INORGANIC CHEMISTRY.IN 8 part 6f the subject, which has practically onIy one fargegeneralisation t o depend upon, it is found difficult to give a generalidea of the progress of a year’s work. A large number of newfacts, of greater or less value, have been accumulated, but how farthey will tend to the advance of the general theory it is impossibleto say. (( Science moves but slowly, slowly, Creeping with invisiblesteps.” Owing to the cust.om which has grown up of publishingresults as soon as any have been obtained, the year’s progress seemsto consist of small bits of research work, often, it is true, of greatimportance in themselves, but the significance of which cannot beappreciated without a careful study of what has gone before. Inan ideal chemical world, nothing would be published until a com-plete account of the subject of research could be presented.But,apart from the general question of publishing carefully worked outinstalments of a large research, the scramble for priority, happilynot common in this country, is often responsible for the appearanceof immature work. I f papers of this kind were withdrawn bytheir authors as soon as they found that they could not confirmtheir results, no great harm would be done, Very frequently,however, the papers are left uncontradicted, to produce great con-fusion in the mind of the chemical student.Some attempt has been made in this report to show, underdifferent headings, some general results of the increase of know-ledge of particular parts of the subject, but naturally they mustbe incomplete.More or less isolated facts are given, as in previousyearg, under the groups in which the principal elements areclassified.I n his Presidential Address, Sir William Ramsay1 gave furtherresults of his experiments on the possible transformation of elements.It wits thought that solutions of thorium shouId, after longstanding, give rise to small quantities of helium. Accordingly, 270grams of thorium nitrate were purified and dissolved in water.The solution was introduced into a flask provided with a, well-greased stopcock, and the flask was exhausted repeatedly. It wasTrans., 1909, 95, 632.REP,-VOL, VI. 34 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.kept for three years. After freeing the gas collected from carbondioxide, nitrogen, hydrogen and oxygen, the residue was testedspectroscopically, but the presence of helium was doubtful.Sub-sequent experiments led to the same result. It was found, however,t h a t carbon dioxide was present in considerable quantity. Afurther experiment wits made in which the grease of the stopcockwas protected from the action of the solution, and still carbondioxide was found in the solution. Further experiments weremade on thorium nitrate, to see if radium emanation would breakdown thorium into carbon. The gas collected showed the presenceof carbon dioxide. Similar experiments with solutions containingother elements of the same family as carbon, namely, silicon andzirconium, also yielded carbon dioxide.A further account ofsimilar experiments by Sir William Ramsay and F. L. Usher 2 givesa comparison of the amounts of carbon obtained when onecubic millimeter of radium emanation wits allowed to act on thevarious solutions :Solution of.. ................ H2SiF6. Ti(BO,),. Zr(NO,),. Th( NO,),. Pb(ClO,)?.Carbon ....................... 0.518 0.982 1.071 0.8i3 2.93 0.968 0.102 itig,I I1 I I1The emanation was collected from a radium bromide solutioncontaining 0.21 gram of metallic radium, and the amount ofemanation used varied from 0.0649 to 0.1120 cubic millimetre. Theexperimental skill required in working with such exceedingly smallquantities is prodigious. It is to be hoped, with the advent of theRadium Institute, that larger quantities of the material may shortlybecome available, arid everyone will look forward with great interestto the repetition of the experiments on a larger scale.A possiblesource of error in the experiments with thorium nitrate has beenpointed out by 0. Angelucci.3 I t was found that a concentratedsolution of thorium nitrate, placed in a dilatometer, showed aconsiderable increase in volume. After some months, thesolution deposited acicular crystals having the composition of6Th(NO3),,Th(C,O,),,48H,O. It is suggested that the carbon di-oxide evolved from the solutions of thorium nitrate is reallyproduced by the decomposition of the very soluble double oxalateand nitrate of the metal.I n connexion with the production of helium by the breakingdown of radium, Sir William Tilden has made a suggestion thatthe other inert gases, neon, argon, krypton, and xenon, may them-selves have been the result of the breaking down of elements of2 Ber., 1909, 42, 2930 ; A ., ii, 850.3 Atli R. Accnrl. Lincci, 1909, [v], 18, i, 526 ; A . , ii, 742ISORGANIC CHEMISTRY. 35higher atomic weight than radium. Such elements may conceivablybe still present in the earth's crust.Of the eighty-one elements, no less than twenty-seven have atomicweights which are multiples of unity to the first decimal place, andit is perhaps natural that many chemists believe, although perhapswith reservations, in the truth of Prout's hypothesis. Manyattempts have been made of late years to reconcile the hypothesiswith facts, of which perhaps the least artificial is the one ofA.C. and A. E. Jessup.4 Another of these attempts which was,on its publication, received with perhaps a little too much enthu-siasm, is that of A. C. G. Egerton.5 This author supposes that thedivergence of the atomic weights from whole numbers is causedby the excess or defect; of electrons. If N is the number of theelement in the order of the atomic weights, the accurate atomicweight is given by the formulae:For " even " elements, ill = A & 0*0978A, where ,4 = 2N.For odd elements, M =d +_0*0078(A - l), where A =3iV+ 1.The weight of an electron is taken as one-thousandth that ofan atom of hydrogen, and 0.0078 would represent a group of eightelectrons.For the first twelve elements, the agreement of the calculatednumbers with the most recent atomic weight determinations isgood.Aluminium and silicon, about the atomic weights of whichthere may perhaps be some doubt, show a difference of 0.1. Forthe elements between sulpb.ur and nickel, the above formuh donot hold, but, by introducing arbitrary factors, agreement betweenthe calculated and observed numbers is obtained.points out that most of the agreement indicated byEgerton is due to mathematical necessity, and he shows that proofof the hypothesis could only be established if the atomic weightswere known to the third place of decimals. Moir himself tentativelyadvances another hypothesis. H e assumes that the fundamentalcause of valency in an element of valency n. is caused by thepresence in it of m atoms of a sub-element of atomic weight 0.0089.Thus the primordial stuff in an atom of hydrogen weighs 0-9989.Sixteen times this weight plus twice the atomic weight of the sub-element gives 16.000 for the atomic weight of oxygen.For thirty-five elements, the difference between the calculated and the usuallyaccepted atomic weights amounts in most cases to less than 0.04.For the remaining atomic weights, an arbitrary number of atomsof a second sub-element of atomic weight 0.1 is assumed. Thisbeing done, an agreement