INORGANIC CHEYISTRT. Inorganic Chemistry. ii. 97 The Boiling Point of Aqueouls Solutions of Nitric Acid at Different Pressures. 11. The Influence of Water-retaining Agents on the Gomposition of the Mixture of Maximum Boiling Point. HENRY JERMAIN MAUDE CREIGHTON and HERSCHEL GASTON SMITH ( J . Franklin Inst. 1915 180 703-709).-1n con- tinuation of earlier experiments (A. 1915 ii 446) which showed that the composition of the maximum boiling mixture changes only from 66.80% of nitric acid a t 110 mm. t o 68.18% a t 760 mm. the effect of the presence of potassium hydrogen sulphate and of sulphuric acid has been examined. The composition of the mixture of maximum boiling point is not appreciably altered by the addi- tion of 10% or 20% of potassium hydrogen sulphate. On the other hand sulphuric acid causes a decrease in the nitric acid content of the maximum boiling mixture the decrease being greater the greater the addition of sulphuric acid.A t 760 mm. the per- centage of nitric acid in this mixture is reduced from 68.18 t o 64.5 in presence of 10% of sulphuric acid and to 59.2 in presence of 20%. A t lower pressures the reduction in the nitric acid content is still larger. The results obtained afford an explanation of the fact that solu- tions of nitric acid can be concentrated t o more than 90% by dis- tillation with concentrated sulphuric acid. H. M. D. Two Reactions of Azoimide with Analytical Importance. FRITZ SOMMER and HEINRICH PINCAS (Ber. 1915 48 1963-1969 2096).-A. Oxidation of Azoimide b y Ceric Salts.-A method for the estimation of azoimide in neutral or acetic acid solutions by measuring the volume of the latent nitrogen was first proposed by Raschig 111 1908 who found t h a t when such a solution is mixed with a slight.excess of iodine and a crystal of sodium thiosulphate the evolution of nitrogen is brisk and complete (compare also following abstract). The authors find that although this method is trustworthy for 0'1N-solutions and therefore for solids it is not so for very dilute solutions the end of the reaction being reached too slowly. On the other hand salts of quadrivalent cerium such as ceric ammonium nitrate ceric sulphate o r the ceric ammonium sulphates immediately effect the complete oxidation of azoimide in neutral o r acid solutions eveln when these are very dilute according to the equation 2N3H + 2Ce0 = 3N + Ce203 + H,O.For 0.1 gram of the sodium salt about 2.3 grams of the cerium com- pound should be used. Free hydrochloric acid or excess of chlorine ions must be avoided however otherwise chlorine gas will be 1 i b era t ed B. Reaction of Azoimide with Nitrous Acid.-According t o Thiele (A. 1908 ii 940) azoimide reacts with nitrous acid thus N3H + HNO = N + N,O + H,O. The correctness of this assumption VOL. CX. ii. 4ii. 98 ARSTRACTS OF CHEMICAL PAPERS. has been proved by adding acetic acid t o a mixture of equivalent quantities of barium nitrite and sodium azide and analysing the gases evolved. The reaction proceeds very smoothly and may be used to estimate simple nitrites. A known excess of sodium azide is added to an acidified solution of the nitrite then the mixture is shaken for a minute or two rendered just alkaline by barium hydroxide and boiled to expel nitrous oxide and finally acidified with acetic acid when the excess of azoimide is estimated as above. Another application of the reaction is the detection of nitric acid in the presence of nitrous.Nitrous acid is completely destsoyed even in a O.OOOlS-solution whilst nitric acid is un- affected when a slight excess of sodium azide is added the solu- tion is acidified and boiled. This method for removing nitrous acid is much better than the old treatment with carbamide or the newer reaction with hydrazine sulphate (Sen and Dey A. 1912 ii 296). J. C. W. Oxidation of Azoimide by Iodine. F. RASCHIG (Bey. 1915 48 2088-2092.Compare Chenz. Zeif. 1908 32 1203).