36 ABSTRACTS OF CHEMICAL PAPERS. Inorganic Chemistry. Hydrate of Iodine Oxyfluoride Fluoroiodates and Iodoxy- fluorides. RUDOLF F. WEINLAND and FERDIKAND REISCHLE (Zeitsch. anorg. Chern. 190P 60,163-172. Compare Weinland and Lauenstein Abstr. 1899 ii 363).-Iodine oxy$uoride hydrate IOF,,5H20 or IF3(OH)2,4H20 is obtained in colourless needles by dissolving iodic acid in a solution of hydrofluoric acid in glacial acetic acid and evaporating. The same compound contaminated with iodine is obtained when alcohol is used as solvent. The crystals fume in the air giving off hydrogen fluoride. Four salts containing pyridine have also been prepared. When 1 mol. each of iodic acid and pyridine are dissolved in an acetic acid solution of hydrofluoric acid and the solution evaporated over sulphuric acid pyridine trzjhoroioclate IF3(OH)2,C,N H separates in colourless prismatic needles.When a concentrated aqueous solu- tion of this salt is allowed to crystallise slowly pyridine dqjhoro- iodate IF,O(OH),C,N H is obtained in colourless crystals. When the pyridine salt first mentioned is dissolved in a mixture of hydro- fluoric acid and glacial acetic acid and 1 mol. of iodic acid 1s added pyridine pentaJEuorodi-iodate IF,O( OH) ,C,NH,,IOF,,H,O is obtained. Finally from a solution of 2 mols. of iodic acid 1 mol. ofINORGANIC CHEMISTRY. 37 pyridine and excess of hydrofluoric acid in alcohol pgridine hepta- Juorodi-iodate IF,( OH),C,NH,,IOF is obtained in colourless needles. All the pyridine salts are stable in air. The following five compounds are obtained by dissolving the corresponding iodo-compounds (iodobenzene and the iodotoluenes) in very concentrated alcoholic hydrofluoric acid.Phenyl iodoxyJtuoride hydroJEuoride C,H,*IOF,,HF the p-bromo-derivative C6H,Br*IOF2 HF and 0- m- apd p-tolyZiodoxy$uoride hydrofluoride C6H,Me*IOF2,H F all form colourless crystals which readily give up hydrogen fluoride. The compound IOF,,C,H,*IOF separates in colourless needles from a solution of 1 mol. of iodobenzene and 3 mols. of iodic acid in alcoholic hydrofluoric acid; it is readily decomposed by water. Ozone. 111. STEPHAN JAHN (Zeitsch. anoi-g. Ckem. 1908 60 292-336. Compare Abstr. 1905 ii 16; 1906 ii 292).-The sources of error in measurements of the electrical potential of ozone are discussed. The velocity of rotation of the electrode is without influence and the anomalies are therefore not to be explained by diffusion (compare Luther and Inglis Abstr.1903 ii 406). The effect is shown to be probably due to the adsorption of ozone by platinum. Measurements of potential were made against hydrogen with platinum electrodes in sulphuric acid with and without the addition of cobaltic sulphate. The results show that over the interval where Co”’ forms 2.5-8*0% of the total Co the potential of the cell corresponds closely with that required by the. ratio Co’” Co”. When Co”’/Co” = 1 the potential is 1.785 volt a value employed in calculating the ozone potential. The E.M.F. of a cell in which the reaction 0,+H,=H20+0 occurs is probably 1-90 volt at Oo. An equation is given by means of which the temperature-limits of stability of ozone may be determined derived from the Gibbs-Helm- holtz equation but the uncertainty of the values for potential and for specific heat is so great that no accurate calculation can be made.It appears however that ozone should be formed from oxygen by the action of heat alone in quite measurable quantity a t a temperature of 2000’ (compare Fischer and Marx Abstr. 1906 ii 224 606). G. S. C. H. D. Ozone. IV. STEPHAN JAHN (Zeitsch. anorg. Clzem. 1908 60 337-357. Compare preceding abstract).-The development of heat in the decomposition of ozone has been measured by a method similar in principle t o that of van der Meulen’s soda-lime being employed as catalyst and a current of ozonised oxygen of constant composition led through the calorimeter. The catalyst was contained in a glass tube connected with a copper spiral tube in a vacuum-jacketed vessel.The heat of dissociation of ozone is found to be 34,100 cal. per gram-molecule with a probable error of 1%. Investigation of the System Sulphur-Iodine. F. OLIVARI (Atti R. Accad. Lincei 1908 [v] 1’7 ii 512-521).-The freezing- C. H. D.38 ABSTRACTS OF CHEMICAL PAPERS. point curve of mixtures of sulphur and iodine is of the type of a cryohydric curve the eutectic point corresponding with 80.8 atom % of sulphur being a t 65.7'. The formation of compounds by fusion of these two elements is hence excluded the supposed melting point of the two iodides S,I and S,T2 being the cryohydric point. Cryoscopic measurements of sulphur dissolved in iodine show that the molecular weight has the value 256 corresponding with 5 in dilute solution but increases as the concentration is raised ; there is therefore no isomorphism between the two elements.The specific volume curve for mixtures of sulphur and iodine is a straight line joining the specific volumes of the two elements. The supposed iodides S21 and SI obtained by crystallisation from solution are mechanical mixtures of sulphur and iodine as also is probably the iodide S&. T. H. P. Absolute Sulphuric Acid its Preparation from Sulphur Trioxide and Water its Specific Electrical Conductivity and that of more Dilute Acid. DAVID 31. LICHTY (J. Amer. Chm. Xoc. 1908 30 1834-1846).-During a study of the decomposition of oxalic acid by sulphuric acid (Abstr.1907 ii 445) it was observed that the rate of decomposition was decreased to a surprising extent by the addition of only 0.01% of water to sulphuric acid supposed to beof 100% strength. Since analytical methods are not sufficiently accurate to determine the presence of very small quantities of water recourse was had to determination of the specific electrical conductivity of the acid. This method however could not serve t o establish the composition of the acid since the exact conductivity of any one concentration was not known. A method has therefore been devised for the preparation of absolute sulphuric acid by the direct interaction of sulphur trioxide and water in stoicheiometricel proportions. The specific conductivity of the acid obtained in this way has been determined a t 25O and found to be 0°01041 reciprocal ohm which is also the minimum specific con- ductivity of concentrated sulphuric acid.It has also been found that the maximum f. p. of concentrated sulphuric acid is identical with th6 f. p. of the absolute acid namely 10.43-10*45°. It is proved therefore that the preparation of absolute sulphuric acid may be based on the minimum specific conductivity and on the maximum f. p. It is shown that for acid of not less than 95% strength the results of analysis by means of conductivity determinations are accurate to within 0.01%. A table of specific conductivities is given which may be used for analytical purposes. Mixed Anhydrides of Sulphuric Acid. AMI~ PICTET and GEORaES KARL (Bull. 8oc. chim.1908 [iv] 3 1114-1126 ; Arch. sci. phys. nat. 1908 26 437-454).-1n preparing the mixed anhydride of nitric and sulphuric acids (S03)*N205 already described (Abstr. 1907 ii 758) there is also formed some sulphonitrous anhydride (S0,),(N20,) m. p. 198-200' b. p. 302-305O/728 mm. which is better prepared by dissolving sulphur trioxide and nitrogen trioxide separately in carbon tetrachloride and mixing the two solutions. It E. G.INORGANIC CHEMISTRY. 