of the same order as the first is obtained.J. Moir4 P11,il. Mag., 1908, [vi], 15, 21 ; A., 1908, ii, 96.1'mns. 1909, 95, 238, &id., 1752.D ( i 36 ANNUAL REPORTS ON TEE PROCXREBlg OF CHEMISTRY.One of the difficulties which will strike the readers of the paperwill be the different valencies ascribed to similar elements, sulphur,for instance, is diad, and chromium, hexad, but the modesty withwhich the whole speculation is put forward tends to disarmcriticism.Atomic Weights,There has been very considerable activity during the year in thisbranch of chemical work, It is unfortunate that there still remainssome doubt as to the exact ratio of the atomic weight of silver tothat of oxygen.Nearly three-quarters of the atomic weights dependon this relation, and until it is established there must always besome uncertainty. Many attempts have been made to determinethis ratio directly by the analysis of silver oxide, but this substanceis difficult t o obtain in a pure state.When dried a t the highesttemperature compatible with safety, it still retains 2.13 per cent.of water. The oxide contains, moreover, less than 8 per cent. ofoxvgen, and on this account presents extraordinary difficulties asregards its accurate analysis. The chief atomic weight deter-minationq are the following.Chlorine.-A complete analysis of nitrosyl chloride, NOCI, hasbeen effected by P. A. Guye and G. FIuss.7 The weighed substancewas distilled over hot silver, which combined with the chlorine; theresidua1 gas was then decomposed by hot copper, which took upthe oxygen, and, finally, the nitrogen was absorbed by metalliccalcium. The direct ratio of oxygen to chlorine, thus found, was16 : 35.468. The value for nitrogen found was 14.006.An interest-ing and valuable paper by R. W. Gray and F. P. Burt gives theresults of the study of hydrogen chloride. The density of this gaswas determined and it was also analysed volumetrically. The hydro-gen chloride was prepared by three methods: (1) from sulphuricacid and sodium chloride, (2) from sulphuric acid and ammoniumchloride, (3) from silicon tetrachloride and water. The gas wassolidified and fractionally distilled. I n the density determinations,two bulbs were used for measuring the gas, one of soda-glass andone of silica. Both were of comparatively small volume, less than500 C.C. The gas, having been measured in one of the bulbs, wasdrawn into an exhausted bulb containing cocoa-nut charcoal cooledbp liquid air.The increase in weight of this gave the weight of themeasured volume of hydrogen chloride. Special experiments wereundertaken to find the volume of gas adsorbed by the surface of*he glass bulb. In spite of the great difficulties attending workon a gas so soluble and hygroscopic, very fair concordance wasobtained in successive experiments.7 J. Chuim. Phys., 1908, 6, 782 ; A,, ii, 135. Trans,, 1909, 95,1633INORGANIC CHEMISTRY. 3’7The second part of the research consisted in the determination ofthe volumetric composition of hydrogen chloride. This was effectedby the use of heated aluminium. Elaborate precautions were takento ensure the purity of the gas and the absence of volatile impuritiesin the metal. The agreement of the ,different experiments was verygood, The final result for the atomic weight of chlorine was 35.460,comparing with the value of 35.463 obtained by Dixon and Edgar 9in their admirable research on the combustion of hydrogen inchlorine, and with 35.461 obtained later by Edgar.10 Anotherpaper by 0.Scheuer11 deals also with the density of hydrogenchloride. Calculating from his numbers, we obtain the atomicweight of 35.464 for chlorine.Carbon.-An important paper by A. Scott 12 on the atomic weightof carbon gives an unexpected value for this constant. Carbonhas hitherto been regarded as one of the elements which mostnearly conform to Prout’s hypothesis, having the atomic weight12.00. Scott, however, finds that this value is a t least one-thou-sandth part too low.The determination was made by the methodwhich is regarded as the most trustworthy of all atomic weightmethods, namely, the estimation of bromine by silver. The saltsused were tetraethylammonium bromide and tetramethylammoniumbromide, and were prepared with the utmost care. They are onlyvery slightly hygroscopic, and, since they are stable up t o 150°,there is no difficulty in drying them. The silver used was the sameas that used in the estimation of bromine in ammonium bromide.Hence, by subtracting the molecular weight of ammonium bromidefrom the molecular weights of the tetra-alkyl bromides, values forthe molecular weights of two hydrocarbons, C8H,, and C,H,, areobtained. Taking hydrogen as 1.007 and silver 107.88, the valuesfor carbon of 12.019 and 12.017 are found.The individual deter-minations agree extremely well, and the author seems to haveexcluded every possible source of error. I n order to explain thiscomparatively large difference between Scott’s value and those ofDumas and Stas, Roscoe, Erdmann and Marchand, and others,Sir Edward Thorpe l3 has drawn attention to a recent paper ofP. A. Guye and N. Zachariades.l* These chemists state that whena very fine powder is weighed, so much air is condensed on it, oroccluded by it, that the weight is seriously affected. I n the caseof finely-powdered potassium chloride, a difference of 32 mg. onPhil. Trans., 1905, A, 205, 169 ; A . , 1905, ii, 696.lo Ibid., 1908, A, 209, 1 ; compare A . , 1908, ii, 577.l1 Compt.rend., 1909, 149, 599; A , , ii, 991.l2 Trans., 1909, 95, 1200.l3 PTOC., 1909, 25, 284.l4 Compt. rend., 1909, 149, 593; A., ii, 98938 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.100 grams was observed. Scott?l5 in reply, has published a seriesof experiments on the point. He finds that, with potassiumchloride, the difference amounts to only 0.3 mg. on 100 grams, andthis error he very reasonably attributes to a hastily constructedcounterpoise. The writer made some experiments of the same kindthree years ago in connexion with another atomic weight deter-mination. I n moist air, all the fine powders tried increased con-siderably in weight, although the substances were not deliquescent;precipitated silica is a familiar example, whilst in dried air no suchincrease can be noticed.Hence the increases in weight observedby Guye and Zachariades were probably due to water and not tocondensed air.A new determination of a different. kind does not confirm Scott’svalue. G. Baume and F. L. Perrot 16 have determined the densityof highly purified methane. They obtain the number 12.004 forthe atomic weight of carbon.lodine.