concludes that the mainquantity of selenium escaping from the pyrites exists as a labilevolatile compound of a low stage of oxidation, possibly SeO, which isreadily reduced to selenium or oxidised t o selenium dioxide. Seleniumcan be readily detected in sulphuric acid by adding a crystal ofpotassium iodide t o the diluted acid and removing the excess of iodineby sodium thiosulphate ; selenium remains suspended in the solution,imparting to it a red coloration, changing to a lemon-yellow.The most favourable conditions for the electrolytic production ofthiosulphates from sulphite solutions have been investigated byVoghera,g who finds that the cathodic fluid should contain poly-sulphides rather than sulphides, the anode should be platinisedplatinum, the anodic solution a concentrated sulphite, faintly alkaline,and further a low anodic density should be maintained.concludes that Caro's permonosulphuric acid is H,SO,, not PriceOH R2S209, and is represented by the constitutional formulat h a t of perdisulpliuric acid being SO,<o o>SO,.OH HOThe statement that the reaction H,S + ZnC1, ZnS + 2HC1 may bedisplaced towards the right or the left by alteration of the externalconditions has been examined by bringing the solutions into contactwith hydrogen sulphide under a pressure of 14-16 atmospheres.Solutions of several metallic salts which under ordinary conditionsgive no precipitate gave precipitates when under pressure ; whereaswhen the pressure is reduced by cooling the mixtures in solid carbondioxide and ether, in no case did the precipitate dissolve, thus demon-strating that the reaction is not displaced to the left.4Gautier 5 finds that metallic sulphides are atkicked by steam at teni-peratures varying from incipient redness to a white heat; sulphurdioxide, hydrogen, sulphur, and in some cases the metals themselvesare formed.Hydrogen sulphide saturated with ste-tm when passedthrough red-hot tubes gives sulphur dioxide, sulphuric acid, andcolloidal sulphur. The presence of sulphur dioxide i n volcanic gases isregarded by Gautier as due to the action of steam on metallic sulphides.The same authority has also studied the action of hydrdgen sulphide onseveral oxides at high temperatures, showing that ferric oxide yields ironsulphide, alumina an oxysulphide, A1,0,, AI,S,, and silica a compound,Zeit.ungew. Clwn., 1906, 19, 1039.Uruiii and Padoa, Atti R. Accnd. Liilte:i, 1905, [v], 14, ii, 525.Compt. reml., 1906, 142, 1465 and 143, 7.Atti R. Accnd. Littcei, 1906, [v], 15, i, 363. ' TI'UIZS., 1906, 89, 5368 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.SiO,,SiS,, whilst carbon dioxide and hydrogen sulphide passed togetherthrough a white-hot porcelain tube yield steam, hydrogen, carbonmonoxide, sulphur, and carbonyl sulphide. The explanation of thepresence of carbonyl sulphide in volcanic gases and in sulphur springsis to be found in this reaction.For the isolation of selenium from sulphuric acid chamber residuesthe treatment with concentrated sulphuric acid at 50-60°, oxida-tion with solid potassium permanganate, and reduction with sulphurdioxide has been patented by Koch;1 by heating amorphousselenium a metallic, crystalline, grey variety is formed, which exists asa labile and stable modification ; both conduct electricity.2 The reduc-tion of selenious and selenic acids to selenium by organic compoundssuch as formic and oxalic acids, formaldehyde, benzaldehy de, dextrose,and others, has been observed by Oechsner de Coninck and C h a ~ v e n e t .