18 ABSTRACTS OF CHEMICAL PAPERS. Inorganic Chemistry. Ignition Temperatures of Hydrogen-Oxygen Mixtures. KAUFMAN G. FALK (J. r1me.r.. Chem. Xoc. 1906 28 1517-1534)- Previous determinations of the ignition temperature of detonating gasINORQANIC CHEMISTRY. 19 have been made either by plunging a sealed bulb of the gas into a bath of known temperature or by passing the gas through a tube heated to a definite temperature. The results obtained by these methods show considerable variation which is due to the fact that the whole of the gas does not reach the ignition temperature at the same instant. The heat would reach the outer layers of the gas first and in the time necessary for the whole of the gas to attain the required temperature an indefinite amount of combination would have occurred.In order to overcome these objections a method has been devised for generating the necessary heat by the adiabatic compression of the gas itself in a steel cylinder fitted with a steel piston. The apparatus and method are fully described with the aid of diagrams. Mixtures of hydrogen and oxygen were employed in the volumetric proportions of 4 1 2 1 1 1 1 2 and 1 4. The whole of the hest involved in the experiments was produced by the adiabatic compression which mas so rapid that radiation to the walls of the vessel was impossible. Two forms of apparatus were employed one of which was larger than the other and probably gave more accurate results. The ignition temperature was calculated by a method based on the equation for adiabatic changes and the following average results were obtained which are expressed as absolute temperatures.I n the smaller apparatus 4H2+0 874O; 2H2+0 811'; H,+O 786'. Jn the larger apparatus 4H + O 893' ; 2H + o 8 19' ; €3 + O 796O ; H + 20 808' ; H + 40 849'. The results show that the ignition temperature is independent of the final pressure of the gas for pressures greater than 39 atmospheres (the lowest pressure a t which the ignition temperature was determined) and under the conditions of the experiments is also independent of the initial temperature of the gas. The ignition temperature of the mixture H,+O is lower than that of any of the other mixtures and this minimum temperature corresponds with a maximum affinity. The probable explanation is that hydrogen peroxide is formed first in the combination of hydragen and oxygen and is subsequently decomposed the degree of completeness of this decomposition depending on the conditions. According to the theory of the explosion wave the ratio of the pressure of the gas a t the ignition point to the initial pressure should be constant for any one mixture and this has been found t o be approximately the case.The ratio found in the case of detonating gas is 36.6 which agrees fairly closely with the value deduced by Jouquet (J. math. 1906 SO). VOGEL VON YALCKEN- STEIN (Zeit. Eleklrochem. 1906 12 763-764).-Asbestos soaked in copper chloride was used as the catalyst a t 450' and platinum tetra- chloride at temperatures above 500'. The equilibrium represented by the equation 2C1 + 2H,O 0 + 4HC1 was attained from both sides.The values of the equilibrium constant K = ( p ~ 2 0 ) ~ . @ ~ 1 2 ) ~ / ( 0 2 ) ( p ~ ~ 1 ) ~ are given below ; they are the mean values of fifty-eight experiments at 450° forty-eight at 650° and five at 600'. E. G. The Equilibrium of the Deacon Process. 2-2ABSTRACTS OF CHEMICAL PAPERS. From hydrogen From chlorine side chloride side Temperature. log K. log K. LogK (cal.). 450" 1 '50 1 -49 1 -52 650 - 0.407 - 0.390 - 0.43 600 - 0 '046 - 0.050 - 0.02 From Dolezalek's experiments on the chlorine-hydrogen gas cell Nernst has calculated 10g[Cl,]~~[H,]~/[HC1]4 = - SS000/4*57 1 T- 1.6. Nernst and Wartenberg calculate from their experiments on the dissociation of steam 10g[H,0]~/[H,]2[0,] = 25050/1'- 1.7510g1'- 0*00013T'+0~2.From these expressions the value of logK in the process studied is found to be 5790/1'- 1.75 log?' - 0*000132'- 1.4. The calculated values in the above table are obtained from this expres- siou . T. E. Sublimation of Sulphur at the Ordinary Temperature. RICHARD J. Moss (Sci. Proc. Roy. DubLSoc. 1906 ii 105-106).-Some fragments of ordinary roll sulphur were sealed up in an exhausted tube tweuty- four years ago. After twentyyeara during which time the tube was placed in a horizontal position in a drawer a minute crystalline sub- limate was observed. H. M. D. The Two Forms of Liquid Sulphur as Dynamic Isomer- ides. ALEXANDER SMITH and CHARLES M. CARSON (Proc. Roy. Xoc. #din. 1906 26 352-356. Compare Abstr. 1906 ii 157).-The liquid obtained by melting sulphur which after recrystallisation has not been in contact with air contains the two liquid forms SX and Sp in equilibrium proportions the equilibrium condition being attained in a few moments.Ammonia bubbled through the melted sul-phnr also facilitates the rapid establishment of equilibrium. The conversion of Sp into SX on cooling is equally rapid so that by plunging into water ordinary monoclinic sulphur is obtained. On the other hand equilibrium is only slowly reached in the fused mass obtained from sulphur which has been exposed to the air. This inert condition is also induced by passing hydrogen chloride or sulphur dioxide through the fused mass obtained from sulphur not exposed to air after recrystallisation. These gases act as simple catalytic agents and the two liquid forms SX and Sp are t o be classed as dynamic isomer- ideP.The equilibrium proportions at any temperature may be rapidly measured by passing in ammonia to accelerate the attainment of equilibrium and afterwards sulphur dioxide before cooling to prevent further change. Iodine retards the adjustment of equilibrium but dis- places the equilibrium to a considerable extent. The solidification of molten sulphur is considered with reference to a temperature concentration diagram. The freezing point of SX is 119.25' the " natural freezing point " at which the two liquid forms are in equilibrium with solid monoclinic sulphur is 114*5O and the liquid contains 96*3°/0 SX and 3.7°/0 Sp. The proportion of Sp in the liquid increases with rise of temperature and the rate of increase is much greater above 160° than it is below this temperature as the fol- lowing numbers show :21 INORGANIC CHEMISTRY.Temperature .. . . . . . . . . . . . . . . . . . . . . . . . . 