-Many chemists have attributed the anomalous positionof tellurium and iodine in the periodic classification to an error inthe determination of the atomic weight of iodine, since there hasalways been a possibility that some chlorine was present in theiodine used. Stas, for instance, dried iodine over calcium chloride,and it has been recently shown that some chlorine is set free inthese circumstances. I n order t o exclude the possibility of anyelement of lower atomic weight being present in the material used,G.P. Baxter and G. S. Tilley17 have carefully crystallised iodicacid. This was converted by judicious heating into iodine pent-oxide, and this reduced by hydrazine. The hydriodic acid wasprecipitated by silver, and by combining the ratio Ag: I with theratio Ag: Ii05, the atomic weight of 126.891 is obtained for iodine,and the low value of 107.85 for silver.Tellurium-The atomic weight of tellurium has been the subjectof two researches. V. Lenher18 has employed &he double bromideof potassium and tellurium, K,TeBrG. The salt is beautifullycrystalline, but it suffers from the disadvantage that it very readilyundergoes hydrolysis with water, so much so that Wills, in usingthe salt for the same purpose in 1879, wits compelled to pick outthe perfect crystals with forceps.Lenher overcame this difficultyby crystallising from dilute hydrobromic acid, and he found thathe could dry the substance satisfactorily without decomposing it.A weighed quantity of the salt wits heated in a current of chlorine.l6 Proc., 1909, 25, 286.16 Compt. rend., 1909, 148, 39; A . , ii, 77.17 J. Amer. Chm. SOC., 1909, 31, 201 ; A., ii, 225,18 Bid., 20 ; A,, ii, 230INORGAXIC CHEMISTRY. 39Bromine and tellurium chloride were expelled, so that the ratioKzTeBr,: 2KC1 was obtained. The atomic weight of tellurium,deduced from sixteen very concordant determinations, was 127.55,being still 0.6 above the atomic weight of iodine.P. E. Browningand W. R. Flint 19 have stated that they have separated two distinctsubstances from tellurium which have the atomic weights 126.49 and128.85. Their method, the fractional precipitation of telluriumtetrachloride by water, had been previously tried. Baker andBennett 20 found identical atomic weights for the element obtainedfrom the first and the last fractions of the precipitation. I n thenew work of the two American chemists, two fractions wereobtained, one of 50 grams, and the other of 13 grams. The atomicweights were obtained by three methods: (1) Conversion of thebasic nitrate into the dioxide, (2) a modification of Brauner’spermanganate method, (3) precipitation of the dioxide from thebasic nitrate by ammonia and acetic acid.The mean results are,as stated above, 126.49 for the least soluble and 128.85 for themost soluble fraction. Their result, if confirmed, would be of greatinterest. It may be pointed out, however, that if the mean atomicweight of both fractions is calculated, which would represent theatomic weight of the tellurium before fractionation, it is found to be126.89, instead of 127.55, as found by other workers. This wouldseem to indicate that the tellurium originally used had beeninsufficiently purified, or that the methods used in the atomic weightdeterminations were not trustworthy. One of the authors is con-tinuing the investigation.Molecular Weights.A very considerable amount of work has been done during theyear on the determination of molecular weights by cryoscopic andebullioscopic methods.Since Beckmann 21 showed that, in allsolvents, iodine has the molecular weight corresponding with I,,it has been assumed that when the solutions are brown, combinationhas taken place between the iodine and the solvent. To test thieview, experiments have been undertaken 2, to determine the freezing-point depression produced by iodine and certain liquids, firstseparately and then together, in the solvents bromoform andethylene dibromide. It was found that it was only with liquidswhich formed brown solutions that the total depression was lessl9 Amer. J. Sei., 1909, [iv], 28, 347 ; d., ii, 996.21 Zeikch. physikal. Chem., 1907, 58, 543 ; A., 1907, ii, 340.!a J.H. Hildcbrand and B. L. Glascock, J. Amer. Chem. Soc., 1909, 31, 36 ;Trans., 1907,91, 1849.d., ii, 22540 ANNUSL REPORTS ON THE PROGRESS OF CHEMISTRY.than the separate depressions, thus indicating that chemical com-bination had taken place. The possibility of chemical actionbetween the solvent and the solute compels one to accept with somereserve some recent determinations of the molecular weight ofselenium in melted iodine 23 and the similar determinations for thesame element dissolved in fused mercuric chloride.24 In the lattersolvent, dilute solutions give depressions of the freezing point whichindicate variations of the selenium molecule from Se8 to Se,. Similarexperiments, with fused mercuric chloride as solvent, gave forsulphur a molecular weight, a t all dilutions used, corresponding with8,.Tellurium, it is pointed out, does react chemically with themercuric chloride, reducing it to calomel.Silent Discharge.The study of the effect of electric discharge on gases is one ofgreat difficulty, because there is little doubt that the action iscomplex. The results may be caused by a t least three differentfactors: (1) heat, (2) light, and (3) a specific influence of the electriccurrent, akin to electrolysis in solution. Until comparativelyrecently the effect was supposed to be due to heat alone, and theincreased yield of ozone obtained when the silent discharge wassubstituted for the spark discharge was ascribed by Brodie to thelower temperature attained.Some experiments by A. Holt, jun.,25give some help towards the elucidation of the problem its to which ofthese factors has the most influence. It has already been shown byDixon26 and by Colliez7 that carbon dioxide is decomposed byelectric sparks whether it is moist or dry, and that the shorter(and hotter) the sparks, and the lower the pressure of the gas, thegreater is the decomposition. It has also been proved 28 that ultra-violet light had no effect on moist carbon dioxide, but that tl;edried gas was decomposed to an extent which varied inversely asthe pressure. Holt’s experiments show that with the silent dis-charge there is decomposition in the dried gas to a very con-siderable extent, 48 per cent. at 30 mm.pressure, and that theextent of the decomposition varies inversely as the pressure. Theresult in this case thus corresponds with the results of the actionof ultraviolet light. I n the case, however, of the moist gas, a verystriking difference is observed. Instead of there being no action,Qs G. Pellini and S. Pedrina, Atti R. Accad. Ldncei, 1908, [v], 17, ii, 78 ;a4 F. Olivari, ibid., 1909, [v], 18, ii, 94 ; A . , ii, 805.95 Trans., 1909, 95, 30.Ibid., 1901, 79, 1063.28 Chapman, Chadwick, and Ramsbottom, Trans., 1907, 91, 942.A,, 1908, ii, 833.26 Ibid., 1885, 47, 571INORGAKIC CHEMISTRY. 