~Oechsner de Coninck has also determined the solubility of seleniumdioxide i n several solvents ; when treated with concentrated sulphuricacid, i t yields SeSO,, hydrogen selenide and some amorphous selenium.Phosphorus pentachloride converts the dioxide into selenium tetra-chloride (SeCl,), whilst phosphorus trichloride gives brown amorphousselenium ; both hydrazine and hydroxylsmine hydrochloride reduceselenium dioxide, the former giving a black amorphous variety andthe latter a red modification, The aqueous solution of the dioxide onexposure to light gives a deposit of selenium, insoluble in carbon di-sulphide.Colloidal solutions of selenium and of sulphur can be obtained byelectrolysing water with a cathode consisting of selenium or sulphurfused on to platinum foil ; the selenium solution is fiery red, that ofsulphur milky-~hite.~ With cathodes of this kind Le Blanc findsthat in the electrolysis of solutions of caustic potash, polysulphides andpolyselenides are formed, but these elements do not dissolve whenemplojed as anodes.Tellurium, however, dissolves both as cathodeand anode, in the first case to form polytellurides, and in the second itdissolves as Te"", the "ions " reacting to form TeO," ions. At thecathode Te" ions and a t the anode Te"" ions, in solutions theequilibrium between these ions is 3Te t 2Te" +Te"". The crystal-lography of the compounds TeBr,Ph2 and SeBr,Ph2, also of therubidium hydrogen selenate and tellurate, shows selenium and telluriumto be isomorphous, a conclusion supported by the study of the solidify-' D.R.-P. 167457.Marc, Ber., 1906, 39, 697 ; also Zeit.anorg. Chent., 1906, 50, 446.Compt. rend., 1906, 142, 571.3 Bull. Acnd. Xoy. Belg., 1906, 67 and 601.5 Miiller and Nowakowski, Ppr., 1905, 38, 3779.ti Zeit. Elektrocheiiz., 1905, 11, 813, and 1906, 12, 619INORGANIC CHEMISTRY. 63ing points of mixtures of these elements in vapging proportions(Pellini).'G~oup 1'11 A.Beyond what has already been referred to in the discussion of the othergroups, there remains to mention Doerinckel's study of the freezing-point curves of mixtures of manganese and silicon, which indicate theexistence of the compounds Mn,Si, MnSi, and MnSi,, conclusionswhich confirm the observations of Lebeau based on the results of theaction of different solvents on alloys of manganese and silicon.Group V l I B.Fluorine dissolves in liquid chlorine, but reacts with chlorine water,producing hydrofluoric and hypochlorous acids (Lebea~).~ D e ~ s s e n , ~continuing the examination of the properties of hydrofluoric acid,referred to in last year's 12epoi.t (p.SU), finds the constant boilingsolution of this compound in water has a boiling point of 111' a t750 mm. and contains 43.2 per cent. of hydrogen fluoride.Attention is drawn by Fabinyi and Forster5 to differences in colour,solubility in water, and reactivity with hydrogen of chlorine pre-pared by allowing sulphuric acid to drop on to a mixture of potassiumdichromate and salt, and of chlorine obtained by adding salt to amixture of potassium dichromate and sulphuric acid. This difference,especially the greater activity of chlorine prepared by the secondmethod, it is suggested, may be explained by assuming the chlorineatoms to be constituted of a different slumber of electrons.Thefunction of the catalyst in the Deacon process for the manufacture ofchlorine appears to be in part catalytic and in some degree chemical,and dependent on the formation of unstable hydrates by the catalyst.GLiquid chlorine can be usefully employed in the preparation of iodinetrichloride by causing it t o react on iodine or metallic iodides; withbromine it forms both the mono- and tri-chloride, according to theproportions employed, Sulphur is not acted on by liquid chlorine, butthe di- and tetra-chlorides of selenium are formed by its action onselenium ; arsenic is converted into the trichloride, but antimony andbismuth are not attacked.Liquid chlorine, further, converts thallousinto thallic chloride, but is without action on carbon disulphide, leadand manganese chlorides, or potassium permanganate. 