80 ABSTkACTS OF CHEMICAL PAPERS. Inorganic Chemistry. Behaviour of the Halogens towards each other. H. W. BAKHUIS ROOZEBOOM (Proc. K. Akad. Wetensch Amsterdam 1906 9 363-364).-1n reviewing the studies that have been made from the pointl of view of the phase rule of the binary systems I + C1 1-1- Br Br+CI the author points out that IC1 is a very stable com- pound ICI and IBr are feeble compounds whilst no compound of bromine and chlorine exists. The combining power is therefore greatest in the most distant elements and greater for the pair Br + I than for the pair Br + C1. J. C. P. Density of Gaseous Hydrogen Chloride Atomic Weight of Chlorine. PHILIPPFI A. GUYX and G. TER-GAZARIAN (Compt. rend. 1906 143 1233-1235).-The recent values obtained for the atomic weight of chlorine are (for Ag = 107.93) 35.473 from the ratio Ag C1 (Richard and Wells Abstr. 1905 ii 450) ; 35.476 from the density of hydrogen chloride (Leduc corrected by Guye Abstr.1905 ii 442) ; 35.460 (Ag = 107-89) from the ratio Ag C1 (Guye and Ter-Gazarian Abstr. 1906 ii 760); and 35.463 from the synthesis of hydrogen chloride (Dixon and Edgar Abstr. 1905 ii 696). I n view of these discrepancies the authors have redetermined the density of hydrogen chloride using the method employed in the determination of the density of nikric oxide (Abstr. 1906 ii ZO) the gas being prepared by the action of sulphuric acid on sodium chloride dried over sulphuric acid and phosphoric oxide liquefied at the temperature of liquid air and submitted t o a series of fractional distillations. The mean of four determinations of the weight of a litre of hydrogen chloride at N.P.T.is 1.6398 grams (Leduc obtained 1*6407) and the corresponding atomic weight of chlorine is 35.461 which is in close agreement with the last two values quoted above but the authors regard i t as provisional only. M. A. W. New Hydrogen Sulphide Apparatus. FERNAND RANWEZ (Ann. Chim. anal. 1907 12 7-9).-The apparatus consists essentially of a broad U-shaped tube constricted at the bend and also a t a little above the bend as regards the limb which contains the iron sulphide. The top of this limb is closed with aperforsted cork through which passes a rectangularly bent tube connected with the delivery tube by means of a small rubber tube which can be closed with a pinchcock.Acid is poured into the other limb and when the rubber tube is not closed the acid comes in contact with the iron sulphide and hydrogen sulphide isINORGANIC CHEMISTRY. 81 evolved. On closing the rubber tube the acid is at once driven back Decomposition of Persulphates. MARIO G. LEVI and E. MIGLIORINI (Gazxelta 1906 36 ii 599-619. Compare Abstr. 1903 ii 474).-The reaction of decomposition of persulphates obeys the LAW of unimolecular reactions in the cases of the sodium and potassium salts but not in the case of the ammonium salt. U p t o a temperature of 30-35’ solutions of persulphates can be kept un- changed for some days the solution of the ammonium salt being most liable to change. Acids and t o a less degree alkalies exert a marked accelerating action on the decomposition of persulphates.Platinum black causes slow catalysis of persulphates in solution the greatest action taking place with ammonium persulphate ; the amount of cata- lysis is a maximum in alkaline and a minimum in acid solution. Lead and certain other metals also decompose persulphates in solution. into the other limb. L. DE I(. T. H. P. Selenates. ENRICO RIMINI and GIOVANNI MALAGNINI (Atti R. Accad. Lincei 1906 [v] 15 ii 561-563j.-The acid selenate of hydr- axine N2H4,H2Se04 is only slightly stable asare also its solutions ir water in which i t dissolves moderately readily. The double hydraxine copper selenate [(N2K,)2,H2Se04,CuSe0,]2 + H,O is isomorphous with the corresponding double sulphate and decomposes gradually when kept. T. H. P. Experiments with the Hot-cold Tube in the Electric Furnace.RAFFAELO NASINI and FRANCESCO ANDERLINI (Gazxetta 1906 36 ii 570-575).-No combination takes place between nitrogen and oxygen when heated in the hot-cold tube in the electric furnace at 2000°. nor does magnesium absorb argon appreciably a t this temperature. T. H. P. Copper Cathodes in Nitric Acid. J. W. ‘CURREXTINE (J. Physical Chem. 1906 10 715-720).-The object of the author’s experiments was to ascertain why nitric acid is reduced t o ammonia when dilute solutions of the acid are electrolysed whereas nitric oxide is the chief product in the ordinary contact action of copper on nitric acid. On electrolysing a 5% nitric acid solution between copper electrodes at a current density of S amperes per square decimetre it was observed t h a t evolution of gas (nitric oxide) did not commence until the more dense copper nitrate solution which collected in the lower part of the apparatus came into contact with the cathode.This result appears to show that the copper ions are the cause of the incomplete reduction of the nitric acid and in a further experiment in which copper was allowed t o act on 10% nitric acid solution means were devised for the removal of the copper ions as fast as these were formed. This removal of copper ions was effected electrolytically and under such conditions it was found that the reduction product of the chemical action between copper and nitric acid is ammonia. The more rapid rate of solution when nitric oxide is the reduction product is attributable t o the accelerating action of nitrous acid.H. M. D,82 ABSTRACTS OB CHEMICAL PAPERS. Preparation of Boron Sulphide from Manganese Boride. JOSEF HOFFMANN (Zeit. angew. Chem. 1906 19 2133-2134. Corn- pare Abstr. 1906 ii 745).-Manganese boride is recommended instead of ferroboron for the preparation of boron sulphide. The substance is heated in a current of hydrogen sulphide a t the melting point of antimony when the boron sulphide deposits in the cold part of the tube as a crystalline mass. It is rapidly decomposed by exposure to moist air. L. DE K. Production of Carbon from Carbides. C. HAHN and ANTON STRUTZ (Metallurgie 1906 3 727-732).-1n connexion with the question as to the possible formation of petroleum by the action of water vapour on mineral carbides the authors have investigated the action of various dry gases on carbides.Heated calcium carbide is decomposed by dry steam hydrogen chloride or hydrogen sulphide yielding amorphous carbon and the corresponding metallic compound. Acetylene is not formed unless moist steam is present. The same result is obtained with aluminium carbide. Manganese carbide on the other hand yields a glistening form of carbon resembling graphite when heated in dry hydrogen chloride. Carborundum reacts with steam at 1 300-1400° yielding silica and carbon in such an intimate state of mixture that it is not possible to determine whether the carbon is amorphous or crystalline. C. H. D. Purification of Crystalline Silicon and some Results of the Presence of Impurities in the Preparation of Copper Silicide.EMILE VICIOUROUX (Bull. soc. chim. 1907 [ivl 1 16-19 Compare this vol. ii 89).-Crgstalline silicon prepared by heating potassium silicofiuoride with aluminium in a clay crucible and purifying in the usual way by treatment first with dilute hydrochloric acid and then with dilute hydrofluoric acid contains iron which reveals itself by the formation of iron silicide FeSi when such silicon is used in making copper silicide (Abstr. 