41considerable decomposition is observed, and this decomposition ISgreater when the pressure is increased. Unfortunately, no memure-ments were made of the amount of ozone formed from the liberate-3oxygen, and until the further experiments are published it wouldbe useless to speculate as to the origin of this very remarkabledifference.It was shown by Brodie29 that in the decomposition of carbondioxide, dried only by sulphuric acid, as much as 85 per cent.ofthe oxygen produced was converted into ozone, so that the change3CO+ 0, = 3c0, can scarcely proceed from left to right, a con-clusion which was confirmed by the work of Remsen andSouthworth.30 Oxygen therefore has little or no oxidising powerwith regard to carbon monoxide. If, however, other oxidisablegases are passed with oxygen through a silent discharge tube, ithas been recently shown that o'xidation does take place.31 Whena mixture of hydrogen and oxygen in the proportion of 2 : 1 byvolume, or with more hydrogen than this, is so treated, water isproduced in considerable quantities, but if the oxygen is in excess,no hydrogen peroxide is formed, but only ozone.Chlorine withoxygen gives chlorine monoxide, and hydrogen chloride gives hypo-chlorous acid. Carbon disulphide is notr oxidised by this method,whilst ammonia gives small quantities of hydroxylamine, but nonitrous acid. It is rather remarkable that carbon monoxide andchlorine appear not to react when passed together through anozoniser.Co mtb us t io n.Notable advances towards the understanding of the theory ofcombustion have been made by H. B. Dixon and his pupils inrecent years. One of the most disputed points in connexion withthis work has been the determination of the temperature a t whichgaseous mixtures undergo inflammation.Experiments by Askenasyand V. Meyer 32 gave for electrolytic gas results varying from 5 1 8 Oto 606O, whilst Relier,= working with the gaseous mixturepassing through a, heated tube, obtained 845O as the ignition-pointof the same mixture. The variation of the temperatures found isundoubtedly due to the influence of the surface of the vesseI used,Most substances act as either positive or negative catalysts, anduntil this influence could be eliminated or allowed for, no trust-worthy results could be expected. By a most ingenious arrange-29 Phil. Trans., 1874, 164, 83.g1 E. Comanducci, R e d . dccad. Sci. Pis. Mat.Nap&, 1908, [iii], 15, 15 ;32 Annalen, 1892, 269, 49 ; A, 1892, 938.33 Ann. Chim. Phys., 1897, [vii], 10, 521 ; A , , 1899, ii, 85.3o Ber., 1875, 8, 1414.A., ii, 47742 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.ment, H. B. Dixon and H. F. Coward% have succeeded in ignitingthe gases out of contact with any solid material. One gas waspassed through a wide porcelain tube, which was heated electrically,while the other passed up a narrow tube fixed along the axis of thowide tube. The temperature was measured by a thermo-junctionplaced either in the narrow tube or just above its orifice. Thstemperature was gradually raised until ignition occurred, and thisignition took place a t a point above the orifice. That the ignition-temperature was not altered by varying the material of the jet wasshown by a series of experiments in which fused quartz, Jena-glass, and soda-glass were employed.I f the rate of passage ofhydrogen through the inner tube, and oxygen through the outerone, was maintained above a certain limiting value, and if thediameter of the outer tube was beyond 40 mm., nearly constantvalues approximating to 600° were obtained. When air was sub-stituted for oxygen, the very interesting result was obtained thatthe ignition-temperature was unaltered. For carbon monoxideburning in oxygen, also, the ignition-temperature of 650° wasobserved, whilst the same gas ignited in air at 6 5 1 O . Theignition-temperatures of methane, ethane, propane, ethylene,acetylene, cyanogen, hydrogen sulphide, and ammonia were alsodeterEined.A valuable summary of t.he work done a t the ManchesterUniversity on the mechanism of the combustion of hydrocarbonswas given as a Friday evening discourse a t the Royal Institutionby W.A. B0ne.35 It was shown that the suggestion of H. E.Armstrong, made many years ago, that the combustion of a hydro-carbon takes place through the formation and subsequent decom-position of hydroxylated molecules, has received very considerableexperimental support. The preferential combustion either of carbonor hydrogen is an idea which must now be abandoned. By theisolation of large quantities of aldehydic products as the result ofthe combustion of hydrocarbons a t low temperatures, it has beenproved that, in these circumstances, Armstrong’s hypothesis is wellgrounded. When the researches were carried on at higher tem-peratures and under conditions approximating to those in ordinaryhydrocarbon flames, it became evident that although the phenomenaobserved were different, the mechanism of the reaction was essen-tially the same at high as at low temperatures.The influence of mere traces of water on the combustion ofsubstances in oxygen and other chemical actions has not yetreceived an adequate explanation.In 1893 Sir J. J. Thornson3634 Trans., 1909, 95, 514.36 Phil. Nag., 1893, [v], 86, 913.as iVature, 80, 82INORGAXIC CHEMISTRY. 43showed that if the force holding atoms together in a molecule wereelectrical in its nature, the presence of liquid drops of high specificinductive capacity would tend to loosen the bonds between theatoms, and so increase the tendency for chemical action to takeplace.The difficulty of accepting this explanation was that itseemed impossible that drops of liquid water could exist in a gas con-taining a very small quantity of water vapour. J. s. TownsendF7however, has shown that in a gas ionised by Rontgen rays thereis a great decrease in the mobility of the ions when a small quantityof water (represented by 0.1 mm. pressure) is present. When morewater is added, visible liquid drops are produced. Hence it isprobable that with still smaller amounts of water than were presentin Townsend's experiments, aggregations of water molecules wouldbe produced if ions existed in the gas, and Sir J.J. Thomson'sexplanation may hold good. H. B. Baker38 has found that anincrease in the ionisation of a mixture of hydrogen and nitrousoxide, a t 530°, produces a corresponding increase in the rate ofaction between the gases. Lime, which has been shown to ionise agas to a considerable extent, increased the normal rate of unionby five times; thoria, which is a, much more powerful ionising agent,increased the rate of action by twenty times; whilst with radiumbromide the gases exploded directly the temperature of combinationwas reached. Similar experiments with the carefully dried gasesshowed &hat the ionising agents had no effect on the combinationof the gases. The hypothesis that, unless water and gaseous ionsare both present, chemical union cannot take place, is tentativelyadvanced.The Combination of Salts with Water.The use of salts containing water of crystallisation for atomicweight determinations has been of late years regarded as impossible.Mallet 39 pointed out, in regard to the alums, that the water wasalways in excess of the calculated quantity, and Richards40 made amost careful investigation of the water in crystallised bariumchloride.I n the same circumstances, the excess of water retainedwas remarkably constant, but by varying the state of division ofthe salt, large differences were observed. I n no circumstances couldthe theoretical amount of water be obtained. Richards drew theconclusion that no atomic weights with any pretensions to accuracycould be found by using salts which contained water ofcrystallisation.'' PTOC.Roy. ~ O C . , 1908, 81, 464.Wilde TJecture, 1909 ; Mem. JIamhcstcr Phil. sbc., 1909, [iii], 53, 16.38 Stas Memorial Lecture, Trans., 1902, 81, 204.lo Zeitsch, physikal. Chem., 1903, 46, 189 ; A , , 1904, ii, 24244 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.P. A. Guye and D. E. Tsakalotos41 assert that it is possible, byworking under proper conditions, to obtain the theoretical loss ofwater from crystallised barium chloride, For this statement theyrely on the constancy of the loss experienced on heating the salt.Since, however, they state that the salt was not sufficiently purefor the atomic weight of barium to be deduced, their results mustbe