INORGANIC CBEMISTRY.MANY circumstances make it a matter of considerable difficulty togive within the limits of a Report such as this anything like a fair,general review of the year's progress in the department of InorganicChemistry. The field of investigation to be traversed is so large anddiversified, the ends sought and the methods of working adopted are SOvaried, that it is scarcely possible to group the results into a moderatenumber of general sections, each of which shall form in itself a moreor less consistent whole. I n inorganic chemistry it is comparativelyrare to find what is so common in the organic division, a considerablenumber of independent workers combined in a common attack on awell-defined problem, such as the elucidation of the constitution ofsome prominent group of substances, or the applications of some generalmethod of synthesis.Even when such is the case, as, for example, inthe investigation of the rare earths, the results are often so muchmere matters of detail, or are so restricted in their interest, that anextended discussion of them would scarcely be justifiable.Hitherto the method adopted in this section of the Report hasbeen to review the year's output of work under the individual elements,grouped according to the periodic arrangement; but, whilst it is inthis way possible to touch on most of the important contributions,the result is very apt to be scrappy and disconnected. On the presentoccasion an attempt has been made to approach more to the method ofgrouping referred to above, by which a somewhat more connected kindof treatment is possible. As a consequence, however, it is not possiblein the same space to refer t o so many individual matters, and it isfrequently difficult to make what might be generally approved as afair or judicious selection, so that to many the result may appear ill-balanced and unsatisfactory.The method adopted in former yearshas not been altogether discarded, and a number of more or less dis-connected facts, which might prove interesting in spite of theirisolation, have been included on the same plan as before.Constitution.It is satisfactory to find t h a t the problems regarding the constitu ionof inorganic compounds (or rather, the question of chemical constitiu32 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.tion in general) are receiving more and more the attention whichthey not only deserve, but require.Even at the present day it is notuncommonly taken for granted that because the compounds dealtwith in organic chemistry are generally much more complex thanthose encountered in inorganic chemistry, therefore the problems ofconstitution are also much more profound; this, however, is not so,the problems in the two cases are different in their nature. I n thecase of organic compounds the tendency to form carbon-carbonlinkings certainly leads to a great multiplicity of substances possessinghigh molecular weight and complicated structure, but even in suchcases the comparatively simple conception of a fixed valency can bemade toaccount for nearly all the observed facts.Carbon forms, forexample, no compouiids of the type M,CX,, although every otherelement in the same group does. I n the study of inorganic chemistry,on the other hand, the ordinary conception of valency proves a brokenreed almost from the beginning, and insistence on it tends only t ohamper progress. The difficulty of giving any generally acceptablerepresentation of the constitution of compounds such asK,O,, KIT, KBF,, etc.,in spite of their apparently simple composition, shows how far we stillare from possessing a workable hypothesis as to the nature of chemicalcombination between elements.Seeing that the next great development of chemical theory may beexpected t o come from the study, not of the compounds of carbon, butof those derived from other elements, it is rather unfortunate that thedifficulties to be overcome are not infrequently slurred over insteadof being boldly faced, and a very large proportion of students maketheir first acquaintance with constitutional problems only in thedomain of organic chemistry.It is somewhat difficult for those whoturn their attention t o the problems of modern inorganic chemistryonly after they have become thoroughly imbued with the spirit oforganic chemistry, to approach the subject with a sufficiently openmind ; indications of the effects of this are apparent in some of thepapers which appear from time to time,During the past year the matter of valency and constitution hasbeen dealt with from the electrochemical point of view in thePresidential Address to the Society,’ and has also been thesubject of several ordinary contributions to the Journal.2 Thesubject is also treated very fully in the latest addition to theseries of “Text-books of Physical Chemistry,” edited by SirW.Ramsay.3 A perusal of this book shows how diverce in someTram., 1908, 93, 774.J. N. Friend, The Theory of Yalcncy.* J. N. Friend, ibid., 260, 1006; H. C. Briggs, ibid., 1564INORGANIC CHEMISTRY. 33respects have been the views propounded from time to time con-cerning the constitution of many of the simplest compounds, andgives an idea of the present somewhat chaotic state of affairs.Even were it only for this reason, the appearance of the work ismatter for congratulation.The Indexing of Inorganic Xubstances.The subject of the nomenclature and classification of inorganic com-pounds is one of great and ever-increasing difficulty, and this has mxdeitself felt very considerably i n connexion with the preparation of ageneral index to the first fifty volumes of the Zeit.schrift f u ranorganische Chemie, in which journal have been published so many ofthe results of recent investigations on complex inorganic substances,A scheme has been adopted which presents many departures fromformer practice, and will doubtless arouse a considerable amount ofinterest, and also of criticism; it is described by A.Rosmheim andJ. Koppel in the preface to the recently-issued index, and also formsthe subject of an appendix to a recent number of the Zeitschrzjl.4 Theauthors point out that, a t first sight, some modification of the methodwhich Richter has so successfully applied to the indexing of organicsubstances might be expected to give equally good results in thisdepartment, but they state reasons why this is not so.For one thing,the chemist who has to establish the identity of some inorganic sub-stance which he has obtained in the course of his work does not., as arule, proceed to make first a complete ultimate analysis of it, butdepends more on the results of qualitative analysis as a guide in hisquest; in preparing an index, thereFore, it is necessary to take intoaccount, not merely the individual elements present in the substsnces,but also the well-mtrked radicles or groups, and an arrangement inwhich exact quantitative composition played an essential part wouldbe of comparatively -slight benefit.To apply Richter’s methodsystematically to inorganic compounds, would practically mean aseparate ‘‘ Lexikon ’’ for each element in turn.The limited space which is here available does not suffice to giveany adequate idea of the scheme as a whole, much less t o discuss i t ;according t o the authors, it is based primarily on : (1) the qualitativecomposition of the s ibstances; (2) the valency of the individualelements ; (3) the electro-a6nity of the individual constituents(elements or radicles) of the compound<. I n order, as far as possible,to prevent allied substances from being separdted by tlie alphabeticorder, the method of printing prefixes, like meta-, pw-, oxy-, etc., initalics and ignoring them in the arrangement is adopted on an exten-4 Ein Yerfahren zur Begistrierunq nnorqnnischer Stofe, 1908, 60, Lfg.3.REP.-VOL. V. 34 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.sive scale. For a similar reason, names of Latin origin with termina-tions to indicate degrees of oxidation, etc., are entirely done awaywith, and in their place the ordinary name of the element is used inconjunction with numbers ; thus, magnelic oxide of iron is to be soughtunder Itisen, and would appear as ‘‘ 3-Eisen-4-oxyd.” From thisexample it will be seen that many names become merely paraphrasesof the formulz, and it seems a pity that the Latin names for theelement have not been retained ; for the present purpose, ‘‘ 3-Ferrum-4-oxyd ” would serve equally well, and, by adopting this method, thesymbol of the element might frequently be employed as a contractionwithout affecting the order, and so many of the longer names could beshortened.A system of indexing for general use ought to be based onnames which correspond with a universnlly-adopted set of chemicalsymbols; but then we have not yet reached even this stage. Whetheror not the system adopted for the present index is a success can bedetermined only by experience in actual use, and there has not yetbeen time for such a t e s t ; in any case, i t may be considered an inter-esting experiment, and if it prove a success, or form a starting-pointfor the ultimate achievement of success, it will certainly constitute animportant contribution Lo the progress attained during the pastyear.New Elements.It was mentioned in last year’s Report (p.38) that “ytterbium ”had been shown by Urbain to contain two distinct elements, to whichthe names of Zutecium and neo-ytteinbium had been assigned. A similarconclusion has been reached by von Wel~bach,~ who proposes for theelement with lower atomic weight the name alclebnraniwn (Ad =:172*9), and for the other, cassiopeiunz (Cp= 174.23). Urbain,6 ingiving further particulars regarding his separation, states thatvon Welsbach’s elements are identical with those obtained by himself,and claims priority.7 Whilst the occurrence of a new element in therare-earth group thus seems definitely proved, Urbain has now shownthat, on the other hand, appearances which seemed formerly toindicate the existence of other elements were illusory ; the phospho-rescent spectra of Crookes’ ionium and incogniturn are reproducedby mixtures of salts of gadolinium and terbium, whilst Bayer’sbauxium, from the bauxite of Tar, was a mixture of vanadium andtungsten with traces of other known elements.Monatsh., 1908, 29, 181 ; A ., ii, 591.Compt. rewd., 1908, 146, 406 ; A., ii, 283.7 The International Committee on Atomic weights has given eiiect to this, andCompt. rend., 1907, 145, 1335 ; Bull. SOC. chi?n., 1907, [ivJ, 1, 1158 ; A . , ii,adopted the name Iwtecium (Lu= 174).108INORGANIC CHEMISTRY.35Although the evidence is at present somewhat meagre, i t seemshighly probable that at least one new element has been discovered inthe mineral thorianite. Miss Evans,g in working through the residuesfrom about five hundredweight of the mineral from Ceylon, obtainedindications of the presence of an element giving a brown sulphidesoluble i n ammonium carbonate, but insoluble in hydrochloric acid,and difficult to oxidise by means of potsssium chlorate and hydro-chloric acid. This sulphide was separated, and by treatment withnitric acid a brown oxide was obtained, which was easily reducible bymeans of hydrogen, yielding, first, a black lower oxide, and finally,a dark grey, non-volatile metal. Only about fifty milligrams of thebrown oxide were obtained, ttie yield being less than at the rate of1 gram from a ton of the original mineral.Owing t o the smallquantity available, little could be done in the way of establishing theprecise nature of the substance, but an attempt to determine theapproximate equivalent weight of the element by reduction of theoxide in hydrogen seemed to indicate a value well above that ofarsenic, with which the element may be supposed to present certainanalogies.A preliminary description of several new elements, or compounds ofnew elements, is also given by M. Ogawa,lo who obtained them, notonly from thorianite, but also from molybdenite and reinite. Forone of these he proposes the name nipponium (Np); i t s equivalentweight is about 50 and atomic weight about 100, and it is thereforesuggested t h a t it may occupy a place above manganese, betweenmolybdenum and ruthenium.It forms a basic lower oxide and a nacidic higher one; the first, after ignition, is brown in colour, and thechloride formed from it gives a green solution. A second elementwith an equivalent weight of about 16.7 is also described as givingtwo oxides ; the higher resembles molybdic anhydride, and formssalts of lead, barium, and silver, which are similar t o the molybdatesof these metals. The oxides can be reduced by hydrogen, yieldinga metal which does not fuse at a red heat, which burns brilliantly inair, and is soluble in hydrochloric acid. Whether or not the elementdescribed by Miss Evans corresponds with either of these elements, orperhaps with a mixture of them, i t is difficult to say at present.Ogawa obtained indications of the presence also in thorianite ofa third new element, which yields a radioactive oxide.TTans., 1908, 93, 666.lo J. CoU.Sci. T6kg6, 1908, 25, xv, 1 ; xvi, 1 ; A., ii, 952, 95336 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY,Atosnic TJ’eiglht8;.I n this department there has been no very great advance duringthe year. I n the case of the more important elements, there havebeen one or two re-determinations, practically repetitions of formerwork with improved precautions, whilst several workers hare extendedtheir investigations to some of the less prominent elements. Thewhole subject of atomic wei hts is being very fully discussed byBrauner in Abegg’s Hccfidbuch, and in the volume, which appearedduring last year,ll he contributes a general article on ‘‘ fundamental ”atomic weights, namely, those of sodium, potassium, silver, chlorine,bromine, and iodine, using the oxygen standard.By this method ofcollective treatment, instead of the individual treatment adopted forthe other elements, Braune is able to give a much clearer and moreconcise review of these interdependent determinations. He discussesthe results critically, and is not infrequently distinctly a t variancewith the decisions of the International Committee ; his articles arewell worth studying by those