-The oxidation of azoimide by iodine in the presence of sodium thio- sulphate might be regarded as a catalytic effect of sodium tetra- thionate as Sommer and Pincas have done (preceding abstract). This is not justifiable however for just as iodine alone causes no evolution of nitrogen so a mixture of sodium tetrathionate with iodine is also without action. If the reaction between iodine and sodium thiosulphate is con- sidered since no more than two molecules can react a t any given instant the first stage in the reaction will be represented thus Na,S,O + I = NaI + NaIS203. It is this intermediate compound which so easily gives up iodine to' the sodium azide (for example) or to another molecule of sodium thiosulphate thus Na,S,O,+ NaIS,O = Na,S,O + NaI. The actual catalyst is therefore the residue NaS,O which takes up iodine if that is in excess and gives i t again to the azide or under normal conditions condenses to form Na,S,O,.Free mineral acids hinder the reaction and since sulphuric acid is liberated as a by-product it' is advisable t o add sodium acetate. Because of this effect of acids it is always advisable too to keep adding sodium thiosulphate t o the solution and this is best effected by using a crystal about the size of a pea which takes sufficiently long to dissolve. Azoimide seems t o be the only substance which whilst un- attacked by iodine alone is oxidised by iodine in the presence of sodium thiosulphate. Bromine water and thiosulphate have the same effect owing no doubt t o the formation of NaBrS203. The only other iodine carrier which has been discovered among sulphur compounds is a sulphide.A crystal of sodium sulphide has the same effect but the liberation of sulphur which also takes place renders the reaction less neat. J. C. W. The Neutralisation Curve of Boric Acid. F,. B. R. PRIDEAUX (TI*nns. Fcrmdcty Soc. 1915 11 76-78).-A discrepancy is foundINORGANIC CHEMISTRY. ii. 99 between the experimental neutralisation curve and that calculated from the single constant of a monobasic acid. This is attributed to the formation of complexes in the borate solwtions the ion H,BO,’ being tlius,weighted by the complexes. A revised constant k =4.5 x 10-10 is given from which the alkalinity of borate solu- tions may be calculated up t o st,rongly alkaline solutions. C.H. D. The Dissociation of Carbongl Sulphide. GILBERT N. LEWIS ( J . Amer. Chem. SOC. 1915 37 2786).-A correction of an error in sign in the receqt paper by Lewis and Lacey (A 1915 ii 767). The modification thus introduced into the van’t Hoff equation brings the calculated heat of formation of carbonyl sulphide from liquid sulphur and carbon monoxide to 11,000 cal. in good agree- nient with the value of Thomsen. D. F. T. The Supposed Formation of Persilicate by the Action of Air on Solutions of Sodium Silicate. HUGO DITZ ( J . 211’. Glmn. 1916 [ii] 92 412-418).-The author regards the state- ment that the solution of a silicate of an alkali metal when exposed t o air undergoes partial oxidation to a persilicate (Jordis A.1914 ii 200) as highly improbable. A solution of such a persilicate should actually contain the hydrolytic products namely a silicate and hydrogen peroxide but the only evidence mentioned (Zoc. c i f . ) is the presence of a little free chlorine in the carbon dioxide liberated by hydrochloric acid. Such a result might be due to the presence of a little manganese as impurity in the silicate solution. However the author (Ditz also Ditz and Kanhauser A. 1913 ii 958) has shown that after prolonged exposure to the air many alkaline substances are found t o contain sniall quantities of nitrite and nitrate and treatment with hydrochloric acid would give rise t o nitrous fumes with the former and if the acid were concentrated to chlorine with the latter. Experimental test has confirmed the possibility of this result with a solution of a silicate of an alkali metal distinct traces of nitrite being present even after exposure for two days the amount being very appreciable after three months.In the same period the quantity of nitrate formed was relatively small whilst titanic acid failed to reveal any trace of hydrogen peroxide. D. F. T. Thermal Analysis of Mixtures of Alkali Hydroxides with the Corresponding Haloids. 111. Lithium Compounde. GIUSEPPE SCARPA ( A t t i R . Accad. Li72cei 1915 [v] 24 ii 476-482. Compare A. 1915 ii 448 633).