7From an elaborate and thorough investigation of the interaction ofAtti a. Accccd. Lincei, 1906, [v], 15, i, 629 and 'ill ; ii, 46.Zed. anorg. Chem., 1906, 50, 117.Zeit. nnorg. Chem., 1906, 49, 297.Levi and Voghers, Gnzxetta, 1906, 36, i, 513 ; Ann. Report, 1905, 59.Compt. rend., 1906, 143, 425.Chem. C'entr., 1906, i, 636.7 Thomas and Dopuis, Compt. r e d . , 1906, 143, 28264 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.chlorine and hydrogen Burgess and Chapman 1 conclude that there isno evidence of a photochemical induction with moist chlorine andhydrogen, in the absence of impurities which are capable of destruc-tion under the conditions of experiment.The photochemical inductionnoted by other experimenters is attributed by the authors to thepresence of traces of ammonia or of compounds yielding ammonia. Itis also maintained that there is no sufficient evidence to justify theview that the action of these elements is contributed to by the forma-tion of condensation nuclei. The action of light in causing the unionof chlorine and hydrogen appears to be due to the light absorbed by thechlorine becoming degraded into hest, and in this degradation a formof chemically active energy is produced. A method for the preparationof hydrogen chloride and bromide has been patented by Hoppe ; 2 thismethod depends on the hydrolysis of certain metallic chlorides result-ing in the formation of the hydracid and a basic salt of the metal,which is then converted into the corresponding halide by suspendingit in water and treating it with a mixture of hydrogen and thehalogen (chlorine or bromine) ; thus from the basic zinc chloride thenormal chloride is regenerated in accordance with the equation :2ZnClOH + H, + C1, = 2ZnC1, + 2H,O.Bray has published several papers dealing with the oxyhalagencompounds, and especially with the reactions of chlorine peroxide andchlorous acid.From the study of the oxidation of a n iodide bypotassium permanganate, hypochlorous acid, ozone, hydrogen peroxide,iodic acid, potassium persulphate,4 and arsenic acid, it is concludedthat in all cases of oxidising agents containing oxygen the first productis hypoiodous acid or a hypoiodite, and probably similar compounds areformed in the oxidation of other halogens.The cryoscopic examina-tion of aqueous solutions of chlorine peroxide indicate that thissubstance forms a hydrate, C10,,8H20( +, H20), analogous to chlorinehydrate; like chlorine, it forms n compound with water and carbontetrachloride. I n the reduction of chlorine peroxide a chlorite is aprimary product ; the reaction may be represented thus :2C10, + H,O + C1’ = 2 HCIO, + CIO’,the chlorous acid further decomposing into chloric acid and hydro-chloric acid, also into chlorine and chlorine peroxide, thus :4HC10, = +Cl2 + 3C10, + 2H,O.Tmns., 1906, 89, 1899.D.R.-P. 166598.3 Zeit. physiknl. Chern., 1906, 58, 463, 569, and 731 ; also Zeit. nnoyg. Chem,See also Merk, Cl~zem. Cent?-., 1906, i, 397. 1906, 48, 217INORGANIC CHEMISTRY. 65The study of the reactions of chlorine peroxide shows that in thedark with water it forms chloric acid and hydrochloric acid :6C102 + 3H20 = 5HC10, + HC1.This action is assisted by the presence of chlorine ions and bymetallic platinum, but not by chlorate ions. I n alkaline solutions Bmixture of chlorate and chlorite is formed. Chloric acid and hydro-chloric acid,l when the concentration of chlorine ions is small, reactas follows :C10,’ + C1’ + 2H = C10, + iC1, + H20,whereas with relatively concentrated hydrochloric acid the reactionproceeds thus :HC10, + 5HC1= 3C1, + 3H20.The action of chlorine peroxide on iodides in solutions of alkalihydrogen carbonates saturated with carbon dioxide is :(1) c10, + 1’ = C10,’ + I ;in neutral solutions the following change takes place :3c10, + 51’ = 210, + 31 + 3C1’,whereas in acid solutions the reaction is represented by the equation :C10, + 51‘ + 4H’ = 51 + Cl’ + 2H,O.In all cases the first action results in the