39 is a hard colourless substance crystallising in needles. Its properties are very similar to those of the nitryl tetrasulphate described previously and like the latter it gradually liquefies but does not fume in moist air. Since Provostaye's nitrosyl sulphate (Alan. Chim. phys. 1840 [ii] 73 362) has m. p. 217-230' and distils at 360° according to various authors it is suggested that it cannot have the formula ( S03)2N203 generally assigned to it but should probably be represented as (S03)6(N203)3 D'Arcy's observation (Trans. 1889 55,157) that boron and sulphur trioxides do not combine when heated together in open vessels is confirmed but combination takes place when the operation is carried out in a closed vessel boryl sulphate SO,(Q*BO) being formed at 115-120° boryl disuZphate (SO,),,B,O at 230° and mixtures of the two a t intermediate temperatures.Both are colourless amorphous bulky brittle hygroscopic substances which have no definite melting point do not fume in moist air but dissociate when heated leaving eventually residues of boron trioxide. They dissolve in water yielding solutions of boric and sulphuric acids.Boryl disulphate reacts with alcohol to form alkyl borates and alkyl-sulphuric acids. I n the last respect its behaviour differs somewhat from that of other mixed anhydrides (compare Friedel and Ladenburg Compt. rend 1865 64 84 ; Pictet and Geleznoff Abstr. 1903 i 309 601 ; Pictet and Bon Abstr. 1906 i 3). When chromium trioxide and sulphur trioxide are heated in closed tubes at 75' or a t a higher temperature a yellow mass of the mixed anhydride SO,ICrO is formed together with some chromium sesqui- oxide from which it could not be separated. F. OLIVARI (Atti R. Accad. Lincei 1908 [v] 17 ii 389-391).-Cryoscopic measurements of solutions of selenium in iodine (compare Timmermans Abstr. 1906 ii 429) give for the molecular weight of selenium the values 150*2-167*7 the number calculated for Se being 158.4; thus selenium has the molecular weight corresponding with Se at 4' (compare Beckmann Abstr.1904 ii 235) with Se at 42' (compare Beckmann and Sfeiffer Abstr. 1897 ii 363) and with Se at 113.9' (compare also Pelliai and Pedrina Abstr. 1908 ii 833). Volumetrio Composition of Gaseous Ammonia and Atomic Weight of Nitrogen. PHILIPPE A. GUYE and A. PINTZA (Compt. rend. 1908 147,925-928).-Ammonin prepared from pure ammonium chloride and lime was decomposed by a hot platinum wire and the nitrogen and hydrogen obtained were passed after purification into a vacuous vessel of about 39 litres capacity until the pressure was about one atmosphere at 0'. The loss in weight of the complete generating apparatus being known the weight and density of the mixed gases could be calculated.One litre of the mixed gases a t 0' and 760 mm. was found to weigh 0.37989 gram when corrected for traces of sulphur dioxide and trioxide. Knowing the densities of hydrogen and nitrogen and making allowance for their slight expansion on mixing the volumetric ratio of nitrogen t o hydrogen in ammonia is 1 3.00172. It follows that the atomic weight of nitrogen is 14.014 (0 = 16). T. A. H. Molecular Weight of Selenium. T. H. P.40 ABSTRACTS OF CHEMICAL PAPERS. Consideration of the sources of error however showed that any value from 14.002 to 14.022 might well have been obtained. The authors consider their method less accurate than direct gravimetric determinations.R. J. C. Oxidation of Ammonia by Potassium Permanganate and the Effect of Ammonium Salts on the Reaction. MORDKO HERSCHKOWITSCH (Zeitsch. physikal. CTLern. 1908 65 93-96). -Dilute solutions of ammonia and of potassium permaoganate in varying concentrations with or without the addition of ammonium salts have been heated for definite periods in a large flask a t 40° and the relative proportions of nitrogen nitrous and nitric acids in the mixture determined. Increase in the proportion of ammonia increases both the nitrogen and the nitrous acid and diminishes the nitric acid; increasing the concentration of both reagents has much the same effect except that the proportion of nitrous acid is not much altered. Addition of ammonium salts diminishes the aniounts of nitrogen and nitrous acid and greatly increases the nitric acid.Ammonium salts are formed in the course of the reaction even when none are present originally and in order to eliminate their action experiments have been made in which excess-of alkali is added previously. I n these circumstances less nitrous acid and more nitrogen are formed. G. S. The Kydrazinates of some Metallic Salts. HARTWIG FRANZEN and 0. VON MAYER (Zeitsch. anorg. Chem. 1905 60 24'7-291).- Hydrazinates of metallic salts have been prepared by Curtius and Schrader (Abstr. 1895 ii 10) and by Hofmann and Marburg (Abstr. 1899 i 486). A number of new methods of preparation are now described. On the addition of hydrazine to a solution of a cobalt- ammine salt reduction always takes place nitrogen being evolved and the hydrazine compound of a cobaltous salt being precipitated.The double compounds form crystalline powders insoluble in water soluble in acids or ammonia. Sulphates and nitrates tend t o combine with 3 mols. of hydrazine whilst chlorides bromides and oxalates with the exception of nickel chloride and bromide and of calcium and strontium bromides retain only 2 mols. The following compounds have been prepared With 3N,H Nickel chloride bromide sulphate nitrate and dithionate ; cobaltous sulphate and nitrate ; zinc nitrate ; cadmium nitrate ; calcium bromide ; strontium bromide. With 2N2H Nickel chloride bromide and oxalate ; cobaltous chloride bromide and oxalate ; zinc chloride bromide iodide oxalate and sulphate ; cadmium chloride bromide iodide oxalate sulphate carbonate and phosphate ; ferrous chloride and oxalate ; manganous chloride sulphate and nitrate ; cupric oxalate ; calcium chloride and nitrate strontium chloride ; barium chloride and bromide.One mol. of hydrazine replaces 2 mole. of ammonia. Full &alyses and descripiions of the compounds are given. C. H. D.INORGANIC CHEMISTRY. 41 Chlorine Azide [Chloroazoimide] N,Cl. FRITZ RASCHIG (Ber. 1908 41 4194-4195).-ChZoroacxoimide N,Cl is liberated as a colourless gas having an odour similar to hypochlorous acid when an aqueous solution of sodium azoimide (1 mol.) and sodium hypochlorite [l mol.) is acidified even with weak acids such as acatic acid or boric acid. It is slightly soluble in water forming a yellow solution and explodes with extraordinary violence when brought i n t o contact with a flame or glowing splinter and sometimes spontaneously with the production of a pale blue flame.It is decomposed by aqueous sodium hydroxide thus N,Cl + 2 NaOH = NaN + NaOCl + H,O. W. H. G. Preparation of Crystalline Silicon. A. J. KIESER (Chem. Zeit. 1908 32 1161-1 162).-With the object of ascertaining whether combination takes place between aluminium and silicon when methods involving the use of aluminium are employed for preparing silicon careful tests have been made for aluminium in the silicon and for ailicates in the slag. For the first tests crystalline silicon was prepared by Vigouroux's method and for the second the slag from Wohler's method was examined. I n each case the quantity of aluminium and silica respectively obtained was so small as t o make it evident that combination between aluminium and silicon does not take place.J. V. E. New Formation of Liquid Alloys of Potassium and Sodium. GEORGE F. JAUBERT (Bull. Xoc. chinz. 