INORGANIC CHEMISTRY.DURING the year, a very large number of papers have been pub-lished which come within the purview of this Report. It isimpossible t o do more than refer to a very few of these, and i t iswith some regret that so many interesting investigations are ofnecessity omitted. A t the same time, there have appeared papersof outstanding interest and importance-papers which seem likelyto mark this year as above other years.There is little doubt that the general trend of inorganic chem-istry is on philosophic rather than purely preparative lines. Lessinterest must surely be found in the reading of the characterisationof new compounds, however novel these may be, than in the read-ing of pioneer work which marches steadily on into the unknown,bearing down old conventions where such oppose it, confident inits sense of conscious.merit, and safeguarded by its sense of trueproportion.Sure of its inherent accuracy and fully aware of itsrevolutionary influence on the treasured tenets of the chemistry ofyesterday, such work must rivet the attention of all. When onthe one side he is confronted with the disintegration of elementsunder electric stresses, and on the other he learns that the1 atomicweight of certain elements can vary by as much as 1 per cent.,according to the parentage of those elements, and when, again, heis told that the physical constants of metals as now known areworthless, because they are perpetually in a state of change fromallotrope to allotrope, the chemist of to-day may be likened to hisforefather of preWAvogadro days.As did his forefathers of thosedays, so also does he now await that great generalisation whichshall co-ordinate and link up all the threads to found a new philo-sophy. Radioactivity, enhanced line spectra, the intra-stellarelements, active nitrogen and oxygen, atomic disintegration,atomic-weight variation, each at the moment a watertight cornpart-ment of research, all will be unified and embodied in the newphilosophy of the twentieth century. Then will a new chelmistryin its greater meaning emerge as a phoenix from the glowingparental fires of the many chemistries of to-day.Little apology is needed for the pages in this Report which have3INORGAPU'IC CHEMISTRY. 35been devoted t o those investigations which are leading to theinception of the new philosophy.It is impossible t o pass by with-out special notice Collie's work on the disintegration of certlainelements, the work of Soddy and Hyman, Richards and Lembert,and of others on the atomic weight of lead. As regards the latterwork, a suggestion as to there being more than one inhabit'ant ofeach gap in the periodic table! was put forward by Sir W. C'rookesmany years ago, before radioactive phenomena were known. Madeas i t was to explain certain facts then under discussion, thissuggstion was forgotten, together with the facts. It was reservedfor Soddy and Fajans independently to enunciate their hypothesis,basing it on sound argument. It is inherently startling in concep-tion, and one is privileged t o ask how many elements are likely tobe concerned and how many atomic weights may be rendered in-secure.One could have wished that by some yet unknown processof radioactive disintegration the formation of the rare earths couldbe explained. It would be far more satisfactory could one lookupon these mysterious elements as the final products of changefrom elements which occupy legitimate positions in Mendelhev'stable. They would then be considered as parasitical growthsarising from elements formed, no doubt, by natnral synthetic pro-cesses. Be that as i t may, there is no doubt that the new workon the atomic weight of lead is thoroughly sound, and that i tsupports the hypothesis of Soddy and Fajans in a most remarkableway.As regards the work of Collie, there now seems little doubt ofits absolute validity. First put forward last year in a somewhattentative fashion, the very definite results recently recorded byCollie, Patterson, and Massori bring conviction in their wake.This work is critically discussed in the Report, and nothing furtherneed here be added.The remainder of the Report follows the lines of those of previousyears, but, as was stated above, some regret is felt that want ofspace compels the omission of much that is interesting.A tomic Il'eights.The year has been signalled by very remarkable work on theatomic weight of lead of radioactive origin.I n last year's Reporton radioactivity, Professor Soddy gave an account of the generalisa-tion which, amongst other facts, involves the existence in some ofthe places in the periodic table of several elements, differing inatomic weight, but remarkably similar in other properties.Itfollows that all the radio-elemsnts which are chemically identicalin character should occupy the same position in the periodic table.0 36 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Such elements are termed isotopic. The unknown end-products ofall the known disintegration series fall into the place occupied bylead. I f these products are entirely stable, the lead from radio-active minerals should differ in atomic weight according t o theamount of thorium or uranium in the mineral. The atomic weightof the thorium isotope should be 208-4, and that of the uraniumisotope should be 206.0. The experimental results obtained infour different laboratories support this in a very remarkable way.In the first case, the atomic weight of the lead in Ceylon thoritewas determined.1 An analysis of this mineral showed thepresence of 61.95 per cent.of Tho,, 0.85 per cent. of U,O,, and0.39 per cent. of PbO. The very small quantities of lead suggestthat it is all of radioactive origin, none being present as anoriginal constituent. If that be so, then the atomic weight of thelead should be about 208.2, since 10 parts of it arise from thoriumto 1 part from uranium.The lead was extracted from a kilogram of the mineral, andafter the most careful purification about 1.2 grams of lead chloridewere obtained.The atomic weight of this lead was compareddirectly with that of the lead in a sample of the chloride preparedfrom the ordinary nitrate in the same way as was the mineralspecimen. The two specimens were compared by titration withsilver nitrate solution, the methods of weighing being those adoptedby Baxter and Wilson in their work, on which the internationalvalue is mainly based. I n one set of experiments the ratio of theatomic weight of the lead from thorite to that of ordinary leadwas found to be 1.0049, whilst in a second set it was 1.0042. Theatomic weight of the lead from thorite was therefore 208.5 and208.3, respectively, in the two sets of determinations. The differ-ence between this and the accepted value is far greater than theprobable error of experiment unless unknown sources of error exist.A still more striking result has been obtained in the Harvardlaboratories,2 for in this case lead of radioactive origin from severalsourcm was used, and the values of the atomic weight were foundto differ very considerably, according to the origin of the lead.Itis not necessary here to enter into a description of the methodsemployed in the purification of the various samples of lead chlorideor of the details of the analysis which evolved the determinationof the ratio P b : Ag. They were carried through with thataccuracy with which the name of Harvard is associated. It mayperhaps be stated that the determinations made with any onesample of lead chloride agreed exceedingly well amongst them-F.Soddy and H. Hyman, T., 1914, 105, 1402.T. W. Richards and M. E. Lembert, J. Avncr. Chem. h'oc., 1914, 36, 1329;A . , ii, 653INORGANIC CHEMISTRY. 37selvee.atomic weights of the lead from the various sources being given:The result may be tabulated as follows, the values of theLead from N. Carolina urnnin i te ..................... 206 '4Leal1 from Joachinisthal pitchblende ............... 206 57L e d t'rorn Colorado carnotite ........................ 206.59Leal1 from Ceyloiiese thorianite ..................... 206.82Lead from English pitchblende ..................... 206.84Coninion 1, ad,.. .................................... 207.1 5Without doubt, these results are exceedingly remarkable, andthey undoubtedly support the generalisation as to isotopic elements.They are entirely contrary to experience, for very careful investiga-tions at Harvard have failed to reveal any difference whatever inthe atomic weights of copper, silver, iron, sodium, and chlorineobtained from the most varied sources possible.Again a spectrographic investigation of the lead of radioactiveorigin showed no difference between it and ordinary lead.Soddyand Hyman, however, found that one line, 1=4760-1, is muchweaker in the case of the " t'horite " lead.The above results have received further confirmation. I n onecase, the atomic weight of lead from pitchblende was found byBaxter's method to be 206.736 as a mean of nine determinations.3I n the second case,4 lead was obtained from various radioactiveminerals, and the atomic weights were determined by the Stasmethod from the ratio1 P b : Pb(NO,),.The lead from uraniumminerals gave an atomic weight between 206.36 and 206.65, whilstthat from monazite gave 207.08, the value for ordinary lead fromgalena being taken as 207.01.As regards the report of the International Committee, no changeis recommended in the values given in the table published in 1913.The report draws attention to certain investigations of varyingimportance that have been carried out. Reference may be madeto one or two of these, and certain papers since published.The density of oxygen has been redetermined after the gas,obtained