interested in this subject. A volume isannounced, which should prove of great value to the same classof reader, and to those specially concerned with analytical chemistry ;this is a collected edition of the results of the atomic weight deter-minations carried out by T.W. Richards and his pupils during thelast twenty-one years. It is being issued in German,’2 and is aresult of the interest aroused by the visit of Professor Richards toBerlin as Exchange ” Professor.The two most important re-determinations during the year are thosefor hydrogen and for chlorine. I n the former case, W. A. Noyes13has renewed experiments on the combining proportions of hydrogenand oxygen in order to test the propriety of, or eliminate the necessityfor, the corrections which he had applied to his previous determina-tions on account of occluded gases, and which brought his value intoclose agreement with that obtained by Morley.The hydrogen wasobtained by olectrolysis, generally from dilute sulphuric acid ;in some experiments it was burned directly, in others itwas first absorbed in palladium; in some it was burnedby means of cupric oxide, in others directly by oxygen (whichwas also prepared electrolytically), using palladium. I n one setof experiments of this latter class, the hydrogen and oxygen wereboth prepared by electrolysis of barium hydroxide solution in place ofdilute sulphuric acid. The results obtained in the different sets of11 Handbzcch d. anorgan. Chemvk, Bd. 11, Abt. 1.12 Expcrimenlnlle Untcrsuchitngcn iil‘w Atonzgezoichte (Voss).13 J. Aqncr. Chem Soc., 1907, 29, 1718 ; A., ii, 100INORGANIC CHEMISTRY.37experiments varied from 1.00771 to 1.00812, m i l give as the probablevalue, 1.00787 ; Noyes, taking:this value in conjunction with that ofMorley (1.00762), consii1ei.s that the mean, 1.00775, would be areasonably trustworthy value.Noyes, along with Weber,l* has also carried out a re-tletermination ofthe Cl/H ratio; the hydrogen was weighed in palladium, passed overweighed potassium platinichloride suitably heated, and the hydrogenchloride produced was collected in water, either directly or, in one setof experirnents, after having first been frozen by means of liquid air.These two sets gave similar results, and in each the two values,HCl/H and Cl/H, were obtainable. The final results give C1= 35.452or C1=35.461, depending on whether the value H=1.00762 orH = 1.00787 (see above) is taken.A set of chlorine determinationshas also been made by Edgar,l5 who burned chlorine in an atmosphereof hydrogen, using silica-ware apparatus, and condensed the hydrogenchloride by means of liquid air ; the mean results of eight experimentswere : for ratio Cl/H, directly 35.194, and from ratio HCl/H, 35,193.For H = 1.00762 these give C1= 35.462 and 35.461. This is in closeagreement with the previous determinations by Dixon and Edgar, andwith Guye’s value deduced from the density.The questions of the homogeneity of tellurium, and of the trueatomic weight of that element, have again been the subject ofinvestigation ; all the recent work goes to confirm the view that thereis not the slightest evidence of lack of homogeneity in the ordinarypurified substance.Marckwald,lG after several hundred systematicfractional crystallisations of 1500 grams of telluric acid, could find nodifference between the first and the last fractions. Lenher17 also,when converting tellurium or its oxide into chloride or double chloride,could find no difference between the element obtained from thecrystallised product and that from the residues, nor could he find anybetween the different specimens obtained by fractional precipitationfrom the tetrachloride by means of ferrous salt. The two observersdo not agree, however, with regard to the atomic weight of theelement. Marckwdd determined the proportion of tellurium dioxideobtained by the regulated heating of telluric acid, specially purified,and carefully dried to constant weight in a vacuum; in this way heobtained the value Te = 126.85, which is very decidedly lower thanthat obtained by Baker,I* namely, 127.60, and is slightly below the14 J.Amer. Chem. SOC., 1908, 30, 13 ; A, ii, 371.15 Nem. Manchester Phil. ii’oc., 1908, 52, No. 7; A., ii, 577.l6 Ber., 1907, 40, 4730 ; A., ii, 33.17 J. Amer. Chm. Soc., 1908, 30, 745 ; A . , ii, 483.l8 Ann. Report, 1907, 3638 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.value for iodine. Lenher, on the okher hand, from the ratio Te : TeO,,obtained the value Te= 127.5, which is in reasonable agreement withBaker's value. I n connexion with Marckwald's low result, it hasbeen pointed out 19 that his method of determination is untrustworthy,owing to the practical impossibility of preparing telluric acid con-taining exactly the quantity of water to correspond with the formulaH,Te0,,2X120.At present, therefore, the balance of evidence seemst o be decidedly in favour of the anomalous positions of iodine andtellurium in the usual periodic arrangement of the elements.A re-determination of the atomic weight of 1ead2O has been madeby the analysis of the chloride prepared by repeated crystallisation inplatinum vessels from solutions containing free acid ; the product didnot discolour on heating, and mas finally fused in an atmosphere ofhydrogen chloride. The proportion of silver, as nitrate, necessary forprecipitation was determined, and also the quantity of silver chlorideformed ; these gave closely concordant results.The mean of thesegives the value P b = 270.190, as compared with the present acceptedvalue, 206.9.Guye's method of calculnting atomic weights from accurate deter-minations of gaseous densities in conjunction with the gas constantshas been applied in the case of hydrogen sulphide21; the result,S = 32.070, thus obtained is an additional indication of the accuracy ofthe method.New determinations of the gaseous densities of krypton andxenon have been made, and from these the values Kr=83*01 andXe= 130.70 are obtained for the atomic weights, assuming the gasest o be monatomic. The former is distinctly higher, and the latterdistinctly lower, than the presently accepted value, but the changesdo not affect the positions of these elements relatively to others.22Other atomic weights re-determined 23 are those of bismuth andpalladium, but these need not be entered upon here; the valuesobtained are not widely different from those in present use.AIEotropy of Elements ; Moleculur Constitution.The relations between the different modifications of elements whichcan exist in more than one form, and the explanations which may beadvanced as to the causes of the wide differences which these oftenl9 Baker, Chew.Arms, 1908, 97, 209 ; A., ii, 483.20 Baxter and Wilson, J . Amer. CAem. Soc., 1908, 30, 187 ; A,, ii, 281.21 Bauine and Yerrot, J. Chinz. Phys., 1908, 6, 610 ; A., ii, 372, 940.22 R.B. Moore, TTCL~S., 1908, 93, 2131.23 Gntbier and Birckenbach, J. pr. Chenz., 1908, [ii], 77, 457; A . , ii, 600.Kemmerer, J. Awzcr. Chern. Li'oc., 1908, 30, 1701 ; A . , ii, 1046INORGANIC CHEMISTRY. 39exhibit, are problems which attract an increasing amount of attention,and there seems to be a growing tendency to refer these differencesto varying degrees of molecular complexity, similar to that so wellknown in the case OF oxygen. As the other elements concerned,however, give only one kind of vapour, evidence as to differencesof molecular complexity is generally wanting, and conjecture insome cases a t least is allowed to take its place.Further investigations into the nature of molten sulphur continueto be made, and, so far, A. Smith’s explanation of the peculiaritiesexhibited by this substance 24 receives at least general c~nfirmation.~~The suggestion is now made by H.Erdmann26 that the formation ofa second distinct liquid phase is due t o a partial splitting up of S,molecules, with formation of S , molecules, The dark variety ofliquid sulphur (Sp) is held to consist of these S, molecules, because itis supposed to be a particularly reactive form of the element, andozone, the particularly reactive form of oxygen, is known to havethe composition 0,. It is even suggested that the name thioxonemight be given to this modification, and that many polysulphidesand organic sulphur derivatives might be looked upon as additiveproducts formed by it, and should therefore be called thioxonidesand polylhioxonides.These assumptions are all made without anyevidence whatsoever as to the molecular weight of S p , and in spiteof the fact that all determinations hitherto made indicate that thesulphur molecule is S , at comparatively low temperatures, butdissociates into 8, molecules when strongly heated, giving noindication of intermediate stages. (From its effect on the freezingpoint of SX, Smith believes that Sp is also composed of S,molecules.)Similarly unfounded assumptions are made regarding variousmodifications of ar~enic.~7 The only molecular-weight determinationsfor this element which have been effected lead to the formulaas4, yet not only is it suggested that various solid modifications ofarsenic must be represented as Ass, As4, As2, and As respectively, butgraphic formulze for the first three are written, with single, double,and treble linkings, and a ring constitution for the first two.Itis assumed that the monatomic molecule must be characteristic ofthe “metailic” variety, for the simple reason that typical metaIsare monatomic; and yet, if chemical evidence counts for any-thing, the ‘‘ metallic ” variety of carbon is presumably highly poly-atomic.24 Ann. Bepod, 1907, 61.2d H. E,. Kruyt, Zeilsch. physiknl. Chcm., 1908, 64, 513 ; A . , ii, 1028.26 Anaalen, 1908, 362, 133 ; A . , ii, 830.37 H. Erdmann, 1-11zncdc?;, 1908, 361, 1 ; A., ii, 55440 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.The relations between the different forms of phosphorus naturallyoccupy a prominent position in this section, and cause considerablediscussion.But, whilst there has been a good deal of wordy warfare,little real progress has resulted. Several papers deal with thecommon white and red forms,28 and a number of others with‘‘ Hittorf’s phosphorus,” which is obtaingd in. a crystalline conditionby the prolonged heating of phosphorus with lead. After a con-siderable amount of discussion as to the exact crystalline form ofthis variety of “phosphorus,” it is now stated that the substancemay contain fully half its weight of lead.29 Supposed amorphousforms of other members of this group-antimony and bismuth-whichhad formerly been described .as being produced by heating thecrystallised metals in an atmosphere of nitrogen, are now stated tobe mixtures produced by partial oxidation of the metal, resulting inconsequence of lack of purity in the nitrogen. On the other hand,the supposed crystalline form of ‘(boron,” obtained by a thermitereduction process,3* always contains aluminium, and corresponds withthe formula AlB,, ; apparently, therefore, the preparation of pureboron in a crystallised form is still unaccomplished.Further attempts to induce the transformation of graphite intodiamond, by crystallisation under very high pressures directly applied,have again been quite unsu~cessful.3~The explosive varieties of the platinum metals, obtained by alloyingthe particular metal with a large proportion of zinc and dissolvingthis out from the alloy by means of acid, have now been shownto contain occluded oxygen as well as hydrogen,33 so that theexplosive effect is not to be referred to a sudden allotropic change,but to chemical reaction between the occluded gases.Ruthenium,however, appeared to be exceptional, as this metal was found to beexplosive even when the greatest care was taken to exclude thepossibility of oxygen being present.From time to time attempts have been made to induce in otherelementary gases a polymerisation analogous to the formation ofozone, but without definite result, It had been supposed thatchlorine gave indications of forming a more reactive variety, but itis found that when highly purified chlorine is subjected t o the28 A. Colson, Compt. rend., 1907, 145, 1167 ; 1908, 146, 71, 401 ; A ., ii, 35,176, 273.29 G. E. Linck and P. Moller, Ber., 1908, 41, 1404 ; A , , ii, 48730 E. Cohen and J. Olie, Zeitsch. physikal. Chem., 1908, 61, 586, 696; A.,31 H. Biltz, Bcr., 1908, 41, 2634 ; A., ii, 762.32 R. Tlirelfall, Trans., 1908, 93, 1333.33 E. Coheri and T. Strengers, Zeitsch- physiknt. Chenz., 1908, 61, 698 ; A . ,ii, 299.ii, 198, 199INORGANIC CHEMISTRY. 