1908 [iv] 3 1126-1131 ; Ber. 1908 41 411G-4120*).-The liquid alloys ob- tained by the action of (1) sodium on potassium hydroxide and (2) potassium on sodium hydroxide are described. When sodium hydroxide is melted in a vacuous Jena flask with potassium reaction occurs a t 200-250° bubbles of gas (hydrogen 0) are evolved the temperature rises to 300° and a layer of the alloy NaK having the appearance of mercury is formed and floats on the excess of alkali hydroxide used. When potassium hydroxide (6 parts) and sodium (11 parts) are melted together in the same way at 250-275" the alloy NaK is formed but by modifying the quantities of the hydroxide and metal used and raising the temperature to 350° the chief product is the alloy NaE,. BQth alloys are liquid and do not solidify even at 10".They oxidise rapidly in cold air especially if fresh surfaces are continually exposed and even when kept under liquid petroleum the potassium slowly oxidises. The alloy NaK is now made commercially and is being applied in a number of industries. T. A. H. Cesium Chromates. FRANS A. H. SCHREINEMAKERS and D. J. MEYERINGH (Chem. TkekbZacl 1908 5 81 1-816).-The following chromates of cmium have been prepared (at 30") Cs2Cr0 ; Cs,Cr,07 ; Cs,Cr,O, and Cs2Cr4018. Their properties are similar to those of the corresponding chromates of potassium ammonium and rubidium. The mono- and di-chromates dissolve in water without decomposition ; * and Ann. Chim.anal. 1909 14 1-5.42 ABSTRACTS OF CHEMICAL PAPERS. the tri- and tetra-chromates are decomposed by water. The solubility of the monochromates of ammonium potassium rubidium and casium increases in the order given ; that of the dichromates di- minishes in the same order. A. J. W. Action of Carbon Dioxide under Pressure on Metallio Hydroxides at 0'. FRANK K. CAMERON and WILLIAM 0. ROBINSON (J. Physical Chem. l908,12,561-573).-The authors have determined the solubility of carbon dioxide in water at various pressures in presence of an excess of the solid hydroxides or carbonates. Succes- sive small quantities of carbon dioxide were forced into a tube con- taining the hydroxides or carbonates vigorously stirred with a small known weight of water and curves were obtained connecting pressure with weight of carbon dioxide in the apparatus.Breaks in these curves at any pressure indicate absorption of carbon dioxide by the solid phase that is the formation of solid carbonates or hydrogen carbonates. The only authentic solid hydrogen carbonates appear to be those of the alkali metals excluding lithium. Calcium ferrous iron and magnesium do not form solid hydrogen carbonates a t Oo even under 5 atmospheres pressure. The hydroxides of aluminium ferric iron and glucinum do not appear to combine with carbonic acid The authors regard the solubility of calcium carbonate in carbon dioxide solution as due to R specific solvent action of the carbon dioxide-water mixture rather than as indicating the formation of a hydrogen carbonate (compare however Raikow,uAbstr.1905 ii 85). R. J. C. The Solvent Action of Carbon Dioxide up the Carbonates of the Heavy Metals. CLARENCE A. SEYLER (Analyst 1908 33 454-457).-1t has been shown by Free (Abstr. 1908 ii 848) that the solubility of basic copper carbonate increases with the concentra- tion of the free carbon dioxide and the author now shows that Free's results exhibit regularities similar to those found for other carbonates and that these can be deduced from the principles of chemical equilibrium. The solubility of the basic carbonates of zinc and lead is governed by the same law. In all three cases the effect of a hydrogen carbonate would be to diminish the solubility according to the law Jn/r x HCO = constant when the carbon dioxide is kept constant.The protective action of calcium in solution as hydrogen carbonate on waters which act on lead and zinc is thus explained. w. P. s. Mixed Barium-Strontium Chromate Precipitate. LIONEL H. DUSCHAK (J. Amer. Chem. Soc. 1908 30 1827-1833).-Hulett and Duschak (Abstr. 1904 ii 616) have stated that the occlusion of barium chloride by precipitated barium sulphate is most easily explained by regarding the precipitate as an isomorphous mixture of barium sulphate with one or more complex barium salts and that cases of isomorphism in which one component owing to its ready solubility is present in only a minute quantity are probably not uncommon.INORGANIC CHEMISTRY.43 A study has now been made of the precipitate of mixed barium and strontium chromates formed as an intermediate product in the separation of barium and strontium by the usual analytical method. The results of the experiments show that a crystalline precipitate of barium and strontium chromates of definite composition is formed in a saturated solution of barium chromate containing given concentra- tions of strontium chromate and acetic acid. It has been found that diffusion takes place within this substance and that therefore it must either be regarded as a solid solution or else diffusion must be recognised as a-possible property of isomorphous mixtures. E. G. Silico-thermic Experiments. PAUL ASKENASY and CHARLES PONNAZ (Zeitsch.EZektrochem. 1908 14 810-81 I).-Barium peroxide and silicon react very violently when the reaction is initiated by means of a fuse. By using coarsely-powdered barium peroxide (50 parts) and silicon (15 parts) a barium silicide with 30% of barium is obtained. The mixture (with less silicon and a little sodium peroxide) may also be used to fuse iron or copper or powdered ferro-silicon or titanium silicide may be used in place of silicon. T. E. Estimation of the Valency of Glucinum by Colloidal Experiments. ANT. GALECKI (Zeitsch. Ekktrochem. 1908 14 767-768).-The quantities of barium calcium potassium glucinum magnesium lanthanum and aluminium salts required to coagulate a colloidal solution of arsenic sulphide are compared. Glucinum behaves in exactly the same way as barium calcium and magnesium very much larger quantities of potassium salts and very much smaller quantities of aluminium or lanthanum salts being required to produce the same effect.The results point to glucinum being bivalent. T. E. Compounds of Lead with Nitrous Acid. ALBERTO CHILESOTTI (Atti R. iiccad. Lincei 1908 [v] 17 ii 377-384 474-483. Compare Abstr. 1908 ii 948)-When treated with water the basic salt Pb(OH),,Pb( NO,),,H,O undergoes hydrolysis yielding the compound 3Pb0,N203,zH20 (Zoc. cit.). The latter then undergoes further hydrolysis during which the ratio Pb:NO increases to a value indicating the formation of the salt 4Pb0,N20,,H,0. The experiments carried out up to the present time show that the following basic salts must be regarded as definite compounds Pb(OH),,Pb(NO,) and its hydrate with 1H,O ; 3PbO,N,O or 2Pb0,Pb(N0,)2 and 4Pb0,N2O,,H,O ; the first two of these salts may be regarded as derivatives of ortho-nitrous acid H,NO,.The salt 3PbO,N,O may be prepared as follows The basic salt Pb(N02),,Pb(OH)2,H,0 is treated with increasing quantities of water the limpid yellow solution formed being replaced by water by which means an almost colourless turbid liquid is obtained from which the suspended substance is gradually deposited ; the substance then remaining undissolved consists of the salt 3P bO,N2O3,xH2O. Study of the electrolysis of the salt 4Pb0,N20,,N,0,,3H20 in44 ABSTRACTS OF CHEMICAL PAPERS. dilute acetic acid shows that the ions NO,’ or the molecules of the neutral nitrite Pb(NO,) do not form complex cations with lead and that the ions NO,’ and NO,’ move for the most part independently of one another and in proportions corresponding with the stronger dissociation of Pb(NO,) and with the greater mobility of the NO,’ ion; the current is carried only in small proportion by lead ions.I n order to ascertain whether lead nitrate combines with the nitrite in small proportion to form complex anions the author uses a method devised by Miolati and Pizzighelli (Abetr. 