in the first place by heating pure potassium per-manganate, had been carefully fractionated.5 The mean valuefrom fifteen determinations of the weight of a normal litre (OO,760 mm., sea-level, and a latitude of 45O) is 1'42906 grams.Incombination with the values obtained by Morley and by Rayleigh,the most probable value is 1.42905 grams.Some further work on the atomic weight of tellurium has beencarried out by the conversion of tellurium hydride, TeH,, into0. Honigschmid and (Mlle.) St. Horovitz, Compt. Tend., 1914, 158, 1796 ;A., ii, 655.31. Curie, ibid., 1676 ; A., ii, 563.A . F. 0. Germann, %id., 1913, 157, 926 ; J. C ' h i ~ . phys., 1914, 12, 66 ; A , , ii,47, 45438 ANNUSL REPORTS ON THE PROGRESS OF CHEMISTRY.tellurium dioxide, Te0,.6 The hydride was prepared fromaluminium telluride, and also by the electrolysis of a 50 pelr cent.solution of phosphoric acid with a tellurium cathode and aplatinum anode.The gas was frozen in liquid air,and then formedwhite crystals, melting a t -57O and boiling a t Oo. The rwultssupport the view that, tellurium is not complex, and that the' pre-sent atomic weight, 127.5, is the correct value.The atomic weight of copper relatively t o t h a t of silver hasbeen determined by an electrolytic methcd,7 the weights of the twometals deposited by the same quantity of current being compared.The mean of ten values gives for the atomic weight of coppelr63.563 f 0.003 (Ag = 107.88).Two determinations have been made of the atomic weight ofmercury. I n the first, mercuric oxide was converted t o mercuryby heating with metallic iron,8 and the mean of nine very closevalues gave the atomic weight as 200-37kO-0'35. This value isconsiderably lower than the; accepted value, and the authors con-sidered that the discrepancy might be explained by the presenceof a small quantity of a higher oxide, but that the result calledfor a further investigation of the atomic weight.This has beencarried out recently by Easley and Brann's method of conversionof mercury into mercuric bromide.9 The values obtained laybetween 200.50 and 200.62, and as a mean of nine experime'nts thevalue of 200.57 +0.008 was deduced, a value which is in close agree-ment with that now accepted.The density of neon after purification from helium has beenredetermined by Leduc.10 This was experimentally found to be0.695, referred to air as unity, from which by his formula, theauthor calculates the atomic weight to be 20.15.Some criticisin has appeared of Auer von Welsbach's work onytterbium (aldebaranium) and lutecium (cassiopeium) , for whichhe found the atomic weights of 173 and 175 respectively.11 Theearths of the ytterbium group, in the form of their nitrates, havebeen submitted t o fractional crystallisation, the stages in theseparation being followed by measuring the coefficients of magnet-isation of each fraction.12 After 4000 crystallisations, eightL.RI. Dennis and Ic. P. Aitdersoii, J. Airier.. Chci~i. Sot., 1914, 36, 882 ;A . , ii, 456.7 A. G. Shiimpton, Proc. T'h?y~~im? SOC., J,ondo?l, 1914, 26, 292 ; A., ii, i 7 4 .G.B. Taylor aiid G. A. Hul~tt, J. Physicd C h i t . , 1913, 17, 755 ; A . , ii, 128.H. €5. Ijakcr a i d W. 11. Watsoii, 5"., 1914, 105, 2630.lo A. Lediic, Conzyt. T e f r d . , 1914, 158, 864 ; A., ii, 361.l1 iJIo?zntsh., 1913, 34, 1713; A . , ii, 130.l2 J. Kluiiieiifeld and G . Urlbdn, C'o?upt, ? * L ' ? L ~ , 1914, 159, 323, 401 ; if., ii,731, 694INORGANIC CHEMISTRY. 39fractions were obtained with the same coefficient, which is acceptedas proof of the existence of a definite substance. The metal of thenitrate corresponding with these sight fractions is called neo-ytterbium, and has an atomic weight of 173.54. The arc spectraof the first and the last' of these eight fractions were observed, andthey were identical except for a few rays in the one fraction duet o thulium, and a few in the other due t o lutecium.A ,?lo t ropy.Two new modifications of phosphorus have been discoveredduring the course of some experiments on the effect of high pressureon the melting point of ordinary white phosphorus.13 The first ofthese is a new form of white phosphorus, which changes reversiblyinto the ordinary modification, and the second is a form obtainedirreversibly from the ordinary form at high pressures and moderatetemperatures, and is 15 per cent.more dense than Hittorff'smetallic red phosphorus. The white modification was first pro-duced by increasing the pressure on ordinary phosphorus t o about11,000 kilogram/cim.2 at 60°. The transition temperature is alinear function of the pressure, and lies between -76.9O a t apressure of 1 kilogramIcm.2 and 64.4O at a pressure of 12,000kilogram/cm.2 By crystallisation from carbon disulphide a t lowtemperatures, the new modification was obtained in the form ofmicroscopic crystals belonging to the hexagonal system.The black modification is readily obtained by heating ordinaryphosphorus to ZOOo under a pressure of 12,000 kilogramjcm.2 Thedensity of this substance is very high, being about 2-69, as against1-83 for ordinary white phosphorus and 2.34 for Hittorff's metalliccrystallised red phosphorus.It is certainly a new form, and ischaracterised by not being spontaneously inflammable; it can withdifficulty be ignited, and may be heated to about 400° in air with-out catching fire.It bears some resemblance to red phosphorus inthat it is attacked by cold nitric acid, but not by sulphuric acid,and i t is not dissolved by carbon disulphide. When heated in aclosed tube it vaporises and condenses on the colder parts of thetube to a mixture of red and white phosphorus. It would seem,therefore, that the vapours of the black and the red forms are, a tleast in large part, identical. I n contradistinction from the red andthe) white varieties, i t is a fairly good conducttor of electricity, thespecific resistance being 0.711 ohm per cm. cube a t Oo. Theteniperature-coefficient has a large negative value, and the relationbetween temperature and resistance is nearly linear between 0'and 75O. The specific heat was determined between 100' and 30°,Id 1'.W. Bridgman, J. Amer. Chem. Soc., 1914, 36, 1344 ; A . , ii, 64740 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.and was found to be 0-170, a value which is considerably lowerthan those recorded for the other modifications. The vapourpressure was determined, and was also found to be lower than thatof red phosphorus. The new form, therefore, appears to be morestable than the red variety.The investigations on the allotropy of the metals have cunsider-ably extended during the year, since copper, cadmium, and zinchave been added to the list of those which possess allotropic modifi-cations.14 The proofs of the existence of the different forms arebased on dilatrometric observations, and as yet the pure allotropesdo not seem to have been prepared.I n the case of copper, tchetransition temperature is 71'7O, and in the case of cadmium thetransition temperature is 64.9O. With both these metals, however,it has been found that the transition temperature depends upon theprevious thermal history of the metal, and therefore it is con-cluded that more than two allotropic modifications exist. Itfollows that these metals normally are metastable mixtures of a tIeast two allotropes, and that they are continually changing slowlyinto the stable modifications a t the ordinary temperature. Theseobservations are of considerable imporkance, for they show that thephysical constants of these metals, and in all probability of manyothers, have no definite significance. It will be necessary to re-determine these, and study the physico-chemical properties bymaking use of the pure allotropic forms.Certain anomalousresults on record are explained by this work, for example,Matthiessen and Bose's observation of the disintegration ofcadmium wires heated a t BOO, and also of changes in the electricalconductivity of copper wires after these had been heated forseveral days a t looo. It is probable that the stable form of zincwas obtained in an almost pure form by Kahlbaum, Roth, andSiedler.ls Again, the specific heat-temperature curves for copper,zinc, lead, aluminium, and silver are not continuous, but changeabruptly a t one or more points. This agrees with the dilatrometricobservations for some of these metals.Further work has also been carried out on the new form ofsulphur, S,, the discovery of which was noted in last year'sreport.16 A method has been worked out whereby the quantities14 E.Cohen and W. D. Helderman, Pruc. K. Aknci. Wetetcnsch. Amsterdam, 1913,16, 628; 1914, 17, 6 0 ; A., ii, 205, 654 ; ibid., 1913, 16, 4 8 5 ; 1914, 17, 54, 1 2 2 ;A., ii, 52, 652 ; ibid., 1913, 16, 565 ; 1914, 17, 59 ; A . , ii, 127, 652; Zeitsch.physikal. Chem., 1914, 87, 409, 419, 426 ; E. Cohen, €'roe. K. Akad. Fetensch.Amsterdam, 1913, 16, 632 ; 1914, 17, 200 ; A . , ii, 202, 799 ; Zeitszh. physiknl.Chem., 1914, 87, 431.15 Zeitsch. anorg. Chcm., 1902, 29, 177 ; A . , 1902, ii, 259.16 A. H. W. Aten, Zeitsch. physikal. Ciienz., 1913, 86, 1 ; A., ii, 121INORGANIC CHEMISTRY.41of S,, S,, and S, can be estimated in a given mixture of the three,and this has been applied to samples of sulphur which had beentreated in various ways. The velocity of change, S, --+ S,, hasbeen investigated, and it is found to be very great at first, but tofall rapidly t o a very small value when the quantity of S, presentis small. The maximum amount of S, which it has been possibleto obtain by heating sulphur to various temperatures is about 6.5per cent. after the sulphur has been heated to 1 8 0 O .A study has also been made of the effect of the three catalystssulphur dioxide, ammonia, and