accepted wit>h some reserve.It may be possible, and it is evenlikely, that salts with water of crystallisation may be eventuallybrought into use in accurate work, but until such accurate workis done with highly purified materials, so that atomic weightsobtained by their use can be compared with those obtained byunexceptionable methods, the results obtained by the employmentof salts with water of crystallisation must be looked upon withsuspicion.The opacity which sometimes makes its appearance in hithertoclear crystals has been investigated in the case of sodium sulphateby D. Gernez.42 A 66 per cent. solution of this salt, cooled below8O in absence of dust, deposits crystals of the heptahydrate,Na2S04,7H20, and the mother liquor remains supersaturated withrespect to the decahydrate, Na2S04,10H20.On adding a crystal ofthe decahydrate, crystallisation occurs through the liquid until itreaches the solid hep tahydrate. This then loses its transparency,and becomes opaque. The opacity is ascribed to the formation ofdecahydrate in the mother liquor occluded in the crystals of hepta-hydrate. This explanation is supported by the fact that when afragment of the porcelain-like mass is added to a supersaturatedsolution of the decahydrate, crystallisation is induced. The factscan be demonstrated in a lecture experiment in the following way.The heptahydrate solution is allowed to crystallise in a U-tube,which has a constriction near the bottom, and the crystals arepressed into a solid cake in the constricted part.On adding acrystal of the decahydrate in one arm of the tube, crystallisation ofthis hydrate takes place down this arm. When the solid plug isreached, opacity is produced in it, and the formation of the decachydrate proceeds without interruption up the other arm of thetube. The following salts behave in the same way: Na&k04,4H20and Na&rO4,10H20, Ca(N0,),,3H20 and Ca(N0,),,4H20,N+S,O,, 2H20 and Na2S20,,5 H20.Since in many cases ammonia seems to be capable of replacingwater of crystallisation, molecule for molecule, as in the well-knowninstances of CuS04,5NH, and CuS04,4NH,,H,0, special interestattaches to the compounds which have been prepared of salts41 J. Chim. Phys., 1909, 7 , 214 ; A., ii, 475.42 Compt.rend., 1909, 149, 77 ; A,, ii, 729INORGANIC CHEMISTRY'" 45with hydrazine. Some of these compounds were prepared in1894,43 such as NiS04,3N2H4, ZnS04,2N&,, ZnC12,2N2H4, andCdC12,2N2H4,H20. A compound with copper nitrate, Cu(N 03)2,N2H4,has also been obtained. I n some new experimentsp4 a large numberof salts have been examined in respect of their combination withhydrazine. The substances formed are in general crystallinepowders insoluble in water, but soluble in acids and in ammonia.The authors have arrived a t the interesting conclusion that onemolecule of hydrazine replaces two molecules of ammonia. Thehydrazine cannot, as a rule, be driven off from the compoundswithout decomposing the salts, which may indicate a more definitecombination than is supposed t o exist between salts and the waterwith which they crystallise.Comparing the compounds formedwith hydrazine with the hydrated salts, the following seem toexhibit a relation : NiS0,,3N2H4 and NiSO,,GH,O, NiC12,3N2H4and NiC1,,6H20, Ni(N0,),,3N2H, and Ni(NO,),,GH,O, CdS0,,2N2H4and CdS0,,4H20, FeC1,,2N,H4 and PeCl2,4H,O, hInC12,2N2H4 andMnC12,4H;0, and others. On the other hand, there are manyinstances in which this relation does not hold: BaC12,2N2H4 andBaC1,,2H2O, SrC1,,2N2H, and SrC1,,6H20. It is probable that afurther study of the three types of association which is foundbetween salts and (1) water, (2) ammonia, (3) hydrazine will bringto light facts which will elucidate the very difficult question of theway in which water of crystallisation is connected with the con-stitution of the salt.Metals and Alloys.The very interesting experiments carried on in the Universityo i Utrecht on the breaking up of metallic tin have shown that thismetal is not alone in being metastable. Most of the metals inordinary use are liable to change, although fortunately this changedoes not commonly take place at the ordinary temperature.E.Cohen and K. Inouye4j have shown that if a pattern is etchedby nitric acid on a sheet of lead and the sheet kept in contact withanother strip at 180° for some hours, the pattern appears on thesecond strip. Contact under pressure, a t the ordinary temperature,does not induce the changc. SimiIar effects have been produced bythe contact, at elevated temperatures, between zinc and zinc, brassand copper, copper and copper, and bismuth and bismuth.Themetals used had all been rolled out into strips, and therefore hadbeen subjected t o severe mechanical strain.43 Curtius and Schrader, J. pr. Chem., 1894, [ii], 50, 311 ; A., 1895, ii, 10.44 H. Franzcn and 0. von &layer, Zeitsch. anorg. Chm., 1908, 60, 247;45 Chem. WeekbZad, 1909, 6, 881 ; A,, ii, 1008.A., ii, 4046 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTXY.A large amount of work has been done on alloys, and in manycases distinct compounds have been isolated, whilst in other casestheir existence has been indicated by the form of the freezing-pointcurves and other methods. A new method has been described forthe production of liquid alloys of sodium and potassium.46 Ifpotassium is melted in a vacuous Jena-glass flask with sodiumhydroxide, and the temperature raised to 250°, a layer of the liquidalloy NaK, is formed, even in presence of excess of sodiumhydroxide.On similarly treating potassium hydroxide withsodium, the liquid alloy NaK is formed, but on largely increasingthe proportion of the hydroxide and raising the temperature to350°, the alloy NaK, is the chief product. G. Masing,47 workingwith metals under pressures of 5000 atmospheres, has come to theconclusion that, contrary to t.he results of the well-known researchesof Spring, pressure has no influence in bringing about the diffusionor combination of metals, but its effect is confined solely to thebringing of the metals into close contact.G. Tammann and Masing48find that after compression the particles of metals can be dis-tinguished by the microscope lying side by side, and combinationgoes on slowly a t the ordinary pressure. For instance, the electricalconductivity of a block of lead and thallium immediately aftercompression was that of a mechanical mixture of the metals. Thisincreased by 60 to 75 per cent. in the course of a month, showingthat combination had taken place.Group I .The existence of a true sodium alum, which has been oftenasserted and denied, has been the subject of an investigation byW. R. Smith.49 This author confirms the work of Aug650 andWadmore,51 and shows that sodium alum forms mixed and layercrystals with other alums.Sodium alum has also been prepared byN. I. Surgunoff,52 who shows that the crystals are cubic if theyseparate from a solution which is supersaturated a t ZOO or below.Above this temperature the crystals formed are monoclinic.It is not generally realised how oxidisable solutions of sodiumsulphite are in the air. The likelihood of impure material havingbeen used has led to the study of this salt by H. Hartley and4G G. F. Jaubert, BUZZ. SOC. chirn., 1908, [iv], 3, 1126; A., ii, 41.47 Zeitsch. anorg. Chem., 1909, 62, 265 ; A., ii, 669.48 Zeitsch. Elektrochern., 