1908 ii 595). Con- ductivity measurements of solutions containing lead nitrite and nitrate show that the specific conductivity curve changes direction when the two salts are present in molecular proportions. It is hence probable that to a certain small extent these two salts combine in the proportion Pb(N03) Pb(NO,) to form a complex salt.Attempts to separate such a complex salt in the solid state from solutions containing the components in molecular proportions were unsuccessful. T. H. P. Optical Investigation of the Copper Complex in Ammonia and Pyridine Solutions. ARTHUR HANTZSCH and PHILIP WILFRED ROBERTSON (Ber. 1908 41 4328-4340. Compare Abstr. 19OS ii 447 646).-The method previously described has been extended t o the coppertetrammine complex and the following conclusions have been arrived at. All cupric compounds whether in aqueous or alcoholic ammonia solutions are optically identical and contain exclusively the coppertetrammine complex Cii(NH3)4 ; the base is coppertetrammine hydrate CU(NH,),(OH)~ and not the diammine hydrate.The colour is not materially altered whether the temperature degree of dissociation or solvent is changed and apparent departures from this are ascribed to chemical change; as for example the formation of a triammine- aquo-complex C U ( ~ H3)3. The solutions in aqueous pyridine of copper salts are likewise identical ; the complex is less stable than that with ammonia and is probably CUE;,. Copper acetate in pure or alcoholic pyridine contains another less intensely coloured complex probably Cu yy2 The chemical influence of the solvent is exhibited in the water and methyl ethyl and amyl alcoholic solutions of copper acetate probably due to the addition of solvent to the unsaturated complex Cu(OAc),. Copper forms in solutions only the complex CLIR and has co-ordina- tion number 4.W. R. (OH,) (OAc),’ Scandium. SIR WILLIAM CROOKES (Phil. Trans. 1908 209 A 15-46. Compare Abstr. 1908 ii 695).-About thirty scandium salts both of inorganic and organic acids have been prepared and analysed and their properties are described in detail. Some of the inorganic salts have already been described by previous observers. The chloride Sc,CI and the corresponding bromide crystallise with 12H,O ; the fluoride is anhydrous. Scandium potassium sulphate,INORGANIC CHEMISTRY. 45 Sc,(S0,),,3K,S04 is contrary t o the statement of Cleve and in agreement with Nilson only slightly soluble in a satura%ed solution of potassium sulphate. The selenate Sc,(Se04),,8H20 rhombic prisms loses 6H,O at looo and becomes anhydrous at 400O; it decomposes at a yellow heat.The nitrate Sc(N0,),,4H20 prismatic crystals is rendered anhydrous on heating over a water-bath and decomposes at a high temperature. When dried in a vacuum at looo for seventeen hours a basic nitrate Sc(OH)(NO,),,H,O is obtained and another basic nitrate Sc,O(NO,) is obtained by heating the nitrate for twelye hours in a hot-air oven at 120°. The majority of the organic scandium salts were prepared by interaction of scandium hydroxide and the corresponding acid. Most of the salts with fatty acids are more soluble in cold than in hot water. The formate (HCO,),Sc*OH,H,O occurs in hard lustrous crystals soluble in hot water and in alcohol. (C2H,02),Sc*OH,2H,0 forms fine white crystals which on being kept for some days lose 1 H20. The propionate (C,H,O,),Sc*OH occurs as a voluminous white powder easily soluble in alcohol ; the butyrate (C,H70,),Sc*OH the isobutyrate with 2H20 and the isovalemte (C,H902),Sc*OH,2H20 are also amorphous and soluble in alcohol Scandium oxulate Sc2(C204),,5 H20 occurs as a crystalline powder which loses 2H20 over sulphuric acid at the ordinary temperature 3H20 in air at looo and 4H,O in air at 1 4 0 O .It is moderately soluble in a solution of ammonium oxalate on which fact a method of separating scandium from other rare earths may be based. The succinate C2H4[ CO,Sc(OH),],,H,O is a white powder insoluble in water and alcohol ; the picrate (C6H,O7N,),Sc*0H,14H2O occurs in yellow needle-shaped crystals which explode faintly on heating. The salt loses 9H,O at 100'. The benzoate (C,H,*CO,),Sc and the o-toluate ( C,H4Me*C0,),Sc*OH,3H,0 form crystalline powders ; a basic o-toluate %C,H70,Sc,Sc(OH) obtained by using excess of the hydroxide occurs as a white powder.The m-and p-toluates both with 3H,O form white curdy precipitates insoluble in water ; the phenylacetate ( C7H7*C0,),Sc*OH,3H20 is also practically insoluble in water. The pyromellitate C H2[ CO,Sc(OH),] 2H20 a white amorphous powder loses 1H,O a t looo and the remaining molecule at 1409 The camphorate C,,H,,O,:ScOH is a white insoluble powder which is very electrical in the dry state. The atomic weights of the closely-related elements scandium yttrium and ytterbium are approximately simple multiples of that of boron and as they occur together in nature it is suggested that the elements of lower atomic weight may be formed from those of higher atomic weight by successive sub-division.Photographs of those parts of the spectrum containing the more prominent scandium lines are given. Scandium. I. RICHARD J. MEYER (Zeitsch. avzorg. Chem. 1908 60 134-151).-The extraction of scandium from wolframite (from Zinnwald Saxony) is described. The mineral contains 0.14-0*16% of rare earths in which scandium oxide predominates. The ore is The acetate G. S.46 ABSTRACTS OF CHEMICAL PAPERS. first fused with sodium hydroxide and extracted with water; the residue (chiefly the oxides of iron manganese calcium and lead) con- tains practically all the scandium (0*30-0*33% of the oxide). Two methods have been used for separating the scandium from the mixed oxides (1) To a neutral or slightly acid solution of the oxides a saturated solution of oxalic acid is added and the mixture kept for twenty-four hours.The precipitate which is richer in the rare earths than the original mixture is washed ignited dissolved in excess of hydrochloric acid and excess of hydrofluoric acid is added. The precipitate thus obtained is quite free from manganese and consists of the fluorides of lead calcium iron and of the rare earths. The remaining metals are removed by special methods and the rare earth residue finally obtained contains about 95% of scandium oxide. According to the second method the mixture of oxides is dissolved as completely as possible in hydrochloric acid solid sodium silico- fluoride is then added and the mixture boiled for half an hour when a precipitate containing all the scandium and practically none of the other rare earths is obtained.Iron calcium etc. are then removed as before and finally a very pure scandium oxide is obtained. The other rare earths except thorium can be separated from scandium by adding sodium thiosulphate to a neutral solution and boiling for some time; the precipitate consists of scandium thio- sulphate. The final product was tested for purity spectroscopically and by an atomic weight determination; it contained about 1% of thorium oxide for which no quite satisfactory method of separation has been found. G. S. Heats of Combustion of Aluminium Calcium and Magnesium. FRANK E. WESTON and HENBY R. ELLIS (Trans. Paraday Xoc. 1908 4 130-133).