production of hypoiodousions, which reacting with iodine and hydrogen ions give free iodineand water, thus :10’ + I’ +2H’ = I, + H20.I n presence of a large proportion of acid, then, reaction (1) is followedby the action of chlorous acid on iodine ions, forming iodic acid andhydrochloric acid :3HC10, + 21’ r= 210,’ + 3HC1.The oxidation of iodine to iodate takes place in the followingreactions :C10, + I + H,O = HIO, + HC15HC10, + 41 + 2H20 = 4HI0, + 5HC1.Brornous acid2 is formed when bromine is added to saturatedsolution of silver nitrate; the formation is probably dependent onthe following reactions : .Br, -I- AgNO, + H,O = AgBr + HNO, + HBrOHBrO + Brp + SAgNO, + H,O = 2AgBr + 2HN0, + HBr02.1 See also Luther and MacDougall, Zeit.physikal. C‘hewi., 1906, 55, 477 ’ Richards, J. SOC. Chem. Ind., 1906, 25, 4.VOL. 1x1. 66 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.For the preparation of aqueous solutions of hydriodic acid, Bodrouxrecommends n convenient method, depending on the action of bariumperoxide on iodine, then removing the excess of iodine by sulphurdioxide, and a t the same time converting the barium into the sulphate,the filtrate from which contains the hydriodic acid, and can be concen-trated in the usual may.The volatility of iron is established by Moissan,2 who has succeededin distilling this and several other metals by heating them in anelectric furnace, and in this manner obtaining crystalline distillates.The metals nearly allied to iron decrease in volatility in the followingorder : manganese, nickel, chromium, iron, uranium, molybdenum, andtungsten.Further, Moissan also finds that the metals indium,osmium, palladium, platinum, rhodium, and ruthenium, when heated incarbon crucibles in an electric arc produced by 500-700 amperes at110 volts, are all fused and boil, giving distillates consisting ofspherules or microscopic crystals, Palladium melts more readily thanplatinum, whilst of these metals osmium is the most difficult todistil.3 In the last Report mention was made of the researches ofDunstan, Jowett, and Goulding on the rusting of iron, of whichphenomenon Moody gives the following explanation. The rusting isdue to the influence of carbon dioxide; neither water nor oxygen freefrom this gas has the slightest effect.A trace of carbon dioxidesuffices to start the action, and freshly formed rust always containssome ferrous carbonate, which is readily attacked by oxygen.Hydrogen peroxide when pure and free from acid has no action onmetallic iron. These conclusions are supported by satisfactory ex-perimental evidence.Treitschke and Tammann record the result oftheir study of the freezing-point curves of mixtures of iron and sulphur ;the microphotographical examination of such mixtures shows thatthe injurious effects of sulphur on iron arise, in mixtures containingabove 2 per cent. of sulphur, from the presence of the fusible sulphidelayered between the particles of iron; the brittleness produced by0.02 per cent. of sulphur is dependent on the properties of themixed crystals, rich in iron. Vigouroux6 has prepared silicides ofiron, cobalt, and nickel by the action at high temperatures of silicontetrachloride on these metals. The compounds so obtained have theformuls Fe,Si, Co,Si, Ni,Si, and Ni,Si; iron silicide is magnetic, theConzpt.rend., 1906, 142, 274. Ibid., 425. Ibid., 189.Tram., 1906, 89, 720. Zeit. anorg. C'henz., 1906, 49, 320.Compt. rend., 1905, 141, 828 ; 1906, 142, 635 and 1270INORGANIC CHEMISTRY. 