1906 ii 168 and this vol. ii 89). The author suggests that crystalline silicon prepared as above should be purified by digesting i t in ft finely-powdered state for some hours with hydrofluoric acid diluted with its own volume of water the operation being conducted in a platinum retort provided with a reflux condenser the residue resulting from this treatment is then washed and heated during one o r two hours in a platinum crucible with sulphuric acid. This treatment should be repeated until a sample of the material is entirely dissipated by treatment with a mixture of nitric and hydrofluoric acids the acids leaving no residue on evapora- tion.When crude crystalline silicon is heated in R leaden dish with hydrofluoric,acid the products evolved include a gas burning with a bluo flame which deposits on a cold surface a yellow substance. The same gas when passed through a heated glass tube furnishes a t the heated portion a similar yellow deposit. T. A. H.INORGANIC CHEMISTRY. 83 Action of Hydrogen on Silicon and Silica. A. DUFOUR (Ann. Chim. Phys. 1906 [viii] 9 433-474).-A detailed account of work already published (Abstr.1904 ii 398 482). The Anhydrous Oxides of the Alkali Metals. ~ T I E N N E RENGADE (Compt. rend. 1906 143 1152-1 153)-The oxides of rubidium potassium and sodium can be obtained in a state of purity by means of the method employed in the preparation of czesium oxide namely partially oxidising the metal and separating the unchanged metal by prolonged distillation in R vacuum (AbEtr. 1906 ii 850) ; rubidium oxide Rb,O forms small crystals which are golden yellow when hot and become pale yellow on cooling ; potassium oxide K,O is a greyish- white confusedly crystalline mass whilst sodium oxide Na,O is white and amorphous. It is probable t h a t in this method of prepara- tion a lower oxide of the metal is first formed which on heating decomposes into the oxide and metal (compare Forcrand Abstr.1899 ii 95). M. A. W. M. A. W. Optical Characters of Isomorphous Crystals. GEORG W ULFF (Zeit. Kryst. Mim 1907 42 558-686).-Determinations of the optical constants of mixed crystals of potassium and caesium sul- phatus of potassium and ammonium sulphates and of rubidium and ammonium sulphates prove that there is no linear relation between the refractive indices of the mixed crystals and those of the end members of the series. L. J. S. Conductivities of Mixtures of Sulphuric Acid with Sul- phates. Formation of Complex Salts of Hydrogen. AUGUSTE HOLLARD (Bull. Xoc. chim. 1906 [iii] 35 1240-1255).-When a 2% solution of sodium sulphate is added to a 6% solution of sulphuric acid in water there is a diminution in electrical conductivity.When the addition is mado to a 2% solution of sulphuric acid in water the con- ductivity no longer diminishes but increases slightly although it is still inferior to the sum of the conductivities of the components of the solution. On the other hand if the addition is made to a 3% solution of sulphuric acid the conductivity remains constant whatever may be the quantity of sodium sulphate added. The conductivity also remains constant when a solution of magnesium zinc or cupric eulphate is added to a 3% solution of sulphuric acid but with ammonium sulphate constancy is only exhibited by an 8% solution of sulphuric acid. These phenomena are due to the disappearance from the sulphuric acid solu- tion on the addition of solutions of the salts named of hydrogen ions which combine to form the complex ions HSO,.The measurements on which these results are based and curves illus- trating them are given in detail in the original. Action of Alkali Silicates on Soluble Metallic Salts. ROBERT DOLLFUS (Compt. rend. 1906 143 1148-1149).-When small crystals of metallic salts such as ferrous copper' or nickel sulphate manganese chloride uranium or cobalt nitrate are thrown into a solu- tion of sodium or potassium silicate a semipermeable membrane of the T. A. H.84 ABSTRACTS OF CHEMICAL PAPERS. metallic silicate is formed round the crystal and owing to h e internal osmotic pressure t h e low density of the metallic salt solution and the air bubbles mechanically adhering to the crystal the '' artificial cell " increases in size and presents the appearance of a vegetable growth. M.A. W. Allotropic Silver and its Colours. F. E. GALLAGHER (J. Physicul Chem. 1906 10 701-714).-The different colorirs of solu- tions of colloidal silver cannot be explained as an effect dire to the diffetence i n the refractive index of the colloidal particles and t h a t of the surrounding medium. The author's experiments on the other hand indicate that thephenomenon is probably due to selective light absorption depending on the thickness of the layer of solution the degree of coagulation and the concentration of the particles in the solution. Light is essential to the changes of colour the rate of change de- pending on the intensity of the light. It is supposed to act by coagu- lating the colloidal suspension or by altering the concentration of the particles by increasing the reduction of the silver in solution.With increasing thickness of the absorbing layer or with increasing size of the particles the amount of blue light transmitted decreases relatively much more rapidly than the amount of transmitbed red light. Ethyl alcohol acetone benzene and other organic liquids prevent the colour changes from taking place in the light. The action appears t o consist in the prevention of coagulation. H. M. D. Distillation of the Alloys of Silver with Copper with Tin and with Lead. HENRI MOISSAN and TOSIO WATANABE (Compt. rend. 1907 144 16-19. Compare Abstr. 1904 ii 617).-About 40 grams of each alloy of silver with copper tin and lead was distilled in the tube form of electric furnace and the residual alloy examined at short intervals. After a fifteen minutes' distillation of an alloy of equal weights of silver and copper the residue contained 95.22% of copper and 3.62% of silver.Distillation of an alloy of the composition 36.98% of silver and 64.04% of tin left an alloy containing 2.83% of silver and 93.65% of tin after ten minutes. Similarly with an alloy of 46*24%of silver and 52.95% of lead an alloy containing 96.81% of silver remained after two and a half minutes. The conclusion is drawn that lead is morevolatile than silver and silver more volatile than copper or tin. The results agree with those obtained by Kraflt ( Abstr. 1896 ii 464 635) who used very small quantities of alloy and dis- tilled in the cathodic vacuum. E.H. Reversible' Action of Oxygen on Magnesium Chloride. FPITZ HABER and F. FLEISCHMANN (Zeit. anorg. Chem. 1906 51 336-347).-The equilibrium constant of the reaction in question has been determined at different temperatures and a formula representing the dispIacement of the equilibrium with ternparatime is suggested. I n carrying out the experiments dry chlorine and oxygen i n varying proportions mixed with nitrogen were led slowly over a mixture of magnesium olride and chloride contained is a porcelain tube heated inINORGANIC CHEMISTRY. 85 an electric furnace the gaseous mixture being analysed on entering and on leaving the tube. The temperatures were such that oxy- chloride was not formed the reaction investigated being represented by the equation MgCl + $02 The equilibrium constant correqponding with this equation is Jcp = p ~ 1 ~ / p o ~ * ; it is the ratio of the partial pressures of the two gases.The most satisfactory results were obtained at 675' and 586O the mean value of kp at the former temperature being 0.25 and at the latter temperature 0.18. From these results by means of van't Hoff's equation connecting the displacement of equilibrium with temperature with & the heat developed in the reaction the value of Q a t the mean temperature of the experiments 630° is calculated as -5736 cal. whilst Berthelot gives Q = - 7800 cal. a t the ordinary temperature. It is known that Q in this case decreases with rise of temperature and on approximate assumptions as to the variation of the specific heatsof the gases with temperature it is calculated t h a t at 630' the value of Q would be - 5540 cal. in satisfactory agreement with the value given above.I n conclusion a formula is given which represents approximately the displacement of the equilibrium with temperature but this problem cannot be solved satisfactorily until the variation of the specific heats of the reacting substances with temperature is known more aecurately. G. S. MgO + Cl,. Action of Oxygen and Water Vapour on Magnesium Chloride. WILHELM MOLDENHAUER (Zeit. anorg. Chem. 1906 57 369-390).