-Both magnesium and calcium reduce aluminium oxide on ignition with a fuse; magnesium also reduces lime.Lime and magnesia are partly reduced by aluminium and calcium respectively at high temperatures. C. H. D. Aluminium Carbide. PAUL ASKENASY W. JARHOWSHY and A. WANICZEK (Zeitsch. Blektrochem. 1908 14 81 1 - 4 1 3).-When pure crystalline aluminium carbide is heated above the melting point of aluminium in air or in an indifferent gas metallic aluminium exudes from it. In making the carbide from alumina and carbon in the electric furnace the best yield of carbide is obtained by rapidly cooling the contents of the furnace. When the cooling is slow the product consists mainly of the apparently unchanged mixture. The authors believe that the reaction Al,C Z 4A1+ 3C has a transition point; at high temperatures the carbide is the stable system at lower temperatures it decomposes into the metal and carbon.At the melting point of aluminium the velocity of decomposition is very con- siderable. These observations explain the presence of hydrogen in the gas evolved by the action of acids on aluminium carbide and also the fact that a copper-aluminium alloy is formed when it is fused with copper. T. E.INORGANIC CHEMISTRY 47 Fused Alumina in the Amorphous State and Reproduc- tion of Blue Colour of Sapphires. LOUIS PARIS (Compt. rend. 1908 147 933-935).-Pure alumina in the oxyhydrogen flame although it is only superficially fused absorbs chromic oxide through- out its mass giving artificial rubies. Oxides of cobalt iron etc. which would give a blue coloration are insoluble in pure alumina.Nickel oxide however dissolves giving a greenish-blue colour. If a small percentage of another oxide such as lime is added to the fused alumina cobalt and other colouring oxides dissolve giving a blue product the optical properties of which show it to be amorphous whereas pure alumina and artificial rubies are crystalline. Stones weighing as much as 20 carats can be obtained indistin- guishable from natural sapphires except by a slight difference in refractive index. R. J. C. Reproduction of the Blue Colour of Oriental Sapphires. AUGUSTE VERNEUIL (Compt. ?*end. 1908 147 1059-1061).-The author contends that the process described by Paris (preceding abstract) does not accurately reproduce the colour of Oriental sapphires.w. 0. w. Alum. ROBERT MARC (Zeitsch. anoyg. Chem. 1908 60 193-207). -Whilst a specimen of alum obtained from Kahlbaum had a solubility in water of about 10 grams per litre at 17" being independent of the degree of supersaturation before cooling and it readily formed super- saturated solutions a specimen obtained from Merck was less soluble had very little tendency to form supersaturated solutions and the solubility depended on the degree of supersaturation before the excess. of salt was caused to separate by inoculation; the higher the super- saturation the greater the final solubility. It was then found that the alum which separated from dilute solutions contained more alkali that is it was more basic than that separating from more concentrated solutions.The original alum of Merck however was of the usual composition as was that of Kahlbaum. No adequate explanation of these phenomena has so far been found. Of the commercial alums examined only Merck's showed the peculiarity in question. G. S. Solidihation Curves of Certain Molten Silicates. EMIL DITTLEB (Monatsh. 1908 29 1037-1074).-The freezingpoint curves for mixtures of aegirite and nephelin labradorite and nephelin and labradorite and diopside have been determined. Comparative measurements were made for mixtures of the minerals and of the artificial substances. In general the melting-point curves for mixtures of the artificial products are higher than those for the minerals. The melting points of the crystalline mixtures lie in all cases approximately on a straight line whilst in the case of the glaases the melting points are considerably lower than those of the two components.The order in which the silicates separate out is magnetite augite q i r i t e labradorite diopside nephelin and sodium augite. This succession is in agreement with Rosenbusch's rule.48 ABSTRACTS OF CHEMICAL PAPERS. The experimental data lead the author to doubt whether the solidification of complex silicate mixtures is even approximately regulated by the factors which determine the same process in the case of simple sptems. RUDOLF F. WEINLAND and P. DINHELACHER (Zeitsch. anorg. Chem. 1908 60 173-17'7).-Meyer and Best (Abstr. 1900 ii 77) have obtained double salts of manganese tetra- and tri-chloride with potassium chloride by leading hydrogen chloride into a solution of potassium permanganate in glacial acetic acid.The authors now show that these and corresponding double salts can be obtained by the action of aqueous concentrated hydrochloric acid on permanganates. The compound MnC1,,2KCl is best obtained by adding calcium permanganate and a concentrated solution of potassium chloride to a 40% aqueous solution of hydrochloric acid cooled with ice and salt ; it forms small dark red crystals. The corresponding rubidium and ammonium salts have been obtained in the same way. From the solutions from which the double salts of quadrivalent manganese have separated double salts of tervalent manganese of the type MnC13,2KC1 are obtained on adding excess of the alkali chlorides in concentrated aqueous solution.The rubidium caesium potassium and ammonium salts have been prepared; all except the rubidium salt have been described by previous observers. H. M. D. Action of Hydrochloric Acid on Permanganates. G. S. Action of Chlorine on Ferroboride and Manganese Boride at High Temperature. JOSEF HOFFMANN (Zeitsch. alzgew. Chem. 1908 21 2545-2546).-When Goldschmidt's ferroboride is heated to 300-400' in a long combustion tube in a current of dry chlorine there is produced near the heated substance a small quantity of a dark green sublimate having the composition Fe,Cl,B. The chief product however condenses farther away on the cooler parts of the tube as a brown solid and is probably a complex substance composed of boron chloride and ferrous chloride. A still more volatile compound was condensed in a U-tube connected to the combustion tube and kept in a cooling mixture; this was a pale yellow highly refractive liquid b. p.24-28' which appears to be impure boron chloride. Manganese boride heated in dry chlorine in a similar manner also gave complex manganese boro-chlorides but in such small quantities that a t present it has not been possible to ascertain their composition. J. V. E. Composition of Colloidal Ferric Hydroxy-chlorides. L~OPOLD MICHEL (Compt. rend. 1908 147 1052-lO54. Compare Malfitano Abstr. 