1909, 15, 447 ; A., ii, 669.49 J. Amer. Chem. SOC., 1909, 31, 245 ; A., ii, 239.5o Compt. rend., 1830, 110, 1139 ; A., 1890, 1059.b1 Proc., 1905, 21, 150.w Bull.Acnd. Sci. St, P&ersbowg, 1909, 1057 ; A., ii, 1001i NORU A N I C CH E M ISTRY, 47W. H. Barrett.53 Three forms of the salt have been previouslydescribed : N+SO,, Na2S0,,7H,0, and Na,SO3,10H2O. The authors,working with extreme care, have been unable to confirm theexistence of the decahydrate. The crystalline form (hexagonalprisms) of the anhydrous salt has been determined, and thesolubility and transition temperature from the heptahydrate toanhydrous salt have been measured.The long-discussed question of the possibility of the existence ofcuprous sulphate has a t last been settled by the preparation of thepure salt.54 The preparation is carried out by the action of methylsulphate on cuprous oxide in absence of water :Cu,O + Me2S0, = Cu,SO, + Me,O.Ethyl sulphate behaves in a similar way.The action is carried ona t a temperature of 160O. The cuprous sulphate is washed withether, moisture being rigidly excluded, and dried in a vacuum.When dry, the new salt is stable in air, but is decomposed rapidlyby water with evolution of heat:Cu2S04 (solid) + Aq = CuSO, (dissolved) + Cu solid + 21 cals.The development of heat in this roaction shows a difference fromthe behaviour of other cuprous salts, and it explains why theprevious attempts to prepare the sulphate in aqueous solution haveresulted in failure.The curiously coloured salt obtained by the action of sulphuricacid on a mixed solution of ferrous and cupric sulphates 55 has beenexamined by A. J. Allmand.56 The red colour of the salt changesto chocolate-brown on heating, and finally to mauve.Only a verysmall quant.ity of ferric salt is present. The author’s experimentslead t o the conclusion that the salt is a solid solution of the generalcomposition (CuFe) SO,,H,O, but, since the colour of CuSO,,H,Ois pale blue and FeS0,,H20 is white, the red colour of the com-bination is still a mystery.The formula ascribed to the black powder which separates onthe anode when silver nitrate solution is electrolysed has beenstated to be Ag7N0,,.57 It has been considered 58 that its formationis preceded by that of a crystalline pernitrate of silver, AgNO,. Theblack powder has been reexamined, and it has been proved that53 Trans., 1909, 95, 1184.64 A. Recoura, Conzpt.rend., 1909, 148, 1105 ; A., ii, 579.65 A. Scott, Trans., 1897, 71, 564.56 Zeitsch. nnorg. Chem., 1909, 61, 202 ; A., ii, 238.57 0. s d c , ibid., 1900, %, 305 ; A . , 1900, ii, 595.58 G. Baborovskjr and B. Kuzma, Zeitsch. Elektrochem., 1908, 14, 196 ; A,, 1908,ii, 378 ; G. Cxborovskq’ and G. Kunna, Zeitsch. physikal. Chenz., 1909, 67, 48 ;A,, ii, 66648 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY,it consists of a true peroxide of silver, Ag8Q4, containing occludedsilver nitrate.One, whichhas the formula 2KBO,,HiO, is obtained by adding cold 3 per cent.hydrogen peroxide solut.ion to a saturated solution of potassiummetaborate, KBO,. The perborate is precipitated by the addition ofmethyl alcohol. After standing, the salt is filtered and washed withice-cold water.It is crystalline and soluble to the extent of 2.5parts in 100 parts of water at Iso. The dry salt is stable in air,but in water it slowly loses oxygen. The second salt, 2KB0,,H,02,is obtained in a similar way, except that the hydrogen peroxide isof 30 per cent. strength. It is stable in air, but deflagrates whenheated to 1 5 0 O . It is less %oluble in water than the former salt.Both the new substances have strong antiseptic properties.Two perborates of potassium have been prepared.59Group I V .The mixture of gases evolved on treatment of magnesium silicidewith hydrochloric acid has been investigated by P. Lebeau. Twohydrides, SiH4 and Si,H,, were isolated in a state of purity, andtt more easily condensable substance, probably Si,H4, was shownto be the source of the spontaneously inflammability of the othertwo hydrides.The case for the analogy between silicon and carbon, as regardsthe formation of chain derivatives, has been summarised byJ.Emerson Reynolds.60 He makes the interesting suggestion thatthe function of silicon compounds, which have long been known asconstituents of living tissue, is not merely to act as strengtheningmaterial, but that they are essential to cell-formation, just ascarbon compounds are. The analogy between these two elementshas been strengthened by the isolation of what are probably five-and six-silicon-atom chains. A. Besson and L. Fournier61 havesubmitted chlorosilicomethane to the action of the silent discharge,and fractionally distilled the resulting l i p i d in a vacuum.Themain products are: (1) Si4Cllo, a colourless, oily liquid, which, whendecomposed by water, gives a white substance resembling silica inappearance, but emits sparks and ignites when gently rubbed;(2) Si,c"ll4, a white solid, which melts a t 170°, and also gives acombustible substance with water; (3) a solid, yellow mass, solublein light petroleum, which apparently consists of a mixture of higherchlorides.59 C. von Girsewald and A. Wolokitin, Ber., 1909, 42, 865 ; A , , ii, 312.6O Roy. Inst. hkprbs, 1909.81 Comp?. rend., 1909, 148, 839 ; 149, 34 ; A,, ii, 399, 663INORGANIC CHEMISTRY. 49Group V.Very different compositions have been assigned to the chlorideof nitrogen obtained by the action of chlorine on ammoniumchloride, many chemists having asserted that it contained hydrogen.The substance, after solution in carbon tetrachloride, has beenproved62 to be NCI,, direct estimation of hydrogen showing thatthere is less than one atom of hydrogen to each hundred atoms ofnitrogen.Another chloro-derivative of ammonia has been prepared in apure state by F.Raschig.63 Equimolecular quantities of ammoniaand sodium hypochlorite react according to the equation :NaOCl+ NH, = NH,C1+ NaOH.By distilling in a vacuum, a faintly yellow, unstable oil is obtained.This substance, to which the author gives the name of chloroamine,reacts vigorously with alkalis, giving ammonia and nitrogen : blacknitrogen iodide is precipita,ted from a solution of potassium iodide,whilst with ammonia a small quantity of hydrazine is produced.Nitrosyl perchlorate, NO-ClO,,H,O, has been prepared by K.A.Nofmann and A. Zedtwitz 64 by passing oxides of nitrogen, preparedby the action of nitric acid on sodium nitrite, into perchloric acid.The salt is obtained in colourless, doubly refracting leaflets, whichare slightly hygroscopic. Water produces decomposition, resultingin a green solution. Alcohol, ether, acetone, and primary aromaticamines ignite when mixed with the new perchlorate, producingviolent explosions.Although ammonium hydroxide exists in small quantity in asolution of ammonia, the compound has not hitherto been isolated.3’. F. Rupert65 has obtained a solid having the compositioncorresponding with NH,OH, which is probably a definite compound.The freezing points of solutions of ammonia of concentrations from4.1 to 100 per cent.have been determined, and the results plottedas a curve, which shows maxima a t 49 per cent. and a t 65 percent. The monohydrate should contain 48.6 per cent. of ammonia,and the hydrate, 2NH,,H,O, 65.4 per cent. The behaviour ofhydrazine under the influence of different oxidising agents has beenstudied by A. W. Browne and F. F. Shetterly.66 It is shown thatthe oxidation can proceed in three ways: (I) with formation offairly large quantities of azoimide and ammonia, when hydrogen6‘2 D. L. Chapman and L. Vodden, Trans., 1909, 95, 188.G;5 Yerh. Ges. deqst. Naturforseh. Aerlze, 1907, 11. i, 120 ; A , , ii, 232.G5 J.Amer. Chenz. h’oe., 1909, 31, 866 ; A . , ii, 726Bli ]bid., 221, 7S3 j A , , ii, 233, 658,Ber., 1909, 42, 2031 ; A . , ii, 668.REP.