-The systems LiOH- LiF LiOH-LiC1 LiOH-LiBr and LiOH-LiI have1 now been investigated. The lithium hydroxide employed contained 98.5% LiOH 0.8% Li,CO and 0.7% H,O and had m. p. 462O; Dittinar (+7.XOC. Chem. Ind. 1888 7 733) and de Forcrand (As 1906 ii 445) found m. p. 445O. Pcr lithium fluoride the author finds m. p. 840O; Carnelley (T. 1878 33 281) gave 801O. Lithium hydroxide and fluoride are 4-2ii. 100 ABSTRACTS OF CHEMICAL PAPERS. completely miscible in the liquid state and form solid solutions containing from about 5 to 85 mols. % of the hydroxide. The eutectic point 430° corresponds with 80 mols. % of the hydroxide. Guntz (Ao 1894 ii 91) gave 600° Huttner and Tammann (A. 1905 ii 229) 600° and Schemtschuschni and Rambach (A. 1910 ii 204) 614O. The curve of primary crystallisation of the system LiOH-LiC1 falls from 605O to a eutectic a t about 290° and exhibits a distinct angle a t 50 mols. % LiOH. The eutectic arrest observed a t about 285O with all mixtures containing from 45 t o 100 mols.% LiOH has its maximum duration a t about 65 mols. % LiOH. Mixtures con- taining 0-50 mols. % LiOH show a second arrest a t 315O this having its maximum duration for 40 mols. % LiOH and corre- sponding with the formation of a compound probably 2LiOH,SLiCl decomposing on fusion. ,. Lithium bromide has m. p. 550O; Carnelley (Zoc. c i t . ) found 547O. The diagram of state for LiOH-LiBr is similar t o that of the pre- ceding system the two branches of the curve of primary crystal- lisation intersecting a t the eutectic point 275O which corresponds with 45 mols. % LiOH. The eutectic arrest is shown with 0-70 mols. % LiOH but not with 75-100 mols. % LiOH; the latter mixtures exhibit however a marked arrest a t 310° this having its maximal duration with 75 mols.% LiOH and corresponding with the formation of a decomposable compound probably 3LiOH,LiBr. Lithium iodide (probably containing a little oxide) has m.' p. 440O; Carnelley (loc. cit.1 found 446O and Sandonnini and Scarpa (A 1914 ii 204) 450O. The curve of primary crystallisation of the system LiOH-LiI exhibits a eutectic a t 45 mols. % LiOH and a pronounced change of direction a t 75 mols. % LiOH. The eutectic arrest occurs a t about 180° and is a maximum a t 45 mols. % LiOH. Mixtures with more than 75 mols. % LiOH show an arrest a t about 310° probably owing to the formation of the com- pound 4LiOH,LiI unstable a t the melting point. Comparison of these results with those previously obtained (A. 1915 ii 448 633) shows that the tendency of an alkali hydroxide to form compounds with the corresponding haloids increases with diminution of the electro-affinity of the cation (compare Abegg and Bodlander A.1899 ii 542). For lithium chloride m. p. 605O is found. T. H. P. Preparation of Phosphorescent Calcium Sulphide. PIERRE BRETEAU (Compt. rend. 1915 161 732-733; J. Yharm. Chim. 1916 [vii] 13 33-37).-Contrary t o the views of Verneuil (com- pare A. 1887 2 539) i t is shown that the presence of sodium chloride and carbonate in small quantities is not essential for the phosphorescence of calcium sulphide. A phosphorescent sulphide can be prepared by heating a mixture of calcium carbonate (100 parts) and powdered sulphur (30 parts) in a crucible a t a dull red heat for one hour allowing the mixture to cool mixing it with alcohol to a paste and adding basic bismuth nitrate in alcoholic solution in sufficient quantity to give 1 part of bismuth to 10,000INORGANIC CHEMISTRY.ii. 101 parts of sulphide. This mixture is dried in the air and then heated a t a dull cherry-red heat for two hours and allowed t o cool slowly. The sulphide so prepared shows a violet phosphor- escence. The bismuth as phosphorogen can be replaced by molybdenum or vanadium or better still by tungsten. W. G. The Derivatives of Perceric Oxide. If. C. C. MELOCHE ( J . Amer. Chem. SOC. 1915 37 2645-2652).-It has not been possible t o prepare perceric sodium carbonate by a method analogous to that used in the preparation of the potassium com- pound (A. 1915 ii 776). A solution of perceric ammonium carbonate may be prepared by this method but the substance has not been obtained in the crystalline form.The solution is less stable than that of the potassium compound and decomposes when kept with the formation of a yellow precipitate presumably ceric carbonate. I f the dark red solution of perceric ammonium carbonate is treated with excess of solid sodium carbonate and the solution evaporated slowly in a vacuum well-formed crystals of perceric sodium carbonate of the formula Ce,0,(C03)2,4Na,C03,30H20 are obtained. It is only sparingly soluble in cold water and effloresces in dry air. Contact with moisture