1906 ii 450)-This paper contains details of the method adopted for studying the composition of colloidal ferric chloride solutions by filtration through collodion membranes. Experiments have been made to determine the ratio Fe Cl in the micro-cells of ferric hydroxy-chloride to which the membranes are impermeable and in the filtrate which constitutes the intercellular liquid.From theINORGANIC CHEMISTRY. 49 analytical data obtained the author arrives at the conclusion that the colloidal solution contains several physical units of variable com posi tion. I n one series of experiments carried out on an ochreous solution of ferric chloride containing a precipitate the intercellular liquid was found to be of uniform composition throughout. The micro-cells suspended in the top layer contained iron and chlorine in the atomic ratio Fe C1= 2.4 1 whilst in the lower layer containing the sediment the ratio was 4.5 1. In the stable colloidal solution obtained by adding ferric chloride to boiling water a separation of micro-cells of different composition was Complex Metal Ammonias. VIII.Transformation of Hexamminetrioldicobalt Salts into Octamminedioldicobalt Salts. ALFRED WERNER (Ber. 1908 41 3879-3884. Compare Abstr. 1908 ii 42 43).-It has been found possible to pass from a hexamminetrioldicobalt salt having the constitutional formula (I) t o an octamminedioldicobalt salt having the formula (11) that is t o replace an “01-bridge” in the former by 2 mols. of ammonia and thus obtain further confirmation of the correctness of these formuls effected by means of the centrifuge. w. 0. w. Hexamminetiioldicobalt sulphate (compare this ’voi. ii 43) is converted by cold 50% nitric acid into dinitratohexamminedioldicobalt nitrate [N0,(NE3),Co(OH),Co(NH,)3N0,](N0,)2,3H20 which crys- tallises in pale violet glistening scales.Its aqueous solutions quickly decompose ; consequently other salts of this series could not be prepared in a pure state. A freshly-prepared solution when treated with sodium dithionate slowly deposits nitratoapuohexammine- dioldicobalt dithionate ~E~o~NH~~~Co~oH~~Co~NH~~~N0,1,(as small pale violet scales. An aqueous solution of the nitrate which has been kept for some time when treated with sodium sul phate yields diaquohexa~nrninedioldicobalt sulphate an intense bluish-violet sparingly soluble salt which could not be obtained pure. The nitrate is converted by liquid ammonia and subsequent treatment with ammonium bromide into octammine- dioldicobalt bromide (compare Abstr. 1907 ii 965).Complex Metal Ammonias. IX. Decammine-p-amino- dicobalt Salts. ALFRED WERNER (Ber. 1908 41 3912-3921. Compare Abstr. 1908 ii 42).-When octammine-p-amino-01-dicobalt nitrate is treated with nitric acid in the cold nitratoaquo-octammine- [ C O ~ ( N H ~ ) ~ ( O H > ( O ~ ~ ) ~ I ( S ~ ~ ) ~ ~ 01 3H2Q W. H. G. p-ccminodicobalt nitrate [(NH,)4Co*NB,Co(N4](~O~)~ = 2 0 is ob- tained quantitatively as a violet salt. This on addition in small quantit& to well-&red liquid ammonia is converted into the decammine-p-amminodicobalt nitrate a small quantity of the octammine- VOL. XCVI. ii. 450 ABSTRACTS OF CHEMICAL PAPERS. p-amino-ol-dicobalt nitrate being regenerated. The constitution of this salt was determined independently of the octammine series as chloro- pentammine-cobalt and hexammine-cobalt salts are obtained by the action of a mixture of hydrochloric and sulphuric acids from which the deduction is made that five molecules of ammonia are attached to each of the cobalt atoms thus [(NH,),CO*NH,*CO(NH,)~]C~~ + HC1= [ c o ~ H 3 ) ~ c l ~ + [Co(NH,),]Cl,.That the decammine salt is not an imino-derivative wf the type [(XH,)5Co*NH(HCl)-Co(NH3)5]C14 is shown by its behaviour towards alkalis the elements of hydrogen chloride not being withdrawn. These salts have all a bluish-red colour and correspond with the rhodochromium salts [Cr,(:?),"JX from which the conclusion is drawn that the hydroxyl in the chromium compound plays *he same part as the p-amino-radicle in the cobalt derivatives. The nitrate Y(NO,),,(Y = [(NH,),*Co*NH,*Co(NH,),lf forms dark bluish-red needles ; the chZoi*ide YCI,,H20! precipitated from the solution of the nitrate by addition of ammonium chloride crystallises in glistening reddish-violet needles and is not decomposed on boiling with concentrated hydrochloric acid. It is only decomposed by the mixture of sulphuric acid and hydrochloric acid by long-continued heating.The bromide Y Br,,H,O forms slender reddish-violet needles ; the dithionate Y2(S20,),,6H20 forms a rose-red powder which on heating a t 60-80° loses 4H20 and is converted into the hexammine and hydroxyl pentammine salts. This was shown by treatment with aqueous ammonium bromide when the hexammine bromide was precipitated. The filtrate on addition of alcohol deposited a bluish-red powder free from sulphur which on boiling with hydrochloric acid gave chloropent ammine-cobal t chloride.The sulphate Y,( SO,) 3H,O crystallises in red needles and thin prisms. W. R. True Peroxide of Nickel. GIOVANNI PELLINI and D. MENEGHINI (Zeitsch. anorg. Chem. 1908 60 178-190).-A nickel peroxide of the formula NiO,,xH,O has been obtained as follows. A dilute solution of nickel chloride is cooled to - 50° an equal volume of 30% hydrogen peroxide added and finally potassium hydroxide in alcoholic solution. The precipitate is washed several times with cold alcohol and then with ether and forms a greyish-green powder stable in the air. The pure peroxide has not been obtained but in one case the ratio N i 0 was as high as 1 1.98. The same peroxide is obtained by the action of hydrogen peroxide on free nickel hydroxide but the reaction is comparatively slow a large proportion of the hydroxide remaining unchanged.The proportion of water associated with the peroxide has not been determined. The greyish-green peroxide gives all the reactions of hydrogen peroxide and its behaviour is quite different from that of the peroxide of the same formula described by Bellucci and CIavari (Abstr. 1905 ii 823) ; it is therefore suggested that the former oxide has the con- 0 0 0' atitution Ni< I and that described by Bellucci Ni< 0 G. s.INORGANIC CHEMISTRY. 51 Chromium Compounds. VII. Hydrates of Chromium Fluoride and an Example of Go-ordinate Isomerism among Hydrates. ALFRED WERNER and W. COSTACHESCU (Ber. 1908,41 4242-4246).-The compounds CrF CrP3,4H,0 2CrF,,7H20 and CrF3,9H,O are described in chemical literature.The authors have examined the compounds with 7 and 9H,O and the following new compounds CrF,,6H20 2CrF,,7H20 and CrF 3H,O. Hexaquochromium fluoride [ Cr(0H2),JF3 forms a violet crystalline powder and is obtained by adding a concentrated solution of potassium fluoride to a similar solution of violet chromium nitrate. Its electrical conductivity (in paraffin vessels) is about half that of the violet chloride and treatment with sulphuric acid gives rise to hexaquo- chromium sulphate. When the moist crystals are kept they yield the violet hydrate CrF,,9H20. This conversion occurs spontaneously when the crystals are washed especially when they are pressed slightly with a spatula. This hydrate yields hexaquochromium sulphate when treated with sulphuric acid and the formula [Cr( OH,),](OH2F) is suggested.