-l!he experiments described in the first part of this paper were made quite independently of those of Haber (compare preceding abstract) and have led to practically the same conclusions.MgO + C1 were carried out in a large porcelain tube into which the magnesium chloride was introduced in concentrated solution and then rendered anhydrous by heating in a current of hydrogen chloride. The magnesium oxide for the reverse reaction was introduced in a similar way. The tube was then filled with oxygen or chlorine as the case might be kept for some hours at a high temperature and the resulting mixture of gases analysed. The equilibrium was reached from both sides at 550° 650' and 700' the values of k=Cc,,/ ,/C< where CclZ and CO represent the respective concentrations of chlorine and oxygen being 17.4 25.3 and 29.5 respectively. Prom these results Q the heat evolved in the reaction at 600° is calculated as -5600 cals. i n good agreement with the calorimetric value when the variation of Q with temperature is taken into account.The action of water vapour on magnesium chloride was investigated by a somewhat similar method. In the first place by heating the chloride with aqueous vapour and hydrogen chloride until equilibrium was established and subsequently analysing the residue it was shown that from 350-505' there is an equilibrium represented by the equation MgCI + HzO MgCl-OH + HCl. At 505-510' the oxychloride is decomposed and above this point equilibrium is established according to the equation MgCI + HzO MgO + 2HC1. The experiments on the equilibrium MgCI + 40,86 ABSTRACTS OF CHEMICAL PAPERS. When equilibrium is attained at 350° 400° 500° 600° and 700° the gaseous phase contains 69.4 62.1 50.2 75.1 and 90.2 volumes of hydrogen chloride respectively.As the proportion of hydrogen chloride thus decreases with the temperature up to the point at which the oxychloride decomposes and beyond that point increases it follows that in the first reaction heat is developed-in other words half the hydrochloric acid is split off from magnesium chloride exothermically whereas the reaction MgC1 + H,O = MgO + ZHC1 is endothermic. The heat developed in the latter reaction calculated from the displace- ment of the equilibrium with temperature is in moderate agreement with the value obtained directly. I n addition to the hydrates of magnesium chloride described by van 't Hoff and Meyerhoffer a seventh hydrate MgCI,,H,O has been obtained by dehydrating the tetrahydrate in a current of hydrogen chloride above 11 1'.The technical bearing of these results is also considered. G. S. Condition8 of Precipitation of Metallic Sulphides. MAURICE PADOA and L. CAMBI (Atti R. Accad. Lincei 1906 [ v] 15 ii 787-7'96. Compare Bruni and Padoa Abstr. 1906 ii 157 ; MacLauclan Abstr. 1903 ii 716).-The authors have determined for solutions of different salts of various concentrations and containing varying proportions of acid the minimum and maximum pressures of hydrogen sulphide between which precipitation of the sulphide of the metal of the salt takes place. The salts examined were (1) cadmium chloride in presence of hydrochloric acid ; (2) zinc sulphate with sulphuric acid ; (3) zinc chloride with hydrochloric acid ; and (4) ferrous sulphate with sulphuric acid.I n the first three cases the pressures employed varied from 15 to 760 mm. and in the last from 200 mm. to 9.2 atmo- spheres. The pressures increase with the acidity of the solution. The two forms of apparatus used one for pressures up to that of the atmosphere and the other for higher pressure are described. T. H. P. Colloidal Plumbic Acid. ITALO BELLUCCI and NICOLA PARRAVANO (Aai R. Accnd. Lincei 1906 [v] 15 ii 542-549).-When potassium plumbate as free as possible from adherent alkali is treated at the ordinary temperature with water it undergoes complete hydrolysis into PbO,,Aq + ZKHO. The plumbic acid remains in solution in the colloidal state since cryoscopic measurements of the liquid indicate for the disaolved substances a molecular weight corresponding with KH0/2.The plumbic acid does not pass through a parchment membrane and the solution which appears turbid in reflected light can be filtered diluted with water heated to boiling and allowed to solidify without in any way changing the plumbic acid. The solution can be evaporated on the water-bath to a syrupy consistency without coagulating and the residual gelatinous mass provided it is not dried is soluble in water,. giving the hydrosol. I n this case the impurity which is essential for the maintenance of the plumbic acid in the condition of hydrosol the so-called " Solbildner," is potassium hydrozide the proportion of which is gradually diminishedINORGANIC CHEMISTRY. 87 by continued dialysis until after 104 hours the percentage of K,O is 1.87 and that of PbO 98.13.This proportion of K,O is of the same order as the amounts of hydrochloric acid essential for the existence of hydrosols of aluminium chromium and ferric hydroxides. T. H. P. Properties of the Plumbic Hydrosol. ITALO BELLUCCI and NICOLA PARRAVANO (Atti R. Accad. Lincei 1906 [v] 15 ii 631-635. Compare preceding abstract).-Two neutral colloidal solutions of plumbic acid containing (1) 0.3174 gram PbO and 0*0081 gram K20 and (2) 0,1288 gram PbO and 0.0330 gram K,O per 100 C.C. were found to cause no depression of the freezing point of water although other solutions containing varying proportions of alkali gave small depressions. The coagulating actions of the anions (potassium salts) decrease according to the series (I) iodide ferrocyanide tartrate fluoride ; (2) acetate chromate iodate nitrate sulphate thiocyanate chloride bromide dichromate pernianganate chlorate perchlorate ferricyanide ; (3) periodate oxalate carbonate arsenate and those of the cations {chlorides) according to the Eeries (1) aluminium ferric calcium strontium barium magnesium cadmium mercury manganese nickel cobalt copper ; (2) ammonium caesium lithium potassium rubidium sodium The anions and cations of series (1) produce total coagulation and those of series (2) partial coagulation whilst the anions of class (3) are without coagulating action.On increasing the volume of the coagulating solution used it was found that the coagulating action was almays increased but not to the same extent for all electrolytes. Solubility and Oxidation Potential of Lead Disulphate and Dioxide.FRIEDRICH DOLEZALEK and KARL FINCKH (Zeit. anorg. Chem. 1906 51 320-327. Compare Abstr. 1906 ii 597).