121" 164" 166" 162" 166" 167" % SEr in equilibrium condition ...... 3.75 7.5 8.0 13.5 15-5 16.7 The marked absorption of heat and fall in temperature at or just above I 60° which are observed when liquid sulphur is heated continuougly at a constant rate is shown to be due to the fact that the proportion of Sp in the liquid lags behind that corresponding to the equilibrium con- dition The observed fact that the liquid becomes viscous at a lower temperature when the rate of heating is diminished is due to the same circumstance. H. M. D. Conditions under which Metallic Sulphides are Precipitated and Redissolved.HENRI BAUBIGNY (Compt. rend. 1906 143 678-679).-A claim for priority against Bruni and Padoa (Abstr. 1906 ii 157) ; compare Abst,r. 1882 805 928 1031 1032 1172; 1883 22 24 25 ; 18S8 113 423; 1889 118 346 652 653). M. A. W. Selenium. WILLXAM OECHSNER DE CONINCK (Compt. rend. 1906 143 682). -The brick-red amorphous variety of selenium obtained when selenious acid is reduced by glucose (? lzevulose) dissolves in concentrat.ed sulphuric acid to form selenium sulphoxide SeSO [com- pare Divers and Shimosk Trans. 1883 43 329 ; 1884 45 194,201 ; Divers and Shimidzu Trans. 1886 49 5831. This slowly decomposes according to the equation SeSO + H,O = Se + H,SO yielding a different variety of selenium to that originally employed. The new variety is pale or deep brown in colour does not become fluorescent when exposed to diffused light or undergo any change in contact with carbon disulphide in which it is very sparingly soluble.On pro- longed exposure to snnlight it is slowly converted into the amorphous black variety of selenium. Vapour Pressure of Carbon Dioxide at Low Temperatures. JOHN ZELENY and ROY H. SMITH (Chem. Centv 1906 1486; from PhyaikaE. Zeit. 7 667-671.)-The vapour pressure of carbon dioxide was measured a t temperatures varying from 7" to - 134'. Solid carbon dioxide was placed in a glass tube closed at the top and connected by a side tube t o a mercury gauge. The tube was immersed in a bath of pentane the temperature of which was measured by means of a nickel-iron couple the pentane bath being itself surrounded by liquid air.The following values were obtained for the vapour pressure in atmospheres 27.8 at - 7' 19.52 at - 20° 9.88 at - 40° 4.35 at - 60° 1 atmosphere at -78.2'; 28-8 mm. at -goo 11.9 a t - loo' 1.4 at - 120° and 0.1 mm. mercury at - 134'. Theaddition of alcohol or ether has very little influence on the vapour pressure. The critical point lies at - 56.4' and 5.11 atmospheres. The value of dp/dT was found t o be 6.35 cm. per degree and the latent heat of vaporisation L was found to be - 140 Gal. M. A. W. P. H. Behaviour of Carbon Disulphide towards Nascent Hydrogen. A. GAWALOWSKI (Chern. Centr. 1906 ii 1248 ; from Zeit. Oesterr. Apoth. Per. 44 460-46 l).-Carbon disulphide when treated with zinc and sulphuric acid yields a gas which produces a pure black pre-22 ABSTRACTS OF CHEMICAL PAPERS. cipitate in a solution of lead nitrate and blackens a filter paper moistened with the same solution.The gas evolved on heating carbon disulphide with zinc and strong potassium hydroxide however yields a bright orange-red precipitate with lead nitrate solution. The colour of the stain produced on filter paper moistened with lead nitrate is orange and fades rapidly. Fractionation of Rare Gases from Mineral Waters. Pro- portion of Helium. CHARLES MOUREU and ROBERT BIQUARD (Conyt. rend. 1906 143 795-797. Compare Abstr. 1905 ii 5 ; 1906 ii 442).-By submitting the rare gases obtained from certain mineral waters to the selective absorption of wood charcoal a t - 185’ (compare Dewar Abstr. 1904 ii 652 728) the authors have separated the helium and part of the neon from the other gases but further treatment with charcoal at - 100’ or at the temperature of liquid air boiling under 40-60 mm.pressure failed to effect the separation of the neon and helium. The former gas therefore is present in quan- tities too small to be estimated. The proportion of helium in the gaseous constituents of the waters of forty-three thermal springs is given in the original and varies from 0.00063 per cent. in Chatel- Guyon spring to 5.34 per cent. in that of Maizibres; the gases from the latter source also contain krypton. Gausticising of Sodium and Potaesium Carbonate8 with Lime. MAX LE BLANC and KABL NovoiwP (Zeit. anorg. Chern. 1906 51 181-201. Compare Lunge and Schmid Abstr.1886 203; Bodlander and LUCBS Abstr. 1905 ii 634).