~exapuochomium chromium hexafluorids [ Cr( OH,),]CrF is obtained as an insoluble green product when the hexaquochromium fluoride is warmed with water. A monohydrate [Cr(OH,),]CrF,,H,O is obtained by double decomposition between hexaquochromium chloride and ammonium chromium hexafluoride (NH4)&rF6. It forms a pale green crystalline powder with a pearly lustre and when heated at 105' loses a molecule of water yielding the insoluble green compound. The monohydrate is isomeric with Poulenc's salt (Abstr. 1893 ii 321) which is regarded as trifluorotriaquochromium monohydrate The hexahydrate loses 3H,O a t 60-70'. [crq;H2)JJ320 ; its aqueous solution has a distinctly acid reaction but does not give the reactions of fluorides. J.J. S. Higher Oxidation Products of Chromium. V. Per- chromates. ERNST H. RIESENFELD (Ber. 1908 41 3941-3951. Compare Abstr. 1908 i 963).-There are two classes of perchromates the salts derived from the penta-acid HCrO are blue in colour those derived from the octa-acid H,CrO are red. The salts of chromato- diperacid do not exist (Abstr. 1904 ii 410 737). A red hydrated potassium perchromate analogous to the sodium salt has been prepared by the action of chromic acid on potassium cyanide in the cold. This salt crystallises in rhombic plates and very quickly loses its water of crystallisation. The constitution of these salts has been investigated by determining their molecular weights. C,H,N,HCr05 is unimolecular and from determinations of the freezing points in aqueous solution the red potassium and ammonium perchromates have also the simple formula M,CrO,.The solutions are good electrolytes. Determination of the mo1.-wt. of the free acid is not possible because of its rapid decomposition in water; this decom- The blue pyridine perchromate 4-252 ABSTRACTS OF CHEMICAL PAPERS. position has been measured in a mixture of ether glacial acetic acid and water and obeys the unimolecular law As this argument for the simple formula in the case of the free acid is not conclusive the electrochemical equivalent of the chromium was determined and shown t o be 1Cr:GAg in the case of the red potassium salt. It is not possible to assign a formula to this salt on the assumption of the univalency of potassium the bivalency of oxygen and the sexavalency of chromium.Although hydrogen peroxide cannot be proved t o be a product of the decomposition of these salts yet alkaline gold chloride and permanganate are reduced during the change and the assumption is made that these salts are analagous to alkali peroxides. The equivalent of permanganate reduced is held to be a measure of the '' holoxide " oxygen groups. The blue pyridine perchromate decomposed very rapidly and so this could not be measured but in the decomposition of the red salts 5.5 equivalents of the permanganate were reduced. The conclusion is drawn that t h i s is due to the formation of 3H,02 and that therefore the constitutions of these acids are /O*OH 0 ECr-O*OH. \O*OH /O*OH Red.W. R. 021c'N0 Blue. Derivatives of Complex Inorganic Acids Alumino- tungststes and Aluminophosphotungstates. LLOYD C. DANIELS (J. Arner. Chem. Xoc. 1908 30 1846-1857).-Balke and Smith (Abstr. 1904 ii 179) have described the aluminotungstates 3(NH4),0,A1,03,9 W0,,4H20 and 1 1 A g,O,:! 1 (N H4!,0,4A1203,36 WO,. The present investigation was carried out with the object of ascertain- ing whether the ratio A1,0 9W0 would exist in other derivatives. The following salts have been prepared and described Copper ccluminotungstate 2Cu0 A1,0,,9W03 16$H,O ; the barium salt 8B;z0,A1203,9 W0,,7H20 ; the mercurous salt 5Hg20,A1,0,,9W0 and the zinc salts 1&Zn0,h1,03,9 WO3,8H,0 and ZnO,A1,O3,9 WO 20H,O. 9(NH,),0,2A1,0,,4P,05 9MT0,,13H,0 ; the silver salt 4Ag,0,2Al,0,,4P20,,9W0,,6H,0 ; the barium salt 4BaO,2Al20,,4P2O5,9 W0,,3K2O and the zinc salt 5Zn0,2A1,0,,4P2O5,9WO3,1 1H,O are also described.The ammonium salt 6(NH4)20,2A1203,3As,05,18W0,,14H,0 ; the barium salt 4BaO,2A1,0,,3A~,C)~ 18 W0,,12H20 and the cadmium salt 4Cd0,2A1,0,,3As205 1 S WO3,17H,O. A description is also given of the following aluminoantimonio- tungstates. The ammonium salt 6(NH4),0,2A1,03,3Sb205,18W03 1 7H20 ; the silver salt 6Ag,0,2A120,,3Sb205,18W03,1 2H,O and the barium salt 5BaO,2Al20 3s b,0,,1 S W0,,6H ,O. Ammonium aZurnino;uhosiohotungstate The following aluminoarsenotungstates have been prepared. E. G.INORGANIC CHEMISTRY. 53 Comppmds of Silicon and Uranium. Uranium Disilicide Usi,. EDOUARD DEPACQZ (Compt. rend.1908 14'7 1050-1052." Compare Abstr. 1907 ii 475 696)-The silicides of tungsten and molybdenum of the type MSi having been prepared previously the author completes the series by describing the preparation and properties of the corresponding uranium compound. Uranium disilicide USi obtained in small yield from silica and the oxide U,O by the aluminothermic method occurs as brilliant microscopic crystals DO 8 belonging to the regular system. Chlorine attacks this silicide at 500° giving the chlorides of uranium and silicon. Cold hydrogen fluoride dissolves it readily whilst oxidising agents have little action. Although stable in air at a red heat it burns slowly in oxygen a t 800° and is decomposed on fusion with alkalis alkali carbonates or potassium hydrogen sulphate.Details are given of the method adopted for the analysis of this Thorium. WERNER VON BOLTON (Zeitsch. Elektrochem. 1908 14 768-770).-The metal is prepared in minute crystals by reduction of the tetrachloride by sodium. By hammering the metal into a thick- walled copper tube 10 mm. in diameter rolling this down to 1 mm. wire and removing the copper with dilute nitric acid a spongy wire of thorium is obtained which can be rolled out to a band. The melting point is 1450' ; it burns in the air ; potassium hydroxide and nitric acid have little action ; sulphuric acid slowly attacks it yielding the sulphate and hydrochloric acid acts quickly leaving about 15% of the metal undissolved in the form of lower oxides which are possibly formed during the action. The density of the powder is 11.32 that of the rolled band 12.16.When the powder is shaken with cold 5% hydrochloric acid the first portion which dissolves (about 10% of the whole metal) forms a complex acid (ThO,),o(HC1),,. the solution of which contains neither thorium nor chlorine ions; its principal reactions are with oxalic acid or sulphuric acid a small precipitate soluble in excess is pro- duced ; ammonium oxalate gives a gelatinous precipitate insoluble on boiling ; ammonium carbonate and citric acid give gelatinous precipi- tates insoluble in excess. The metal reduced from this compound is ordinary thorium. T. E. RICHARD J. MEYER (Zeitsch. EZekts.ochem. 1908 14 809-8 lo).-The reduction of thorium oxide by carbon tetrachloride yields a metal containing considerable quantities of oxide (Abstr.1901 ii 106). The oxide left by Bolton's thorium when dissolved in acid (preceding abstract) was therefore doubtless present in the metal. This also explains the action of dilute hydrocbloric acid ; the dissolved substance was probably metathorium chloride derived from the oxide. Chlorides and Oxychloridels of Thorium. EDOUARD CHAUVENET (Compt. rend. 1908,147,1046-1048. CompareMatignon Abstr. 1901 ii 106).