iodine, which are known t o havegreat influence oc the equilibrium S, S,. Ammonia was foundconsiderably to shift the equilibrium, with the result that thequantity of S, was reduced t o a very small value.The velocity ofthe change, S, - S,, is about the same whether iodine is presentor not. The bearing of the existence of S, as a factor in theequilibrium on the properties of sulphur was considered, and itwas shown that, in general, these are explained more satisfactorilythan is possible on the lines of the old two-component system.Thus the change in viscosity with temperature, and the variationin crystallisation power with previous thermal treatment, are shownto be far better explained when the presence of S, is taken intoaccount.Finally, analyses of solid sulphur, after treatment in variousways, are given, and it is shown that the quantities of S, presentare always exceedingly small, and never greater than 0.1 per cent.,and this only in sulphur that has been previously melted.Similarly, S, is only t o be found in sulphur that has previouslybeen melted, and the maximum amount formed was 0.5 per cent.These results are very different from those given by Kruyt,l7according to whom, sulphur which has been heated to 90° and 6 5 Ocontains 3 and 2.4 per cent.of S, respectively. Kruyt, however,estimated the amount of S, by the change in melting point, andnot by direct analytical methods.Electric Bischcirge.The work of Collie and Patterson and of Sir J. J. Thomson onthe production of helium and neon in vacuum tubes during thepassage of the electric discharge was dealt with a t some length inlast year's Report. During the year, several papers have been pub-lished upon this work, and there would seem to be' little doubt thatthe evidence in favour of the actual production of these gases bythe discharge ha5 been materially strengthened.A very obviouscriticism can be made of the detection of minute quantities of neonl7 Zeitsch. physikal. ChenL., 1908, 6%) 513 ; A . , 1908, ii, 102842 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.in the residuum after all the diatomic gases have been frozen outby means of charcoal cooled in liquid air, namely, that this gasarises from very small quantities of air which had leaked into theapparatus during the experiment. Such criticism might a t firstsight seem to be justified, because two indelpendent observers havebeen unable to detect the production of either helium or neon whenthe possibility of leakage of air had absolutely been precluded.l* l9The fact that an exceedingly small quantity of air is necessary t ogive enough neon f o r spectroscopic detection was pointed out byStrutt, who found that 0.01 C.C.of air is amply sufficient. Thefailure of Strutt and of Merton t o observe the formation of eitherhelium or neon might therefore readily be interpreted as a some-what effective criticism of the apparatus used in those experimentsdescribed in last year’s report.The essential improvement in the new apparatus described byStrutt and by Merton lies in the fact that the gas, after it hasb&n subjected to the influence of the discharge, is not transferredt o a second apparatus for analysis.The whole operation is carriedout in one self-contained piece of apparatus, and in Merton’s ex-periments, after the gas to be experimented with had once beenintroduced, it did not come into co’ntact with any stopcocks. Asingularly ingenious method of introducing hydrogen into theapparatus was devised by Merton in the shape of a small palladiumtube sealed into a glass tube. The outer end of the palladium tubewas hermetically closed, and by heating this tube with a Bunsenflame, small quantities of hydrogen diffused into the apparatus,which had previously been exhausted. There is no need to describethe actual experiments, since they gave negative results, despitethe fact that they were similar in character to those previouslycarried out by Collie, Patterson, and Mason.Quite apart, however, from the later results published by Collie,i t does not seem possible t o accept such a simple explanation aswould seem to be suggested by Strutt and Merton’s work.Itmust not be forgotten, in the first place, that if the neon foundwere due to an exceedingly small air-leak, the amount of argonwould be enormously greater (700 times), and could not possiblyescape detection. I n the second place, the production of helium,which was so clearly noted in the early work, lies in a differentcategory. Whilst neon might conceivably be due t o air-leaks, theformation of helium in even greater amounts than the neon, whichwas definitely observed in several cases, can hardly be susceptibleof the same explanation.The amount of helium present in the airis far less than that of neon, and therefore an air-leak would give18 (IIon.) R. J. S t l u t t , Pmc. Koy. Suc., 1914, [ A ] , 89, 499 ; A . , ii, 201.19 T. R, Merton, ibid., [A], 90, 549 ; A,, ii, 726INORGANIC CHEMISTRY. 43the latter gas in considerable excess. With regard to the absenceof argon, there is just the possibility that it might have been re-moved by the charcoal along with the diatomic gases. Be thathow it may, there does not seem t o be the faintest possibility thatreasonable amounts of helium and little neon could be due to anundiscovered leak of air.The problem gained considerable interest when Collie,20 usingMerton's own apparatus, obtained considerable quantities of heliumand neon by the cathode-ray bombardment of powdered uraniummetal in an atmosphere of hydrogen.Five grams of the uraniumwere finely powdered in a stetel mortar, and then heated to red-ness in a vacuum f o r half an hour. The metal was then trans-ferred to a small bulb, in which it could be bombarded by thecathode streams, and this tube was sealed on to the Mertonapparatus. A small bulb containing charcoal, a hard glass tubecontaining copper and cupric oxide, and a small bulb containingphosphoric oxide, were1 also sealed to the apparatus. The wholewas washed out several times with pure oxygen, and then exhausteduntil the discharge would not pass. The tube containing theuranium was then heated as strongly as possible, and the gaseswere pumped off and examined.Carbon gases and hydrogen werepresent, but only just sufficient helium and neon to be detected inthe usual way. The tube was then again washed out with pureoxygen and re-exhausted to the utmost limit. A small quantityof hydrogen was admitted by heating the palladium tube for twentyto thirty seconds, so as to allow the discharge t o pass. The bom-bardment of the' uranium lasted for two hours, and during thistime carbon gases and hydrogen were evolved. These wereabsorbed by cooling the charcoal with liquid air and by heatingthe cupric oxide. At the end, considerable quantities of heliumand neon were found t o be1 present, and the experiment was re-peated several times with the same result. By slight variation inthe treatment of the gases, nitrogen was found in small quantities,but this disappeared after sparking above the mercury in the con-taining vessel.Only very small traces of argon were noted, sinceit showed only the blue spectrum. During these experiment8 thegases did not come into contact with any stopcocks, and thereforethe arguments against air-leaks as the origin of the monatomic gasesappear absolutely sound, and the results would seem to discountthe criticism that might have been based on Strutt's and Merton'swork.I n describing these experiments, mention is made for the firsttime of a fact which may give the key to the divergent results'Lo J. N. Collie, Prbc. Ii'oy. A'oc., 1914, [ A ] , 90, 554 ; A . , ii, 72744 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.obtained by the different observers, for it must be rememberedthat positive results were also obtained by Thomson, as well as byCollie and Patterson and by Masson, in 1913.I n one or two experi-ments Collie found that by using a larger induction coil with amercury break in place of the original 10-in. coil with a platinumbreak, no helium or neon was produced. On repetition of the ex-periments under exactly the same conditions, but with the old coil,i t was found that. the helium and neon were again produced to thesame amounts as previously. The very small traces of nitrogendetected in the residual gases after the charcoal had been cooled isattributed by Collie to the presence of some nitride in the uranium.The presence of this trace of nit”rogen and its spectroscopic detec-tion is of some considerable importance.It is a well-known factthat very minute quantities of nitrogen can be detected by meansof its spectrum, and, further, it is perfectly evident that onlyminute traces were present, since the gases had been for some timein contact with the charcoal cooled in liquid air. It may safely bededuced, therefore, that those positive results of helium and neonproduction when no nitrogen was observed (and these form byfar the h g e r majority) could not in any way be due t~ air-leaks.I n a recent paper the previous positive results are confirmed ina still more striking way, and also the question of possible air-leaks is very critically considered.21 The precautions taken againstsuch leaks are described in detail, and these leave no doubt that,whatever may be the origin of the production of helium and neon,it does not arise from faulty apparatus.I n this paper the detailsof the various types of apparatus used by these authors in theirseparate and individual experiments are described more minutelythan was the case in the previous papers. Very many differentdesigns of discharge tubes were used, some with outer jackels andsome without. In the case of the former, the jackets were eitherexhausted, filled with neon, or filled with water. Since these threealternatives made no apparent difference, it follows that the soleremaining possibility of external contamination is excluded,namely, the porosity of the glass of the discharge tube to heliumand neon only during the discharge. This indeed savours of hyper-criticism, but clearly even the most far-fetched hypothesis must betested.As stated in last year’s Report, both helium and neon werefound in the outer jacket when the experiments were carried outwith this vessel completely exhausted. This result, which has beenconfirmed, is perhaps the most extraordinary of all that have beenrecorded. It may be remarked here that numbers of control ex-periments were made, that is to say, the apparatus was left stand-J. N. Collie, H. S. Pattcrsoii, and I, Maason, Proc. Roy. Soc., 1914, [ A ] , 91,30 ; A . , ii, 847INORGANIC CHEMISTRY. 