41influence of the silent electric discharge in an apparatus capable ofindicating a change of volume of 1 in 8000, no appreciable changetakes place ; 34 the increased activity sometimes observed may be due.to the presence of oxygen, which could give rise to the production ofozone or of oxides of chlorine.Alloys.Possibly no branch of inorganic chemistry has during recent yearsmade such rapid strides or accumulated such a mass of new and dis-tinct observations as that which deals with the alloys ‘of the metalsand with those allied substances which for convenience may be in-cluded under the more general heading.One reason for this rapidprogress is doubtless to be found in the technical demand for steels,and similar products, possesing special properties, but perhaps thechief reason has been the great improvement in appliances and pro-cesses (as, for example, in connexion with electrical heating andpyrometry), which have made investigations of this kind more accurate and more suitable for ordinary laboratory work, together with thephysico-chemical developments which facilitate a proper interpretationof the observed phenomena.As is pointed out by G. Charpy 35 ina lecture dealing with the recent progress in connexion with one singlesection of the work, the system Fe-C, investigations in this fieldinclude the following branches : (a) deduction of an equilibriumdiagram by the application of the phase rule, ( 6 ) thermal investiga-tions, (c) isolation of definite compounds from the alloys by chemicalprocesses, (d) microscopic examination, (e) observation of the physicaland chemical properties of the alloys. The discussion of these variousmatters, as given in the lecture, is of general applicability, and mightprove useful to anyone desiring a concise resume of the subject. Cer-tain of these branches are, of course, interesting more particularly tothe physical chemist, the metallurgist, or the engineer, but the isola-tion of definite compounds comes very directly within the scope of theinorganic chemist, and the interest becomes still more pronouncedwhen, as has not infrequently been the case, the systems dealt withare not those of two typical metals, b u t include a metal and a non-metallic or ‘‘ semi-metallic ” element with which distinct compoundsare formed in the ordinary course.From the numerous results ofquite recent work, a few illustrative examples may be selected, moreor less at. hazard.Some pairs of elements, although exhibiting more or less completemiscibility in the liquid state, show no formation of definite compounds34 E. Briner and E. nurand, Zeitsch.Elcktrochem., 1908, 14, 706 ; A . , ii, 940.36 Bull. Soc. ckim., 1908, [iv], 3, 1 ; A . , ii, 69742 ANNUAL REPORTS ON TITE PROGRESS OF CHEMISTRY.and not even of mixed crystals to any noteworthy extent ; an exampleof this is given by the system A1-Si.36 These elements mix com-pletely in all proportions when fused, and the system gives a simplefreezing-point curve of two branches, with eutectic point. I n othercases the complete miscibility extends to the solid state, resulting inthe formation of homogeneous alloys ttiroughout, as in the systemFe-V.S7 I n contrast t o these, some elements which, being closelyallied, might have been expected to form mixed liquids or mixedcrystals, do not do s o ; this is the case with arsenic and bismuth,which exhibit only slight miscibility when fused, and separatecompletely on solidifying.38There are numerous examples ol the formation of ingredients which,from the point of view of chemical composition and otherwise, appearto be definite compounds, even in the case of elements where thiswss hardly to be expected; an interesting example is providedin the system Molten thallium easily dissolves spongyplatinum, and when mixtures which contain an excess of thallium arecooled, the resulting alloy shows a fine eutectic surrounding largercrystals.I f the proportion of platinum is kept below 10 per cent.,these crystals can easily be isolated by means of dilute nitric acid,and, when analysed, they prove to have the composition correspondingwith the formula TIPt.This compound loses part of its thallium onstrong heating, but retains some even a t the temperature of the oxy-hydrogen flame; it displays great similarity to the analogous leadcompound PbPt. Examples of this kind, of which there are many,show that in dealing with alloys the ordinary chemical equivalents(that is, electrochemical equivalents) have no general significance.The isolation of compounds is often a very difficult problem, owicg t othe lack of suitable solvents, which, while attacking the other in-gredients, will not also attack the compound itself. A good exampleOF such a difficulty being successfully overcome is provided by t h eisolation of magnesium silicide, Mg,Si (which is decomposed even bywater), from an alloy rich in magnesium; the excess of metal in thiscase was removed by treatment with a mixture of ethyl iodide andether.40I n the case of ‘‘alloys” derived from elements like sulphur,selenium, phosphorus, etc., which with metals form compounds of theordinary type, the results obtained by the methods under considera-tion are often much less simple than might be expected ; a t the high36 W.Fraenkel, Zeilsch. mtorg. Chem., 1908, 58, I54 ; A . , ii, 592.37 R. Vogel a d G . Tammann, ibid., 73 ; A . , ii, 502.38 K. Friedrich and A. Leroux, Metallz~~gic, 1908, 5, 148 ; A . , ii, 148.39 1,. Hackspill, Cmnpt. rend., 1908, 146, 820; A . , ii, 504.P. Lebeau and R. Bossiiet, ibid.,‘252 ; A., ii, 184INORGANIC CHEMISTRY. 43temperatiires of fusion, known compounds may become unstable andnew compounds be produced, with proportions totally different fromthe others.Thus, the only nickel sulphide that is stable in contactwith the liquid phase of the system NCS is the compound Ni,S, (atatmospheric pressure the maximum amount of sulphur present is31 per cent., owing to volatilisation), but, on cooling, transformations,take place in the solid, and the existence of the three compoundsNiS, Ni3S4, and NiS,, is clearly indicated; there is no Ni,S,although a compound, 2FeS,Ni,S, can be formed when iron ispresent.41A few years ago it was found that certain alloys, although formedfrom non-magnetic elements, w e very distinctly magnetic, and con-siderable attention is now devoted to the observation of this property ;as noted elsewhere, it has now been found that some nitrides, oralloys of nitrides with excess of the metal, are magnetic in a verypronounced degree.A branch of investigation, very closely allied to the study of alloys,deals with the problem of the separation of silicates and similar sub-stances from fused magmas, and a fair amount of work is beingsteadily accomplished in this connexion.The subject is of greatinterest petrologically, from the bearing which it has on the eolutionof problems concerning the formation of igneous rocks, etc. Theszlmegeneral methods are also occasionally used to throw light on otherquestions ; thus it has recently been shown that selenium and iodineform neither compounds nor mixed crystals ; 42 also, that sulphur andiodine form no compounds, and give only one series of mixed crystals,containing only moderate percentages of sulphur. This is a matterof some interest in connexion with the effect of iodine on theequilibrium between SX and S P .~ ~Xeduction of Oxides, etc.The reduction of certain metallic oxides by mere raising of thetemperature is a chemical process to which the student is very earlyintroduced, but until quite recently very little attention was paid tothis matter as a general phenomenon, and i t was only in isolated casesthat the reversibility of oxidation processes and the part played by theoxygen concentration in the surrounding atmosphere were properlytaken into account ; as a rule, also, the cases specially considered fromthis point of view were those in which reduction takes place from a higherI<.Bornemann, illctallwqie, 1908, 5, 13, 61 ; A . , ii, 292.*2 G. Yellini and S. Pedrina, Atti B. Accnd. Lincei, 1908, [v], 17, ii, 78 ; d.,43 F. Ephraim, Zeitsch. a x o y g . C h n . , 1908, 58, 338 ; A., ii, 581.ii, 83344 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.to a lower oxide only, the case of mercury being treated as exceptionalwith regard t o the reversibility of the action between metal andoxygen. I n view of a number of recent experimental investigations,however, i t is becoming necessary to deal with the question in a moregeneral and comprehensive manner ; the results obtained show thatmany " stable " oxides: can easily undergo dissociation at not veryhigh temperatures if the other conditions are suitable, that is,provided the pressure or concentration of oxygen be kept sufficientlylow. For example, in the cathode light vacuum,44 cadmium oxideat 1000' dissociates into metal and oxygen, lead oxide does so at 750°,and bismuth oxide a t a still lower temperature; some sulphides alsodissociate, giving sulphur and metal or lower sulphide.For anumberof oxides the dissociation pressures at various temperatures have beenmeasured,45 and where possible also the temperatures have been ascer-tained at which the pressure reaches that of oxygen in the air, thisbeing, of course, the temperature a t which decomposition would setin under ordinary conditions. I n air,cupric oxide would begin to decompose in the neighbourhood of 1025'with formation of oxygen and cuprous oxide; but this is a much morestable compound, for even at that temperature its dissociationpressure is not more than 1 mm.Other determinations of a more orless similar kind have been made by other observers regarding theoxides of chromium (alone, and in presence of copper 0xides),4~1nanganese,~7 iron,4s amd iridium.49I n view of the dissociation phenomena just referred to, themechanism of the reduction of oxides by means of, say, carbon wouldappear to be uncertain; is the oxide directly attacked by the reducingagent, or does this merely act as a kind of absorbent for the oxygenliberated by dissociation ? This problem also has been recentlyinvestigatedy50 and it, would appear that, in many cases at least, carbonacts directly on the oxide ; evolution of gas takes place at tempera-tures f a r below those at which direct dissociation can be detected inabsence of carbon.It has recently been satisfactorily proved that evenmagnesium oxide can be reduced by carbon at a temperature of 1700" ;this had previously been surmised from observed facts,51 but has now44 Damm and Krafft, Ber., 1907, 40, 4775 ; A., ii, 39 ; see also W. von Bolton,45 H. W. Foote and E. K. Smith, J. Amer. Chcm. SOC., 1908, 30, 1344 ; A., ii,46 L. Woliler and P. Wohler, Zeitsch. physikal. Chem., 1908,:62, 440 ; A., ii, 387.47 R. J. Meyer and K. Rotgers, Zee'tsch. anoyg. Chem., 1908, 57, 104 ; A., ii, 191.48 P. T. Walden, J. Amer. Chem.Soc., 1908, 30, 1350 ; A., ii, 852.49 L. Wohler and W. Witzmann, Zeitsch. Elektrochem., 1908, 14,97 ; A . , ii, 301.50 H. C. Greenwood, Trans., 1908, 93, 1483.51 Lebeau, Compt. rend., 1907, 144, 799; A., ii, 1907.One example may be cited :Zeitsch. angew. Chem., 1906, 19, 1537.847INORGANIC CHEMISTRY. 45been demonstrated very conclusively 52 by dissolving the resultingmagnesium vapour in metallic copper (which itself has no action onmagnesia), and, still more conclusively, by condensing it as a metallicmirror on tho walls of an evacuated glass vessel.A study of the rates of reduction of the oxides of lead, cadmium,and bismuth by means of carbon monoxide demonstrates clearlythat lower oxides of these metals exist as definite compounds, a matterwhich formerly wits in some doubt.53 When the results are plotted,breaks in the resulting curves indicate very clearly the formation ofthe compounds Pb,O, Cd,O, and BiO. From results obtained by atotally different method, it appears, further, that these are basicoxides, yielding ions in solutioii; this method consists in showingthat the ordinary salt solutions can dissolve and re-deposit theappropriate metal.It is known, for example, that when copper isbrought into contact with solution of cupric sulphate, an equilibrium,Cu + CuSO, ZZ Cu,S04, is established ; rise of temperature shifts thisquite appreciably towards the right, and a hot solution deposits copperon cooling.54 By arranging a circuln tion apparatus, in which solutionof a suitable salt was brought into contact with metal at a heatedpart and mas then cooled at another, deposits of metal were obtainedin the case of lead, cadmium, bismuth, and thallium.55 These experi-ments, whilst pointing distinctly to the formation of lower salts, giveno evidence as to the composition of these, but it may be taken forgranted that in the first three cases they correspond with the oxidesreferred to above.With regard to thallium, however, the matter isnot so simple; the ion of the thallous salts is Tl', and a lower saltwould therefore involve the assumption of R compound ion, Tl,' ; thiswould present a certain analogy to the anion of the periodides, Is'.The preparation in ihe solid state of halogen compounds derived froma lower oxide of bismuth has actually been effected by the action ofthe metal on the ordinary halides.It is stated56 that in this way adichloride, BiCI,, can be obtained in distinct crystals, which are lessdense than an equivalent mixture of metal and txichloride mould be,and therefore must consist of a true compound ; the analogous bromideand iodide were also obtained. On the other hand, a physico-chemicalstudy of the whole system, bismuth-chlorine, leads to the conclusionthat BiC1, does not exist, but that BiCl and BiC14 do ; 57 on similargrounds, the existence of a bromide, BiBr, is upheld.52 R. E. Slade, Proc., 1907, 23, 152 ; Trans., 1908, 93, 327.6o F. J. Brislee, Trans., 1908, 93, 154.6d Ann. Beport, 1907, 45.55 H. G . Denham and A. J.Allmand, Trans., 1908, 93, 424, 833.s6 W. Hcrz and A. Guttniaiin, Zeitsch. nnorg. Chenz., i908, 56, 422 ; A , , ii, 198,5' B. G. Eggink, Zeitsch. physikal. Chew., 1908, 64, 449 ; A., ii, 104346 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Although they scarcely fall within this section, two other cases ofnew derivatives being obtained may be mentioned here. A loweroxide of titanium, TiO, has been recorded, but analogous derivativeswere practically unknown; the di-iodide, TiI,, which has now beenprepared as a distinctly crystalline substance,5s is the first definiteexample of these. It is obtained by passing the vapour of the tetra-iodide over heated mercury in an atmosphere of hydrogen, and, beingless volatile than the tetra-iodide and mercuric iodide, is easily obtainedseparately from these in the form of black lamella.It is insolublein organic solvents, and is decomposed by water and aqueoussolutions.The other new compounds referred to are a hydrated sesquioxide ofpalladium (which is thus brought more into line with its two neigh-boure, rhodium and ruthenium), and substances derived from it ; 59 inthe meantime, although the stage of oxidation is known, the exactcomposition of the hydroxide itself is not known, and it must be repre-sented by the formula Pd,03,xH,0. It is obtained by electrolyticoxidation of palladous nitrate, and separates as a brown precipitate atthe anode; the temperature must be kept low, and certain otherprecautions observed. The process can easily proceed beyond thedesired stage, not so much by direct oxidation as by the action of acid,giving dioxide and palladous salt.The new oxide dissolves easily inhydrochloric acid, forming a chloride which is unstable; but bysuspending the oxide in ether along with rubidium chloride or caesiumchloride, cooling with solid carbon dioxide and ether, and then passingin hydrogen chloride, crystalline compounds of the compositionN,PdCl, can be prepared. These are decomposed by water, givingderivatives of the lowel- chloride ; no potassium derivative could beobtained.A striking and peculiar phenomenon involving oxidation and reduc-tion has been observed in connexion with certain vanadium compoundsof the alkali metals and of silver.60 When any of these compounds isheated to a high temperature in air and then allowed to cool, it isfound that " spitting " takes place, similar t o what is so well known inthe case of molten silver; here, also, the effect is due to the escape ofoxygen.It has been proved that, at the high temperature, vanadatesare formed corresponding with the general formula M,O,xV,O, (wherex ranges in value from 2 t o 6), but that these, on cooling, change intodmble vanadyl vanadates of various compositions, part of the V,O,becoming V,O, with loss of oxygen ; this oxygen is re-absorbed whenthe vanadyl derivatives are again heated in air.68 E. Defacqz and H. Copaux, Cornpt. rend., 1908, 147, 65 ; A., ii, 699.69 L. Wohler and F. Martin, Zeitsch. anorq. Chm., 1908, 57, 398 ; A., ii, 392.6o W.Prandtl and H. Murchhauser, ibid., 1907, 56, 173 ; A . , ii, 46INORGANIC CHEMISTRY. 