-The solubility of plumbic sulphate in solutions of sulphuric acid of different strengths has been determined directly at 22". The concentration of the acid is conveniently expressed as the number of mols. of acid present per mol. of water. The curve representing the solubility rises rapidly from zero a t 0.3 mol. t o a maximum a t 0.6 mol. of acid slowly falls to a minimum value at 1 mol. of acid and beyond that point gradually rises. These results are in complete accord with the observations in a previous paper that below a concentration of 0.3 mol. of acid a higher sulphate cannot exist being hydrolysed to the dioxide that between 0.3 and 0.6 mol.of acid the solution is in equilibrium with the basic sulphate PbOSO,,H,O and beyond the latter point with the normal sulphate Pb(SO,),. The solubility of the former is increased that of the latter at first diminished by increasing acid concentration so that the maximum solubility coincides with the transition point. The gradual increase of solubility beyond a concentration of 1 mol. of acid is due in all probability to formation of complexes. Lead dioxide prepared by hydrolysis of the salts of quadrivaleat lead has approximately the salqe solubility in concentrated sulphuric T. H. P.88 ABSTRACTS OF CHEMICAL PAPERS. acid as the disulphate and has therefore been changed to the latter. The dioxide prepared by electrolysis or by chemical oxidation of lead salts is much less soluble.Contrary to the observations of Elbs and Rixon (Abstr. 1903 ii 427) and in agreement with these results lead dioxide is very slightly soluble in an acid solution containing less than 0-3 mol. of the latter per mol. of water. By means of a formula connecting the change of solubility with the acid concentration the solubility of the peroxide a t lower acid concentrations hss been calculated; it falls off very rapidly with dilution of the acid. The oxidation potential of the disulphate with different concentra- tions of sulphuric acid has been measured at 11.5' against a hydrogen electrode; the E.M.F. attains a maximum value of 1.9 volts at a concentration of acid corresponding with the maximum solubility of the disulphate.A formula has also been deduced showing the strength of acid which produces the most powerful oxidising effect at different temperatures. G. 8. Bismuth-thallium Alloys. MASUMI CHIKASHIG~ (Zeit. anorg. Chem. 1906 51 328-335. Compare Heycock and Neville Trans. 1892 61 S88 ; 1894 65 31).-As a result of his investigation of this system by Tammann's method of thermal analysis the author draws the conclusion that the two metals form a compound Biil?13 whilst there are indications of the formation of a second compound BiTl under certain conditions. The freezing point curve of the system shows three maxima at 212' and 37% 303.5' and 88-7% and 302' and 99.2% by weight of thallium respectively the former of which corresponds with a compound Bi,Tl,; and there are three eutectic points at 197' and 20% 186" and 53% and 297' and 93% by weight of thallium respectively.The first maximum and the second eutectic point were not observed by Heycock and Neville. T 10 compound Bi,TI forms a series of mixed crystals from 35-5-35 5% of thallium; another serics extends from 66*3-88*75% of thallium. The cooling curve of alloys from 60-76% of thallium bhows a transition point at 90° which may indicate that at that temperature the mixed crystals last mentioned react to fozm a compound BiTl,. As regards the maxima at 88.7' and 99.2% of thallium the alloys in this region behave as mixed crystals and not as chemical compounds ; and it is therefore considered that no chemical combination has taken place Heycock and Neville who obtained similar results regarded them as being due to the presence of impurities in their thallium but i t is pointed out that this cannot be the case as the thallium used in the present experiments was quite pure.The compound Bi,Tl is soft and difficult to polish the fresh surface is light grey in colour but r<ipidly becomes yellow and the compound is oxidised by long contact with air. EMILE VIGOUROUX (BUR SOC. chim. 1907 [iv] 1 7-lO).-Copper turnings freed from any visible solid impurity are heated with hydrochloric acid a few drops of nitric acid being added to the mixture from time to time to facilitate G. S. Preparation of Pure Copper.INORCtANIC CHEMISTRY. 89 solution. The black liquid so produced is filtered in the absence of air into a large excess of recently-boiled cold distilled water.The mother liquor is decanted from the precipitated cuprous chloride and a fresh supply of distilled water added to the latter. Excess of aluminium in large fragments is added to this mixture which is then set aside until the cuprous chloride is completely reduced. The excess of aluminium is then picked out with wooden tongs and the precipitated copper washed by decantation with water digested with hydrochloric acid again washed with distilled water and finally reduced by heating in a current of hydrogen. Copper may also be prepared by reducing cuprous chloride with soft iron but the latter unlike aluminium does not attack cupric chloride. Cupric sulphate is reduced by aluminium but the action proceeds slowly especiitlly in concentrated solutions of the sulphate even when these are warmed.T. A. H. Commercial Copper Silicides. ~ I L E VIGOUROUX (Bull. Xoc. chirn. 1906 [iii] 35 1233-1237. Compare Abstr. 1906 ii 168 and Lebeau ibid. ii 29).-1t is pointed out that Lebeau's statement (Zoc. cit.) that chemists who have worked on copper silicides have assumed that a silicide represented by the formula c'u,Si exists is inaccurate since the author had shown (Proces-verb SOC. Xci. yhys. nat. Bordeaux July 18th 1901) that the maximum quantity of silicon which could be combined with copper was about 1 O% corresponding with the formation of a compound Cu,Si. Three specimens of commercial copper silicides reputed t o contain 20% 15% and 10% of silicon respectively have been analysed and found to contain as impurities free silicon iron silicide FeSi man- ganese silicide MnSi in addition t o iron calcium and aluminium present in the form of silicides or aluminosilicides.The method of analysis adopted depends essentially on treating the raw material with nitric acid and is described in detail in the original. T. A. H. FRITZ FOERSTER and F. BLANKENBERG (Bey. 1906,3@ 4428-4436).-Acid cupric sulphate solutions dissolve copper forming a cuprous salt until an equilibrium is established as expressed by the equation 2 C u ' z - C ~ " + Cu. At the ordinary temperature the con- centration of cuprous sulphate is small but increases on heating. A similar equilibrium is established in amrnoniacal cupric sulphate solu- tion which in presence of copper is decolorised if the copper concentra- tion is below &/lo or when heated in stronger solution ; this becomes coloured again on cooling.From a solution containing 0-5 mol. cupric sulphate and 8.7 mol. ammonia a colourless crystalline compound Cu,S0,,4NH3,H2O is obtained which in presence of moisture decom- poses into copper and a cupric compound. Cuprous chloride forms a similar crystalline compound with ammonia. Cuprous Sulphate. E. F. A. Some Reactions of Mercuric Iodide. N. A. ORLOFF (Chem. Zeit. 1906 30 1301).-A methyl alcohol solution of palladous chloride converts mercuric iodide into the chloride with formation of palladous iodide. Freshly precipitated silver chloride turns yellow with a methyl90 ABSTRACTS OF CHEMICAL PAPERS. alcohol solution of mercuric iodide but in this casea double compound of the two salts is formed.An aqueous solution of thallium chloride gives with a methyl alcohol solution of mercuric iodide a red pre- cipitate of the latter but an aqueous solution of bismuth chloride remains quite clear and colourless on adding an alcoholic solution of mercuric iodide. L DE K. Adsorption Compound formed by Iodine with Basic Praseo- dymium Acetate. N. A. ORLOFF (Chem. Zeit. 1907 31 45).