45ing under exactly the same conditions as in actual experiments, butwithout any discharge being passed.I n every such control nohelium or neon was found.A great number of different types of experiments have now beenmade, including the discharge between various metallic electrodes,the bombardment of various metals, oxides, and salts, and themercury arc lamp in silica tubes, with and without a water jacket.I n all three types, both helium and neon were found in varyingamounts. Many isolated experiments gave negative results, owingt o reasons as yet unexplained. The most important part of therecent paper, a t any rate to those who still maintain an attitudeof destructive criticism, lies in a detailed discussion of the possiblesources of error and the precautions that were taken against sucherrors.The question of atmospheric contamination must now betaken as being definitely decided, for the controls show that thereis no leak in the apparatus, and the sole remaining possibility ofthe permeation of air through the walls of the discharge tubeduring the experiment is negatived by the absence of the nitrogenand argon spectra, and also by the fact that the use of an outerjacket, which entirely covers the discharge tube, makes no appreci-able difference in the result.I n the writer’s opinion, the strongest evidence yet advanced forthe reality of the result is to be found in the fact that negativeresults are sometimes obtained, and that these have been traced insome casm to differences in the type of induction coil used.If,as now seems probable, the helium and neon are really disintegra-tion products arising from the effect ot the discharge, such negativeresults are bound to be obtained unless the discharge is tuned, asit were, to the tube used. I n the earlier experiments the authorswere fortunate enough in most cases to effect such tuning un-consciously, but when many and various types of tubes were triedthe natural result happened. Success and want of success followedone another in an apparently inexplicable way, owing to want oftuning. One thing a t least may besaid, namely, that the negativeresults must surely rule out air-leaks as contributory causes unlesssuch leaks exhibit a discontinuity in their behaviour, which wideexperience shows to be entirely foreign to their nature.It is a familiar fact to anyone who has carried out experimentson the phenomena taking place in vacuum tubes and electric dis-charge generally that the results obtained vary in a most perplex-ing way.The principal source of such variations seems to lie inthe induction coil, the capacity of the discharge tubes, the natureof the discharge, etc. Such results afford keen disappointment tothe experimenter when first met with, but they fall into line a 46 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.once when their origin is explained. I n the present work, theirvery existence and their explanation, at any rate in many cases, bythe variation in electrical conditions do more t o establish thevalidity of the positive results than destructive criticism can everbe relied upon to effect.As regards the origin of these gases, i t was stated above thatthere can be little doubt but that they rise from atomic disintegra-tion under the influence of the discharge.Sir J. J. Thornson, aswas detailed in last year's Report, favoured the view that thehelium obtained had previously been occluded in the electrodesused or the substance bombarded. Against this view certain rigidtests must be recorded. I n the first place, the positive resultsobtained with electrodeless discharge tubes prove that metallicelectrodes are not necessarily the source, and, further, the produc-tion of helium and neon in the arc between mercury surfaces isagainst such a view, seeing that neither gas was obtained on boil-ing the mercury in a vacuum.Again, Collie, Patterson, andMasson did not experience the steady fall in the amounts of heliumand neon obtained when the same tube was used several times insuccession, as was noticed by Thomson. Further, the followingtests were applied in connexion with certain sets of experiments.Positive results of helium and neon were, for example, obtainedwith a jacketed tube with aluminium electrodes and with hydrogenas the gas. All air-leaks were precluded by the jacket and by thefact that every control experiment gave no trace either of heliumor neon. A quantity of the aluminium wire used in making theelectrodes was melted in a vacuum, but gave off no trace of eithergas, and the same was true of the glass out of which the tube hadbeen made.A further portion of the aluminium wire was thendissolved in air-free potassium hydroxide solution. The hydrogenevolved was passed over heated cupric oxide, and the residue wasexamined in the same apparatus as used for the discharge experi-ments. Similarly, a large quantity of the glass was powdered andacted on by potassium fluoride and concentrated sulphuric acid inthe absence of air. About 300-400 C.C. of silicon fluoride werecollected, and frozen out by means of liquid air. I n neither casewas any helium or neon found in the residual gases. Similarnegative results were obtained with old Ro'man and Chinese glasses.Some valuable support t o this is obtained from some tests carriedout by Sir J.J. Thornson, who found that aluminium salts gavethe same quantity of helium on bombardment by cathode rays,whether they were made from ordinary aluminium metal or fromthe same metal which had originally had helium forced into i t bybeing made the electrode of a helium tube. As Thomson says, thilNORGANIC CHEMISTRP. 47shows that solution can be trusted to eliminate adsorbed gas. Itis unreasonable, therefore, in the extreme to suppose that if heliumand neon were present as such in the materials used, they wouldnot be set free on dissolution which is accompanied by gas evolu-tion. Solution of aluminium in potassium hydroxide, for example,entails a far more thorough physical disintegration of the piece ofmetal than does the bombardment of its surface. The only tenablesupposition is that any inert gases physically admixed must be setfree; it is not possible t o assume that anything more than a physicaladmixture in the case of inert gases pre-exists in aluminium or inglass.The proved absence of neon and helium from the resultingproducts of chemical reaction must prove, therefore, that they areabsent altogether.I n conclusion, the case in favour of these gases being due toatomic disintegration seems overwhelmingly strong, and the writerfeels that no apology is needed for so full an account of this work,which must rank as one of the most remarkable investigations ofmodern times.Some noteworthy contributions have been made t o the problemof active nitrogen, first discovered by Strutt, and further dealt withby him in papers describeld in the Reports for 1912 and 1913.It,would seem that some doubt has been thrown on the mechanismof the production of the active nitrogen, that is t o say, whether i tis produced from absolutely pure nitrogen or whether a trace ofoxygen is a necessary factor. It might a t first be thought that i tis a matter of secondary importance as t o whether the oxygen isa contributory factor or not, seeing that the essential point hasclearly been established, namely, that an active nitrogen does exist.It must not be forgotten, however, that the mechanism of theformation of this gas is not understood, nor has the reason of itsabnormal reactivity ever been explained.I n last year’s Report mention was made of some experimentswhich seem to prove that the presence of oxygen is a necessity forthe activation of the nitrogen, but this statement was controvertedby Strutt himself. The original experiments were again repeated,and, in spite of Strutt’s contradiction, the authors maintained theirposition.22 The greatest possible care was taken in the making ofthe apparatus and in the filling of it with absolutely pure nitrogen.No fat was used on the stopcocks, and all traces of mercury vapourwere removed by means of gold-leaf cooled in liquid air.Severalseries of experiments were carried out, and in the last the nitrogenwas prepared in the apparatus by heating barium azoimide to 1 7 0 O .I n each case no sign of the Strutt phenomenon was observed unless22 E.T i d e and E. Domcke, Bw., 1913, 46, 4095; A., ii, 12248 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.small quantities of oxygen were mixed with the nitrogen. Theoxygen was obtained by heating silver oxide contained in a smalltube sealed to the apparatus.A slightly different view of the plienomena has been put forwardby Koenig and Elod,23 for they differentiate between the after-glow(Strutt’s phenomenon) and the