above Oo results in hydrolysis and with larger quantities of water the crystals are completely decomposed with the formation of a gelatinous orange- red precipitate. A perceric rubidium carbonatel was obtained in crystalline form by the method which served for the preparation of the potassium compound but this substance has not been analysed.The action of hydrogen peroxide on a saturated solution of potassium acetate in presence of cerous nitrate yields a solution which probably contains perceric potassium acetate. It liberates iodine from potassium iodide gives a blood-red solution with potassium carbonate and a reddish-brown precipitate with hydrogen peroxide. I n all these perceric compounds half the total oxygen combined directly with the cerium is available for oxida- tion under certain conditions. H. M. D. Soma New Rare Earth Compounds. A. J. GRANT and C. JAMES (1. Amer. Chem. SOC. 1915 37 2652-2654).-With a view to the improvement of the methods employed in the separation of the rare earths a number of new salts have been prepared.Terbium pyromucate (C,H,0.C0,)3Tb,5H,0 radiating crystals very soluble in water ; terbium l-bromo-2-nitrobenzene-4-sulphonate ( N02*C,H3Br*S0,)3Tb,10H20 very small crystals the solubility of which differs little from that of the gadolinium salt; terbium propionate (C,H,*C02),Tb,2H,0 white powdery voluminous crystals ; ferrous lanthanum nitrate 3Fe( N0,),,2La(N03),,24H,0 flat green hexagonal crystals fairly stable' in the1 absence of air. Lanthanum pyromucate ( C,H30~C0,),La,2H20 ; and yttrium pyro- mucate (C,H,O*CO,),Yt,SH,O were also prepared. H. M. 9.ii. 102 ARSTR 4CTS OF CHEMICAL PAPERS. The Separation of Yttrium from the Yttrium Earths.111. J. P. BONARDI and C . JAMES ( J . Amer. Cheni. Soc. 1915 37 2642-2645. Compare A. 1914 ii 370 657)-Further experi- ments have been made with the object of finding a rapid efficient and economical method for the separation of yttrium from the yttrium earths The results of fractional precipitation by the addi- tion of various reagents are briefly described. These reagents include ammonium sebacate azcbenzenesulphonic acid potassium sulp hite sodium citrate tartrate tungstat e p henoxyacetate and m-nitrobenzoate ammonium camphorate and potassium cobalti- cyanide. The use of a cobalticyaiiide solution appears the most promising for the purposes of the required separation and this method is t o be further examined. H. M. D. Decomposition of Mineral Sulphides and Sulpho-salts by Thionyl Chloride.H. B. NORTH and C. B. CONOVER ( A w r . J . Sci. 1915 [iv] 40 640-642. Compare this vol. ii 28).- Mineral sulphides react in general in a similar manner t o the pre- cipitated sulphidea with thionyl chloride according t o the general scheme MS + 2SOC1, MC1+ SO + S2C12. The reactions take place a t 150-175O and require from a few hours' heating in the case of easily decomposed sulphides such as galena pyrites cinnabar orpiment stibnite and arsenical pyrites to one or two days in the case of pyrargyrite proustite covellite sphalerite and tetra- hedrite. Of the minerals examined only argentite molybdenite and cobaltite were notl attacked by thionyl chloride under the above conditions. G. F. M. Reaction of Gases with Lead and Silver.W. STAHL (Clic m. Zeit. 1915 39 885-886).-A summary of published work on the behaviour of lead and silver a t high temperatures towards oxygen sulphur dioxide hydrogen nitrogen the oxides of carbon hydrocarbons argon and helium. G. F. M. Supposed Allotropy of Copper. G. K. BURGESS and I. N. KELLBERG ( J . IVashington Acad. Sci. 1916 5 657-662).-From dilatometric observations Cohesn and Helderman (A. 1914 ii 205 654) have drawn the conclusion that there are two enantiotropic forms of copper with a transition temperature a t 69.2O t o 71'7O. I n support of this view the change in the electrical conductivity of copper after heating a t looo has been put forward. Comparative measurements made with copper and platinum wire resistance thermometers wound on the same frame and exposed alternately t o Oo and looo show that platinum exhibits similar changes in conductivity and thatl in both cases constancy is obtaineld after a few alternations of temperature. A detailed examinatioln of the resist.ance1 of copper ovelr the range 00 t o 1000 by the method described in a previous papelr (A.1914 ii 794) lias also giveii entirely negative results in respect of the alleged allotropic transformation in the1 neighbourliood of 70°. Within the above range the rwistance of copper changes in a continuousIh'OltGANIC CU EMISTHY. ii. 103 niaiiner and there is no evidence of the copper being in a iiieta- stable condition. H. M. D. Sulphides of Copper. EUGEN POSNJAK E. T. ALLEN arid H. E. MERWIN (Economic Geology 1915 10,491-535).