47Peroxides, 66 Per-scc Its," etc.The various substances which may be conveniently classed togetherunder the above heading present many points of general interest, andprovide problems for many of which no very satisfactory solution hasbeen propounded. This is reflected in the fairly considerable amountof recent work in connexion with them.One had thought that the constitution to be assigned to themembers of the two groups of higher oxides, commonly designated asperoxides and exemplified by, say, MnO, and BaO,, had been fairlysatisfactorily settled, as expressed by the general formula: M< 0 and0BE<! ; the former are oxides of the ordinary typz, possessing feeblybasic properties (capable of yielding an ion M"") or feebly acidicproperties, or possibly both, whilst the latter are " salts " of hydrogenperoxide.(The expression pel-oxidate seems to be coming into use,and is in many respects convenient.) It is now suggested, homever,61that the two best-known members of the first group, namely MnO,and PbO,, cannot possess a similar constitution, and that the leadcompound should be represented by the formula P b e ? It is rather0'difficult to see what particular character this formula is intended t oexpress; it would seem to indicate a novel kind of peroxidate, butlead dioxide does not yield hydrogen peroxide. The only reasonadduced for assiguing different constitutions is the fact thatmanganese dioxide yields dithionate when treated with sulphurousacid, whilst lead dioxide does not.I n view of the great resemblancesbetween the two oxides, this reason appears totally inadequate, unlessthere is no other way out of the difficulty; one simple explanationwould be to assume that, whilst manganese dioxide forms a normalsulphite, Mn(SO,),, which rearranges to form manganous dithionate,MnS,06, lead dioxide forms a basic sulphite, PbOSO,, which rearrangesto form lead sulphate, PbSO,.The possible existence of a defiuite higher oxide of silver has oftenformed the subject of investigation, and still continues to do so; inlast year's Report (p. 46), facts in favour of assuming an oxide, Ago,yielding a cation, Ag", were discussed.According to the evidenceobtained from electrochemical determinations,G2 it would now appearthat whilst only the oxide, AgO, can be obtained by electrolytic oxida-tion in presence of alkali, with acid electrolyte both this and a stillhigher oxide, Ag,O,, can be produced, each corresponding with aMarino, Zeitsch. a?torg. Chsm., 1907, 56, 233 ; A . , ii, 106.e2 Luther and Pckoriif, ibid., 1908, 57, 290 ; A., ii, 27748 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.definite electrolytic potential. F he establishment of the existence oftwo higher oxides of silver-not peroxidates-is interesting, and bringsthis metal into closer agreement with its neighbours in the first groupof elements, with copper through the oxide Ago, and with goldthrough Ag,O,.This oxide shows a great tendency t o adsorb silversalts, and so gives rise to the various indefinite substances which atdifferent times have been described as silver peroxide. According t oother workers,63 the dark anodic deposit formed during the electrolysisof silver nitrate solution is silver pernitrate, AgNO,, which decom-poses with formation of these indefinite '' peroxides."Reference may here be made to a new oxide of thallium which hasbeen prepared,G4 corresponding in composition with the formula TI0 ;this substance illustrates very well the problems in isomerism whichmay arise even in connexiori with comparatively simple inorganicsubstances. The oxide is obtained a t a low temperature by the actionof hydrogen peroxide on a solution of thallous sulphate andpotassium hydroxide; a red precipitate forms at first, but thisrapidly changes into a bluish-black substance having the above com-position.Three possible views might be held as to the nature ofsuch a mixture. I n view of the known molecular weight of indiumdichloride, MCl,, the existence of a simple oxide, T1:0, is a t first sightnot improbable ; taking into account more1y;the m e h d of preparation,a thallous peroxide, T1,0,, analogous t o sodium peroxide mould seema likely product ; whilst the generally accepted views regarding manythallium compounds with halogens, etc., would point to a thallous-thallic oxide, TI*O*TI:O, as being highly probable. The most generallysatisfactory assumption to make would be 'that thallous peroxidateis produced in the first instance, but rapidly rearranges into thedouble oxide ; this would he analogous t o the change which thallouspersulphate, TI,S,08, undergoes into the isomeric thallous-thallicsulphate, TL'Tl"'(S04),.Further evidence regarding the action of hydrogen peroxide onmercury and the formation of mercury peroxide65 is given by vonAntropoff 66 ; he also comes to the conclusion t h a t HgO, is mercuricperoxidate, and suggesh that probably a still less stable mercurousperoxidate, Hg,Oe, is also formed when hydrogen peroxide and mer-cury interact.The formation and decomposition of hydrogen peroxide itselfcontinue to be invest'igsted.Nernst'a view that its formation inflames is not due to the combustion process, but is merely ttie resultof the exposure of water vapour (with or without free oxygen) to highBaborovskf and Kuzma, Zeitsch.Elektrochmn., 1908, 14, 196 ; A . , ii, 378.84 Rabe, Zeitsch. nnorg. Ch,ern., 1908, 58, 2366 Ann. Report, 1907, 49.A., ii, 498.66 J. pr. Chem,, 1908, [XI, 77, 273 ; A., ii, 383INORGANIC CHEMISTRY. 49temperature, followed by rapid cooling, is confirmed by the observa-tions of F. Fischer and 0. Ringe,G7 who succeeded in showing the forma-tion of peroxide when the heating was effected by meam of ( a ) a Nernstfilament, ( 6 ) a capillary tube of magnesia, heated in a flame, (c) ahydrogen flame directly, (d) electric spark discharge ; this last resultis contrary to Nernst’s statement, and was secured by maintaining agreater velocity of the current of steam and oxygen so as to ensuremore rapid cooling.A patent 68 has been secured for the actual pre-paration of hydrogen peroxide by the method of blowing variousmixtures of steam, hydrogen, and oxygen through flames or othersuitable sources of heat, the velocity being not less than one metreper second; or the source of heat may be caused rapidly to rotate inthe mixture.According to Abe1,Gg the catalysis of hydrogen peroxide by iodine oriodides is due to the occurrence of the two reactions : (1) H,O, + I, =2H’ + 21‘ + 0, and (2) H,O, + 21’ + 2H’ = 2H,O + I,. The first isgreatly accelerated by alkalis, the second by acids, and by maintainingin the solution a suitably small concentration of hydrogen ions(addition of acetic acid and sodium acetate), the two may be made t oproceed at the same rate, so that then the only apparent change is theliberation OF oxygen.The second of Rbel’s equations might be re-presented as ( a ) H,O, + 21’ = 2HO’ + I, and ( b ) 2HO’ + 2H’ = 2H,O ;(1) and ( a ) together, then, represent more clearly the two chief modesof action of hydrogen peroxide, reducing or oxidising, according asdivision of the molecule takes place between the hydrogen and oxygen,or between the oxygen atoms themselves (mldst the alternation ofboth results in apparently simple decomposition). This behaviour cant o a certain extent be compared to that of an ‘‘ amphoteric ” nietallichydroxide, such as that of zinc, towards alkali and acid respectively :O,H HO:K OH HC1ZnO,H H O K and OH H C1’The catalysis of hydrogen peroxide is apparently not influenced bythe concentration of the dissolved oxygen present, a t least when acolloidal solution of a noble metal serves as catalyst; the rate ofdecomposition has been measured under very considerable oxygenand up to as high as 200 atmospheres no difference couldbe found.The formation of ozone (which we may here treat as a peroxide)from oxygen by means of the silent electric discharge can be renderedmore and more complete by lowering the temperature, and it has beenBer., 1908, 41, 945 ; A ., ii, 370.6* D.R.-P. 197023 ; A . , ii, 829.CD Zeitsch. Elcktrochem., 1908, 15, 598 ; A., ii, 939.70 Spear, J. Amer. Chem.. Soc., 1908, 30, 195 ; A., ii, 370.REP.-’VOL.V. 50 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.found possible,71 by maintaining the apparatus a t the temperature ofliquid air (at which temperature the vapour pressure of ozone isexceedingly slight), to convert 99 per cent. of the oxygen into ozone.Even a t this temperature the spark discharge gives less than 1 percent. of ozone, and this small amount may be due really to acmmpany-ing silent discharge.I n last year’s Report (p. 61) mention was made of the preparatiouof a new barium percarbonate, BaCO, ; additional particulars concern-ing this salt have now been published, and also a deecription of thepreparation of a series of related sodium percarbonates.72 The forma-tion and decomposition of the barium salt play a n important part inthe preparation of hydrogen peroxide by the interaction of hydratedbarium peroxide and carbonic acid; if this action is brought about inpresence of a suitable proportion of mater and at a low enoughtemperature, almost no hydrogen peroxide is liberated until morecarbonic anhydride has been passed in than is equivalent to thebarium present, but thereafter the formation is very rapid.Thecarbon dioxide apparently first unites directly with the bariumperoxide, and forms a yellow solid of the composition representedby the above formula, but it has not yet been obtained free fromwater; the compound is not rapidly decomposed by water, nor doeseither alcohol or ether remove hydrogen peroxide from it, so thatit does not seem to be merely an additive compound of the lattersubstance.It is quickly decomposed by acids, including carbonicacid, with formation of hydrogen peroxide. The sodium salts havebeen obtained in a somewhat similar manner from sodium peroxides, inpresenceof a suitable proportion of water and a t low temperatures, byaddition of gaseous or solid carbon dioxide. The peroxides used wereNa20a, Na20,, and NaHO, ; i t is stated that this last can be obtainedin two isomeric forms, namely, NaO*OH and O:Na*OH, the former bythe action of hydrogen peroxide on sodium ethoxide,and the latter, asdescribed by Tafel,73 by the action of absolute alcohol on Na20,. Thesalts obtained, having the composition indicated by the formulae, aredesignated as follows : sodium dioxide carbonate, Na2C0, ; sodium di-oxide bicarbonate, Na,C20, ; sodium trioxide carbonate, Na,C05 ; sodiumtrioxide bicarbonate, NaHCO, (two isomerides, one from each of theisomeric sodium hydrogen dioxides).The names are evidently takenmerely from the name of the oxide used in the preparation ; what theprobable constitution of the compounds is remains to be seen, andBriner and Durancl, Compt. rend., 1907, 145, 1272 ; A , , ii, 101.72 Wolffensteiri and Peltner, Ber., 1908, 41, 275, 280 ; A . , ii, 18@, 183 ; D.R.-P.73 Ber., 1894, 27, 2297 ; A . , 1894, ii, 448.138569, 196369 ; A . , ii, 180, 830INORGANIC CHEMISTRY. 51there must at present be considerable doubt even as to the truecomposition. Substances of so unstable a nature as these are,which cannot be crystallised or otherwise purified, may quite well bemixtures containing carbonate and a percarbonate in various pro-portions ; the substance NaHCO, might be Na,C,O, + H20,, and so on.The assumed existence of two isomeric substances corresponding incomposition with the formula NaHCO, is based merely on observeddifferences in the degree of stability, and this by itself is, of course,not very conclusive evidence; this also constitutes the only evidencefor the existence of isomeric peroxides.No definite indication can begiven as to whether Na2C,0, is the analogue of the potassium saltobtained electrolytically. Possibly the further investigations inwhich the authors are engaged may enable them t o clear up some ofthese points.The preparation of crystalline sodium perborateT4 can be con-veniently effected by first saturating with carbon dioxide a 50 percent.solution of sodium peroxide, adding a sufficiency of a saturatedsolution of sodium metaborate, and then cooling almost to zero.Indications have been obtained of the formation of small quantitiesof perstannates by the electrolysis of concentrated solutions of alkali~ t a n n a t e s , ~ ~ but the compounds are very unstable, and the solutionssoon decompose.I n a paper dealing with columbium and its compounds,76 C. W. Balkeand E. F. Smith describe several new well-crystallised saltsof percolumbicacid, H3Cb0,, which represents one of the most highly oxidised sets ofcompounds known. OF the salts mentioned, the most interesting fromthe theoretical Standpoint are the rubidium and caesium salts,Rb3Cb0, and Cs,CbO, ; being anhydrous, they indicate quite clearlythat the high degree of oxidation of this class of salt cannot bereferred to additive compounds with hydrogen peroxide.I n thepresent state of our knowledge, these substances are most convenientlylooked upon as columbates in which every atom of oxygen has beenreplaced by the peroxide group, -0.0-, an assumption which is inaccord with their mode of preparation by the action of hydrogenperoxide. The salts lose exactly half of their oxygen on ignition, andleave ordinary ortho-columbates.The formation and reactions of the persulphates continue to formthe subject of a considerable amount of original work, and the specialapplicability of the salts t o certain oxidation processes becomesmore and more evident. In some respects, however, there seems tobe misconception regarding the typical mode of action of these55 Coppdoro, Gazzetta, 1908, 38, i, 489 ; A ., ii, 596.76 J, Amer. Chem. ,S’oc., 1908, 30, 1637 ; A., ii, 1043.74 D. R.