-1n the reversible re- action Na,C03 + Ca(OH) = 2NaOH + CaCO when the solution is in equilibrium with solid calcium hydroxide and carbonate the value of the equilibrium constant k = (OH’)2/C0,” can be calculated when the solubilities of the two solid substances are known. In aqueous solu- tion calcium carbonate is partially hydrolysed and to obtain the solubility of the non- hydrolysed salt necessary for the calculation of the constant determinations have been made in excess of sodium hydroxide (by which the hydrolysis is depressed compare Gardner and Gerassimoff Abstr. 1904 ii 544) directly at 18’ and 95-10C0 and by conductivity measurements a t the former temperature. By comparison of the solubility in water with that in sodium hydroxide the degree of hydrolysis has been determined ; according to the conductivity measurements 67% and by direct measurement SO0/ of the salt is hydrolysed a t 18’.From the solubility of the non-hydrolysed salt thus obtained the value k - 16040 a t 1 8 O has been calculated. The equilibrium has also been reached directly from both sides with 1 2 and 3N’sodium and potassium hydroxides and lime a t intervals of temperature from 15-150’ in iron bombs the relative proportions of hydroxide and carbonate being determined by titration. The reaction is most complete with small concentration of carbonate; not only is the yield decreased by increase of carbonate I 2 and 3N-sodium carbonate and excess of lime giving 99 97 and 93*6O/ of hydroxide respectively at loo” but the equilibrium is somewhat displaced in favour of the carbonate.The yield is not appreciably increased by rise of temperature. P. H. M. A. W.INORGANIC CHEMISl'RY. 23 The technical bearing of these results is discussed. It is pointed out that the patents for increasing the yield by working under pressure are useless ; efforts should be directed to facilitating the attainment of equilibrium which is favoured by high temperature thorough stirring and presence of excess of lime. The possible loss of part of the soda as an insoluble double salt is also discussed. The above results are in satisfactory agreement with those of Lunge and Schmid but not in all respects with those of Lucas (Zoc. cit.). G. S. Solubility of Lithium Sulphate in Mixtures of Water and Alcohol. FRANS A.H. SCIIREINEMAKERS and WTLLEN A. VAN DORP junr. (Chem. Centr. 1906 ii 1235 ; from Chem. Weekblad. 3 557-561).-The following conclusions are arrived a t (1) hydrated lithium sulphate is not altered by mixtures of water and alcohol con- taining less alcohol than a limiting value whereas the anhydrous salt under these conditions becomes hydrated ; (2) in solutions of alcohol and water containing more alcohol than the upper limiting value the anhydrous salt is not affected whilst the hydrated form loses its water; (3) in solutions intermediate between these two both salts remain unchanged. The solubility of lithium sulphate diminishes rapidly with increase in the amount of alcohol ; it is practically insoluble in SO0/ alco holm P.H. Influence of Lithium Sulphate on the Formation of Layers in the System Water-Alcohol-Ammonium Sulphate FRANS A. H. SCHREINEMAKERS and J. TH. BORNWATER (Chem. Centr. 1906 ii 1306 ; from Chem. TVeekbhd. 3 569-575. Compare preceding abstract ; Abstr. 1906 ii 855).-In continuation of their earlier work the authors have found that whereas there are two liquid phases in the system water-alcohol-ammonium sulphate there is only one phase in the system water-alcohol-lithium sulphate or water-alcohol-lithium- ammonium sulphate. The effect of lithium sulphate on the forma- tion of two layers in the system water-alcohol-ammonium sulphate has been studied and the results are recorded in tabular form. At first the addition of lithium sulphate increases the amount of alcohol in the alcoholic layer and diminishes the amount in the aqueous layer; then the reverse phenomenon takes place and the composition of the two layers becomes more and more alike.P. H. Silver Chromate. 11. BENJAMIN M. MARGOSCHES (2ei.f. anorg. Chem. 1906 51 231-235. Compare Abstr. 1904 ii 731).-The behaviour of silver chromate towards certain inorganic acids has been described in the previous paper and the investigation has now been extended t o some weak acids more particularly acetic acid with regard to which there are contradictory statements in the literature. Silver chromate is practically insoluble in water and in glacial acetic acid but dissolves to a considerable extent in dilute acetic acid. It behaves in a similar manner with propionic lactic and other organic acids Chromates can be detected most readily by conversion into the24 ABSTRACTS OF CHEMICAL PAPERS.lead stilt which is insoluble in acetic acid. When the red modification of the silver salt is dissolved in dilute acetic acid and the solution con- centrated the greenish-black modification separates. Action of Alkali Chlorides on the Double Silicates of Calcium and Aluminium. F. H. CAMPBELL (Lmadw. Versuchs.-Xtat. 1906 65 247-252).-The object of the experiments was to ascertain the amounts of calcium liberated from calcium aluminium silicate (prepared by precipitating a solution of the chlorides with sodium silicate) by shaking for some days with solutions ( N / 5 to 8N) of lithium sodium potassium and ammonium chlorides. Lemberg (Zeit. deut.geol. Ges. 28 579) and Dittrich (Abstr. 