formation of the active nitrogen.The presence of the former is not necessary t o prove the existenceof the latter, a statement that is of some importance, since i t wasthe after-glow which Tiede and Domcke used as criterion of theformation of the active nitrogen. The real test for the presenceof active nitrogen lies in the reactions which can be carried out byits means.It was found, further, that active nitrogen can readilybe prepared by passing the pure gas, a t 15 mm. pressure, througha direct-current arc. The after-glow was then exceedingly wellmarked, and a number of new reactions of the activated nitrogenwere observed. For example, ethylene and acetylene react vigor-ously, the aftier-glow disappears, and the region where the twogases meet is marked by a lilac flame showing the spectrum ofcyanogen. Hydrogen cyanide is formed in a quantity correspond-ing with that found by Strutt. Pentane gives ammonia, amyleneand hydrogen cyanide.It was also found that when pure oxygen is passed through thearc, a weak, bluish-green after-glow is formed. I f the glowingoxygen is mixed with glowing nitrogen from a second apparatus,both glows are immediately extinguished, and oxides of nitrogenare formed in quantity.If the arc is extinguished in eitherapparatus, the formation of the nitrogen oxides ceases a t once. Itis concluded from these experiments that an active form of oxygenis produced, differing from ozone and capable of existing only f o ra short time.It may be pointed out that these results are in agreement with,and entirely confirm, those obtained by Lowry, and fully describedin the Report for 1912. This seems t o have escaped the notice ofKoenig and Elod, for they make no reference t o the fact that thephenomenon had previously been discovered by Lowry.I n a further paperF4 the results of Tiede and Domcke areadversely criticised.If mercury vapour is mixed with the nitrogen,the after-glow is not observed, although the active gas is still pro-duced. The presence of alkali metal vapour has a similar effect,and thus the suggestion is made that Tiede and Domcke’s resultswere due to the admixture of either mercury vapour o r potassiumor barium vapour (from the azoimides used to prepare the nitrogen).Tiede and Domcke, however, carried out experiments with23 A. Koenig and E. Elod, Ber., 1914, 47, 516 ; A . , ii, 264.lbid., 523 ; A., ii, 266INORGANIC CHEMISTRY. 49nitrogen fr9m which all the oxygen had been removed by means ofheated copper.25 They state that with this nitrogen the Struttphenomena are not obtained with iodine, sulphur, sodium, andthallium chloride, but the addition of the least trace of oxygen a tonce gives rise to the phenomena.I n this work, the alkali metalvapours were certainly absent, and the precautions for removingmercury vapour would seem amply to be sufficient, so that Koenigand Elod’s criticism does not appear t o be sound.Baker and Strutt, on the other hand, found themselves quiteunable to confirm these results, for on repeating the experimentsthey found the after-glow very pronounced, and also very decidedevidence of the activity of the nitrogen.26An interesting conclusion to the argument is arrived a t in ajoint paper by all four authors.27 It is explained that the contra-dictory results are due very largely t o differences in the apparatusused.Whereas with Tiede and Donicke’s apparatus the brightnessof the after-glow was increased by the addition of traces of oxygen,in Baker and Strutt’s apparatus no appreciable diminution in theglow was observed when absolutely pure nitrogen was used. Theconclusion is drawn that i t is probable that the production of activenitrogen is favoured by traces of oxygen, although in a suitableapparatus the result can be obtained even with the purest nitrogen.It is curious, however, that a real explanation of no activenitrogen being obtained by Tiede and Domcke is not given, unlessi t be attributed entirely to differences in the apparatus. If thisis so, it is another and mwt interesting instance of the necessityof tuning the discharge t o the type of tube used, such as wasreferred to above under Collie, Patterson, and Masson’s work.A tany rate, the criticism of Koenig and Elod is effectively met by thefacts published in the joint paper; this is pointed out in a finalcommunication.28Possibly this dependence on the relation between discharge andapparatus used is similar to that observed in the formation ofammonia from its elemenk under the influence of the silent dis-charge.29 I n this case, also, the amount of ammonia formed de-pends upon the dimensions of the apparatus and the density andthe oscillation frequency of the current.zi E. Tiede and E. Domcke, Ber., 1914, 47, 420 ; A . , ii, 196.26 H. B. Baker and R. J. Strutt, ibid., 801, 1049 ; A., ii, 357, 457.27 H. B. Baker, E.Tiede, R. J. Strutt, and E. Domcke, ibid., 2283; A . , ii, 724.28 E. Tiede and E. Domcke, ibid., 2284 ; A., ii, 724.M. Le Blsnc, Ber. K. Sachs. @:as. Wiss., Math.-phys. KI., 1914, 66, 38 ;A., ii, 809.REP.-VOL. XI. 50 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Group I .I n last year’s Report reference was made to the preparation ofpure anhydrous sodium monosulphide by the action of heat on thehydrosulphide. The action of sulphur on the hydrosulphide hasnow been investigated with the view of studying the polysulphidesof sodium.30 Approximately saturated solutions in alcohol of thehydrosulphide were heated on the water-bath with equivalent pro-portions of sulphur, and the product was either precipitated aftercooling by the addition of ether or recovered by evaporationof the solution.Amounts of sulphur corresponding with thedi-, tri-, tetra-, and penta-sulphides, as well as a large excess ofsulphur, were used. The results are interesting, because the onlypure solid compound obtainable by the above method was the tetra-sulphide, Na,S,. With smaller quantities of sulphur, mixtures ofthe tetrasulphide and hydrosulphide were formed, and with greaterquantities of sulphur the tetrasulphide, mixed with free sulphur,was obtained. On the other hand, there was definite evidence t oshow that higher polysulphides exist in the solution, and someindication was obtained that when a great excess of sulphur wasused, small quantities of higher polysulphides were mixed with thetetrasulphide when the latter was separated in the solid phase.Thesolution in such cases would seem to contain equilibrium mixturesof the tetrasulphide and higher polysulphides, and confirmatoryevidence of this was obtained by the determination of the hydrogensulphide set free by the mutual action of weighed amounts ofsulphur and sodium hydrosulphide in solution. The latter experi-ments certainly prove that the first action of the sulphur is to givethe tetrasulphide, and if insufficient sulphur is used to convert thewhole of the sodium hydrosulphide into the tetrasulphide, mixturesof the hydrosulphide and the tetrasulphide are obtained.By the action of metallic sodium on a solution of the tetra-sulphide in alcohol, the disulphide has also been prepared in a purestate. Sodium tetrasulphide, Na2S4, is a dark yellow, crystallinepowder with an olive-green tinge.It is extremely hygroscopic, anddissolves in water to give a deep orange solution, which turns redon heating. The disulphide, Na,S,, is a bright yellow, micro-crystalline powder, readily soluble in water t o a deep yellowsolution.By heating known quantities of water and finely pulverisedglasses of known constitution to high temperatures in a gold cruciblecontained in a special bomb, certain new crystalline alkali silicateshave been prepared.31 The temperature was measured by means30 A. Rule and J. S. Thomas, T., 1914, 105, 1’77.31 G. W. Morey, J. Amer. Chern. SOC., 1914, 36, 215 ; A , , ii, 202INORGANIC CHEMlSTRY, 51of a thwmo-couple, and was accurate to &5O.The following com-pounds were obtained : potassium hydrogen disilicate, KHSi20,,orthorhombic crystals which do not lose water at 350O. Thecrystals are very little affect'ed by water, and may be boiled forseveral hours a t looo without any appreciable decomposition.Potassium disilicate, K,Si,O,, a compound which is hygroscopic andvery readily acted on by water. Sodium disilicate, Na2Si20,, whichresembles KHSi20, in optical properties; the crystals may beleached with water, but on prolonged contact they are decomposed.Sodium metasilicate, Na2Si0,, forms crystals which are very readilydecomposed by water. This compound was previously known onlyin the form of somewhat indefinite hydrates.Some interesting work on nitrites may be referred t o in view ofthe bearing it has on the fixation of nitrogen.32 This reference mayconveniently be included here, in spite of the fact that much of i trefers to the next group.The nitrites of the alkali and alkalineearth metals can be prepared in the pure state only by doubledecomposition of silver nitrite with the chlorides of these metals,and in this way two new nitrites, LiNO,,H,O and Ca(N0,)2,4H,0,have been obtained. The hydrated nitrites generally can be de-hydrated in a vacuum over phosphoric oxide without decomposition.The solid nitrites and their saturated solutions are not oxidised byoxygen a t atmospheric pressure. Oxidation takes place only in thepresence of acid, when the free nitrous acid plays an importantpart in the reaction.The following double salts also wereobtained :N+&z(NOZ)OH~O, K,Ag,(NO2)4,H20, and BaAg2(NO,),,H20.I n the action of heat on sodium nitrite, the following reactionstake place in addition to the ordinary reactions:NaNO, + NO, = NaNO, + NO, ZNaNO, + NO2= 2NaNO,+ N.This explains the actions which occur when nitrogen peroxide actson calcium