67otlms are non-magnetic. The interaction of iron and silicon, and ofnickel and silicon, has also been investigated by means of the freezing-point curves of fused mixtures of these elements. Quertler andTammannl conclude from these results that iron and silicon formcompounds of the formulae FcSi and Fe,Si, and that nickel and siliconform five compounds, namely, Ni,Si, Ni,Si, Ni,Si,, Nisi, and Ni,Si,.A ferric hydrogen sulphate, FeH(S0,),,4H20, is produced as a whitecrystalline solid by evaporating a solution of ferric sulphate insulphuric acid., Rubidium and caesinm iron-selenium alums areformed by dissolving ferric hydroxide in selenic acid and treating thesolution with rubidium or cmium carbonate. The formula of therubidium salt is Rb,Fe,(Se04)4,24H~20 ; they both form violet crystalsbelonging to the regular system.3 Crystals of ferric hydroxide,Fe20,,2H,0, and of the oxide Fe203, are formed by acting oncrystalline ferric sulphate with caustic soda ; the crystals arts pseudo-morphous with ferric sulphate.4Several researches are recorded dealing with the commercial alloysof iron ; Hiorns has investigated the influence of varying quantities ofsilicon, phosphorus, manganese, o r sulphur on the formation ofcrystals of cementite, ferrite, or graphite in cast iron.Fettweis 6finds, experimenting with iron phosphide, that increase in theproportion of phosphorus diminishes tho proportion of carbondissolved by iron.Itoozeboom’s theory is shown by Wust * not toapply to iron-carbon alloys containing a large proportion of carbon.When the freezing point of iron has been depressed to 1130’ by theaddition of carbon, further increase in the proportion of carbon hasno effect, and graphite does not separate out from the liquid solution.Microscopic examination of the alloys shows cementite to be formed firston solidification ; it is not stable below 1000°, is metastable below 700’,and does not break up on prolonged heating. The final conditionis ferrite and graphite. The influence of foreign elements on theseparation of graphite from cast iron has been studied by adding aweighed amount of the element to a molten cast iron of known carboncontent, and determining the proportion of graphite to total carbon inthe alloy so produced.Thus tin is found to reduce the solubility ofcarbon and increase the separation of graphite ; iron can, in presenceof an excess of carbon, dissolve 16 per cent. of tin. ulphur reducesthe solubility of carbon in iron, but does not promote its conversionZeit, anorg. Chenz., 1905, 4’7, 163, and 1906, 49, 93.Komar, Chem. .&if., 1906, 30, 15.J. ChmL. SOC. I7d, 1906, 25, 50.3 Roiicngliolo, Gnxzettn, 1905, 35, ii, 553.7 Zeit. physiknl. Chew.., 1900, 34, 437.Ber., 1906, 39, 2270.(j i~fctnlla~rgie, 1906, 3, 60.a Mdalliwgie, 1906, 3, 1.F 68 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.into graphite ; in fact, it neutralises the influence of silicon promotingthe separation of graphite.I n a silicon-free iron, phosphorus in amountsbelow 2.5 per cent. has no effect on the carbon, but above this amountit causes the separation of graphite. I n the presence of 0.9 per cent.of silicon the separation of graphite is produced by 3 per cent. ofphosphorus (Wii&t).l The reduction of mixtures of ferric oxide andtungsten dioxide with aluminium yields homogeneous alloys, fromwhich by treatment with hydrochloric acid the ferrotungsten Fe,W,can be obtained (Vigouroux).2 Steels of lom-carbon content can bealloyed with copper, forming homogeneous alloys which contain amaximum of 32 per cent.of copper, whereas with steels of high-carboncontent alloys containing a maximum of 16 per cent. are formed. Thephysical and mechanical properties of a series of these copper steelshave been examined byNickel and antimony alloy readily, and appear to form four com-pounds, of the following compositions, NiSb, Ni,Sb,, Ni,Sb,, andNi,Sb. The first is the colour of copper, and is hard and brittle ;Ni,Sb, is grey, and although harder than the former is not so brittle.Some of these alloys are magnetic (Loss~w).