-T’. 193722 ; A., ii, 689.E 52 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.compounds ; although they are derivatives of hydrogen peroxide(from many points of view, some name, such as paroxydisulphonicacid, indicating this relation more clearly would be preferable topersulphuric acid), they do not in aqueous solution tend so much togive up oxygen directly as to take up metal; that is to say, thestandard mode of decomposition is comparable to the second kind ofaction observable with hydrogen peroxide, as previously noted. Thisis well exemplified by.the action of persulphates on metals, metallichalides, thiosulphates, lower salts of metals which form two basicoxides, etc., and is most conveniently expressed by the use ofionisation formuls ; in all such cases, the action is then merely theassumption of two extra negative charges by S20i', forming2SO,", for example, Zn + S,O, = Zn" + 2SO,", 2Br' + S,O, = Br, +ZSO,)', 2Fe" + S,O," = 2Fe"' + 2S0,".I n one at least of recent paperson the interaction between metals and persulphates,77 there is atendency to look upon the action as being in many cases an indirectone, through co-operation of water.It is true that not infrequentlythe final result is not just the simple one indicated above, but in thesecases the cause is to be sought in subsequent secondary actions.When ammonium persulphate is the salt used, as is generally the case,the possibility of complications occurring is considerably greater thanis the case with the potassium salt. It has been long known thatpersulphate oxidises ammonia and the ammonium radicle withproduction of nitrogen and of nitrate,7* and therefore the action of thepersulphate radicle itself on any particular substance is better observedby using the potassium salt. But even, with this salt, complicationsmay arise, due, for example, to hydrolysis of the primary product; anacid solution may be developed, resulting in hydrolysis of persulphuricacid and formation of hydrogen peroxide, which in its turn may be thesource of the free oxygen which has sometimes been observed.Theextent to which this latter hydrolysis may take place is evidenced bythe fact that in a recent; patent79 this is claimed as a mode ofpreparing hydrogen peroxide, special conditions and precautions, ofcourse, being necessary. I n circumstances where this kind of actiondoes not come into play, however, the reactions into which per-sulphates most readily enter are undoubtedly those which can bereferred to the simple ionic change already mentioned.One or two points concerning sulphur analogues of the class ofy7 M. G. Levi and others, Cazzetta, 1908, 38, 583 ; A. , ii, 581. See also Turrentine,78 See also Levi and Migliorini, Gazxettn, 1908, 38, ii, 10 ; A., ii, 535 ; KenipfJ.Physical Chcm., 1907, 11, 623 ; A . , ii, 104.and Oehler, Ber., 1908, 41, 2576 ; A , , ii, 764.D.R.-P. 199958 ; A . , ii, 1028INORGANIC CHEMISTRY. 53substances dealt with in this section may also conveniently be referredto here.From time to time conflicting statements have been made as to thetrue composition of ‘‘ hydrogen persulphide.” After first havingbeen determined to be somewhat complex, it was later assumed to berepresented by the formula H,S,, from analogy t o hydrogen peroxide ;then various higher sulphides were supposed to exist, correspondingwith the polysulphides of the metals, but this was contradicted byRebs,so who stated that whatever polysulphide might be added to anacid, the resulting hydrogen compounds were H2S and H,S, only.This in turn is now shown to be incorrect, on the authority ofindependent investigators.By fractional distillation (under lowpressure) of the oil obtained by pouring alkali polysulphidesolution into hydrochloric acid, the disulphide, H2S,,81 and the tri-sulphide, H2S3,*2 have been isolated as unstable liquids. The com-position has been exsctly determined by improved methods of analysis(the hydrogen being driven out as hydrogen sulphide, which couldbe accurately estimated), and the molecular weight by cryoscopicmethods. Evidence is also published for the existence of compoundsfrom H,S, to H2S9.83 I n discussing the constitution of these poly-sulphides, Bloch is not unfavourable to Mendelheff’s conception of apossible ‘‘ homologous series ” formed from HSH by successive re-placements of H by SH ; the general formula for such a series wouldof course be S,H,.I n view of various investigations carried out in recent years,it has been suggested that tetrathionic acid has a peroxidic con-stitution, ( H02S,)*O*O*(S202H,), and not the persulphidic constitution,(HO,S)*S*S*(SO,H), generally assumed, the reason being thatalkaline reducing agents apparently abstract oxygen directly from itssalts ; thus, alkaline arsenite solution forms arsenato as well as mono-thioarsenate.It has been shown,s5 however, that there are seriousobjections to the assumption of a peroxide union in the tetrathionates,and that Mendelheff’s persulphide formula, which fits in so well withthe general behaviour of the salts, can perfectly weil account for thesenew facts also.8O Annalen, 1888, 246, 356 ; A., 1888, 1155.81 I.Bloch and F. Hohn, Ber., 1908, 41, 1961, 1975 ; A., ii, 579.83 Eloch and Hohn, ibid., 1971 ; A., ii, 579. R. Schenclc and V. Falcke, ibid.,83 G . Rriini and A.. Borgo, Atti R. Accnd. Liiicei, 1907, [v], 16, ii, 745 ; A . , ii, 102.*4 Ber., 1908, 41, 1980 ; A , , ii, 580.Price and D. F. Twiss, Trans., 1907, 91, 2021 ; J. E. Mackenzie and H.2600 ; A . , ii, 762.Marshall, Trans., 1908, 93, 172654 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Complex salt&Although a short section under the above heading seems in manyways desirable, it is a difficult matter to treat collectively the widelydifferent mbstances which are thus brought together.I n the absenceof general guiding principles, also, any decision as to what are, andwhat are not, included under the term must appear very arbitraryand, in what follows, no attempt is made at any really systematictreatment.I n view of the great development. which certain branches of thesubject have now attained, the appearance of any collective account ofany of these branches is t o be welcomed. During the period under,review the -cobaltammine compounds have thus been treated byP. Pfeiffer,sG and the fact that the description of these substancesextends over 212 large pages gives some idea of the extent to whichthis field has been cultivated.The same author has tabulated thenumerous cases of isomerism which have been observed amongst thecomplex chromium derivatives.s7 A. Colson, who has published manypapers dealing with the green chromic sulphates, has recently given acollected statement of the results obtained by himself.s* A class ofcomplicated compounds which so f a r has not received a great deal ofsystematic investigation is the series of complex acids and saltsderived from molybdenum trioxide and analogous substances ; moreattention is now being devoted to these, and a discussion of their probableconstitution has been published by A. Miolati and R. Pizzighelli.89The polyhalides of the alkali metals may fairly well be calledcomplex salts ; their probable constitution is certainly quite obscure.It is therefore highly desirable that there should be no doubt as towhat members of the group actually exist, and at present the mattercannot be considered as settled.According to the most recent in-vestigation:* the only iodine derivatives of the metals potassium,rubidium, and caejium which can be obtained a t 25’ are : KI,, RbI,,&I,, CsI,, and K17. The compounds, RbI,, Rbl,, Cs17, CsI,,mentioned by Abegg and Hamburger could not be obtained.Complex halogen derivatives of two metals have come in for furtherexamination in the case of several members of the platinum group.I n the case of the iridium compounds, the iridi-salts {M2TrU6) presentno novel characteristics, but the iridio-salts exhibit several features OFinterest from the point of view of Werner’s theory of co-ordinated36 Gmelin-Kraut’s Eandbuch der anorgnnischen Chemie, Bd.V, Abt. 1.87 Zeitsch. nnorg. Chem., 1908, 58, 317 ; A,, ii, 594.83 Ann. Chim. Phys., 1907, [viii], 12, 433 ; A . , ii, 45.89 J. pr. Chem., 1908, [ii], 77, 117 ; A., ii, 595.90 H. W. Foote and W. C. Chalker, Amer. Chcm. J . , 1908, 39, 561 ; A ii. 586INORGANIC CHEMISTRY. 55corn pound^.^^ They can easily be prepared from the iridi-salts byreduction with normal oxalates. The salts are of two types, namely,M,IrC16 and M21rC15, the best example of the former being thesodium salt, Na31rC16,1 2H,O. The corresponding salts of the potassiummetals and of ammonium readily change into those of the other typeby the action of water ; these contain a molecule of water, which is notdriven off at 150°, and are therefore t o be looked upon as aquoiridio-pentaclilorides, M,( H20,Cl51r).The hexachlorides are easily de-hydrated.A number of analogous derivatives of rhodium have been preparedand investigated in connexion with a search for material suitable foruse in determinations of the atomic weight of that element.g2 These“rhodipentachlorides ” 93 also fall into two classes similar to thosementioned above for iridium, and in the second class the molecule ofwater again plays an important part. The corresponding brominecompounds, however, all crystallise in the anhydrous form.I n a series of papers, A. Werner continues the discussion of thepreparation and the constitution of various groups of complex cobalt-ammine compounds.94 Amongst these are a series of violeo-salts ofdichlorotetramminecobalt, [CI,Co( NH3)4] X, stereoisomerides of thepraseo-salts formerly known ; they form intensely blue crystals, andare presumably cis-compounds (1 : %constitution).To the red saltsisolated from Vortmann’s insoluble sulphate (which is a mixture of-NH red and green salts), the constitution Co(NH,),]X,is assigned, the two cobalt atoms being apparently bound throughinactive amino- and hydroxyl groups. The designation “ p-amino- ” isproposed for the former type of union, and the name of octammine-p-amino-ol-dicobalt salts is adopted for these compounds. A consider-able number of them have been prepared. Another series of di-cobaltderivatives has also been elucidated, and is represented by the formula[(NH,),COI(OH),~CO(NH,)~]X~ ; these hexamminetrioldicobalt saltsare isomeric with the dodecamminehexoltetracobalt salts,and are red in colonr.A new set of iodo-salts, obtained by Sand andBokman, has been shown to be the iodopentamminecobalt series,[CoI(NH,),]X, ; these salts are green in colour.[CO(OH)~{CO(~H~>~)3I x699l M. Delepine, Conapt. rend., 1908, 146, 1267 ; A., ii, 702. * A. Gutbier and A. Huttlinger, Ber., 1908, 41, 210 ; A., ii, 200.93 There is room for the adoption of some more systematic nomenclature forsubstances of this kind ; it seems unfortunate that whilst K,IrC15 is potassizsmiridiochloride, K,RhCI is potassium rhodipentachloride.94 Bcr., 1907, 40, 4605, 4817, 4834 ; 1908, 41, 3007 ; also J. Saud and G.Bok.Iiian, ibid., 40, 4497 ; A . , ii, 42, 43, 45, 95056 ANNUAL REPORTS OM THE PROGRESS OF CHEMISTRY.Physiochemical investigations regarding the hydrolysis and themolecular weight of the different chlorides and sulphates of chromium,with a view t o the further elucidation of their constitution, have beencarried out by various cherni~ts.~5 The results do not seem to giveconcordant indications as to the molecular complexity of the green andblue varieties. A green chlorosulphate of chromium,has been prepared, isomeric with that formerly known,[ CrC1,5H20] SO,, 3H20.96As indicated by the formulae, the new salt easily gives C1 ions insolution, but not SO, ions, just the opposite of what is the case withthe other salt.Complex salts, in which iron forms the central constituent, havebeen dealt with by several workers. New pyrophosphate derivativesare described,97 some related to ferric oxide, some to ferrous ; the best-defined of these are the ferric ones of the type M6Fe2(P207)8, andcrystallise with varying proportions of water ; the acid itself has alsobeen isolated.The analogy to the ferricyanides is pointed out (P,O,equivalent to 4CN), and it is proposed to call the salts ferripyro-phosphates ; the ferropyrophosphates, M4Fe2(B207)s, are powerfulreducing agents (see p. 58). Analogous salts derived from metalsother than iron have also been prepared.There has been a good deal of further investigation of the ferro-nitrosulphides (Roussin’s salts) referred t o in last year’s Report(p. 71 Q S ) , in the hope of clearing up their constitution, and severalco-ordination formulae are suggested for the two classes of compounds.When the salts are decomposed in various ways, the nitrogen isobtained in different stages of oxidation-nitrous or nitric oxide,hyponitrite, nitrite-according to the circumstances.99Many of the complex salts derived from molybdenum and tungstencontain twelve atoms of the respective element in the molecule.I nthe paper by Mioletti and Pizzighelli, already mentioned, it issuggested that the acids may all be represented by co-ordinationformulae of the type [R(MO~O~)~]H,, where n is 7 if R represents anelement like phosphorus, and 8 when R is an element like silicon.The formulae stated by various other investigators for substances pre-pared by them do not, however, correspond with acids of this basicity.95 J.Sand and F. Grnmmling, Zcilsch. p7~ysiknl. Chem,., 1908, 62, 1, 28 ; H. G.Denham, Zeitsch. anorg. Chem., 1908, 57, 361 ; A . , ii, 293, 294, 389.96 R. F. Weinland and T. Schumann, Zeitsch. anorg. Chem., 1908, 58, 176 ; A.,ii, 595.98 The formula HFe,(NO)$, is there printed incorrectly.99 L. Caiiibi ; I. Bellucci and P. de Cesaris, Atti R. Acead. Lineei, 1907,[v], 16, ii, 658, 740 ; 1908, [v], 17, i, 202, 424, 5 4 5 ; A . , ii, 41, 111, 388,499, 593.