-The precipitate obtained by the addition of ammonia t o a solution of praseodymium acetate gives with iodine a violet-blue adsorption com- pound similar in properties to the corresponding lanthanum compound (Biltz Abstr. 1904 ii 339). W. H. G. Preparation of Pure Neodymium Oxide and Two New Methods of Separation of Rare Earths.OTTO HOLMBERCI (Chem. Centr. 1906 ii 1595-1597 ; from Chem. Lab. Univ. UjxaIu. Pamphlet 114 pages).-The following organic salts of rare metals have been prepared. The figures in brackets give the solubility of the anhydrous salt in 100 parts of water at 15'. Yt(C,H40,),,5H20. Malonates Ce2(C,H204),,6H20 ; Yt2(C,H,0,),,SH,0. Citrates 2cec6H50,,H,o ; 2LoC6H5O7,7H,O ; 2YtC6H,0,,5H,0. All three salts are amorphous but become crystal- line gradually. Di(C6H,S0,),,9H,0 ; Yt(C H so,)3,9H,o. m-Nitro benzenesul-phonat e: $h[ C6H4( N02)S03] 7 H20 (6 1 -0) ; (25.5) ; Ls[C6H,(N02)R0,],,6L120 (16.0) ; Pr[C6H,(N02)S0,],,6H20 (33.9) ; Nd[C,H4(N02)S03],,6H20 (46.1) ; Di C,H,(N02)S03 ,,6H,O (47.8) ; Yt[C,H4(N02)S0,]3,7H20 (48.3) ; Sa~~6H4(N02)80,33,7H2~ (5 0.9) ; Gd [ C6H,( NO,)SO,] 7 H,O (4 3 .S).m- Chlorobenzenesul- phonates La(C6H,C1S03) 9H20 ; Pr(C,H4C180,),,9H,O ; all crystallise in needles (1 3.0). La(C6H,BrSO3),,9H,O ; Di(C,H,BrS0,),,9H20 ; crystallises i n needles (13.0). Chloronitro- benzenesulphonates [Cl NO SO,H = 6 3 13 La[C6H3C1(N0,)S0,],,8H20 ; Pr[C6H3Cl(N0,)t%)3]3 1 4H,O ; Di[C6H3Cl(N02)S0,18,1 6H,o crystal- lise in flat plates (25). Pr(CloH7S0,),,6H20 ; Di( C,,H7S03),,6H20 crystallise in slender needles (5-6). 1 5-Nitronaphthalenesulphonates ; Pr[ c~oH6(No2)~03]3 6H20 ; Di[ C,,H,(N 0 2 ) S 0 ] 6 H20 ; crys t allise in prismatic needles (0.5). Tartrates Ce2(C,H,06),,6H,0 ; La,(C4H,0,),,gH20 ; Lq(C3H204) 395 2 0 ; Benzenesulphonates La(C6H,S0,)3,9H,0 ; Pr(C6H,S0,)3?gH,0 ; All crystalline (53-63). Ce[ c6H4( N02)S03]3,6H20 ui( c 6 ~ 4 c 1 s 0 3 ) 3 ~ ~ 2 0 m-Bromobenzenesulphonates a-Naphthalenesulphonates La(C,0H7S0,)3> ,O ; LaCC,,H6(N0,)S0,1a,6H,0 ; 1 6 -Nitronapht halenesulphonates La[ H 6 (NO,) s 0 3 1 3 9 H ~ o ;INORGANIC CHEMISTRY.92 Pr[C,oH,( N02)S0,],,9H20 ; Di[C,,H,(N0,)S03],,9H20 crystallise in needles (0.2). 1 7-Nitronaphthalenesulphonates Pr[C,oH,(NO,)SO,],,l 1H,O ; Di[C,oH,(N02)S03]3,9H20 crystallise in slender needles (about (1 -2). The salts of m-nitrobenzenesulphonate show the greatest differences of solubility and are also best adapted for separation on account of their slight isomorphism. Whilst the ratio of the solubilities of didymium and lanthanum ammonium nitrates is 1.6 1 that of the m-nitrobenzenesulphonates is 3 1. When 200 grams of didymium oxide which contained some praseodymium and samarium (at wt.142.4) mere dissolved in nz-nitrobenzenesulphonic acid and fractionated into twelve portions samarium gadolinium and yttrium accumulated in the mother liquors; on further separation the twelve original fractions yielded fifty-two. The differences in the atomic weights determined by the oxide-sulphate method gave an indication of successful separa- tion. The fractions containing the most neodymium were again treated and the middle portions still further fractionated; fractions 7-13 then gave a constant atomic weight = 144*0-144*1. The absorption bands of a solution of neodymium chloride were measured for eight different concentrations varying from 2 N to N/64 (Forsling) and the maxima between 689.5 and 342 pp are tabulated in the original paper.The spectra showed no trace of absorption bands of samarium lanthanum yttrium or praseodymium but the first and last fractions of the last series contained a small quantity of samarium or praseodymium. Four series of atomic weight determinations gave values varying from 144.01 to 144.13 or a mean value= 144.08. A mixture of samarium and gadolinium has also been submitted to the m-nitrobenzenesulphonic acid and to the picric acid methods of separation. Gadolinium separated first and then samarium and yttrium whilst cerium accumulated in the mother liquors. After continued fractionation the potassium sulphate method eventually yielded almost pure samarium oxide which gave an atomic weight = 150. Pure gadolinium compounds were prepared from other fractions.Gadolinium could not be readily separated from terbium by the m-nitrobenzeuesulphonic acid method but by means of picric acid and partial precipitation with ammonia gadolinium of atomic weight = 156 was finally obtained. Samarium may be separated from gadolinium by Urbain-Lacombe's bismuth magnesium nitrate method. Experiments have also been made on a terbium material which contained yttrium gadolinium samarium and neodymium by means of m-nitrobenaenesulphonic acid and picric acid. Paucity of material however prevented the complete isolation of terbium. Absolute Atomic Weight of Dysprosium. GUSTAVE D. HINRICHS (Compt. rend. 1906 143 1143-1145).-The author has applied his method of calculating atomic weights (Abstr.1893 ii 163 277 316 317 ; 1894 ii 39 87 276 ; 1900 ii 534 539; 1905 ii 517; 1906 ii 450) to Urbain's analytical data for the atomic weight of dysprosium (Abstr. 1906 ii 855)) and finds the value Dy=162*5 (S=32 H = l ) . La[CloH,(N0,)S0,13,9H,0 ; E. W. W. M. A. W.92 ABSTRACTS OF CHEMICAL PAPERS. Solubility of Carbon in Manganese Sulphide. M. HOUDABD (Compt. rend. 1906 143 1230-1233).-When manganous sulphate is reduced by means of excess of sugar carbon at the temperature of the electric furnace the product has a black metallic appearance and consists of manganese sulphide containing dissolved graphite (compare Mourlot Abstr. 1896 ii 25) the quantity varying from 0.002% to 3.20% according as the heating was prolonged during 2 or 25 minutes respectively and similar results are obtained when pure manganese sulphide is fused in the electric furnace with 10% of sugar carbon.The graphite obtained from the fusions is non-intumescent ; yields with Brodie's mixture the pale yellow crystalline oxide (Abstr. 1873 348) which on burning gives black pyrographitic acid; its ignition temperature is 640° (Moissan Abstr. 1896 ii 165 ; 1903 ii 141) and it contains 98.01-98*92% of carbon and 0.209-0*218% of ash. Manganese sulphide a t the temperature of the electric arc converts a particle of diamond almost completely into that form of graphite which is insoluble in fuming nitric acid. GUSTAV GIN (Compt. rend. 1906 143 1229-1230).-When rhodonite is reduced in the electric furnace a crystalline silicomanganese is obtained having the per- centage composition Mn = 68.64-69.26 ; Fe = 4.40-3.18 ; Al- 0*74-0.60 ; Si = 25A02-25*51 ; C = 0*16-0*18 ; S = 0.01 ; P= 0.01 ; and containing 95% of a crystalline manganese silicide Mn,Si (compare Lebeau Abstr.1904 ii 343 ; and Vigouroux Abstr. 1905 ii 822) which forms brilliant prismatic crystals sometimes 5 cm. long DI5 6.05 m. p. 125O0-13OO0; is attacked by chlorine with in- candescence at a red heat is not changed by oxygen a t the ordinary temperature but superficially oxidised a t 500' ; is readily attacked by the gaseous halogen acids by nitric or hydrochloric acid or aqua regia or by fused alkali carbonates or mixtures of carbonates and nitrates whilst sulphuric acid has no action on the compound. M. A. W. A New Manganese Silioide. &I.A. W. Influence of Chromium on the Solubility of Carbon in Iron and on the Formation of Graphite. PAUL GOERENS and A. STADELER (MetaEZurgie 1907 4 18-24).