1903 ii 176) have shown that the power of parting with calcium is greatly reduced or even lost altogether after ignition. It is now shown that even drying at 100’ considerably reduces the amount of - calcium dis- placed by 3-sodium chloride. The action is very slow at 25* but the amount of calcium replaced is very little less. than a t 32.5’ the tem- perature finally adopted. The amount of calcium displaced rises in each case with the increase in concentration of the alkali salt ; but less with higher than with lower coacentratione. With sodium potassium and ammonium chlorides the maximum is reacbed with 3s-solutions whilst in the case of lithium chloride the maximum is with a 6N-solution. With low concentrations the displacing power of lithium sodium and potassium chlorides increases with the atomic weight; it is greatest in the case of ammonium chloride.A t higher concentrations sodium and potassium chlorides have nearly the same maximum; lithium chloride has a higher and ammonium chloride a much higher maximum. N. H. J. M. Molecular Weights of Inorganic Compounds in Boiling Quinoline. ERNST BECKMANN [with WERNER GABEL] (Zeit. anorg. Chern. 1906 51 236-244).-The quinoline was dried over phosphoric oxide or potassium hydroxide and fractionated from ‘ $ active ” aluminium. The mean value of the molecular elevation of the boiling point determined with benzil benzoin and other organic compounds is 56.1. The chlorides bromides and iodides of zinc and cadmium and the chlorides and bromides of cobalt and nickel have the normal molecular weights in this solvent whilst cuprous chloride gives results which indicate that the CuC1 molecules present in dilute solution associate to Cu,Cl molecules in more concentrated solution.Isomorphous Crystals of Lead and Barium Nitrates. P. GAUBERT (Compt. rend. 1906 143 776-777).-The investigation of selective absorption exhibited by the cubic and octahedral faces of lead nitrate crystals towards organic colouring matter (Abstr. 1906 ii 152 343) is extended also towards isomorphous nitrates such as barium nitrate. The so-called mixed crystals of lead and barium nitrates are not homogeneous but vary in composition with the nature of the face to which they correspond. G. s. G. S.INORGANIC CHEMISTRY. 25 The experiments were conducted as follows from the crystals deposited at 18-25' by a saturated solution of 500 grams of barium nitrate and 10 grams of lead nitrate 1 gram each of the crystals below the octahedral faces and the cubic faces respectively were separated and the amount of lead in each determined by means of a colorimetric method ; the results showed that the parts of the crystals corresponding with the octahedral faces contained more lead than those under the cube faces and similar results were obtained with crystals of lead nitrate which had been grown in a solution containing a little baryta and nitric acid. M.A. W. Influence of Small Quantities of Elements in Copper on its Reactions with Nitric Acid. JOHN H. STANSBIE (J. SOC. C h m . Ifid. 1906 25 1071-1075).-The investigation is a continua- tion of experiments described in a former paper (Abstr.1906 ii 166) the apparatus having been modified somewhat to facilitate the deter- mination of the amount of nitrous acid formed in the reactions. For this purpose the solution was not allowed to come into contact with air until the free acid had been neutralised by addition of a measured quantity of sodium carbonate in excess. The nitrous acid was then estimated by means of ferrous ammonium sulphate and potassium permanganate solutions and the unneutralised sodium carbonate gave the amount of acid used in the reaction. Experiments with pure copper a t temperatures from 15-85O indicate that the secondary change by which nitric oxide is produced is practically uninfluenced by a rise of temperature whereas the amount of nitrous acid formed decreases considerably. The nitrous acid is supposed to result from the oxidation of nitric oxide according to the equation 2N0 + H20 + HNO i= 3HN02.A t all temperatures a certain amount of nitric acid disappears in changes other than the formation of nitric oxide and nitrous acid. This may be ascribed in part to the formation of ammonia but the quantities of ammonia observed are not large enough to explain the entire loss. Pure bismuth dissolves according to the equation 2Bi + $HNO,= 2Bi(N03),+4H,0+2N0 and the whole of the nitric acid used is accounted for by the nitric oxide and nitrous acid formed Arsenic also reacts in a simple manner and the nitric acid used corresponds with the equation 6As + 1 OHNO = 3As,O + 5 H,O + 10NO.Experi- ments with copper alloys containing 0-3"/ of arsenic or antimony at 6 5 O show that the decrease and increase in the amount of nitrous acid formed as the amount of the second element increases follow pretty closely the decrease and increase in the volume of -the liberated nitric oxide (Zoc. cit.). I n the case of copper-bismuth alloys the amount of nitrous acid formed increases slightly until the percentage of bismuth reaches about 0.25 and then decreases exactly as was observed for the volume of nitric oxide liberated. The author considers that the conclusions drawn in the first paper are supported by these further experiments. H. M. D. Mercury Chromates. A. GAWALOWSKI (Chem. Centr. 1906 ii 1307 ; from Pfmwn. Post 39 602).-Nercuric perchomate is obtained26 ABSTRACTS O f CHEMICAL PAPERS.as it scarlet-red granular precipitate on adding a solution of potassium dichromate and some 25O/ sulphuric acid to a solution of a mercuric salt. The dried substance on ignition evolves water oxygen and mercury vapour and leaves a brownish-black modification of chromic oxide. If sulphuric acid is not present R brick-red earthy pre- cipitate of mercury dichromate is obtained which on ignition yields a finch-green modification of chromic oxide. When metallic mercury is left for several days covered with a cold solution of potassium dichro- mate and an excess of concentrated sulphuric acid a bright red variety of mercuric chromate is formed which is probably a poly- chromate ; this substance on ignition leaves about 4 2 O / of a leaf-green variety of chromic oxide.P. H. Chemistry of the Rare Earths. 