[ CrS0,,5H20]C1,97 P. Pascal, Compt. rend., 1908, 146, 231 ; A . , ii, 193INORGANIC CHEMISTRY. 57Mercuric cyanide reacts with a number of other mercuric salts (forexample, the perchlorate) to form complex derivatives ; from a physico-chemical study of the properties of the solutions of these compounds,lproof is obtained that the mercury in them forms part of a complexunivalent cation, (HgCN)'.A number of the salts have beenprepared in the crystalline condition ; their dilute solutions give noprecipitate on the addition of sodium hydroxide. Even the compoundof mercuric cyanide with mercuric oxide appears t o be the basic oxide,(HgCN),O, and t o give a hydroxide, (HgCN)OH, when it dissolves inwater. Similar complex cations apparently are formed from mercuricperchlorate and mercuric iodide, bromide, and chloride, the tendencyto form them decreasing in the order given, Crystalline compoundsare obtainable in the first and second case, but not in the last; thesecompounds are decomposed by treatment with water.Colloids.During recent years much interesting work has been done in con-nexion with colloids generally, and a considerabIe proportion of thenew developments have to deal with the inorganic branch of the sub-ject.Many of the papers which have appeared in the course of thepast year fall rather within the domain of physical chemistry, butothers of a more descriptive character may be referred to here.Colloidal sulphur has already been prepared by the method of usingelectrical discharges ; it is now shown that it can also be obtainedfrom the sulphur which separates when sodium thiosulphate solutionis added to cold concentrated sulphuric acid.The liquid is somewhatdiluted, heated, filtered through glass wool, and allowed to cool ; theseprocesses are repeated so long as any sulphur, which will notredissolve, is precipitated. A cloudy, yellowish-coloured mass is t, husobtained, which, on warming, forms a liquid, which is apparentlyperfectly clear ; it separates again on cooling. This mass is separatedas completely as possible by means of a centrifuge, and is partlywashed in a similar manner, after which the remaining acid is neutral-ised with sodium carbonate. The product is soluble in water, but ifthe attempt is made to remove the sodium sulphate from the hydrosolby dialysis, insoluble sulphur is very soon deposited.Dilute solutions,in which the sulphuric acid has not been neutralised, can be preservedfor a long time, but the hydrosol very easily precipitates insolublesulphur on addition of various salts.The preparation of a more or less colloidal form of graphite, speciallysuitable for lubricating purposes, has been effected by prolonged1 V. Borelli, Gnzzetta, 1908, 38, i, 261 ; ii, 421 ; A., i, 515 ; ii, 1039.2 M. Raffo, Zcitsch. Chem. Ind. Kolloide, 1908, 2, 358 ; A . , ii, 68358 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.agitation of the graphite with a solution of tannic acid ; 3 the materialthus obtained passes through filters and remains in suspension formonths, but it becomes flocculated on the addition of hydrochloricacid. The method of preparation was based on the known action oftannic acid and other substances on china-clay, by which the latter isrendered much more workable for certain pottery purposes; afluid “slip,” suitable for casting in forms, can by such means beprepared with much less water than formerly had to be used.Convenient reducing agents for the preparation of colloidal metalsand some other reduction products are found to be provided by theferropyrophosphates of the alkali metals *-salts which can be easilyprepared in solution by adding a solution of ferrous salt to a solutionof an ordinary alkali pyrophosphate.(Solutions of these and similarcomplex pyrophosphates are useful also in connexion with the electro-deposition of certain metals for purposes of quantitative analysis,)The colloidal solutions of gold and silver obtained in this way are ofa very intense colour, and.it is stated that the reaction may be used asa colorimetric method of estimating these metals.Strange to say,platinum compounds give no result with this reagent, but cupric andmercuric salts can be reduced first to the lower stage of oxidation andthen to colloidal metals. The colloidal cuprous hydroxide, which canbe obtained in this way, appears yellow by reflected light, and is saidto provide avery sensitive reaction in testing for trsces of copper.I n the state of colloidal solution, palladium still exhibits theproperty of absorbing hydrogen,5 forming a hydrosol of the ‘‘ hydride ” ;and in this state its power of absorption is much greater than is thecase when it is in the form of palladium black This absorbs, at most,somewhat less than nine hundred times its volume of the gas, but thecolloidal form takes up from nine hundred to nearly three thousandtimes its volume.(No explanation can be given for the great fluctua-tions of the value found in different experiments.) The colloidalhydride reacts rapidly with any free or loosely-combined oxygen whichmay be present, and the numbers given have been corrected for thehydrogen used up in this way.The bearing which colloidal silver and siIver compounds have inrelation t o photographic images, and to the photo-chemistry of silvergenerally, is discussed in a series of papcrs by Luppo-Cramer ; 6 heconsiders that the ‘( sub-halides ” of silver are not definite compounds,but adsorption products, and that the different appearances of the3 E.G. Acheson, J . Franklin I n s t . , 1907, 164, 375 ; A., ii, 375.4 P. Pascal, Compt. rend., 1908, 146, 862; A . , ii, 500.6 Zcitsch. Chem. Ind. Xolloide, 1907, 2, 135 ; 1908, 2, 360; 3, 33, 135 ; A,, ii,C. Paal and J. Gerum, Ber., 1908, 41, 805 ; A . , ii, 392.378, 691, 841, 945INORGANIC CHEMISTRY. 59silver images obtained with different developers are due to the more orless colloidal state of the silver and the different substances adsorbedby it from the developer. (The other view receives favourable con-sideration from T r i ~ e l l i , ~ who gives a full discussion of both theories.)A specially interesting development in the study of colloids recentlyhas been the preparation in colloidal form of a number of what wereformerly considered the very antitheses of such substances-thetypical I‘ crystalloid ” salts, such as the halides of the alkali metals ;organosols of these can be obtained by the interaction of suitablecompounds, such as ethyl ethylsodiomalonate and ethyl chloroacetate,8dissolved in anhydrous ether, benzene, etc.Similar results have alsabeen obtained with a number of salts of magnesium and the metals ofthe alkaline earths ; it appears more and more likely, therefore, thatsnbstances generally may be obtained in the colloidal form if only theycan be produced by interactions in solvents in which the solubility ofthe ordinary form of the product is sufficiently slight.Colloidalbarium sulphate can be obtained by double decomposition in glycerolsolutions of the reacting salts; the hydrosol thus obtained isparticularly stable in presence of barium nitrate, although salts ofmost other metals precipitate it ; it is also coagulated by polybasic acids,but not by monobasic. A newly-devised general method for obtainingcolloidal salts of the metals of the alkalis and the alkaline earthsconsists in actiug with the appropriate acid on tho thiocyanate of themetal, each substance being dissolved in a mixture of ether and amylalcohol.1° A number of colloidal salts of magnesium and the metals ofthe alkaline earths can also be prepared by the interaction o€ theappropriate acid on the basic oxide of the metal, dissolved in methylalcohol.11I n general, the various salts prepared by these methods can beprecipitated, as gels, by the addition of suitable liquids, and re-dissolvedin the colloidal form, by means of others which act as appropriate‘ 6 solvents.”Some interesting experiments which had for their aim the prepara-tion of gelatinous aluminium silicates, of definite composition, by theinteraction of sodium silicate and aluminium acetate, gave entirelynegative results ; 12 from their behaviour towards solvents, etc., theprecipitates obtained in this may appear to be merely mechanical7 Zeitsch. wiss.Photograph. Photophysik. Photochem., 1908, 6, 358 ; A . , ii, 1036.e C. Pad and G. Kiihn, Bey., 1908, 41, 51, 58 ; A., ii, 179.9 A.Recoura, Compt. re?&., 1908,146, 1274 ; A., ii, 692.10 P. P. von Weimarn, Zeitsch. Chem. I n d . Kolloide, 1908, 3, 89 ; A., ii, 842.11 C. Ncuberg and B. Rewald, Sitzungsber. K. A k a d . Wiss. Berlin, 1907, 820 jA , , ii, 39 ; Biochem. Zeitsch., 1908, 9, 537 ; A., ii, 495.12 H. Stremme, Ccdr, Mix,, 1908, 622, 661 ; A , , ii, 104160 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.mixtures of hydrated alumina and hydrated silica. This observation,taken in conjunction with the widely variable composition of thenatural non-crystalline and clay-like substances known as halloysite,aliophane, etc., leads to the conclusion that these are not definitecompounds, but variable mixtures.:Bure Earths.As in forrher years, there has again been a considerable amount ofinvestigation carried out in connexion with this complicated group ofsubstances, but there is not much %,f it that calls for special commentin a geyeral rQsum6; the most noteworthy result, the separation ofytterbium into two components, has already been mentioned (p.34).The search for improved means of fractionation, of course, continues,and several new or modified processes are suggested. Amongst thesalts recommended for the purpose are the brornates ; l3 starting fromthe oxalates, they are obtained by treatment with sulphuric acid,followed by double decomposition with barium bromate. Other com-pounds which have proved serviceable for certain groups of elementsare the malonates, obtained directly from the hydroxides by the actionof malonic acid, and the l-naphthol-8-~ulphonates,~~ obtained From thecarbonates by the action of the sulphonic acid; the solubilities ofmany of the salts of these acids have been definitely measured.A considerable number of new compounds of the rare-earth metalshave been prepared and described, but generally they exhibit nostriking or exceptional characters.Several sulphides have beenexamined,15 and some of these have been found to give a distinct odourof hydrogen persulphide when acted on by hydrochloric acid; it istherefore suggested that, say, cerium disulphide is not a simpleS sulphide, Ce<<$, but a persulphide, and might be written Ce,S,S.Several substances, partly oxide and partly sulphide, have also beenobtained.The Argon Group.The most interesting contribution to our knowledge of this class ofelements which has been made during the year has undoubtedly beenthe liquefaction of helium by Onnes,lG regarding which a communica-tion was also made to the Chemistry Section of the British Associationat Dublin by Sir James Dewar.Premature announcements of success13 C. James, Chent. News, 1908, 97, 61, 205 ; A . , ii, 190, 498.14 H. Erdmann and F. Wirth, AmKden, 19OS, 361, 190 ; A., i, 621 ; ii, 694.15 A. Duboin, Compt. refid., 1908, 146, 815 ; A., ii, 502 ; W. Biltz, Ber., 1908,16 Proc. K. Akad. Welemch. Amsterdam, 1908, 10, 744 ; 11, 168 ; d., ii, 490,41, 3341 ; A., ii, 1037.944INORGANIC CHEMISTRY. 61had been made, but it was found that the results on which thesewere based had been obtained with impure material, and that onemilligram of- hydrogen, diffused through a space of 7 c.c., broughtabout peculiar phenomena in the highly compressed and cooledhelium, which had caused it to be assumed that liquefaction of thiselement had taken place.For the thorough purification of the material-two hundred litres ofhelium were used for the final condensation-Dewar's method of absorp-tion in charcoal at low temperatures was employed in conjunctionwith the usual chemical methods for eliminating active elements.Preliminary study of the isotherms for helium at the temperaturesproducible by liquid hydrogen had shown that the Joule-Kelvin effectwould probably be sufficient to secure condensation, so that the Linde-Hampson method of working should prove successful, and theapparatus by which the result was actually obtained was similar tothat used for the liquefaction of hydrogen.Considerable quantitiesof liquid air and hydrogen for auxiliary cooling were prepared before-hand, and three hours' work on the helium sufficed to produce fully60 C.C. of colourless, transparent liquid, having a density of only 0.15.It boils at 4 . 5 O Abs., which is within one degree of its critical point ;attempts to freeze it by rapid evaporation under diminished pressurewere unsuccessful, although the temperature reached mas probablyabout 3' Abs.Attempts have been made t o obtain compounds of argon by meansof arc and spark discharges in liquid argon, using electrodes of variousmetals, but without S U C C ~ S S .