oxide and on the alkali and alkaline earth carbonates.When nitrogen peroxide acts on calcium oxide, there is always aloss of nitrogen, whatever be the conditions of temperature, inaccordance with the equation 2Ca0 + 5N02 = ZCa(NO,), + N, a factthat is of industrial importance.An investigation has been made of the composition of the solu-tions in equilibrium with the four salts potassium chloride, sodiumnitrate, sodium chloride, and potassium nitrate, which affords usefulinformation as to the most economical production of " conversion "saltpetre.33 I f a solution containing sodium nitrate, potassiumchloride, and water in the molar ratio 0'80: 0.62: 1.81 is evapor-32 M.Oswald, Ann. Chim., 1914, [ix], I, 3 2 ; A . , ii, 197.33 W. Reinders, Proc. K. Aknd. Wetensch. Amslerdam, 1914, I?, 1065; A., ii, 549.E 52 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.ated a t looo, 0.42 mol. of sodium chloride is deposited. On cool-ing to 5O, the salts separate out in the proportion 0.575 mol. ofpotassium nitrate: 0.038 mol. of sodium chloride : 0.10 mol. ofsodium nitrate. If, however, a suitable amount of water is added,the separation of solid sodium chloride and sodium nitrate can beavoided, and 0.563 mol.of potassium nitrate obtained in crystals,which represents a yield of 90.8 per cent.Group IZ.Metallic strontium has been obtained by the electrolysis of thebinary mixture, strontium chloride-potassium chloride.84 A eutecticpoint occurs with the mixture containing 15 per cent. of potassiumchloride, and when the same method is used as in the case ofcalcium, sticks of strontium 10 cm. long and 1-2 cm. in diametercan be obtained. The mixture of chlorides melts a t 628O, which is220° below the melting point of strontium chloride. The currentdensity must be 20-50 amperes per sq. cm. of cathode, and theefficiency is then 80 per cent. Metallic barium has also beenobtained in a similar way.The investigation of the phase systems of calcium nitrate hasbeen extended to a study of the three-component system, calciumnitrate-lime-water.35 The nature of the solid phases capable ofexisting in equilibrium with aqueous solutions of calcium nitratecontaining lime was investigated a t 2 5 O and looo.Only onedefinite basic nitrate, Ca,N,O,, was found, which forms severalhydrates, containing 8, 1, 2, 3, and 4 molecules of water re-spectively. No series of solid solutions, Ca0,zN,0,,yH20, werefound such as postulated by Cameron and Robinson,36 nor does thehydrate, Ca,N,0,,3&H20, exist, as stated by these authors.Some interesting work on calcium sulphate may be mentioned inconnexion with its dehydration and commercial use as plaster ofParis.87 When gypsum is heated a t temperatures not above 210°,it becomes anhydrous, but the product is much less dense and moresoluble than natural anhydrite.The conversion of the hemihydratet o the soluble anhydrite, which is the last stage in the dehydrationprocess, is reversible, each process only requiring a few minutes at1100 in a dry and a humid atmosphere respectively. Since the hydra-tion to the hemihydrate takes place so easily, this salt naturally34 B. Neumann and E Rergve, Zeitsch. E'lcklrochem., 1914, 20, 187 ; A . , ii, 365.35 H. Rassett, jun., and H. S. Taylor, T., 1914, 105, 1926.36 F. K. Cameron and W. 0. Robinson, J . Physical Chenz., 1907, 11, 273;G. Gallo, Gazzetta, 1914, 44, 1, 497 ; A ., ii, 561 ; C. Gaudefroy, Compt. rend.,A,, 1907, ii, 444.1914, 158, 2006 ; 159, 263 ; A., ii, 650, 728INORGANIC CHEMISTRY. 53forms the principal constituent of plaster of Paris. The besttemperature for dehydration f o r commercial purposes depends uponthe humidity of the air in the oven, thus explaining the, variableresults that have been obtained. The hemihydrate dso absorbswater from the air a t the ordinary temperature up to a total water-content of about 8 per cent. Microscopic and calorimetric observa-tions confirm Le Chatelier's theory of the mechanism of the settingof plaster of Paris.Group IZZ.In the Reports for 1912 and 1913 reference was made to theisolation of the three hydrides of boron, B,H,, B4H10, and B,HI2.The first two compounds dissolve in alkali hydroxide, giving solu-tions which do not smell of the hydrides, but are unstable, and tendto give up hydrogen3* Since the lattar reaction takes place leasteasily with B4H10, this hydride was used in an investigation of theproducts formed when it is passed into aqueous alkali hydroxides.The first action takes place between one molecule of the hydride andfour molecules of a monacid and two molecules of a diacid base,according to the equationB4H10 + 4KOH = 4KOBH3 + H2.By treatment of potassium hydroxide dissolved in one and a-halftimes its weight in water with excess of B4H10 a t Oo, the solid hypo-borate was obtained in the form of colourless, octahedral crystals.The formula KOBH, was proved by analysis.This substance isstable in the absence of moisture; it is deliquescent, and its solu-tions decompose slowly at room temperature in accordance withthe equation2KOBH3 + 2H20 = ZKBO, + 5Hz.This reaction is brought about immediately by the addition of acid.The aqueous solution of the hypoborate is a very po,werful reducingagent.With copper salts it gives a precipitate of copper hydride,and with nickel salts a black, insoluble nickel boride, NiB,. Thiscompound is non-magnetic, but on heating to 500° it becomes mag-netic, a t the same time sintering to a grey, metallic form.When the compound KOBH, is heated at 500°, metallic potass-ium, hydrogen, and water are expelled, and the reaction is probablyexpressed by the equation5KOBH3= K3B503 + 2K + 2H2O + 11H.The residue, K3B503, is somewhat hygroscopic, and dissolves inwater to an alkaline solution, which has only a weakly reducingaction on potassium permanganate. It gives a characteristic yellowt o brownish-red colour when warmed with nitric acid, Whens8 A.Stock and E. KUSS, Ber., 1914, 47, 810; A., ii, 35954 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.treated with sulphuric acid, the compound gives hydrogen and boronh y dr ides.Sodium hypoborate, NaOBH,, was also prepared, and found topossess properties similar t o those of the potassium salt. No in-soluble hypoborates could be obtained by mixing a solution of thepotassium or sodium salt with those of other metallic salts.The reactions described in this paper are exactly analogous t othose described by Travers and Ray.39Two papers have appeared dealing with work on scandium."*41.Many ney salts of this element are described, and also the extrac-tion and purification of the oxide from wolframite.It is hardlynecessary to specify the new salts that have been prepared, but apoint of some interest is raised in the fractional separation of thescandium from thorium, yttrium, and ytterbium. An examinatioiiof 300 fractions showed that scandium is a uniform substance.The rose-yellow precipitate in the second analytical group was in-vestigated, and i t was found that; the claim previously made, thatthis is due to a new element, is not justified.42~43 This sulphide wasobtained in large quantities, and was proved to consist of coppersulphide with small quantities of tin and tungsten sulphides,together with much sulphur.Aluminium nitride 44 has been obtained in colourless, hexagonalneedles by heating the crude compound a t 2010-2030° in a tubeheated in a carbon resistance furnace in the presence of excess ofnitrogen.The cornpound dissociates when heated a t atmosphericpressure at 1850O.Mention was made last year of a paper in which doubt wasthrown on the existence of the aluminates as definite chemical com-pounds. This was controverted in a latea paper. Two furtherpapers453 46 have since appeared by the same authors, and thO ques-tion as to the existence of aluminates seems t o have been decidedin a third paper.*' Conductivity measurements show that suchcompounds must exist, for the measurements are definitely againstthe assumption of colloidal aluminium hydroxide. Moreover, whenii, 938.Zeitsch.anorg. Chevn., 1914, 86, 257 ; A , , ii, 269.39 M. W. Travers and R. C. Ray, Ptoc. -Roy. Soc., 1912, [ A ] , 87, 163 ; A., 1912,40 R. J. Meyer and, in part, A. Wassjuchnov, N. Drapier, and E. Bodlander,41 J. Sterba-Bohm, Zeitsch. Elektrochem., 1914, 20, 289 ; A . , ii, 565.42 M. Ogawa, J. CoU. 8ci. T6ky6, 1908, 25, xv, 1 ; xvi, 1 ; A . , 1908, ii, 952,45 A. Skrabal and P. Artmaon, Chcm. Zeit., 1909, 33, 143 ; A . , 1909, ii, 243.(4 J. Wolf, Zeitsch. anorg. Cheyn., 1914, 87, 120; A . , ii, 568.45 E. G . Mahin, J. Anzer. Chern. Soc., 1914, 36, 2381 ; A . , ii, 850.J6 W. Blum, ibid., 2883 ; A . , ii, 850.47 R.E. Slade and W. G. Polack, Trans, Faraday Soc., 1914, 10, 150 ; A , , ii, 811.953INORGANIC CHEMISTRY. 