-A general physico-chemical investigation of the sulphides of copper and of the various copper-iron sulphide minerals carried out with the object of ascertaining tlie conditions of formation and alteration of the more important minerals of the copper sulphide deposits.Cuprous sulpliide prepared in a vacuum furnace has m. p. 1130O f lo. No dissociation occilrs when cuprous sulphide is heated up to 1200O. Synthetic cuprous sulphide has I):? 5.785 a value which is almost identical with t h a t of the' purest mineral sulphide examined. The copper sulphides formed by fusing together copper and sulpliur are of v::riable composition and always contain more sulphur than is demanded by tlie ,ratio( 2Cu:S. They are micro- scopically homogeneous and vary continuously both in colour and specific gravity with composition.As the sulphur increases they become darker in colour and have smaller specific gravities. The specific volumes of these products lie within the limits of experi- inental error on a straight line connecting tlie specific volumes of cuprous and cupric sulphides. It is shown that all such products are solid solutions of the two sulphides. Cuprous sulphide melts a t 1096O in an at'mospliere of hydrogen sulphide and a t 1057O in an atmosphere of sulphur vapour. The lower melting point as com- pared with that of cuprous sulphide in a vacuum is clue to dis- solved cupric eulphide which increases in amount with the increase of pressure of the sulphur vctpour surrounding it. When cuprous sulpliide is heated in an atmosphere of hydrogen sulphidel a t various temperatures below its melting point the sulphur content increases.For each temperature the1 products contain a definite amount o f sulphur. This sulphur content increases with decrease of temper- ature until a t 358O the product becomes cupric sulphide. Solid solutions of cuprous-cupric sulphide can be prepared by heating compressed powde'rs of tho two sulphides a t about looo. The analysis of a number of natural chalcocites showed that such solid solutions sometimes occur in nature. I n this connexion the nature of the mineral pyrrhotite has been discussed. Cuprous sulphide is dimorphous and has an inversion temperature1 a t 91O. The inversion temperature is considerably influenced by the size of the grains of the sulphide.Increasing amounts of cupric sulphide dissolved in cuprous sulphide raise the inversion temperature. This takes place until a concentration of cupric sulphide of 876 is reached after which an inversion is no longer observed. Crystals of chalcocite were repeatedly formed in the dry way a t various temperatures but only above about 250° were they large enough f o r crystallographic investigation. These crystals were isometric. Isometric crystals of cuprous sulphide were also obtained from solutions; a t 250° by the reaction of cuprous chloride with sodium sulphide'; and at ZOOo 170° aiid 1 2 5 O by Frystallisation of cuprous sulphide from solutions of hydrogen sulphide. Crystals formed byii. 104 ABSTRACTS OF CHEMICAL PAPERS. the action of ammonium sulphide on copper which have generally been described as orthorhombic cuprous sulphide are in reality a double polysulphide of the formula Cu,(NH,)S and are tetragonal. Covellite has been prepared in many ways and a crystallographic and optical study of its properties carried out.It is hexagonal and optically positive; the optical dispersion of o is extremely high wLi<l.0 wNa = 1-45 and orl = 1-80. Two very pure samples of covellite have been found to have DY 4.683 and 4.676 respectively. This value is higher than any value previously given for this sub- stance. The highest value found for synbhetic covellite was 4.652. Covellite can be heatled in an atmosphere of hydrogen sulphide up to 358O; i t is then in equilibrium with the gas and a t this temperature and below it chalcocite can be completely converted into covellite.J. F. 8 The Nature of Subsidiary Valencies. XII. Ammines of Copper. FRITZ EPHRAIM and EDOUARD BOLLE (Ber. 1915. 