-1t is known that a part of the chromium in iron-chromium-carbon alloys is present as the carbide Cr,C,; it is not known however whether definite double carbides of iron and chromium are formed or whether the two carbides Cr,C and Fe,C merely form solid solutions with one another. The authors have melted iron with ferro-chromium in presence of an excess of carbon and have determined the freezing point and composition of the resulting alloys. It is necessary to heat t o about 1600' to insure saturation with carbon. The presence of chromium increases the solubility of carbon in iron so that when 62% of chromium is reached 9.2% of carbon is dissolved.The freezing point of these alloys increases with increasing chromium content; the results are recorded in a diagram with triangular co- ordinates. The position of the martensite-cementite eutectic point is displaced in the direction of higher carbon by addition of chromium 60INORGANIC CHEMISTRY. 93 that an alloy containing 5% of chromium and 4.7% of carbon is entirely composed of eutectic. Microscopic examination confirms the abave results. When the eutectic point is passed crystals appear which a t first have the appearance of cementite but with increase of chromium resemble more and more closely the crystals of chromium carbide. The formation of graphite is hindered by the addition of chromium even when silicon is present.C. H. D. Copper and Iron [Alloys]. HERNANN WEDDING and WiLHELnz MULLER (Stcdd. u. Eisen 1906 26 1444-1447).-Copper alloys with iron in all proportions. The alloys are conveniently prepared by Goldschmidt’s aluminothermic method. The malleability of iron is reduced by addition of copper but the hardness and strength are increased. A n alloy containing 7.77% of copper and 0.14% of carbon does not shorn any separate micrographic constituent but the crystal grains are more curved and interlocked than in pure iron. I n alloys containing more carbon the presence of copper hinders the formation of pearlite causing segregated cementite to appear. The transformation tem- perature is lowered to 620-640” by the addition of copper. C. H. D. Soluble Iron Sulphide. A.KONSCHEGCJ and HANS MALFATTI (Zeit. ccncd. Chenz. 1906 45 747-751).-A preliminary paper on the behaviour of iron sulphides towards alkali sulphides. Ferrous sulphide does not dissolve in aqueous potassium or sodium hydroxide but the mixture of ferrous sulphide and sulphur formed on adding ammonium sulphide to ferric chloride yields a dark green solution. The authors assume the presence of a polysulphide of iron with acidic properties which in presence of alkali hydroxides forms a true solution and not one cobtaining colloidal iron. I n the absence of ammonium chloride ferric hydroxide or basic phosphate dissolves i n ammonium sulphide to a dark green liquid which is precipitated by ammonium chloride and also by a large excess of ammonium polysulphide (compare also de Koninck Abstr.1906 ii 397). L. DE K. Ferrous Compounds of Nitric Oxide. WILIIELM MANCHOT and K. ZECHENTMATER (An?wZen 1906 350 368-339. Compare Gay Abstr. 1885 1109 ; Thomas Abstr. 1899 ii 368 426).-The authors have investigated quantitatively t h e absorption of nitric oxide by aqueous and alcoholic solutions of ferrous salts under varying condi- tions. The results are tabulated and expressed in curves and the following conclusions are drawn. The absorption of nitric oxide by a ferrous salt results from the formation of a chemical compound contain- ing Fe and NO in the proportion 1 1. Under all conditions therefore the limit of absorption is reached when 1 mol. of nitric oxide has been absorbed for each ferrous atom present and corresponds with the con- version of Fe” into Fe”’.The formation of the compound is a reversible reaction; the degree of dissociation varies with the ferrous salt but is dependent also on the pressure temperature concentration of the ferrous VOL. XCII. ii 794 ABSTRACTS OF CHEMICAL PAPERS. salt presence of indifferent solutes and nature of the solvent t o an extent which is in agreement with the laws of chemical dissociation. G. Y. Cryoscopy of Colloidal Solutions of Ferric Hydroxy- chloride. G. MALFITANO and LEOPOLD MICAEL (Compt. rend. 1906 143 1141-1143. Compare Abstr. 1906 ii 450 526 647).- Attempts to determine the molecular mass of the granules of colloidal ferric hydroxychloride from the difference between the freezing point of a colloidal solution and the filtrate obtained by filtering the solution though a collodion membrane (Abstr. 1906 ii 526) were unsuccessful as the differences were sometimes negative and sometimes positive.M. A. W. Acid Functions of Nickel Dioxide. ITALO BELLUCCI and S. RUBEGNI (Atti A?. ACCCC~. Lincei 1906 [v] 15 ii 778-787. Compare Bellucci and Clavari Abstr. 1905 ii 823).-The authors have confirmed the existence of the compound Ni,0,,2H20 prepared by Dudley (Abstr. 1897 ii 171). The alkali which is only removed by very thorough washing with water exists in combination with the nickel oxides in the form of the nickelite Na,Ni,O or 2Ni0 NiO Na,O which the authors have prepared and analyeed. This compound when washed with boiling water is hydrolysed t o the compound NiO 2Ni0,2H20 whilst the potassium compound NiO,,NiO,K,O (compare Hoff mann and Hiendlmaier Abstr.1906 ii 747) on hydrolysis with iced water yields the compound NiO2,Ni0,2H,O. I n both cases the product of hydrolysis retains unchanged the crystalline appearance of the original alkali nickelite. T. €€* P. t Nickel Chromates. MAX GIU~GER (Zeit. anorg. Chem. 1906 51 348-355. Compare Abstr. 190G ii 45 l).-Normal aqueous solutions of nickelous chloride were mixed with normal solutions of sodium pot'assium and ammonium chromates left for some days and the resulting precipitates washed and dried. All the products were reddish-brown in colour and proved t o be basic chromates of nickel containing varying proportions of alkali chromatee. They are hydrolysed by water alkali chromate and acid chromates of nickel going into solution and highly basic chromates of nickel remaining behind. Unlike the behaviour of the corresponding manganese and cobalt compounds no higher oxide of nickel was found in the residue.By interaction of stronger. solutions three crystalline double chromates were obtained. The compourd N iCr04,K2Cr04,2H20 occurs in yellow needles which are stable in the air but undergo hydrolysis with water. Tho compound NiCr04,(NH,),Cr04,6H20 was obtained in green crystals (compare Briggs Trans. 1903 83 391). By a slightly different method a second nickel ammonium chromate of the probable formula (NHJ2Cr04 NiCr0,,NH3,HS0 was isolated in well-formed orange-brown prisms. This compound isINORGANIC CHEMISTRY. 95 stable in the air but is hydrolysed by water with formation of basic nickel chro m ates.0. s. The Reduction of Oxides by Aluminium. Preparation of Chromium. EMILE VIGOUROUX (Bull. SOC. chim. 1907 [iv] 1 10-1 3. Compare Abstr. 1905 ii 823).