55hydrolysis takes place, the aluminium hydroxide is deposited in thecrystalline form. The arguments, therefore, are strongly in favourof the existence of aluminates as true chemical compounds.Group ZV.In the Report for 1912, the preparation of carbon subsulphide,C,S,, was referred to, this compound being produced by maintain-ing an arc bet'ween a graphite cathode and an anode of graphiteand antimony under carbon disulphide. By somewhat similarmethods, carbon sulphidotelluride, CSTe,48 and carbon sulphido-selenide, CSSe,49 have been obtained. I n the case of the former,the graphite and antimony electrode is replaced by one containing10 or more parts of graphite to 100 parts of tellurium.The carbondisulphide is then found to contain non-volatile decompositlion pro-ducts of carbon disulphide, and the compounds C,S, and CSTe,which are volatile with the carbon disulphide vapour. The separa-tion of these two compounds proved to be a very difficult matter,owing to the extreme ease with which the sulphidotelluride decom-poses. It was effected by a combination of two methods, namely:(1) the repeated fractional extraction of the C,S, from the solutionby carbon disulphide vapour, and (2) the conversion of the C,S,into thiomalononaphthylamide by its reaction with 8-naphthyl-amine. The dilute solution of the sulphidoklluride, after dryingwith phosphoric oxide, was concentrated on the water-bath to astrength of 5-10 per cent., using a Hahn fractionating column.The solution was fractionally distilled a t -30° under verydiminished pressure in a specially designed apparatus,50 and thepure sulphidotelluride was obtained in yellowish-red crystals, meltring a t - 5 4 O to a brilliant red liquid of high refractive power.Molecular-weight determinations with benzene and carbon di-sulphide as solvents gave values between 176 and 181, theory 172.The compound decomposes very rapidly a t room temperature, andis exceedingly sensitive to light, decomposition taking place even a t-5OO.No evidence could be obtained of the existence of carbonditelluride, CTe,.The sulphidoselenide, CSSe, was obtained in a similar way, theanode consisting of 17.5 parts graphite to 100 parts selenium. Thecompound was more readily isolated, owing to its greater stability.At room temperature the sulphidoselenide forms an intense yellowliquid, stable in air, and possessing a pungent odour of onions.Itdoes not take fire when brought into contact with the flame,4x A. Stock and P. Praetorins, Ber., 1914, 47, 131 ; A., ii, 199.49 A. Stock and E. Willfroth, ibid., 144 ; A., ii, 200,A. Stock, ibid., 1 5 4 ; A., ii, 17156 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.although the vapour of the boiling liquid burns with a beautifullilac flame. It melts at - 8 5 O , and boils a t 8 4 O , the vapour pressurea t loo being 45 mm. It is decomposed by light and on long keep-ing a t room temperature, but is much more stable than the sulphido-telluride.No evidence was obtained of direct union betveencarbon aria selenium by passing selenium vapour over carbon a t1000°, neither was i t found possible in any way to prepare carbondiselenide, CS%.Some doubt has been thrown on the constitution of white lead,51the view being expressed that it is not a true basic carbonate ofthe formula 2PbCO,,Pb(OH),. I n one paper a description is givenof a process by which white lead is obtained by spraying basic leadacetate solution a number of times through an atmosphere ofcarbon dioxide. The solid which Beparates a t first has the constitu-tion PbCO,,Pb(OH), ; on continued action of the carbon dioxide,this solid absorbs the gas, passing through the stage2PbC0,,Pb(OH)2,and finally becomes PbCO,.The increase in the percentage ofcarbon dioxide in the solid is gradual, and there is no evidence ofany break in the continuity at the stage 2PbCO,,Pb(OH),. Leadhydroxide is soluble in an aqueous solution of sucrose, but inno stage in the above process is there any loss in weight when thesolid is treated with sucrose solution. This dismisses the possibilitythat the solid phase is a mixture of lead carbonate and leadhydroxide. Further, a mixture of the compound PbCO,,Pb( OH),and lead carbonate in the correct proportions exhibits the sameproperties, both physically and as a pigment, as does ordinary whitelead.Further evidence is adduced in the second paper, for it is shownthat lead carbonate, when agitated with a solution of basic leadacetate, withdraws lead hydroxide. Again, in the case of the basiccarbonate, PbCO,,Pb( OH),, the percentage of carbon dioxide canbe increased or decreased a t will by agitation with neutral or basiclead acetate solution.Analogous results were obtained with basic zinc carbonate, pre-cipitated calcium carbonate, precipitated barium sulphate, andprecipitated barium carbonate, for these compounds on agitationwith basic lead acetate solution gave ZnC03,Zn(OH),,3Pb(O13)2,2CaC03,Pb(OH),, 3BaSO,,Pb(OH),, and 3BaC03,2Pb(OH), re-spectively.The barium carbonate and zinc carbonate compoundsequal white lead in their properties as pigments. Finally, the factthat white lead loses lead hydroxide when treated with ammoniumchloride solution shows that i t is more loosely combined than would51 E. Euston, J.Ind. Eng. Chem., 1914, 6, 202, 382; A., ii, 366, 465INORGANIC CHEMISTRY. 57be expected of a true basic salt. The view is put forward thatwhite lead is an adsorption compound of lead carbonate and leadhydroxide.An interesting method of decomposing natural phosphates andsilicates has been described.52 The powdered mineral is heated incarbonyl chloride vapour, the anhydrous metallic chloride beingformed in each case. The natural phosphates, vivianite, pyro-morphite, uranite, and monazite, are attacked at temperaturesbetween 300° and 500°, whilst the silicates, such as tho'rite,gadolinite, cerite, and zircon require temperatures of above 1000°.Emerald, however, is not decomposed a t 1400O.Group V .An investigation of the dissociation of gaseous nitrogen trioxide 53shows that the equilibrium conditions are somewhat complex, andnot to be represented by the two equationsThe pressure-volume relation has been determined for varioustemperatures and volumes of the gas, and h n alternative sets ofequations are considered in reference to the data thus obtained.The final conclusion is drawn that liquid nitrogen trioxide a t lowtemperatures consists mainly of N40, molecules, together with someN204, NO,, and NO molecules, according to the extent of the dry-ing.The vapour consists of a mixture of N406, N203, NO,, andNO molecules, reacting according to the equationN,O, zz N,O, + NO, + NO,mixed with some wet N204, NO,, and NO molecules.The effectof the prolonged drying of the liquid appears t o be to enable theNO, and NO molecules t o combine to give N406, which in the drystate dissociates to N,O,, NO,, and NO in equal volumes, the N203not further dissociating. It further seems likely that N406 mole-cules in the liquid state are blue, and that in the gaseous stateboth N406 and N203 are colourless, or nearly so. Ordinary wetnitrogen trioxide is green, owing to the mixture of blue N406 witha relatively large amount of wet NO,. The green colour is due tothe yellow molecules of NO, mixed with the blue N,06. At verylow temperatures all specimens of the liquid become quite blue,which is no doubt due to the fact that a t these temperatures theyellow NO, molecules completely associate to give the colourlessN204, which would no longer produce the green colour with theblue N406.Some measurements have been published of the vapour pressureN406 Z N204 + 2N0 and N,O, 2N0,.52 J.Barlot and E. Chauvenet, Compt. rc?Lcl,, 1913, 157, 1153 ; A . , ii, 49.ja B. M. Jones, T., 1914,105, 231058 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.of 60lid nitrogen peroxide between -looo and -34O.W Thedynamical method was employed, the amount of vapour diffusinginto a known volume of hydrogen being determined. The vapourpressures lie between 0,0023 mm. a t -looo and 39.24 mm. a t-30°, and they are expressed by the curveMeasurements were made between -40° and -60° by a staticalmethod, and these are in good agreement.When a solution of ammonium phosphate (1.6 grams in 25 C.C.of water) is added to a solution of 1 gram of ferrous chloride in20 C.C.of alcohol saturated with nitric oxide a t Oo, the whole opera-tion being carried out in an atmosphere of nitric oxide, a blackish-brown, viscid oil is precipitated, which crystallises when cooled ina freezing mixture.55 The substance was obtained pure in the formof brown, flaky crystals, and has the formula Fe(NO)HPO,.Freshly precipitated ferrous phosphate also absorbs nitric oxide togive the same compound. Copper sulphate, chloride, and bromidecombine with nitric oxide according t o the reversible actionCUR, + NO Z CuR,NO.The combination takes place in alcoholic solution, the presence ofwater greatly diminishing the combination. Migration experimentsshowed that the brown, ferrous nitric oxide compounds containedthe complex cation (FeNO)", whereas in the green compounds acomplex ferrous anion exists.The combination with nitric oxideseems to be characteristic of the normal ferrous and cupric com-pounds, and not of those in which the metal is present in a stable,complex ion.A study of the phase diagram of the system ammonia and waterhas revealed the existence of two compounds, 2NH3,H,0 andNH,,%O, which melt a t - 78'9O and - 79*0° respectively.56 Theeutectic point a t which ammonia and the compound 2NH,,H,Oco-exist as solid phases lies a t 81-4 mols. per cent. of ammonia and-92'5O. The point a t which 2NH,,H,O and NH,,H,O co-exist liesat 58.5 per cent.of ammonia and -86O, and that correspondingwith the co-existence of NH,,H,O and ice a t 34.7 per cent. ofammonia and - 100.3O.Solutions of alkali or other chlorates, bromates or iodates, and ofhydrazine salts, may be mixed, and even boiled, without any reaction taking place.57 I f , however, a piece of tarnished copper wirelog p=14.9166 + e(o.0604).s4 A. C. G. Egerton, T., 1914, 105, 647.55 W. Manchot, Ber., 1914, 47, 1601 ; A . , ii, 567.56 A. Smits and S. Postma, Proc. K. Akad. TVetemch. Amsterdam, 1914, 17,57 W. R. E. Hodgkinson, J. SOC. Chem. I&., 1914, 13, 815 ; A., ii, 771.182 ; A., ii, 809INORGANIC CHEMISTRY. 