~The composition of Fischer’s salt (potassium cobaltinitrite) isshown by Riiy to be K6C02(N0,),,,3H20, and its decomposition byheat to be represented by the following equation :CO,(NO,)~,~KE(NO,,~H~O = Co20, + 6N0 + 3EN0, + 3KN0, + 3H20.Continuing his investigations of double chromates, mentioned in lastyear’s Report, Groger has succeeded in preparing from potassiumchromate and cobalt chloride a potassium cobaltous chromate,K,CrO,,CoCrO,, 2H20,which crystallises in dark brown needles, and is hydrolysed by water,forming potassium chromate and a basic cobalt chromate, Thecorresponding sodium and ammonium compounds could not, by reasonof their ready hydrolysis, be obtained. The investigation of the salts ofcobaltammine is still continued by Werner, who with Bindschedler7 showsthat trichlorotriamminecobalt, CoCl,(NH,),, is the first, and trixquo-triamminecobalt chloride, [Co(OH,),(NH,),]CI,, is the last of a series,of which the intermediate members have the compositionand [CoCl(OH,),(NH,),]CI,, and are formed by the action of causticsoda on dichloroaquotriamminecobalt chloride. By the action ofCCOC’ ,(OH,) ( NH,),IClMetallurgie, 1906, 3, 169.Z’mns., 1906, 89, 551.Comnpt. rend., 1906, 142, 1197.Zed. nnorg. Qhcm., 1906, 49, 58.o: Zcit. anorg. Chcrn., 1906, 49, 196.8 B i d . , 1421 ; 143, 346 atid 377.i B e y . , 1906, 39, 2673INORGANIC CHEMISTRY. 69acetic acid and hydrochloric acid on trinitratotriamminecobalt,lthe triaquotriamminecobalt chloride [Co(OH,),(NH,),]CI, is formed.The blackish-brown solution produced by the action of carbonmonoxide on dilute aqueous solutions of palladous chloride (from0.005 t o 0.05 per cent. of palladium) contains colloidal palladium(Donau).2An alloy of platinum and iridium containing 10 per cent. of iridiumis dissolved slowly by hot concentrated sulphuric acid ; the solutionboiled with ammonium sulphate gives a precipitate of spongyplatinum, the filtrate from which has yielded a number of iridiumcompound^.^The observations of Quennessen* show that oxygen takes part inthe dissolution of platinum by sulphuric acid, as there is little or noaction when these are heated together in a vacuum. A doublesulphate of potassium and iridium of the composition Ir(SO,K),,H,Ois obtained in bluish-green crystals by the action of sulphuric acidon potassium iridochloride. The solutions give coloured precipi-tates with many metallic salt solutions (Dele~ine).~Magnus’s green salt is formed by the addition of potassium platinouschloride to platodiammine chloride, Pt(NH3)4C12 ; if potassiumplatinichloride is entirely absent then a red isomeride of thegreen salt is formed; to this the formula P t < ( ~ ~ ~ ] ~ ~ : : > p t isassigned, whilst the green salt is represented by the formulaHydrazine and hydroxylamine platinocyanides are formed by theinteraction of the sulphates of these bases and barium platinocyanide ;they are remarkably fluorescent and exhibit different colours witha1 tered hydration .7Werper and Dinklage 8 describe potassium nitrilobromo-osmonate(OsNBr4)K,2H,0; i t is formed by the action of hydrobromic acid onpotassium osmiamate; and is obtained in dark red prisms, thetreatment of the mother liquors of which with ammonium bromidegives dark brown prisms of the composition [OsNBr,](NH4),,H,0.Rubidium and caesium hydrogen nitrilobromo-osmonat es are alsodescribed.A series of alloys of platinum and silver containing from 10 t o 60Jorgensen, Bey., 1882, 15, 1900.3 Jfonatsh., 1906, 27, 71.DelBpine, Compt. rend., 1906, 142, 631. Ibid., 1841.Ibid., 1525.Jorgensen and Siireosen, Zeit. ccnorg. Chent., 1906, 48, 441.7 Levy and Sisson, Tmns., 1906, 89, 125. BeT., 1906, 39, 49970 ANNUAL REPOKTS ON THE PROGRESS OF CHEMISTRY.per cent. of the former. metal has been prepared by Tbompqon andMiller.1 The action of nitric acid on these alloys is irregular, andit is evident that nitric acid cannot be employed t o effect a separationof gold or iridium from platinum when alloyed with silver.P. PHILLIPS BEDSON.J. Awter. Chcrn. Soc., 1906, 28, 1115