~ ~ A polymerisation of the element couldnot be expected from a monatomic gas (the formation of ozone bysimilar processes presumably depends on the preliminary rupture of thediatomic molecule). Apparently there is no tendency to form evenunstable compounds with the material of the electrodes ; suchcompounds, if formed at the temperature of the discharge, mightpossibly become fixed by the very rapid cooling.Most of the other work of the year deals almost entirely with thenatural occurrence of the various gases, and with unsuccessful attemptst o discover other members of the group. It would now appear to befairly certain that no other gases are present in the atmosphere i nappreciable quantity; the residues from about 120 tons of liquid airhave been subjected to a searching examination,lS but no indications ofany unknown element could be obtained ; other experiments have alsogiven negative results.It has been suggested by Sir WilliamRamsay that the emanation from the radioactive elements mayrepresent the higher members of this group.l7 F. Fischer and G . Iliovici, Bcr., 1908, 41, 3802 ; A . , ii, 1034.R. B. Moore, €'.roc. Roy. Soc., 1908, 81, A, 195; A . , ii, 84062 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Some very striking figures regarding the total quantity of noble gasesevolved from certain thermal springs are given by C. Moureu and R.Biquard ; l9 according to their results, one spring, the Bourbon-Lancy,must evolve in the course of a year fully sixteen thousand litres ofnoble gases, of which quantity no less than ten thousand litres consistof helium.All the waters of these springs are radioactive, a point ofconsiderable interest in connexion with the production of helium fromradium emanation.Group I.As was mentioned in last year's Report (p. 47), calcium hydridehas been suggested as a convenient and portable material for use inthe preparation of hydrogen, which is obtaiDed merely by the actionof water on it; a competitor now appears in the form of aluminiumfilings which have been treated with small quantities of potassiumcyanide and mercuric chloride.20 The hydrogen is obtained by droppingwater on the prepared metal, and the action takes place rapidly if thesupply of water is so regulated as to permit of the temperature beingmaintained a t 70" by the heat evolved in the action.One gram givesabout 1.3 litres of gas at the ordinary temperature, as compared withabout 1 litre in the case of calcium hydride.21The possible utilisation of natural silicates as sources of potassiumcompounds on the large scale has been investigated ; although potash-felspar when finely divided may be decomposed by water or aqueoussolutions of various salts, the methods so far tried do not promise t obe sufficiently economical to be practicable.22 Leucite, which is alsoricher in potassium, can be fairly easily decomposed by sulphuric acidor nitric acid, and various processes for separating potassium saltsfrom the solutions thus prepared are sugge~ted.~3The relative solubilities of metallic silver in molten lead and zinc,which are so widely different that the desilverisation of lead by zinc(Parkes' process) is the best method known, have been determined atvarious concentrations.24 At 500° the partition coefficient Zn : P b isabout 300 : 1.I n Parkes' process the separation is effected a t amuch lower temperature than this, the zinc being allowed to solidify,and under these conditions this r,ttio is increased greatly in favour ofthe zinc.l9 Compt. rend., 1908, 146, 435 ; A . , ii, 277.2o Nauricheau-Beauprd, ibid., 1908, 147, 310 ; A., ii, 829.21 I n last year's Report, the vo1ume:of hydrogen stated is only one-tenth of the2a A. S. Cushman and P. Hubbard, J.Amcr. Chem. Soc., 1908, 30, 779 ; A., ii,* C. Manuelli, Qamettn, 1908, 38, i, 143 ; A . , ii, 386.24 G. N. Potdar, J. Coll. Sci. T6ky6, 1908, 25, ix, 1 ; A . , ii, 945.correct amount.586IN ORG A N IC C HEM ISTRY. 63Finelg-divided gold is energetically attacked by fused sodiumperoxide, with formation of sodium aurate, and from this salt, bymeans of dilute sulphuric acid, auric acid, H3Au03, or, more probably,HAuO,,H,O, can be obtained. To prepare auFates satisfactorily fromthe acid, it should be treated with the pure appropriate hydroxide inan atmosphere free from carbon dioxide, and the resulting solutionshould then be evaporated in the dark.25The properties of metallic calcium and its capabilities as a chemicalreagent continue t o receive a fair amount of attentioo, and severalpapers dealing with various reactions in which it takes part haverecently appeared.These include an investigation of its action onammonia alid on amino-derivatives under various conditions ; 26 it isparticularly reactive with arylamines, producing compounds of thetype (NHR),Ca. Further information regmding the use of the metaland its hydride as reducing agents in " thermite '' processes has alsobeen published ; 27 reactions with the hydride are less violent thanwith the free metal.Within the last fern years an extensive series of investigations hasbeen carried out on the compounds formed by union of calciumsulphate and other metallic sulphatey, and a considerable numberof new salts have been obtained; during the year this work hasbeen continued, and additional double and triple sulphates, suchas Rb,Ca,(SO,), and K,Ca~Cd(SO,),,2H2O, have been prepared.28When water of crystallisation is not taken into account, salts ofthis kind are easily enough represented by ordinary constitutionalformula Anhydrous calcium sulphate (the mineral anhydrits) isfrequently treated as being isomorphous with the sulphates ofstrontium, barium, and lead; like them it forms rhombic crystals,but its axial ratios differ considerably from those of the othersulphates; it is now shown fairly conclusively that it is not iso-morphous with these.The varions salts have been preparedartificially in crystals of moderate size by deposition from solutionin hot concentrated sulphuric acid, and it is found that calciumsulphate does not form mixed crystals with the others, mhicb,however, form mixed crystals among themsel~es.~~25 F.Meyer, Compt. rend., 1907, 145, 805 ; A., ii, 47.26 H. Erdmann and H. van der Smissen, AnnaEen, 1908, 361 32 ; A., ii, 587.27 F. M. Perkin and L. Pratt, [l'mns. Farnday Soc., 1908, 3, 179 ; A, ii, 379.2* J. D'Ans, BcT., 1907, 40, 4912 ; 1908, 41, 187, 1776, 1777 ; A . , ii, 104, 182,890.P. Gaubeit, Cwnpt. I-cud., l?07, 145, 877 ; A., ii, 3864 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.It would appear that bicarbonates of this group of metals can reallyexist in the solid state, although they are exceedingly unstable;30thus, when ammonium or potassium bicarbonate solution is added toa solution of calcium chloride, both being cooled to zero, a white,crystalline precipitate is obtained, the composition of which ap-proximates t o that expressed by the formula Ca( LICO,),.An interesting description is given of the working up of theresidues from 30,000 kilograms of pitchblende residues for radium.31The operations extended over two years, and the total radium con-tained in the products was equivalent t o rather more than 3 grams ofradium chloride.A final fraction of radium bromide was found tolose bromine on keeping. An atomic weight determination on thechloride gave the value Ra = 225.Group 111.There has been comparatively little work done in connexion withthis group, beyond that dealing with ‘‘ thermite ” reactions and thepreparation of alloy-like derivatives of boron and silicon.Indiumselenate has been prepared, and i t s ability t o form alums demonstratedby the isolation of the cEsium salt, CsIn(SO4),,12H,O, in octahedralcrystals .32 A number of crystallised complex silico-tungstates con-taining indium have also been described, analogous to similarcompounds formed by aluminium, iron, chromium, etc.33Group IV.The absorption of carbon monoxide by solutions of cuprous chlorideand other salts has been investigated on somewhat similar lines tothose referred t o in last year’s Report (p. 57) with regard t o nitricoxide and varius metallic salts.34 The proportion of gas absorbednever exceeds the ratio CO : Cu, and in all kinds of aqueous solutionsa definite compound, CnC1,C0,2H2O, is formed with the chloride.No absorptiov takes place except in the presence of water, ammonia,aniline, or other substance capable of forming additive compoundslike t h a t above formulated.I n addition to a fair amount of work on the preparation of silicidesand silicon alloys, and on ‘‘ thermite ” processes involving silicon,30 E.H. Keiser and others, J. Amer. Chem. Soc:., 1908, 30, 1711, 1714 ; A., ii,1036, 1037.L. Haitinger and K. Ulrich, Monatsh., 1908, 29, 486 ; A., ii, 857.32 F. C. Mathers and C. G. Schluederberg, J. Ainer Chem. Xoc., 1908, 30, 211 ;93 G. N. Wyrouboff, BUZZ. SOC. franc. MZ’?L., 1907, 3 , 277 ; A., ii, 386.34 W. Manchot and J. N. Friend, Annalen, 1908, 359, 100; A., ii, 375.A ., ii, 386INORGANIC CHEMISTRY. 65there has been a good deal of investigation as to the formation ofsilicic acids and soluble silicates. As a result, there appears to bevery considerable doubi; as to the existence of any definite hydrate ofsilica what~oever.3~ From a determination of the amount of hydrogenevolved when certain silicides and titanides are decomposed by suchagents as hydrofluoric acid, sulphuric acid, or potassium hydroxide,W. Manchot 36 seeks to show that in compounds of that class siliconand titanium form chains, but the evidence does not seem t o justifythe construction of constitutional formulae for such substances.Pure zirconium fluoride, which can beeeasily prepared by the actionof dry hydrogen fluoride on zirconium chloride, forms a snow-white,crystalline powder, which, as regards chemical character, is sur-prisingly inerL37 It is sparingly soluble in water, and does notundergo hydrolysis, but can be re-precipitated as a hydrate,ZrF,,3H20.Group KA very large amount of work in connexion with the members ofthis group has recently appeared.A considerable proportion of theinterest centres around problems connected with the technicalutilisation of atmospheric nitrogen, either by oxidation to nitriteand nitrate or by reduction to ammonia.A number of metallic nitrides have been studied, their preparationbeing effected by heating the metals in an atmosphere either ofnitrogen 38 or of arnm0nia.3~ The temperatures at which metals absorbnitrogen vary greatly in different cases : with magnesium, calcinm,aluminium, and chromium the process begins at about 800°, but withiron, copper, and some others no action occurs below 1250’.I n thefirst three cases, definite compounds are formed, namely, Mg,N2,Ca3N2, and AlN; in the other cases the proportion of nitrogenabsorbed is not sufficient t o form compounds of this “ammoniatype,” and it is doubtful whether the resulting substance is not ofthe nature of a solid so1ution;of nitrogen or metallic nitride, in themetal. An interesting character of several of these products is thatthey are magnetic; in some (chromium and titanium) this property isdistinctly noticeable, and in the case of manganese (12 per cent.nitrogen) it is almost as intense as in the case of iron.It variesgreatly with the composition, however, and attains a maximum0. Miigge, Centr. Mi?&,, 1908,-129 ; J. M. van Eemmelen, Zeitsch. anorg. Chem.,1908, 59, 225 ; H. Le Chatelier, Compt. rend., 1908, 147, 660; A., ii, 277, 838,1033.Annalen, 1907, 357, 129, 140 ; A., ii, 40, 46.I. I. Shukoff> J. Xuss. Phys. Chem. Soc., 1908, 40, 457; A., ii, 484.37 L. Wolter, Chem. Zeit., 1908, 32, 606 ; A., ii, 701.39 G. G. Henderson and J. C. Galletly, J. Soc. Chem. Ind., 1908, 27, 387 ; A,,ii, 485.REP.-VOL. V. 66 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.according to other observers 40 at a composition which would correspondwith the formula MnpN,. When metals are heated in ammoniainstead of nitrogen, the results obtained are somewhat similar; incases where no distinct fixation of nitrogen takes place (for example,tin), there is nevertheless a considerable change in the physicalcondition of the metal, and a considerable proportion of the ammoniaundergoes decomposition into its elements.The formation of definite compounds when metals dissolve in liquidammonia, assumed by Moissan and others, is denied by Kraus41 so faras lithium, sodium, and potassium are concerned, but confirmed in thecase of calcium; the compound is not represented by the formulaCa(NK,),, however, as given by Moissan, but by Ca(NH,),.The direct formation of hydrazine from ammonia can be broughtabout, with a good yield, by means of sodium hyp~chlorite,~~ pro-vided the viscosity of the solution has been increased by the additionof a small quantity of some suitable organic substance, such as glue.The further condensation of hydrazine to azoimide (hydrazoic acid)by means of oxidising agents is well known ; oxidation by means ofnitric acid forms a simple lecture experiment.A quantitative studyof the behaviour of several allied oxidising agents under certainconditions has disclosed several peculiar results.43 In presence ofsulphuric acid, potassium chlorate gives a large yield of azoimide,bromate gives much less, and iodate gives none at all ; in presence ofsilver sulphate, as well as sulphuric acid, all three give about the sameyield (averaging about 12 per cent.), which is, however, much lessthan that obtained with chlorate in absence of silver (more than22 per cent.).The actions with the halogens themselves also exhibitpeculiarities. A convenient method for the actual preparation ofazoimide is suggested 44 in the interaction of ethyl nitrite and hydrazinesalts under certain conditions; thus, a yield of more than 80 per cent.can be obtained by shaking the ester for six hours with an aqueoussolution of hydrazine sulphate and sodium hydroxide.Ebler and Schott publish a long paper 45 dealing with hydroxylamineand its derivatives, in which they give a full discussion of its probableconstitution; they accept the view that it is tautomeric, towardsalkalis acting as a weak acid, NH,*OH, and with acids acting as aH weak base, NH,:O, forming oxonium salts, NH,:O<X. Certain4u E. Wedekind and T, Veit, Ber., 1908, 41, 3769 ; A., ii, 1041.41 J.Amer. Chem. Soc., 1908, 30, 653 ; A . , ii, 486.J2 F. Raschig, D.R.-P. 198307 ; A., ii, 1029.43 A. W. Browne and F. F. Shetterly, J. Amer. Chem. Soc., 1908, 30, 53 ; A.,44 J. Thiele, Ber., 1908, 41, 2681 ; A., ii, 940.45 J. pr. Chem., 1908, [ii], 78, 289 ; A., ii, 1029.ii, 373INORGANIC CHEMISTRY. 