11. GREGOIRE N. WYROUBOFF and AUGUSTE VERNEUIL (Ann. Chim. Phys. 1906 [viii] 9 289-361. Compare Abstr. 1906 ii 88).-This paper is chiefly a det.ailed account of work already published relating t o the nature of the oxides of cerium and the separation and purification of cerium compounds (compare Abstr. 1897 ii 452 ; 1898 ii 222 ; 1899 ii 224 423 598,613) ; also polemical against Brauner regarding the atomic weight of the element (compare Abstr. 1897 ii 492; 1898 ii 294; and Brauner Abstr. 1903 ii 295). The red sulphate of cerium (com- pare Rammelsberg Abstr. 1873,601 ; Muthmann and Stutzel Abstr. 1900 ii 544 ; Meyer and Aufrecht Abstr. 1904 ii 175) forms hexagonal crystals which are strongly dichroic appearing a beautiful red when viewed across the base and deep orange or yellow in a per- pendicular direction; the composition of the salt is represented by the formula Ce,0,,3Ce0,8S0,,S0,H2,28H20 and it is to be’regarded as the sulphate of the violet oxide Ce70 (Ce30,,4CeO).M. A. W. Elements of the Ytterbium Group. CARL AUER VON WELSBACII (Monatsh. 1906 2’7 935-945).-Most of the supposed elements of the ytterbium group which have been described are compounds. The present paper contains a description of a part of the author’s work on the isolation of the true elements of t h i s group. The material for the in- vestigation consisting of crude ytterbium oxdate obtained from monazite and containing cerite earths thorium and phosphoric acid was treated in the manner previously described for the working up of crude earths (Abstr.1884 717; 1885 350 1113j and the elements of the ytter- bium group separated finally as the basic nitrates. These were purified by recrystallisation as follows one-half of the substance was stirred with water and neutralised with nitric acid; the other half was then recrystallised from the neutral solution of the nitrate and a series of fractions of the basic nitrate obtained the final liquid giving only a weak erbium spectrum. This process was repeated twenty-one times when the fractions which were free from yttrium all contained erbium ytterbium holmium and dysprosium. To separate the elements of the ytterbium group advantage is taken of the differences in the solubility in a saturated solution of ammonium oxalate of the double oxalates which are formed by precipitation of the moderately concentrated solution of the nitrates with a hotINORGANIC CHEMISTRY.27 saturated solution of ammonium oxalate. The solubility of ytterbium ammonium oxalate in a saturated solution of ammonium oxalate is ten times that of holmium ammonium oxalate. The method of fractional crystallisation is described in detail. Of the seriesof about one hundred fractions obtained the first is almost colourless about the next fifteen are light yellow then orange-yellow passing through the rose-coloured erbium to the colourless ytterbium salts at the other end. Observed directly by transmitted light the first fractions show a changed dysprosium with traces of the neodymium spectrum ; the light yellow fractions give the normal dysprosium the orange-yellow the holmium and the intensely rose-coloured the characteristic erbium spectrum.As the rose-colour weakens the erbium spectrum changes into a modified spectrum which is termed Ery after which the thulium lines appear to disappear in the colourless ytterbium fractions. The spark spectra of the first fractions are complicated and contain lines which do not coincide with those of any known element. The orange-yellow fractions which give the most intense holmium absorption spectrnm give a yttrium spark spectrum in which the holmium lines are absent but in which other groups of lines appear The rose- coloured fractions give the erbium the colourless soluble fractions the ytterbium spark spectra.The mother liqnors from the crystallisation of the ammonium oxalates were further investigated and found to contain ytterbium traces of scandium thorium uranium iron and substances such as alumina and silicic acid which are derived partly from the water and partly from the ammonium oxalate. The fraction which should have contained pure ytterbium gave a spark spectrum which shows marked difierences from the ytterbium spectrum; this is considered to be the first certain indication of the degradation of ytterbium G. Y. Catalytic Action of Muminiurn Chloride. PAUL ROHLAND (Cimn. Zeit. 1906 30 1173-1174).-The addition of 0.2 molecule of alumi,nium chloride per litre to a mixture of quicklime and water con- siderably accelerates the setting of the lime.The rate of hydration of the two sulphates of calcium containing respectively l&H,O and 2H,O is also increased whereas that of the anhydrous sulphate is decreased. The setting of Portland cement is accelerated by small quantities of aluminium chloride but is retarded by quantities above 7"/,. The accelerating effect of aluminium chloride is increased by the presence of sodium carbonate or of aluminium sulphate. A summary is given of the other reactions in which aluminium chloride acts as a catalytic agent. P. H. Crystallised Aluminium Sulphate. A. GAWALOWSKI(C~~. Centr. 1906 ii 1236 ; from Zed. Oesterr. Apoth. Verein. 44 460).