48 1770-1777. Compare A 1915 ii 166 441 454).-Details are given of the preparation of various ammines of copper salts in addition to those previously described ; the ammonia tension has also be'en measured a t various temperatures. The cupric salts of dibasic acids in no case give a hexammine compound even a t the te'mperature of an ice-salt freezing mixture the highest ammine formed being the pentammine. Many of the salts of mocobasic acids give a hexammine compound. I n many cases where the hexammine is capable of existence only a t very low temperatures tlie triammine is nevertheless stable a t fairly high temperatures.The order in which the anions exert an effect on the stability of the ammine is quite different from that observed in the case of the nickel salts; it resembles rather that observed with the zinc salts (A. 1915 ii 454) although with many differ- wcee. The theoretical discussion of the results is t o be given later. The following new compounds are described Copperhexammine nitrate blue crystals; the ammines with 5 and 23/4 molecules of ammonia which have been described by other investigators could not be obtained. Copperhexammine perchlorate. Copperpent- ammine dithionate; the ammine with 9 molecules of ammonia could not be obtained. Coppertetmmmine tetrathionate light blue crys- tals ; copperhexammine thiocyanate ; coppertetrammine formate ; coppertriamrnine benzoate copper-red crystals ; coppertetrammine benzoate dark blue crystals; and copperpentammine benzoate greyish-blue crystals.The ammonia tensions of the following salts were also deter- mined coppertetrammine nitrate coppertetrammine thioimlphate coppertetrammine dithionate coppertetrammine thiocyanate and copperpentammine oxalate. A table is given shotwing the absolute dissociation temperatures and the heats of formation of all the copper ammines hitherto examined. T. S. P. .Thermal Reactions sf Aluminium. RAFAEL LUNA NOGUERAS (-4naZ. Fis. Quim. 1915 $3 390-420).-A historical account oflNORGANIC CHEMISTRY. ii. 105 the production of aluminium and of its applications as a thermal agent. A. J. W. Investigation of the Carbides of Aluminium Nickel and Copper. E.BRINER and R. SENGLET ( J . Chim. Phys. 1915 13 351-375).-A study of the conditions under which carbon com- bines directly with the three metals and of the dissociation of the resulting carbides. The formation of aluminium carbide from the elements has been observed a t 750° and 900°. This compound is exothermic and represents theref ore the stable form a t lower temperatures. Dis- sociation of the carbide was found to occur howevex a t 540° and when heated in a current of air carbon dioxide and aluminium oxide are obtained. This change has been followed a t various temperatures between 540° and 900°. Nickel carbide Ni,C is an endothermic compound and the optimum temperature of formation is in the neighbrourhoiod of 2100O.A t lower temperatures the carbide dissociates relatively rapidly a t 1600° and slowly a t 900°. The carbide must conse- quently be cooled rapidly if a good yield is required. Some evidence has been obtained of the direct combination of copper and carbon to form a carbide a t high temperatures. It is endothermic and dissociates rapidly a t about 1600° more slowly a t lower temperatures. Identification of Tervalent Manganese in Glass. S. It. SCHOLES ( J . Ind. Eng. Chem. 1915 7 1037).-The author coii- siders that the violet tint of manganese glass is due to compounds of the trioxide MnzO3! deducing as evidencie in support of this theory the fact that pirlk solutions were obtained by extracting a lead glass of low m. p. and a water-soluble potash glass with hydrofluoric or 30% sulphuric acids.On diluting the extracts with water a light brown precipitate separated out similar to that obtained by diluting a manganic sulphate solution. Their behaviour towards oxalic acid was also the same. H. M. D. G. F. M. A Magnetic Study of the A3 Transformation in Pure Iron. K ~ T A R ~ HONDA and HIROMU TAKAGI (Sci. Rep. Tohoku Imp. Unb. 1915 [ii] 4 261-269).