67metals, such as zinc and calcium, can displace hydrogen from itdirectly, forming more or less unstable hydroxykbmcctes, for example,Cit(O*NH,). It is also stated that hydroxylamine reacts likehydrogen peroxide with titanium solutions, giving the same yellowcolour.I n connexion with the oxides of nitrogen, much of the recentwork bears on the synthetic production of these substances at hightemperatures, attained elec trically or otherwise ; a considerable pro-portion of t,his deals with the various equilibria which are establishedduring the formation of the oxides 40 or during their absorption by waterand ~olutions,~7 and cannot be entered upon here.Under suitableconditions, oxidation of the nitrogen may proceed as far as nitricanhydride (for example, nitrogen peroxide is rapidly oxidised into thisby ozone), and in such a case synthetic nitric acid of high concen-tration may be prepared directly. The direct production of purenitrites is another problem of great technical importance, and it isclaimed that this result can be attaiaed by oxidising atmosphericnitrogen to the appropriate extent in the electric arc, and main-taining the gases at a temperature not below 300' until absorptionin solution of alkali hydroxide has been effected; in this way theformation of higher oxides is almost entirely prevented.48 Theanalysis of mixtures of nitrogen oxides can be effected by examina-tion of their ultra-red absorption spectra ; 49 all five oxides, and alsoozone, can be detected by their different maxima OF absorption.I nthis may the nature of the products obtained under various conditionsof electrical discharge, etc., can be conveniently examined.Some interesting observations on the glowing of phosphorus andsome of its compounds have been described. The glowing of phos-phorous anhydride when mixed with oxygen has been investigated byS~harff,~o who finds that it is influenced by temperature, oxygen-concentration, presence of other substances, etc., in very much thesame way as is the glowing of phosphorus itself ; to a certain extentthis is also the case with the glowing of the sulphide P,S,.Theseresults seem to point to the conclusion that the phosphorescenceassociated with the slow oxidation of phosphorus is due, not t o theprimary oxidation of the element, but to the subsequent oxidation ofthe phosphorous anhydride so produced. This view seems to be sup-46 E. H. Keiser and L. McMaster, Anzer. Chenz. J., 1908,39, 101 ; A., ii, 223. 3'.Haber and A. Koenig, Zcitsch. Elektrochem., 1907,13, 725 ; 14, 689 ; A., ii, 34, 940.47 F. Foerster and M. Koch, Zeitsch. angew. Chew,., 1908, 21, 2161, 2209 ; A , ,ii, 941, 1031.a Radische Anilin- & Soda-Fabrik, D.R.-P. 188185 ; A., ii, 175.49 E.Warburg and G. LeithSuser, Sitzzengsjer. K. Akad. Whs. Berlin, 1908, 148 ;80 Zeitsch. physikal. Chem., 1908, 62, 179 ; A., ii, 373.A., ii, 175.F 68 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.ported by a striking phenomenon observed by L. and E. Bloch 51 duringan investigation of the ionisation effects which are produced when airis passed over phosphorus. They found that if the speed of the aircurrent is increased, then the phosphorescence is drawn out in thedirection of the current, and, on further increasing the speed, itbecomes completely detached from the phosphorus ; with a suflicientlylong tube and suitably regulated current, a steady phosphorescentcolumn may be obtained, separated from the phosphorus by a darkregion several metres in length.I n this dark space there is neitherproduction of ozone nor ionisation, so that all three phenomena aredirectly connected. It is possible, of course, that the phosphorescenceis due to phosphorus vapour which is carried forward, but moreprobably it is due to vapour of the oxide.Several corrections of what are said to be erroneoug statementsregarding phosphorus compounds may be noted. Only three definitesulphides of phosphorus exist,’namely, P,S,, P,S,, and P2S5,52 other so-called sulphides being merely mixtures. The various ‘< polyphoe-phates ” are merely mixtures of pyrophosphate and metaphosphate.53Hypophosphoric acid corresponds with the simple formula H,PO,, asis shown by the molecular weights of its esters,54 determined ebullio-scopically on solutions in alkyl iodides.A number of papers dealing with the subject of arsine have recentlyappeared, and yield a considerable amount of detailed informationregarding its decomposition by heat,b5 its behaviour with solutions ofvarious metallic salts,56 and its oxidation.57 With silver nitrate solu-tion there is a certain amount of precipitation of silver arsenide,Ag3As, and not of metallic silver alone, as is frequently stated ; withammoniacal nitrate solution, arsenide is also precipitated, but under-goes oxidation with formation both of arsenite and arsenate, anddeposition of silver.A n interesting method of purifying, say, concentrated sulphuric acidfrom arsenic compounds has been patented ; 58 i t consists in convertingthe arsenic into chloride or‘ fluoride by addition of the appropriateacid, and then extracting with benzene, which completely removes thearsenious halide ; the benzene in turn can be purified by agitation withwater. Dichlorobenzene acts in the same way, and can be applied inb1 Compl. rend., 1908, 147, 842; A ., ii, 1032.52 A. Stock, Ber., 1908, 41, 558, 657 ; A . , ii, 274.53 N. Parravano and G. Calcagni, Atti $. Accud. Lincei, 1908, [v], 17, i, 731 ;54 A, Rosenheim and M. Pritze, Ber., 1908, 41, 2708 ; A., ii, 942.55 A. Stock and others, Ber., 41, 1319 ; A., ii, 488.56 H. Reckl9ben and others, Zeitsch. anal. Chetn., 1907, 46, 671 ; A . , ii, 36.57 H. Reckleben, ibid., 1908, 47, 105 ; A ., ii, 176.58 Chemische Fabrik Griesheim-Elektron, D.R.-P. 194864 ; A., ii, 686.A., ii, 838INORGANIC CHEMISTRY. 69connexion with a scrubbing process for gases contaminated witharsenic ; in such cases its lower volatility is an advantage.The rarer members of this group:have recently received a consider-able share of attention, .and a large number of new compounds ofthem are described, many..of them being of a complex kind. Theyinclude derivatives of so-called hypovanadic acid, V2O,,2H,O,59 whichin these compounds exhibits basic functions ; whilst most of its deriv-atives are more or less decidedly blue in colour, the double nitritesare colourless or yellow, so that complex derivatives are evidentlyformed by it.I n Balko and Smith's paper already mentioned (p. 51),various columbates, fluoro-columbates, and percolumbates are described,along with analogous tantalum compounds; many have a complexcomposition. The columbium material used by them, as well as thatsimilarly obtained from many different sources, was subjected torigorous spectroscopic examination,60 and it is shown that pure colum-'bium solutions, quite free from titanium, give a yellow coloration onapplication of the hydrogen peroxide test, so that the presence ofcolumbium vitiates this as a reaction for titanium.Group TI.The elucidation of the reactions involved in the contact processfor the manufacture of sulphuric acid, mlien metallic oxides are used ascatalysts, has given rise to a considerable amount of investigation,and, in particular, the equilibrium conditions in the dissociation ofthe sulphates of iron and of a good many other metals has been veryfully studied.61 Some of the results obtained indicate that the catalyticaction of ferric oxide is not due to alternate reduction and oxidation,since sulphur dioxide by itself is without effect at the temperaturesinvolved; nor is the action due to the union of ferric oxide, sulphurdioxide, and oxygen, forming a ferric sulphste, since the concentrationof sulphur trioxide at a given temperature is higher than whatis produced by dissociation of ferric sulphate.The action would there-fore appear to consist in the direct union of sulphur dioxide andoxygen, induced by some sort of condensation effect at the surface ofthe catalyst.The dehydration of the ordinary crystallised thiosalphates is amatter of some difficulty, which formerly gave rise to divergent view.59 Ann.Report, 1907, 59 ; G. Gain, Cumpt. rend., 1907, 146, 403 ; Ann. Chim.Phys., 1908, [viii], 14, 224; A . , ii, 284, 599.TV. M. Barr, J. Amer. Chern. Soc., 1908, 30, 1668; J. H. Hildebrand, ibid.,1662 ; A!., ii, 1045.61 G. Keppeler and others, Zeitsch. physikd Chem., 1908, 62, 89 ; Zeitsch. nngew.Chem., 1908, 21, 532, 537 ; L. Wijhler and others, Ber., 1908, 41, 703 ; Zeitsch.physikal. Chem., 19OS, 62, 641 ; A,, ii, 289, 290, 482, 58170 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.regarding the constitution of these salts ; apparently, however, thedirect preparation of the anhydrous salts is comparatively simple.Bypassing hydrogen sulphide over powdered sodium sulphide at 3003, drysodium hydrosulphide is obtained, and from this the thiosulphate canbe obtained by the action of air or oxygen at 100-150°.62Most of the important work on chromium deals with the complexsalts and similar substances derived from it. The complications andisomerism apparently extend to the hydroxides ; the pure substance of'' Guignet's green " is a hydroxide, Cr,O,(OH),, which has the samecomposition as the greyish-violet hydroxide, but, whilst the formerexbibits distinct vapour pressures of 13-26 mm. at temperaturesbetween 7 5 O and S4", the latter has practically none.63G~oup VII.There is little of moment to be noted here in connexion with thehalogens; several papers have been published which deal with thedetails of electrolytic processes of preparation on the large scale, andcall for no comment.One technical matter of general interest may,however, be referred to, namely, the preparation of dry calciumhypochlorite.64 According to one process, this can be effected byacting on milk of lime with chlorine, evaporating the clear solution a ta low temperature under diminished pressure so as to crystallise outthe hypochlorite as quickly as possible, and dehydrating the crystal-lised salt, also under diminished pressure. An improvement consistsin adding fresh quantities of lime and chlorine alternately, after thefirst chlorination, by which means it is possible to obtain a good yieldof hypochlorite without evaporation.It is claimed that the dried saltis not deliquescent, and keeps well. It is, of course, much moreefficient than bleaching powder, and, on treatment with hydrochloricacid, it gives nearly its own weight of chlorine. A substance of thiskind should prove of considerable use in laboratory work.The propriety of placing manganese in the seventh group ofelements has been questioned,@' on account of its almost complete lackof relationship with the halogens, and its great similarity to membersof the eighth group. The difficulty is analogous to that which resultswhen copper, silver, and gold are classed along with the alkali metals,and shows that Mendeldeff's original arrangement is not altogether ahappy one ; although the arrangement may occasionally serve for arough classification (as in the present case), its use nowadays is hardly62 D.R.-P. 194881, 194882 ; A., ii, 689.'63 L. W'Viihler and W. Becker, Zeitsch. nngew. Chcm., 1908, 21, 1600 ; A., ii, 765.(i3 D.R.-P. 188524, 195896; A., ii, 280, 692.65 H. Reynolds, Chcna. News, 1907, 96, 260 ; A., ii, 41INORGANIC CHEMISTRY. 71justifiable.avoided by the use of longer periods.66Many difficulties, such as those mentioned above, can beGroup VIII.Interest in the metals of this group centres largely in the complexsalts formed by the various members, and in the alloys and similarsubstances of which they form constituents; in both these depart-ments there has been considerable activity during the year.The rusting of iron seems to provide a never-ending problem, anddifferent investigators continue to draw diff erent conclusions as theresult of their researches ; several additional papers have recentlyappeared.67 Two different problems are really involved in thediscussion : one the behaviour of pure iron in presence of pure waterand oxygen, and the other the behaviour of the different kinds ofiron and its alloys in ordinary use. Owing to the great difficulty ofgetting reasonably homogeneous samples of pure iron (slight differ-ences of composition may produce pronounced differences in theelectrical conditions, etc.), the first problem is one regarding whichit is very difficult to obtain a satisfactory decision, and it is one ofrelatively slight practical importance. The other, however, is amatter of the very greatest practical importance, and necessitates astudy of the effects which variations in composition and in theattendant circumstances produce on the rate of rusting underordinary conditions; the second of the papers to which reference hasbeen made is an extensive investigation of this kind.There is further evidence that platinum, when used as an anode incertain electrolytic processes, or otherwise submitted t o the action ofpowerful oxidising agents, is not quite so resistant as is generallyassumed,6s so that the possibility of traces of platinum passing intosolution in this way should not be overlooked. I n consequence ofthe highly refractory qualities of iridium and rhodium, these metalsseem to be almost ideal materials (except as regards price) for theproduction OF crucibles and other apparatus.69 '' Boiling aqua regia,fused microcosmic salt, or other phosphates with frequent additions ofcarbon, strongly heated silica or silicates with a reducing agent, boil-ing lead at a white heat, boiling zinc, and molten nickel, iron, orti6 See, for example, J. Walker, Introduction to Physical Chemistry ; A. Werner,Neucre Anscharcungen a?$ dem Gebietc der aizorganischen Chemie.67 J. N. Friend, J. I r o n and Steel Inst., 1908, 77, i, 5 ; E. Heyn and 0. Bauer,Mitt. k. i!lat.-pruf.-Aint., Gi.oss-Liziiterfeldc, 1907, 26, 1 ; -4., ii, 698, 849 ; W. A.Tilden, Trans., 1908, 93, 1356.6y Awn. Report, 1907, 73 ; C . Marie, Coinpt. rewd., 1908, 146, 475 ; R. Ruer,Zeitsch. EZektroche?iz., 1908, 14, 309, 633 ; A . , ii, 299, 601, 954.6g Sir W. Crookes, PYOC. Boy. SOC., 1908, 80, A, 535 ; A . , ii, 70272 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.gold, are without effect on the crucible, which, after cleaning, retainsits weight unchanged ” ; it is also passive under many other forms ofchemical torture.A very full investigation of the preparation and properties of theoxides of iridium has been carried out.70 No monoxide, IrO, could beobtained, although such has been said to exist; the sesquioxide,Ir203, and the trioxide, Ir03, could not be prepared in a pure con-dition; the dioxide, Ir02, however, was obtained practically pure as asolid, and also in colloidal solution. When heated, the sesquioxidedecomposes into dioxide and metal, some oxygen also being evolved.HUGH MARSHALL.7O L. Wohler and W. Witzmann, Zeitsch. anorg. Chem., 1908, 57, 323; A., ii300