-Crystal- 1 ised aluminium sulpliate A12(S0,)3,1 8H20 is obtained by dissolving " aluminium carbonate " in sulphuric acid adding nitric acid and evaporating until no more nitric oxide is evolved; it crystallises in smooth tetrahedra with truncated edges.The salt obtained by heating the carbonate with sulphuric acid alone differs from aluminium sul-28 ABSTRACTS OF CHEMICAL PAPERS. phate in evolving nitric oxide when treated with nitric acid; the author considers that the aluminium in the " carbonate " is combined in the form AlO*OH. P. H. Revision of the Atomic Weight of Manganese. GREGORY P. BAXTER and MURRAY A. HIKES (J. Amer. Chem. Xoc. 1906 28 1560-t580).-From a computation of the values of the atomic weight of manganese obtained by previous workers Clarke has found the most probable value to be 54.987. The present investigation was carried out with manganous chloride and bromide.These salts have been analysed previously by Dewar and Scott (Abstr. 1883 S56) who obtained for the atomic weight the values 54-91 from the chloride and 54.97 from the bromide. Four specimens of manganese bromide were employed. Two of these ( A and B) were prepared from pure potassium permanganate in the following manner. The solution mas treated with sulphur dioxide ammonium carbonate was added to the resulting solution of manganous sulphate and the precipitate was well washed and then dissolved in nitric acid. The manganous nitrate thus obtained was crystallised repeatedly from a solution acidified with nitric acid was afterwards treated with ammonium carbonate and the precipitate was thoroughly washed and dissolved in hydrobromic acid in a quartz dish.The solution of the bromide was heated to expel bromine and the salt was repeatedly crystallised and finally dried in a vacuum over potassium hydroxide. The third sample (6') was prepared from pyrolusite by dissolving it in hydrochloric acid boiling the solution to expel chlorine passing hydrogen sulphide into the solution and fractionally precipitating with sodium hydroxide until the precipitate was obtained free from iron. The manganese was then precipitated with ammonium carbonate and the manganous carboncte was treated as in the preparation of the previous samples. The fourth sample (D) was prepared from manganous sulphate by treating the solution with hydrogen sulphide converting the sulphide into carbonate and heating the latter as in the former cases.Two samples of manganous chloride were prepared one from the carbonate obtained in the preparation of sample B and the other from the mother liquors of the manganous nitrate resulting from the preparation of samples 12 and D. The hydrobromic hydrochloric and nitric acids and the silver employed were carefully purified by methods which are described. The analysis of the bromide and chloride was effected by titrating weighed portions of the salts after fusion in hydrogen bromide or chloride against weighed portions of pure silver ; the precipitated silver salts were collected and weighed. I n some experiments the silver nitrate was added to the manganous salt whilst in other cases the operation was carried out in the reverse manner. Various precautions were taken t o ensure accuracy and vacuum corrections were applied.Manganous bromide bas Dj6 4.385 and manganous chloride Di5 2.977.INORGANIC CHEMISTRY. 29 Thirty-one analyses were made of the bromide and fourteen of the chloridg and the results are tabulated. The average of the results obtained with the bromide gives Mn=54*957 and the average of those obtained with the chloride gives Mn = 54.958LAg = 107.930 ; Br = 79.953 j C1= 35.4731. E. (3 Corrosion of Iron by Acids. C. F. BURGESS and S. G. ENGLE (Trans. Amer. Electrochem. Soc. 1906 9 199-206).-The rate of solution of different kinds of iron in normal solutions of sulphuric and hydrochloric acid has been measured. Electrolytic iron deposited from a solution containing sulphate and chloride the same iron which had been heated to about 1000° and allowed to ccjol slowly soft sheet-iron containing little carbon tempered steel and ordinary cast-iron were used in the experiments.The electrolytic iron before heating dissolved about four times as rapidly as the steel six times as rapidly as the cast-iron and about forty times as rapidly as the soft sheet-iron. After heating the rate of solution of the electrolytic iron diminished to about one-fortieth of its original valua. The authors suppose that this is indirectly due to the removal of hydrogen by the thermal treatment but suggest that change of the crystalline structure is the primary cause. A coarsely crystal- line sample of electrolytic iron was found to dissolve more rapidly than a more dense deposit.Electrolytic iron is suggested as a means of obtaining pure hydrogen by the action of acids. Traces of arsenic exhibit a very marked influence in protecting iron from the corroding action of acids. No definite relationship between the difference of potential ekhibited by the various samples of iron in acid solution and the rate of corrosion could be detected. H. M. D. Roussin’s Salts. ITALO BELLUCCI and C. CECCHETTI (Atti R. Accud. Lincei 1906 [v] 15 ii 467-4’74. Compare Abstr. 1905 ii 253).