59or a fragment of cupric oxide is introduced into a cold solution ofpotassium chlorate and hydrazine nitrate, nitrogen is evolved. Thereaction is accelerated by warming, and is quantitative in accord-ance with the equation2KC10, + 3N2H,,HN03= 6H20 + 3N2 + 3HN03 + 2KC1.The method may be used for the estimation of chlorates, bromates,and iodates.A slight exces of hydrazine salt is used, and theexcess is destroyed by potassium permanganate and nitric acid, whenthe resulting chloride, bromide, or iodide can be estimated.A considerable amount of work has been done on the nitrogenderivatives of sulphurous and sulphuric acids. By the action ofsulphur dioxide on ammonia in ethereal solution, amidosulphinicacid, NH,*SO,H, or its ammonium salt, NH2*S02N€I,, is formed,according as t o whether the sulphur dioxide or ammonia is inexcess.58 The free acid is a pale yellow compound, whilst the am-monium salt is a white substance.The compound N4H12S5010,previously dwribed by Divers and Ogawa, has been found not t oexist ; the substance originally prepared was impure ammoniumtrithionate.69Hydroxylamineisomonosulphonic acid, NH2*O*S02-O€I, was firstindicated by Raschig in the product formed by hydrolysis ofhydroxylamineisodisulphonic acid by hydrogen chloride.60 Thiscompound has now been obtained by the action of chlorosulphonicacid on hydroxylamine hydrochloride a t room temperature.61 Itseparates from a mixture of ether and methyl alcohol as a micro-crystalline powder; it liberates iodine from potassium iodide, andis hydrolysed in acid solution to hydroxylamine.Hydrazine hydrazinesulphonate has been obtained by aspiratingair first through a flask containing fuming sulphuric acid of highanhydride content, and then through a flask containing anhydroushydrazine.62 Considerable quantities of the salt are formed inaccordance with the equation 2N2H4 + SO, = H,N=NH=S0,H,N213,.The product is dissolved in water and heated with excess of bariumhydroxide in order to expel the unchanged hydrazine, the excessof barium is precipitated with carbon dioxide, and the filtrate fromthe barium carbonate is evaporated in a vacuum.A residue ofbarium hydrazinesulphonate, (N2H3*SO3),Ba,2H2O, is left, whichmay be obtained in glistening needles by solution in water andprecipitation with alcohol. The calcium salt, (N,H3*S03),Ca,H20,58 M.Ogawn. and S . Aoyama, Sci. Beports T6Roki~ Imp. Univ., 1913, 2, 121 ;A . , ii, 264.E. Divers and M. Ogawa, T., 1901, 79, 1102.F. Sommer and H. G. Templin, Ber., 1914, 47, 1221 ; A., ii, 458.6o F. Raschig, Annulen, 1887, 241, 161.62 W. Traube and 0. Vockerodt, ibid., 938 ; A., ii, 35860 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.and the strontium salt, (N,H,*S0,),Sr,2H20, can be obtained in asimilar way.By double decomposition, using the barium salt and ammoniumcarbonate, ammonium hydrazinesulphonate, N,H,*SO,NH,, wasobtained as a deliquescent, crystalline mass. This salt is isomericwith hydrazine amidosulphonate.63.N,H,*SO,E,and the sodium salt, N,H3*SO3Na,H2O, were prepared in a similarway from potassium and sodium sulphates. The former crystallisesin needles and the latter in monoclinic tablets. All these salts arestable in alkaline and in neutral solution, but are decomposed byacid into hydrazine and sulphuric acid.From the barium saltand silver nitrate, the silver salt, N,H,*SO,Ag, was obtained inbeautiful listening needles. From the barium salt, by the actionof sulphunc acid, the free acid, N,H,*SO,H, was obtained inglistening needles, which melt and decompose a t 217O.I f a strongly cooled, aqueous solution of potassium nitrite istreated with finely powdered hydrazinesulphonic acid, the follow-ing reaction takes place:NH,*NH*SO,H + KNO, = 2H,O + N3S03K.From the resulting solution the salt, potassium azoimidesulphonate,has been obtained.It crystallises in long, flat prisms, which ex-plode on heating. The corresponding sodium, ammonium, andbarium salts have been obtained in a similar way.The potassium salt,' gGroup V I .Pure perchromic acid has been prepared by the interaction ofchromium trioxide and 97 per cent. hydrogen peroxide in methylether solution a t -3OO.64 The reaction takes place according tothe equationZCrO, + 7H,o,= 2H,CrO, + 4H,O.The blue solution was poured off the excess of eit,her chromiumtrioxide or hydrogen peroxide, dried over phosphoric oxide, andthen evaporated in a vacuum at -30°, when the acid was obtainedas a dark blue, crystalline mass, which decomposes a few degreesabove -3OO. Analysis shows that it has the formula H3Cr0,,2H,0,the water being water of constitution. The red perchromates areconsidered to be anhydro-salts of the blue perchromic acid, theirconstitution being E>Cr(O=OM),, whilst that of the acid is(HO),Cr(O*OH),.It is found that by the action of sulphuric acid on calcium63 A.P. SabanBev, J. Buss. Chem. Soc., 1899, 31, 375 ; A., 1900, ii, 13.64 E. H. Riesenfelcl and W. Mau, Ber., 1914, 47, 548 ; A , , ii, 279INORGANIC CHEMISTRY. 61fluoride it is impossible to obtain hydrogen fluoride of a greaterstrength than 95-96 per cent.6560. The best yield is obtained with90 per cent. sulphuric acid. When pure sulphuric acid is used,fluorosulphonic acid may be produced, and a quantitative yield ofthis acid may readily be obtained by heating calcium fluoride withfuming sulphuric acid containing 60 per cent.of the anhydride.The reaction takes place according t o the equationCaF, + H,SO, + 2S0, = CaSO, + 2F*SO,-OH.Sodium fluorosulphonate may readily be obtained by heating theacid with sodium chloride in a platinum retort under reflux. Boil-ing alcohol extracts the sodium fluorosulphonate, and, on cooling,this salt separates in iridescent plates or needles.Fluorosulphonic acid is decomposed on boiling with sulphuraccording to the equation2F*S02*OH + S = 3 S 0 , + 2HF.Group VZZ.A thermal analysis of mixtures of bromine and water leads t othe conclusion that the composition of the hydrate of bromineis Br2,8H,0.67 The crystals were centrifuged on kaolin a t 0" inorder to free them from any adhering bromine or water, and werethen analysed. The analytical results confirm the correctness ofthe above formula.An interesting study has been made of the relative stability ofpure calcium hypochlorite, bleaching powder, and a mixture ofcalcium hypochlorite and calcium chloride.68 Towards dry air, freefrom casbon dioxide, calcium hypochlorite and bleaching powdershow very little difference in behaviour a t 90°, there being only aslight loss of chlorine in each case.I n moist air containing carbondioxide a t room temperature, bleaching powder loses more of itsavailable chlorine than does the hypochlorite. The mixture ofhypochlorite and chloride behaves similarly to bleaching powder.I n an atmosphere of dry carbon dioxide, calcium hypochlorite losesa small percentage of its available chlorine after five hours' ex-posure; the mixture of hypochlorite and chloride loses a slightlygreater percentage, whilst bleaching powder loses all its availablechlorine. I n moist carbon dioxide, both the hypochlorite andbleaching powder lose all their available chlorine; the former, how-ever, gives hypochlorous acid as well as chlorine, whilst the latteronly evolves chlorine. With ammonia, both calcium hypochloriteand bleaching powder give an almost quantitative yield of nitrogen.0. Ruff and H. J. Braun, Ber., 1914, 47, 646; A, ii, 263.H. Giran, Compt. rend., 1914, 159, 246 ; A., ii, 723.66 0. Ruff, ibid., 656; A . , ii, 263.68 K. A. Hofmann and K. Ritter, Ber., 1914,47, 2233 ; A . , ii, 61262 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Group VIZl.Some attention has been paid to the question of the sorption ofhydrogen by 'O, 7 1 ~ 72. The curves obtained by plottingthe rate of solution as a function of the quantity of gas occludedconsists of two portions, indicating the occurrence of two distinctprocesses.70 The suggestion is made that there are two modifica-tions of palladium, which differ very considerably in respect oftheir activity towards hydrogen. The first portion of the velocity-concentration curve, on this view, is mainly determined by theactive form, whilst the second portion refers to the sorption of thehydrogen by th0 less active form. When palladium-black is used,a smooth curve is obtained, and the suggestion follows that thismaterial consists almost entirely of the active form. A somewhatsimilar view is put forward in the idea that palladium consists ofa crystalline form and an amorphous form, the latter being themore active.71Analogous results were obtained in a series of measurements ofthe electrical conductivity and the density of palladium wires con-taining measured amounts of occluded hydrogen.72Some experiments have been carried out on the nature of pre-cipitated nickel sulphide.73 It appears that there are threedifferent modifications, which differ in respect of their solubility inacids. The precipitates from a neutral solution of nickel chlorideand soluble sulphides consist mainly of a modification which dis-solves readily in acids. The less soluble modification is formed byboiling with water, or by long contact with a cold solution con-taining nickel, or with cold dilute acetic acid. All the modifica-tions have the same composition, NiS, and the most soluble is themost readily oxidised by air.a-Nickel sulphide is soluble in mineral acids down to O-O~LV,whilst P-nickel sulphide is fairly rapidly dissolved by 2N-hydro-chloric acid, and the y-modification is not appreciably dissolvedexcept on addition of oxidising agents. The differences in solu-bility are too great to be explained by colloidal conditions, andmust be due to polymerisation.E. C. C. BALY.69 F. Halla, Zcitrch. physikal. Chcm., 1914, 86, 496 ; A., ii, 178.7O A. Holt, Prx. Roy. Xoc., 1914, [ A ] . 90, 226; A., ii, 452.71 A. Sieverts, Zeitsch. physikal. Chem., 1914, 88, 103; A . , ii, 626.72 G. Wolf, ibid., 1914, 87, 575 ; A., ii, 517.A. Thiel and H. Gessner, Zeitsch. anorg. Chem., 1914, 86, 1 ; A., ii, 27