-Commercial chromium prepared by the " thermit " process has the advantage of being free from carbon but frequently contains silicon aluminium iron and portions of slag. A much purer material can be prepared on a small scale by a modification of this process which consists in mixing from 10% to 20% of dried chromium trioxide with the dried chromium sesquioxide and then incorporating the necessary quantity of aluminium powder. The mixture is heated in a crucible lined with magnesia. Under these conditions a vigorous reaction ensues which when 20% of chromium triovide has been added lasts about a minute the contents of the crucible becoming thoroughly melted and the slag separating readily f r o 9 the metal.The chromium produced contains as impurities 0.36-0.4% of silicon and 0*74-0*85% of aluminium and iron. T. A. H. Definite Compounds formed by Chromium and Boron. ARNAND BINET DU JASSONEIX (Compt. rend. 1906 143 1149-1151). -Chromium-boron alloys (this vol. ii 30) contain the two definite compounds Cr,B and CrB dissolved respectively in a medium less rich in boron which can only be isolated in a state of purity from the nearly homogeneous alloys the composition of which is close to that of their own. The tioride Cr3B2 isolated from an alloy containing 11.6% of boron by the action of hydrogen chloride below a red heat has D'j 6.7 burns in fluorine when gently heated and is attacked with incandescence by chlorine below a red heat; the action of bromine is less vigorous whilst iodine vapour has only a superficial action a t a red heat.When Eeated in air or oxygen it becomes covered with a vitreous layer of the borate; it is converted into a mixture of boron and chromium sulphides together with unchanged boron by the action of boiling sulphur; it is completely soluble in concentrated or dilute hydrofluoric hydrochloric or sulphuric acid and is oxidised with incandescence by fused alkali hydroxides or carbonates ; it is not acted on by nitric acid or alkali solution or by nitrogen a t a white heat. The boride CrB isolated by the action of hydrogen chloride or chlorine below a red heat on the alloy containing 16% of boron has DI5 6.1 and whilst its chemical properties closely resemble those of the lower boride Cr3BrZ i t is in general more readily attacked; it takes fire in cold in contact with fluorine and when heated in nitrogen at a white heat forms a greyish-black substance which liberates ammonia on treatment with fused potassium hydroxide.M. A. W. The Violet and Green Varieties of Chromium Chloride. F. JOST (Ber. 1906 39 4327-4330. Compare Recoura Abstr. 1886 508 ; Piccini Abstr. 1895 ii 229 ; Werner and Gubser Abstr. 1901 ii 453).-The author shows by determinations of the molecular 7-296 ABSTRACTS OF CHEMICAL PAPERS. weight by the ebullioscopic method of the violet and green varieties of chromium chloride respectively in methyl alcoholic solution that the water plays a different r6Ze in the green chloride and violet chlorides.The violet salt gives values for its molecular weight varying from 54.7 to 60.3 whereas the green salt gives values varying from 33.1 to 33.4. The molecular conductivity of a solution of the violet salt in methyl alcohol varies from 67.7 to 97.7 whereas the green modification gives values varying from 43.6 to 64 2. The experiments quoted are in accordance with Werner’s theory in so far that in the green hydrate the water is more loosely bound than in the violet. A. McK. Cobaltimolybdates. CARL FRIEDHEIM and F. KELLER (Ber. 1906 39 4301-4310).-The cobaltimolybdates described by the authors are distinguished from cobaltomolybdates in that chlorine is evolved when they are boiled with hydrochloric acid.They are characterised by possessing a colour varying from brilliant green t o dark green. The compound 3(NH4),0,Co0,Co0,,1 2Mo03,20H,0 is obtained by adding a 30% aqueous solution of ammonium persulphate t o a mixture of cobaltous acetate and ammonium paramolybdate acidifying with dilute acetic acid and then slowly heating. It may alho be prepared by the action of cobaltous acetate on the ammonium permolybdate 3(NH4),0~5M00,,2M00,,6H,0 It forms green rhombic plates and dissolves in water to a green solution. With concentrated sulphuric acid oxygen is evolved ; ammonia is evolved by the action of potassium or sodium hydroxide. It is reduced by sulphurous acid. Hydrogen peroxide causes the evolution of oxygen whilst permolybdatss are formed. The compound 2(NH4),0,Co0,Co0,,10M00~~,1 2H,O best obtained from cobaltous acetate ammonium paramoly bdate and hydrogen peroxide forms dark green rhombic crystals.The compound 3K20,Co0,Co0,,10M00,,1 1H2O7 obtained either from cobaltous acetate potassium persulphate and potassium paramoly b- date or from cobaltous acetate and ammonium permolybdate and potassium chloride or by the addition of potassium chloride to a cold saturated aqueous solution of the compound 3(NH,),O,CoO,CoO 1 2M00,,20H20 forms greenish-white rhombic crystals. The compound 3K20,Co0,Co0,,1 2M00,,15H20 prepared. by the addition of potassium chloride to an aqueous solution of the compound 2(NH4)20,Co0,Co0,,10Mo0 12H,O forms olive-green crystals. The compouud 3 Ba0,Co0,Co0,,9M003,25H20 prepared by the addition of barium chloride to an aqueous solution of the compound 2(NH,)20,Co0,Co0,,1 OMoO 22H,O is bright green.The compound A( NH4),0 1 ~Ba0,Co0,Co0,,10M00,,18~H20 pre- pared from barium chloride and the compound is bright green. A. McK. 3(NH4)207 Co0,Co02 1 2Mo03,20H20,INORGANIC CHEMISTRY. 97 Preparation of Titanium Tetrachloride. MILE VIGOUROUX and G. ARRIVAUT (BUZZ. SOC. chim. 1907 [iv] 1 19-21).-Com- mercial ferrotitanium containing about 55% of titanium is employed RS a raw material. This i n a coarsely powdered condition is placed in a wide porcelain tube heated in a Mermet furnace. A current of dry chlorine is passed 'over the heated alloy and the reaction commences when the latter is a t a dull red heat. This method has the disadvantage that considerable quantities of ferric chloride are formed which are liable to block the tube unless they are periodically removed.This difficulty may be overcome by digesting the finely powdered ferro- titanium with dilute hydrochloric acid which leaves a product contain- ing S0-90% of titanium with small quantities of titanium dioxide. The latter may be removed by levigation and the residue is treated as before when it furnishes crude titanium tetrachloride and leaves in the porcelain tube a small residue of titanium dioxide. The titanium dioxide obtained as a by-prodnct mny also be converted into the tetra- chloride by mixing it with carbon and heating in a current of chlorine. The crude titanium tetrachloride is filtered to remove suspended ferric chloride and the filtrate on fractional distillation yields the pure tetra- chloride b.p. 136". T. A. H. Colloidal Compound of Thorium with Uranium. B ~ L A SzILARD(Compt. rend. 1906,143,1145-1 147).-Thoriumhydroxide,pro- cipitated by meansof ammoniaand well washed partially dissolves in a hot solution of uranium nitrate forming a deep red colloidal mixture which on separation in a centrifugal apparatus and drying in a vacuum over sulphuric acid forms small brilliant carmine red transparent plates which are amorphous and have the following percentage composition Th = 72-74 U = 4-5.5 H,O = 1%- 13,O = 10-1 1. The dry compound is stable towards heat and light dissolves without decomposition in hot water and yields the corresponding salts when dissolved in dilute acids; it has D18'3 4482-5*45 and when thrown into water produces a sharp hissing with evolution of a gas the nature of which is at pre- sent not determined.The compound is more readily prepared by heating the mixture of thorium hydroxide and uranium nitrate in an autoclave at 200' under 15 atmos. pressure. When uranyl chloride replaces uranium nitrate in the above reaction the colour of the colloidal compound is reddish-yellow ; the reaction proceeds very slowly when uranium acetate or sulphate is employed in place of the nitrate whilst thorium nitrate has no action on thorium hydroxide. M. A. W. Some Sulphates of Quadrivalent Vanadium. GUSTAVE GAIN (Compt. rend. 1906 143 1154-1156. Compare this vol. ii 32).