-Roussin’s salts or nitrosulphides of the first series when treated with alkali hydroxide pass into nitrosulphides of the second series Fe(NO),SR‘. When the sodium salt Fe4(NO),S,Na,2H20 is treated with hydrazine o r hydroxylamine in either acid or alkaline solution i t yields the corresponding hydrazine or hydroxylamine compound whilst with phenylhydrazine hydrochloride it gives the phenylhydrazine derivative (compare Hofmann and Wiede Abstr.1896 i 391). The hyldrazine salt Fe,(N0)7S,H,N,H4 dissolves slightly in water and readily in alcohol or ether. The hydroxylamine derivative is readily soluble in water alcohol or ether. The phenylhydraxine compound Fe4(N0)7S,H,N,H,Ph melts under hot water ia which and in benzene it is sparingly soluble and dissolves in alcohol or ether. The semicarbaxide salt Fe,(N0)7S,H,N,H3*CO*NH2 dissolves in water alcohol or ether. All these new compounds form shining black crystals and are Fe*( N 0)7S,R’ Fe,(NO)$3 H,N H30,30 ABSTRACTS OF CHEMICAL PAPERS. markedly stable. unaltered for a long time in the absence of light.They can be crystallised from water and remain T. H. P. Behaviour of Chromium towards Sulphuric Acid. ALFRED BURGER (Ber. 1906 39 4068-4072. Compare Doring Abstr. 1902 ii 660; 1906 ii 451 ; Mazzucchelli Abstr. 1905 ii 57Oj.-The author has attempted without success to estimate chromium by measurement of the hydrogen evolved on solution of the metal in an acid. When dissolved in boiling sulphuric acid a sample of chromium prepared by Goldschmidt's process and containing 99"/ of chromium forms a blue or a t lower temperatures a greenish-blue solution. I n both experiments more hydrogen is evolved than corresponds with the formation of the chromous salt a pure solution of which cannot be obtained. On concentration of the solution a further evolution of hydrogen takes place but is not complete on addition of dilute sulphuric acid and re-evaporation until fumes of sulphuric acid are given off.The rate of oxidation of the cbromous to the chromic sulphate is found to be greatest in a 48°/0 solution. The oxidation takes place more rapidly in hydrochloric acid solution. The chromous salt present in the solution is estimated best by addition of an excess of ferric sulphate and titration of the resulting ferrous sulpbate with potassium permanganate and the total chromic salt by oxidation to chromic acid by means of persulphate and titration with ferrous ammonium sulphate and potassium permanganate. A sketch is given of the apparatus employed. G. Y. Reduction of Chromium Oxide by Boron. ARMAND BINET D; JASSONEIX (Compt.rend. 1906 143 897-899).-Moissan (Abstr. 1894 ii 454) first prepared a crystalline chromium boride by heating the two elements in a carbon crucible in an electric furnace; Tucker and Moody (Trans. 1902 81 14) showed that the compound has the composition represented by the formula CrB; and Wedekind and Fetzer (Chem. Zeit. 1905 29 No. 98) obtained the same compound by the alumino-thermic process ; the author however finds that when chromium oxide is reduced by boron in magnesia crucibles heated in an electrical furnace the fused masses thus obtained contain 5 to 17'1" of combined boron the boride CrB (B= 17*4"/,) representing the limit of saturation of chromium by boron. These borides are very hard and scratch glass or quartz; those containing about 7°/0 of boron have a well-marked crystalline structure which disappears as the proportion of boron increases and the boride CrB is not crystalline.The densities vary from 6.8 (7'1 boron) to 6.1 (26O/ boron) and all the compounds are attacked by hydrofluoric hydrochloric or sulphuric acid even in the cold with the formation of boric acid. Nitric acid or alkaline solutions have no action on them; chlorine attacks them with in- candescence but incompletely below a red heat forming a mixture of chromous and chromic chlorides ; hydrogen chloride attacks th-em under similar conditions liberating hydrogen and forming chromousINORGANIC CHEMISTRY. 32 chloride and they are oxidised with incandescence by the action of fused alkali carbonates or hydroxides. When the reduction of chromium oxide by boron is carried out in carbon crucibles the fused products have a crystalline structure but always contain carbon.M A. W. Derivatives of Quinquevalent Chromium. 11. RUDOLF F. WEINLAND and M. FIEDERER (Bey. 1906 39 4042-4047. Compare Weinland and Fridrich Abstr. 1906 i 37)-Chromium oxychloride forms double salts with chlorides of the alkali metals. The yotnssium salt CrOCI8,2KC1 forms rhombic prisms of a dark garnet-red colour. Its oxidising power gradually diminishes when it is kept in a desiccator. Corresponding rubidium ccesium and ammonium salts have been prepared and analysed. Their crystalline forins are similar to those of the corresponding double salts derived from molybdenum oxychloride (Klason Abstr. 1901 ii 162; Nordenskjold ibid.454) and from columbium oxychloride (Weinland and Storz Abatr. 1906 ii 764). The cssium compound has been shown to be isomorphous with the compound CbOC12,2CsC1. J. J. S. Reduction of Metallic Sulphides. OLIVER W. BROWN (Trans. Amer. Electrochelm. h'oc. 1906 9 109-1 15).-The product obtained by heating molybdenite in graphite crucibles at a high temperature contains considerable quantities of sulphur. Complete reduction can however be effected by heating the sulphide with a mixture of lime and coke. An excess of lime is advantageous but an excess of carbon leads to the formation of carbide which retains a considerable quantity of molybdenum. This method of reduction is superior to the ordinary method of roasting and subsequent reduction with carbon in that it prevents loss of metal by volatilisation of the oxide.When galena is similarly heated in the electric furnace with lime and coke very little lead is obtained and this is probably due to the formation of a stable double sulphide of lead and calcium. The reduction of st,ibnite is also very incomplete. H. M. D. New Molybdenum Silicide. OLIVER P. WATTS (Trans. Amer. Electrochem. Xoc. 1906 9 105-1 07. Compare Vigouroux Abstr. 