-VOL. VI. 50 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.peroxide, potassium chlorate, or potassium persulphate are used inacid solution; (2) with formation of little or no azoimide but largequantities of ammonia, when the oxidation is effected by potassiumpermanganate, manganese dioxide, or ferric oxide in presence ofsulphuric acid; (3) with formation of little azoimide or ammonia,as is the result of the use of potassium iodate, mercuric oxide, ormercuric chloride. Persulphuric acid gives about 40 per cent. ofthe possible yield of azoimide, but Thiele's method,"7 which employsa mixture of hydrazine and ethyl nitrite, is still the most effective.Continuing his patient study of the nitrites, P.C. RSy68 hasshown that if a. solution of ammonium nitrite is heated in avacuum very little gas is evolved below 40°; on cooling, most ofthe salt crystallises. If the temperature is raised to 70°, slowdecomposition takes place, but a considerable quantity of the saltappears as a sublimate. There is as yet insufficient evidence toshow whether the salt has sublimed unchanged, or whether dis-sociation and recombination have taken place.Two new solid hydrides of phosphorus have been prepared.69Calcium phosphide, on treatment with water, gives a mixture which,on passing over granular calcium chloride, gives a yellow deposit.On treatment with cold dilute hydrochloric acid, the calciumchloride is dissolved, and the solid left is washed with water andthen with alcohol and ether.The hydride is a canary-yellowpowder, which, when first prepared, has no odour. On standing inthe air, it rapidly changes, especially in sunlight, giving off aspontaneously inflammable phosphine. It is remarkably insolublein reagents, and on analysis it is stated to have the compositionP,,H6. On heating in a vacuum, it turns red and evolves purephosphine, the residue being the second new hydride,5P,,H6 = 6P9H2 + GPH,.The second hydride is stable in dry air, but in presence of moistureit is converted into phosphine and phosphoric acid.Group VI.The ordinary ozone tube used in the laboratory rarely gives anylarge percentage of ozone, 3 to 8 per cent.being the usual yield.Chemists will be glpd to know of the simple and convenient processdescribed by F. Fischer and K. Bendixsohn.70 The ozone is preparedby the electrolysis of dilute sulphuric acid, the anode exposing very67 Bw., 1908, 41, 2681 ; A . , 1908, ii, 940.6y A. Stock, W. Bottcher, and W. Lenger, Ber., 1909, 42, 2839, 2847;70 Zeitseh. nnorg. Chem., 1909, 61, 13 ; A , , ii, 136.68 T,rans., 1909, 95, 384.A . , ii, 7 2 i INORGAXIC CHEMISTRP. 51little surface to the liquid. The most effective anode is made byimbedding platinum foil in glass and grinding away the edge so thatonly a line of 0.1 inm. breadth is exposed. The proportion ofozone in the oxygen evolved amounts to as much as 23 per cent.The methods adopted have been hitherto insufficient to decidebet.ween the two formulze assigned to Caro's acid, H,SO,71 andH,S,O,.72 The analysis of salts is inconclusive, since a salt, MHSO,,is indistinguishable from M,S,O,,H,O, as in the historic instanceof sodium hyposulphite.Two new researches, however, support theformer formula, showing that the acid is monobasic. H. Ahrle73has prepared a nearly pure (92.3 per cent.) acid by the interactionof 100 per cent. hydrogen peroxide and sulphur trioxide. Thisprocess was preferred to that of the action of hydrogen peroxideon sulphuric acid, since the equationwas found to be reversible. The experiments show that the acidis monobasic, confirming those of T. S. P r i ~ e .7 ~ The strong acid isstable below Oo, but decomposes slowly at the ordinary temperature.Catalytie agents, such as finely-divided platinum, produce explosivedecomposition. R. Willstatter and E. Hauenstein75 have prepared thetwo acyl derivatives, C6H,*CO*O*O*S03K and C,H,*S02*O*O*S03K.These acyl derivatives are not per-acids, but behave as mixedperoxides of persulphuric acid and acyl. The benzoyl derivative,for example, is deconiposed by alkalis into Caro's salt and abenzoate, thus :C,H5*CO*'O*O*S03H + H20 = H*O*O*S03 + C6H,*C0,H,and by acids into sulphuric acid and benzoyl hydrogen peroxide:C,H,*CO*'O*O*SO3H + H20 = H*O*O*CO*C,H, + H,SO,.The acyl derivatives are monobasic acids, and thus the formulaH2S05 for Caro's acid is confirmed.Many chemists have doubted the existence of sulphur dichloride,but the question may be considered as settled by the work ofE.Beckmann.76 The liquid, produced by the action of chlorine onsulphur monochloride, S,Cl,, can be distilled a t low pressures withpractically no decomposition. It solidifies at -8OO. I n solution inliquid chlorine, it has a molecular weight corresponding with theformula SCl,, and the same formula is arrived at from cryoscopicH,O, + H,SO, H,SO, + H,OBaeyer and Villiger, Ber., 1901, 34, 853 ; A , , 1901, ii, 380.Armstrong and Lowry, Proc. Roy. SOL, 1902, 70, 94 ; A., 1902, ii, 558.73 J.pr. Chem., 1909, [ii], 79, 129; A,, ii, 395 ; Zeitsch. angew. Chem., 1909,74 Trans., 1906, 89, 53.76 Ber., 1909, 42, 1839 ; A , , ii, 566.7ti Zeilsch.physikal. Chern., 1909, 65, 289 ; A., ii, 137.22, 1713 ; A , , ii, 804.E 52 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.measurements in xylene, ethylene dibromide, acetic acid, andbromine.Many formulze have been ascribed t o perchromic acid. Accordingto E. H. Riesenfeld,77 there are two such acids: HCrO,, whichis blue, and H,C'rO,, which is red. These two acids give rise tosimilarly coloured salts. The molecular weight of the blue pyridinesalt has been determined by the cryoscopic method, the solution inwater having been used. Since, however, considerable dissociationtakes place in this solution, the amount of dissociation had to beseparately measured. The molecular weight corresponds with theformula C5H5N,HCr05.The decomposition of the free acid takesplace too rapidly for its molecular weight to be determined. Thered potassium perchromate is obtained by the action of chromicacid on a cooled mixture of potassium cyanide and hydrogendioxide.On the other hand, W. Oechsner de Coninck78 maintainsMoissan's formula H,CrO, for perchromic acid, and compares theperchromates with the peruranates which he has obtained by theaction of alkalis on uranyl chloride in presence of air, for example :2U02C1, + 4K20 + 0, = 4KC1+ 2K,U05.Group VII.Nothing can be of more value to the science than the exhaustivestudy of one particular action. Such a study, if complete, couldnot fail to throw light on the most difficult of problems, the natureof chemical attraction. The period of induction in the union ofhydrogen and chlorine has been investigated by D.L. Chapmanand his co-workers for some years past, and it is now certain thatthe speculation that this period was occupied by the moleculesgetting up a certain type of vibration must finally be abandoned.It was shown79 that a trace of ammonia is responsible for a longinduction period. The ammonia would produce nitrogen chloride,which itself strongly inhibits the action. It has also been proved 6othat, contrary t o the statement of Bunsen and Roscoe,*1 excess ofeither of the.reacting gases, hydrogen and chlorine, to the extent of6 per cent. has no influence on the rate of union. The influence ofoxygen as inhibiting the action was first pointed out by Bunsenand Roscoe (Zoc.cit.). Their observation has been confirmed, andfrom quantitative measurements Chapman and NacMahon 82 have77 Ber., 1908, 41, 3941 ; A . , ii, 51.78 Bull. Acud. roy. Belg.. 1909, 175 ; A., ii, 318.79 Burgess and Chapman, Trans., 1906, 89, 1433.