- By the action of dill'erent quantities of sulphuric acid on the deep blue liquid obtained when a mixture of the oxides V,O and V,O is treated with a saturated solution of sulphur dioxide the author has prepared three acid uccnccdyl su@hates belonging to the same group as those prepared by Koppel and Behrendt (Abstr.1903 ii 551) whilst two other members of the same series have been prepared by the action98 ABSTRACTS OF CHEMICAL PAPERS. of sulphuric acid on red hydrated hypovanadic acid V20,,2H20. These new acid vanadyl sulphates are crystalline and possess blue colours 2VOS0,,3H2S0 1 5H,O is azure blue 2VOS0,,4H2S0 16H,O is tur- quoise blue 2VOSO4,5H,80,,1 5H,O is greenish-blue 2VOSO4,7H,S0 1 5H20 is bright blue and 2VOS0,,8H,S0,,16H20 is very bright blue. M. A W. Preparation and Chemical Properties of Antimony Penta- fluoride. OTTO RUFF [and in part with GRAF HELLER and KNoCH] (Bey.1906 39 4310-4327. Compare Ruff and Plato Abstr. 1904 ii 265)-Additional details for the preparation of antimony penta- fluoride by the action of hydrogen fluoride on antimony pentachloride are given. I n the course of the preparation it is probable that a compound SbF&HF is formed from which the hydrogen fluoride must be completely eliminated before the antimony pentachloride is distilled. Antimony pentafluoride has h. p. 149-150' instead of 155' as formerly stated by Ruff and Plato (Zoc. cit.). Chlorine apparently has no action on antimooy pentafluoride ; bromine forms a viscid dark-brown mass possibly containing the com- pound SbF,Br. Iodine reacts with antimony pentafluoride with development of heat and the formation either of a bluish-green liquid or of a dark brown or bluish-green solid according to the amount of pentafluoride used.When an excess OF pentafluoride is used and the mixture is hea.ted at 160-220° the compound (SbF,),I is formed as a dark bluish-green crystalline mass m.p. 110-115" which does not lose iodine at 240'; by prolonged heating in a current of carbon dioxide practically no iodide is given off but the excess of antimony pentafluoride distils over. Water decomposes i t with hissing and separation of iodine. If an excess of iodine is used and the temperature is raised above the b. p. of iodine the compound SbF,I is formed by the action of iodine on antimony pentafluoride after the excess of iodine is evaporated off. When heated above 260' iodine is evolved.The compound SbF,I m. p. slightly below SO" is dark brown and differs from the compound (SbF,),I in undergoing slow decomposition with water. The solution of sulphur in antimony pentafluoride is dark blue and yields the compound SbF,Sr m. p. about 230'. I t is very hygroscopic and when acted on by moist air yields sulphur ; antimony trichloride and thionyl fluoride are also formed. The action of water probably proceeds according to the equations 2SbF,S + H,O = 2SbF + S + SOF + 2HF and SOF + H,O =SO + 2HF. Antimony pentafluoride is acted on by phosphorus with production of flame and the formation of yellow vapour which sublimes. The majority of metals when dry do not act on antimony penta- fluoride. When heated with sodium rapid combustion occurs with the formation of white vapour.Antimony dust reduces it to the trifluoride. With little water i t forms the solid hydrate SbF,,2H20 (loc cit.). I t s aqueous solution i> nnt decomposed in hhe cold byINORGANIC CHEMISTRY. 99 hydrogen sulphide or by potassium iodide ; action does not occur until the temperature is raised. On the addition of sodium hydroxide or of sodium carbonate it forms sodium bydrogen pyronntimonate in quantitative yield. When its aqueous solution is evaporated the syrupy residue con- tains a hydrate of antimony pentafluoride which when heated is decomposed with formation of water hydrofluoric acid and antimonic acid; the syrup however does not form anhydrous antimony penta- fluoride when heated with sulphuric acid or phosphoric oxide.When a current of hydrogen sulphide is passed over it snlphur hydrogen fluoride and an antimony thiofluoride are formed. When a current of dry ammonia is passed over it a vigoroiis action . occurs and the pentafluoride becomes coated with a yellowish-red crust which prevents further action. When heated however with liquid ammonia a t 100" in a platinum tube it forms diccmino-dic~ntintony-tri- Ruo~oimide dihycZroJuoride NH(SbF,-N H,-HF) a white powder which is very readily decomposed by atmospheric moisture ; it4 is slowly acted on by water with formation of antimonic acid ; its aqueous solution is acid towards litmus. Antimony pentafluoride is acted on by nitrogen sulphide sulphur dichloride and chromyl chloride respectively. When heated to above 100' with molybdenum pentachloride it forms R molybdenum fluoride and antimony pentachloride the latter parting with some of its chlorine to form don ble compounds with the excess of antimony pentafluoride. With tungsten hexachloride it forms tungsten hexafluoride.It reacts vigorously with phosphorus trichloride with the formation of phosphorus trifluoride. With phosphoric oxide it forms phosphorus oxyfluoride. It forms a series of crystalline compounds with arsenic trifluoride. With antimony trifluoride i t forms compounds varying in composition from SbF5,2SbF to SbF5,5SbF,. Hydrogen chloride is evolved when it is mixed with tin tetrachloride titanium tetrachloride or silicon tetrachloride. Colloidal silicic acid when warmed with antimony penta fluoride forms antimonic acid and silicon tetrafluoride.The behaviour OF antimony pentafluoride towards various carbon compounds was also investigated Filter paper cork mood and india- rubber are attacked a t once. The action with benzene is vigorous hydro- gen fluoride being evolved. Ether alcohol acetone glacial acetic acid ethyl acetate carbon disulphide light petroleum toluene arid chloro- form all react with antimony pentaflnoride. With carbon tetrachloride at about 45" a gas is evolved which readily liquefied and is probably The compound SbF3,2NH is obtained by the action of liquid ammonia on antimony trifluoride as tt yellow powder which loses ammonia in contact with moist air. It is less stable than the compound obtnined from antimony pentafluoride. CC1,F. A . McK. Alloys of Palladium and Silver. RUDOLF RUEE (Zeit. anorg. Ghern. 1906 51 315-319. Compare this vol. ii 32).-The freezing- point curve falls regulaxly from the melting point of palladium 1541" to that of silver 961.5' ; unlike the system palladium-copper the curve is slightly concave to the axis of concentration. I t follows that the100 ABSTRACTS OF CHEMWAL PAPERS. metals form a complete series of mixed crystals with no indication of chemical combination and this conclusion is confirmed by microscopic observations. The hardness of the alloys lies between that of the components and increases with the proportion of palladium. The micro-structure of the alloys is described and illustrated by six microphotographs. G. S. Alloys of Palladium and Gold. RUDOLF RUER (Zeit. anorg. Chenz ,1906 51,39 1-396. Compare preceding abstract)-The freezing- point curve of the system Falls continuously from the melting point of palladium to that of gold ; like the corresponding curve for palladium- silver alloys it is concave to the axis of concentration. The metals form a continuous series of mixed crystals and there is no indicatlion of chemical combination. Alloys containing more than 10% of palladium are white. The hardness increases with increase in the amount of palladium until 700/ of the latter element is present but beyond this point decreases. The fact that none of the three elements copper silver gold enter into chemical combination with palladium is in accordance with Tammann’s rille ( Abstr. 1906 ii 346) according to which either all or none of the elements of x natural group in the narrower sense enter into chemical combination with another element. G. S.