-Four different specimens of pure iron fro'm the Bureau of Standards Washington have been examined thermomagnetically. The specimens weighing about 0.4 gram are tested in an atmosphere of nitrogen. Both the beginning and end of the transformation may be readily traced on the curves and the critical points thus determined are in good agreement with those found by thermal analysis.The magnetic value of Ac3 for pure iron is 908-911° and of Ar3 889-898O. The number of magnetons is not usually an integer and varies for different speci- niens. These facts are opposed t.0 Weiss's theory. The Magnetic Transformation of Cementite. K ~ T A R ~ HONDA and HIROMU TAKAGI (Sci. Rep. Tohoku Imp. Unniv. 1915 [ii] 4 161-167).-A thermaI and magnetic examination of a series C. H. D.ii. 106 ABSTl1ACTS OF C11EX1 ICAL l’.\t’EltS. of white pig jroiis containing from 2.90% to 3.49‘;; of combined carbon shows that the magnetic transformation of cementite begins on cooling a t 215O and ends on heating a t the same temperature. The temperature range is about 50°. The magnitude of both the thermal and magnetic changes increases with the proportion of cementite.The critical temperature is the same in steels containing as little as 0.14% of carbon so there is no indication of the presence of more than one carbide. It is small in a grey iron. C. H. D. Presence of Platinum in Spain. S. P I ~ A DE RUBIES ( A d . Pis. Qciina. 1915 13 420-433).-A summary of the records as to the occurrence of platinum deposits in Spain. A. J. W. The Series of Triamino-aquo Salts of Bivalent Platinum [Pt3NH,,H2O]X2. L. TSCHUGAEV and I. TSCIIERNJAEV (Compt. relid. 1915 161 792-794). Compare this vol. ii 42).-0ii pass- ing a current of air through a solution of the chloride [OH.&P‘<JyH h’H S H,.OH 1% containing ammonia and ammonium. sulphate or any other sulphate and a small amount of any copper salt oxidation occurs a colourless crystalline precipitate being obtained which is soluble in warm dilute sulphuric acid.Addition of potassium platino- chloride to this acid solution gives a precipitate of the plcitino- chloride [Pt3NH3,H,0]PtCl green needles which when warmed with hydrochloric acid or a soluble chloride is converted into Cleve’s platinochloride [Pt3NH3,Cl],PtCl,. Addition of potassium platino- bromide to the acid solution gives the ylatii~obron~ide [Pt,3NH3,H,0],PtBr green needles which is converted by hydrobromic acid or soluble bromides into the platinobromide [Pt3NH3,Br],,PtBr,. Analysis of the original oxidation product points to its composition being remesented by the formula L in which case i t is allied to Werner’s binuilear diol complexes the index of co-ordinat,ion of the platinum being equal to 6.W. G. The Series of Rydroxy-pentarnminoplatinic Salts. L. TSCHUGAEV and W. CHLOPIN (Compt. rend. 1915 161 699-700. Compare Tschugaev and Vladimirov A. 1915 ii 569).-Tlie authors have prepared a series of salts having the general formula [Pt,5NH3,(OH)]X,. The carbonate is readily prepared by pass- ing a current of ozone through a mixture of Peyrone’s chloride (1 part) and ammonium carbonate (2 parts) in an excess of ammonia for two or three hours with continual stirring. The carbonate is precipitated being insoluble in water and is readily decomposed by acetic acid giving t’he rrcetafe [Pt,5NH,,(oH>](CH3*Co,),.MINERALOGICAL CHEMISTRY. ii. 107 This salt in solution is decomposed by mineral acids to give tlie corresponding salts. In this way the chloride [Pt,5NH3 (OH)]%H20 rhombic plates and the nitrate [Pt,5NH3,(OH)](NO,) crystal- lising in anhydrous needles were prepared. Both these salts are soluble in water and are ionised. The hydroxyl group like the chlorine atom in the series [PtC1,5NH3]X is completely masked and apparently not capable of being directly replaced by a halogen or any other negative atom o r group. [ Pt,5NH3,(OH)]”! is tervaleiit is shown by a study of the coagulating power of the chloride the results of which agree with those of the luteocobalt- aminines which give by dissociation tervaleiit ions. Like the members of the series [Pt75NH,,C1]X and the series [Pt,GNH,]X the salts of this series give carbonates and sulpliates wliich are practically insoluble in water but soluble in the fixed alkalis. Further the chlorides of all these three series are readily reducetl by zinc in dilute hydrocliloric acid giving in each case the chloride of Reiset’s base I [Pt,4NH3]CI,. That the group W. G.