*O D. L. Chapniaa and P. S. MacMahon, ibid., 1909, 95 135.Phil. Trans., 1857, 147, 390. 82 Tmt.-s., 1909, 95, 959INORGANIC CHEMISTRY, 53drawn the conclusion that if the relation of oxygen present to thesensitiveness of the mixture, which they have established for smallquantities of oxygen, holds with infinitely small quantities ofoxygen, then the sensitiveness of pure hydrogen and chlorine wouldbe infinite, in other words, these gases would combine in the dazk.Nitrogen, carbon dioxide, and nitrous oxide only act? as diluents,whilst nitric oxide acts as an inhibitor. It is pointed out that theinhibiting gases are all capable of reaction with the constituents ofthe explosive mixture.Iodine dioxide, obtained first by Millon in 1844, has been pre-pared in a pure condition by M.M. Pattison Muir.83 By actingwith concentrated sulphuric acid on iodic acid until iodinebegins to be evolved, and allowing the mixture to cool, a yellow,crystalline crust is obtained. This, after washing with water, andthen with ether, gives the pure dioxide of iodine. Owing to itsinsolubility and its decomposition on heating, its molecular weightcould not be found. By the direct action of sulphur trioxide oniodine dioxide, a solid compound, I;O4,3S0,, was obtained.A substance which is possibly another oxide of iodine has beenobtained by F.Fichter and F. Rohner 84 by the action of ozonisedoxygen on iodine dissolved in chloroform. A yellowish-white pre-cipitate is produced, which is very sensitive to moisture and readilydeliquesces to a black syrup. This ready absorption of water seemsto distinguish it from the dioxide described above. The per-centages of iodine and oxygen determined by the analysis of thisproduct add up to only 97, but on the results of these analyses theauthors ascribe the formula I,O, to the substance. There mustremain some doubt as to its real constitution until a purer producthas been obtained. Chloroform very strongly retained seems t obe the impurity.Group VIII.The mechanism of the process of the cementation of iron has beenlong doubtful.Contributions towards the solution of this problemare made in two recent papers. G. CharpyB5 has shown that irontakes up carbon from carbm monoxide, but whether iron carbonylwas first formed was not investigated. It has been believed thatcarbon can be taken up directly when heated with iron, but theexperiments on which the belief was based have not been unexcep-hionable. A new series of experiments has been published, in whichgreat care has been taken to exclude the possibility of gaseouss3 Trans., 1909, 95, 656.84 Ber., 1909, 42, 4093 ; A., ii, 991.Compt. rend., 1909, 148, 560; A., ii, 40554 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.compounds either being present or being formed.Strgar-charcoal,purified by heating in chlorine, was heated with iron t o 1000° ina vacuous glazed porcelain tube, the iron being freed from occludedgases by a preliminary heating in a vacuum. I n these circum-stances, no cementation took place. Since, however, it was possiblethat the contact of the two elements was insufficient, they werepacked into a steel cylinder provided with a finely perforated screwstopper. I f the scopper was loose and the cylinder and its contentswere heated in the vacuous tube, it was again impossible to findany evidence of union. When, finally, t.he cylinder was treatedin the same way, with the screw stopper exerting great pressure onthe mixture, as much as 0.3 per cent. of carbon was taken up bythe iron.The conclusion, therefore, must be drawn that undermanufacturing conditions, the cementation of iron takes placeentirely through the agency of gaseous compounds. The passimstate of iron is a phenomenon which, like the rusting of the samemetal, seems to be of perennial interest. The existence of a thinfilm of oxide on iron, which has been rendered passive by any of thewell-known methods, has been denied on optical grounds.86 It isshown,87 however, that by making bright iron the anode in theelectrolysis of strong sodium hydroxide, the passive state is inducedby the formation of a layer of oxide so thin as not sensibly tointerfere with the reflecting power of the surface.I f the force which binds water of crystallisation to salts is distinctfrom ordinary chemical attraction, the same thing probably appliesto the union of substances with hydrogen peroxide. Many com-pounds have been obtained by means of hydrogen peroxide, in whichit is hard to say whether a real peroxidised compound has beenprepared, or if the hydrogen peroxide is in a state of associationsimilar to that of water of crystallisation. A peroxide of nickel88has been prepared by the addition of a cooled alcoholic solution ofpotassium hydroxide to a mixture of nickel chloride and hydrogenperoxide at -5OO. It is a greyish-green powder, having the com-position NiO,,sH,O, which gives all the reactions of hydrogenperoxide, easily liberating this substance when treated with diluteacids. For this reason, the authors believe that it has a different;constitution from that of the black nickel dioxide, which is obtainedby the action of sodium hypobromite on nickel hydroxide,S. Tanatar,8g however, finds that the dioxide prepared by the latter86 W. J. Miiller and J. Konigsberger, Zeitsch. Elektrochem., 1907, 13, 659 ;A., 1907,ii, 924.87 P. Krassa, ibid., 1909, 15, 490 ; A . , ii, 738.88 G. Pellini and D. Meneghini, Zeilsch. anorg. Chem., 1908, 60, 178 ;88 Ber., '1909, 412, 1516; A., ii, 484.A., ii, 50INORGASIC CHEMISTRY. 55process will reduce potassium permanganate, and will also givehydrogen dioxide when treated with hydrogen cyanide. There are,however, other differences between the substances, and Tanatarregards the greyish-green dioxide as a molecular compound,NiO,H,O,. At present there is no evidence to show which of theformulze is the true one.A new oxide of platinum has been discovered by L. Wohler andF. Martin.90 By the electrolysis of a solution of platinic hydroxidein 2N-potassium hydroxide, a golden-yellow crust forms on theanode. This was found to have the composition K20,3Pt0,. Bythe action of cooled acetic acid, the trioxide of platinum wasobtained as a brown powder. The trioxide readily loses oxygen,and it slowly liberates chlorine from dilute hydrochloric acid. Itis without action on dilute sulphuric, nitric, and acetic acids, whilstconcentrated sulphuric and nitric acids convert it into the dioxide.It does not decompose hydrogen dioxide.H. B. BAKER.Ber., 1909, 42, 3326 ; A., ii, 898