-Thorium tetrachloride is most conveniently prepared free * and Bull. Soc. chim. 1909 [ iv] 5 3-5. compound. w. 0. w. Thorium. T. E.54 ABSTRACTS OF CHEMICAL PAPERS from oxychloride by heating thorium dioxide in a current of carbonyl chloride. A deliquescent white powder which is formed at the same time has been proved t o be identical with the crystalline chloride by analysis and by measurement of its heat of solution in water (56.75 Ca!.at 12.6O). An aqueous solution of the tetrachloride deposits ThC14,7H,0 on evaporation (compare Kruss Abstr. 1897 ii 456). The heat of dissolution of this hydrate is 14.75 Cal. at 13.5'. When heated at 120-160' in hydrogen chloride the oxychloride OH*ThC13,H20 is obtained as a very hygroscopic powder of which the heat of dissolu- tion is 47.63 Gal. at 13.5" (compare Rosenheim Abstr. 1900 ii 351). The chloride ThCI 7H,O is converted into the oxychloride ThOCI at 250'; the heat of dissolution of the latter is 28.15 Cal. at 13'. w. 0. w. Basic Sulphates of Thorium and Cerium. OTTO HAUSER and F. WIRTH (Zeitsch. anorg. Clhem. 1908 60 242-246).-Dilute solu- tions of thorium sulphate deposit flocculent thorium hydroxide con- taining some adsorbed sulphate on boiling ; thus resembling zirconium sulphate (Abstr.190'7 ii 626). Concentrated solutions heated in sealed tubes at 160-180° deposit crystals of ThOSO,,H,O. A t 105-115O the same salt is deposited mixed with the normal hydrated sulphate. A dilute ice-cold solution of ceric sulphate deposits after twenty- four hours pale yellow crystals of 4Ce0,,3S0,,12H20 completely analogous t o the basic zirconium sulphate. Boiling saturated solu- tions of ceric sulphate deposit crystals of 2Ce0,,3S0,,4H20 which however could not be obtained in a pure state since they rapidly redissblve on cooling and some reduction to cerous sulphate takes place. C. H. D. Derivatives of Complex Inorganic Acids Phosphovanadio- molybdates.WILLIAM BLUM (J. Amer. Chem. SOC. 1908 30 1858-1862).-Wolcott Gibbs (Abstr. 1884 713) described two phosphovanadiomolybdates 8( NH,),O 2P20,,8V2O5 1 4Mo03,50H20 and 7(NH4)20,2P205,V,0,,48Mo03,30H,0 and suggested that it was probable that the vanadium pentoxide considered as V,0,*03 might replace the molybdenum trioxide of the phosphomolybdate. He was unable however to obtain a salt containing a higher proportion of vanadium pentoxide than that in the first compound mentioned. The present author is of opinion that this was probably due to the fact that an excess of ammonium vanadate was used in all the experiments. The formation and composition of the ammonium phosphovanadio- molybdates has now been further studied and the following corn- pounds have been obtained 6(NH4),0,P,0,,4V,05,1 8Mo03,52H,0 and the corresponding silver salt ; 5(NH,),0,P205 2&V,05,21 &MoO3,50H,O ; 8(NH,),O,P,O5,5V,O5 1 8M00,,45H20 ; 7(NH,),0,P,05,5~V,05,1 64Mo0,,50R2O ; 8( NH,),O,P,O W,05,1 5Mo03,50H,0,INORGANIC CHEMISTRY.55 and the corresponding silver salt ; 8(NH,)20,P205,8V20 1 4M00,,50H20 ; 8( NH,) 20,P,05 1 O& v,o 1 1 +Moo ,5 OH@ ; and 8(NH,),0,P205,1 lV205,1 1 MoO,,!XH,O. All these ammonium salts crystallise in the tetragonal system and are soluble in water. E. G. Action of Antimony Trichloride on Cobalt and on its Alloys with Antimony. F. DUCELLIEZ (Compt. vend. 1908 147 1048-1050. Compare Abstr. 1908 ii 853).-When cobalt is submitted to the action of antimony trichloride at '700-1200° cobalt monoantirnonide CoSb is formed as a crystalline non-magnetic powder Do 8.12 m.p. about 1200'. This undergoes slight oxidation in air and burns readily in oxygen ; hydrogen chloride has little action on it but hot concentrated sulphuric acid dissolves it rapidly. At 700-1 450" it attacks antimony trichloride forming antimony and cobalt chloride ; the composit'ion of the residual substance however remains unaltered. At 1200' cobalt is attacked by antimony tri- chloride and forms magnetic alloys containing less antimony than the foregoing antimonide ; this compound remains as a non-magnetic powder when these alloys are treated with sulphuric acid. Three classes of alloys of cobalt and antimony have been prepared by direct union of the elements at 500" in a current of hydrogen.(1) Magnetic alloys containing less than 67.04% antimony. When heated with antimony trichloride at SOO" or when treated with sulphuric acid these lase their magnetic properties and leave a residue of the monoantimonide CoSb. (2) Alloys containing 67.O4-8Oa27% of antimony which decompose readily on heating and form the monoantimonide when heated at 1200' in hydrogen. Substitution of antimony trichloride for hydrogen gives the same compound together with antimony and cobalt chloride. (3) Allqys containing more than 80027% antimony which on treatment with nitric acid followed by hydrogen chloride give cobult diantimonide CoSb a grey crystalline powder Do 7.76 m. p. about 700'; this closely resembles the monoantimonide in its chemical properties.w. 0. w. Atomic Weight of Bismuth. 111. Analysis of Bismuth Bromide. ALEXANDER GUTBIER and HANS MEHLER (J. pr. Chem. 1908 [ii] 78 409-420. Compare Abstr. 1906 ii 92 ; 1908 ii -600).-The bismuth bromide employed in this investigation WRS pre- pared by the action of bromine on four specimens of bismuth. The percentage of bromine present in each of the four samples of bromide was! determined by dissolving the ,latter in 2N-nitric acid and pre- cipitating with silver nitrate. As the result of eight experiments it was found that 34.81207 grams of bismuth bromide gave 43.80773 grams of silver bromide. The extreme values obtained were Bi= 208.18 and Bi = 207.88. The value 208.0 +_ 0.06 (0 = 16) for the atomic weight of bismuth determined by this method is in agreement with the result (Bi = 208.0) obtained by Gutbier and Birckenbach (Abstr.1908 ii 600) by the synthesis of bismuth oxide. W. H. G.56 ABSTRACTS OF CHEMICAL PAPERS. Atomic Weight of Bismuth. IV. Synthesis of BiBmuth Sulphate. ALEXANDER GUTBIER and RUDOLF L ~ O N JANSSEN (J. pr. Chem. 1908 [ii] 78 421-436. Compare preceding abstract).-The atomic weight of bismuth was determined in this investigation as follows a known weight of the element was dissolved in nitric acid (D 1*05) the nitrate obtained evaporated with sulphuric acid and the sulphate heated to constant weight in an electric-oven at 380". Two specimens of bismuth were used ; three experiments with one sample gave the mean value Bi= 208.04 whilst two experiments with the second sample gave Bi = 208.1'7. The extreme values obtained were Bi = 208.26 and 207.92. The atomic weight of bismuth determined by this method is 208.1 (0 = 16). From the results of the investigations of Marignac (Abstr. 1884 813) Lowe (Abstr. 1884 558) Schneider (Sbstr. 1895 ii 114) and of Gutbier and his co-workers it follows that bismuth has the atomic weight 208.0 (0 = 16). Carbonatels and Oxalates of Bismuth. LUDWIG VANINO and EMILIE ZUMBUSCH (Bey. 1908 41 3994-3999).-Attempts t o pre- pare normal bismuth carbonate and oxslate show that these substances are extremely sensitive to the hydrolytic action of water. Only by working under strictly defined conditions can substances of definite composition be obtained. The normal carbonate has not been isolated but the oxalate is obtained when a solution of bismuth nitrate and mannitol in equal molecular quantities (0.1 gram-mol. per 100 c.c.) is treated a t 25" with the calculated quantity of a saturated solution of oxalic acid and the precipitate washed with cold water. I f the mashing is performed with hot water the precipitate after dry- W. H. G. ing has the composition OH*Bi:C,O,. c. s.