1900 ii 144).-A molybdenum silicide of the probable formula MoSi has been obtained by heating a mixture of the oxides of molybdenum silicon and boron with metallic copper and aluminium in an arc furnace with the addition of cryolite as a flux and lime as a retarder to prevent the reaction from becoming too violent By successive treat- ment of the product with nitric acid and dilute hydrofluoric acid dark coloured crystals with a metallic lustre were obtained.These corre- spond closely with the formula RloSi but contain small quantities of boron and iron. The silicide is not acted on by boiling nitric acid boiling aqua regia or boiling hydrofluoric acid but reacts with incandescence with fused sodium carbonate and slowly with fused sodium nitrate D2"' 6.31. H. M. D.32 ABSTRACTS OF CHEMICAL PAPERS. Reduction of Molybdic Acid in Solution by Molybdenum and the Titration of the Reduced Solution by Permanganate. MARCEL GUICHARD (Compt. rend. 1906 143 744-746. Compare Abstr. 1901 ii 659).-When a solution of the compound Mo0,S08 containing not less than 560 grams H,SO per litre is placed in contact with a large excess of powdered molybdenum in a closed vessel for a period of several months a brown solution is obtained which contains a salt of the oxide &Io,O and not of the oxide MOO as stated by Rammelsberg (Avzn.Phys. Chem. 1864 127 281). This fact taken in conjunction with the results obtained by Bailhache (Abstr. 1902 ii 243) and Klason (Abstr. 1901 ii 162) render it probable that the oxide MOO does not form salts. The titration of the reduced solution was conducted in an atmosphere of hydrogen by means of a perman- ganate solution which had been standardised against iron reduced in hydrogen. M. A. W. Preparation of Hydrated Hypovanadic Acid. GUSTAVE GAIN (Cornpt. rend. 1906,143 823-825).-The sulphite 2V20,,3S0,,10H,? prepared by dissolving in a saturated solution of sulphur dioxide in air- free water the mixture of oxides V,O and V,O obtained by calcining ammonium metavanadate at a dull red heat forms fine silky clear blue needles and when an aqueous solution of the salt is boiled sulphur dioxide i d evolved and crystals are deposited which after drying on porous tile form a pale red crystalline powder of hydrated hypovanadic acid V,O 2H,O.M. A. W. Alloys of Palladium and Copper. RUDOLF RUER (Zed. anorg. Chem. 1906 51 223-230).-The freezing-point curve of the palladium-copper alloys falls regularly from the melting point of palladium (which in accordance with the most recent determinations is taken as 1541O) to the melting point of copper 1084O the part representing alloys rich in copper being nearly horizontal.Although the course of certain parts of the curve is somewhat uncertain owing to a tendency to super-cooling there is no evidence of chemical com- bination the metals forming a complete series of mixed crystals. On etching alloys containing 30-70°/0 of palladium with dilute aqua regia slender needle-shaped crystals were observed. The pro- portion of these crystals was not appreciably altered by heating for two hours a t 1500O or for the same period at 1180° just above the melting point of the mixture and did not attain a maximum with a definite composition of the alloy so that they do not indicate the formation of a chemical compound. of palladium the alloys retain the structure of copper. Those containing 10°,lo of palladium are red in colour; when more than 2O0l0 of the same element is present they are white.The alloys are rather harder than the metals themselves and this property attains its maximum when the components are present in equal parts by weight. Even up to G. S.MINERALOGICAL CHEMISTRY. 33 Solid Solutions in the Dissociation of Palladous Oxide and Gupric Oxide. LOTHAR WOHLER (Zed. Elektrochem. 1906 la 781-786).-Further experiments have been made on the dissociation of palladous oxide (compare Abstr. 1906 ii 94). The principal results are that the pressure observed depends on the relative quantities of oxide and metal present and also on the time of heating. The greater the relative amount of metal the lower is the pressure. The results are best explained by supposing that palladium slowly dissolves in the oxide forming a solid solution the dissociation pressure of which is lower than that of the pure oxide. The dissolution takes place very slowly so that the pressure observed depends on the time of heating. The dissociation pressures of cupric oxide were measured at tempera- tures between 9603 and 1084'; its m. p. was found to be 1064'. That solid ,solutions are formed in this case was shown by heating 2 grams of the oxide in a vacuous quartz tube (volume 20 c.c.) at lOOO' equilibrium is reached in a few minutes at 111 mm. The tube was again evacuated and equilibrium re-established ; the repetition of these operations gave pressures of 106 104 101 95 and 91 mm. Similar results were obtained at other temperatures. An experiment with 7.3 grams of the oxide gave a dissociation pressure of 50 mm. at 960'; taking this as the dissociation pressure of the pure oxide and using the beat of dissociation of copper oxide 2CuO = Cu,O + 0 - 33.6 Calu. the dissociation curve of the pure oxide is calculated from van't Hoff's equation. The curve so obtained gives higher pressures than any of those measured with mixtures of copper and copper oxide but the experimental curves approach it more nearly the purer the copper oxide used. T. E.