INOltGANIC CHEMISTRY, Inorganic Chemistry. Biological and Chemical Purification of Water. By A. TIXIER (J. Plzawn., 1899, [vi], 10, 297--300).-1n order t o avoid the liberation of free alkali which OCCU~S when potassium permanganate or calcium permanganate is used for the purification r>f water, the72 ABSTRACTS OP Crf EM1CAL PAPERS. author uses a solution containing aluminium permanganate and barium permanganate. The solution employed has a specific gravity of 35" B., and contains 290 grams of permangnnic acid per litre, and 7 per cent. of alumina. It is added to the water to be purified until a persistent pink coloration is produced ; the water is allowed to remain for 24 hours, and after filtration through a carbon or other filter is fit for consumption, By Nrcor,A TECLU (S.p. Chem., 1899, [ ii], 60, 402--403).-Several new forms of apparatus for exhibiting the phenomenon of ozone production during electric discharge are figured. By JOSEF M. EDER and EDUARD VALENTA (Chenz. Centr., 1899, ii, 358).-The spectrum of chlorine was ex- amined a t pressures varying from 10 to 100 mm., a concave grating being employed. The authors confirm Ciamicinn's observation of the broadening of many lines and increase of brightness of the continuous spectrum by increase of pressure, but no band spectrum was obtained. Very characteristic lines were those of wave-length 4132 in the violet and 3860 in the ultra-violet, and the wave-length of many lines are given which a t low pressures are seen to be doubled or trebled ; a line a t 3750 does not become broader by increased pressure.By E. H. SARLES (J. Amer. Chenz. Xoc., 1899,21, 1038).-When chlorine is passed into ethyl alcohol, the liquid finally separates into a lower yellow layer and an upper layer having a grass-green colour. The green solution has bleaching proper- ties which it, however, loses when uncombined chlorine is removed by a current of air or carbon dioxide. The solvent, which gives a green solution of chlorine, has not been isolated ; it is either easily decom- posed or very volatile, since when the liquid has been fractionated several times, the green colour is completely lost and cannot be re- gained by passing fresh chlorine into the liquid. Theory of the Electrolytic Formation of Hypochlorite and Chlorate. By FRITZ FOERSTER (Zeit.anorg. Chem., 1899, 22, 1-32).-1nvestigation of the formation of hypochlorite and chlorate as a purely chemical process (Abstr., 1899, ii, 278) has prepared the way for a theory of the electrolytic preparation of these substances. A review of the work dealing with the electrolysis of alkali chlorides (compare Oettel, Abstr., 1896, ii, 517; Haber and Grinberg, Abstr., 1898, ii, 215 and 365; Wohlwill, Abstr., 1899, ii, 213) shows that the following conclusions may be regarded as established. Hypo- chlorite is formed chiefly by the interaction of the chlorine liberated at the anode with the alkali produced during the electrolysis or previously present : in alkaline solution, hypochlorite can be oxidised t o chlorate by a primary anode process ; the formation of chloric acid in the electrolysis of dilute hydrochloric acid is due to the direct oxidation of C1 ions a t the anode.A theory of the secondary electrolytic formation of chlorate is discussed, based on two suppositions ; (1) that in the electrolysis of H. R. LE S. New Ozone Apparatus. A. L. Spectrum of Chlorine. L. M. J. Colour of Chlorine Soh tions. J. J. 8. -INOKGANIC! CHEMISTRY. 73 neutral alkali chloride solutions, chlorine is continually being liberated at the anode, (2) that the hypochlorite takes part in the electrolysis, and that its ions are discharged at the anode with evolution of oxygen. The known facts regarding the influence of current density, temperature, acidity, and alkalinity, on the formation of chlorate, are largely in harmony with the theory ; further experiments to determine the exact nature of the process have been made by Muller (see fol- 1 owing abstract).Formation of Hypochlorite and Chlorate in the Electrolysis of AIkaIi Chlorides. By ERICH MULLER (Zed. anorg. Chem., 1899, 22, 33-90. Comp:we Abstr., 1899, ii, 742).-The supposition that in the electrolysis of neutral alkali chloride solutions the formation of hypochlorite and chlorate depends on the same purely chemical process as in the non-electrolytic formation of these compounds, has several consequences. Experiment on the whole confirms these consequences, and shows that the electrolytic formation of chlorate in neutral alkali chloride solutions consists in the following processes ; chlorine is discharged at the anode, and unites with the alkali from the cathode to form hypochlorite ; even while the hypochlorite con- centration is still small, ClO ions are discharged a t the anode, forming hypochlorous acid, which in its turn oxidises hypochlorite and chloride t o chlorate; a t low temperatures, hypochlorite is to a small extent oxidised by anode oxygen.Experiments in acid and alkaline solutions of the alkali chlorides lead t o the conclusion that the electrolytic process in these does not differ essentially from that in neutral solution. The electrolytic formation of chlorate may thus be referred in general to four equa- tions : (1) 6 + 3HC10 = 610, + 3 H + 3El; (2) 610 + 2HC10 = 610, + 2 H + 2C1; (3) C1+ 3 0 = C10, ; (4) C10 + 2 0 = C10,. I n dilute hydro- chloric acid, the existence of C10 ions is excluded, and in this case the formation of chlorate must take place according to equation (3).I n other cases, the processes may be represented in varying degree; thus, when there is no appreciable evolution of oxygen, the chief pro- cesses are those indicated by equations (1) and (2); when the evolu- tion of oxygen is considernbls, as i n strongly alkaline solutions, the formation of chlorate will take place chiefly according t o equations (3) and (4). Molecular Weights of some Elements and their Derivatives. By GIUSEPPE ODDO and E. SERRA (Gccxzetta;, 1899, 29, ii, 343-353).- The molecular weight of iodine has been determined by the boiling point method, the solvents employed being tetrachloromethane, carbon disulphide, benzene, and ethyl alcohol.The numbers obtained with the first two solvents are sensibly the same as those of Beckmann and Stock (Abstr., 1895, ii, 382). Benzene and ethyl alcohol yielded the respective values 273-279 and 265-327, Beckmann and Stock's numbers being 335-360 and 330-342; the latter results, being corrected for the volatility of iodine, give the values 233-255 and J. C. P. -_ + - - - - + - - J. C. P.74 A4 13 Y T B ACTS 0 P C k1 EM IC A L P APE H S . 233-241 respectively. The authors' object to this very large correction being applied, as with slow boiling, the amount of iodine volatilised is very small and scarcely sufficient to colour the vapour of the solvent. I n boiling tetrachloromethnne, the sulphur molecule contains 8 atoms up to concentrations of about 3 per cent.; beyond that strength of solution, the results obtained are not concordant. Phosphorus pentachloride has the normal molecular weight in boiling tetrachloromethane. Solutions of iodine mono- or tri-chloride in tetrnchloromethane give lower boiling points than that of the pure solvent, I n the case of the trichloride, this is due to the fact that it sublimes a t 70-75". With the monochloride, dissociation takes place according to the equation 9ICl= I, + I, + 3IC1,, and the effect of the trichloride in lowering the boiling point is greater than that of the iodine in raising it. Action of Arsenious and Antimonious Oxides on Sulphur Monochloride. By GIUSEPPE ODDO and E. SERRA (Gaxxetta, 1899, 29, ii, 355).-When arsenious oxide and sulphur monochloride are heated together in a reflux apparatus, they react according to the equation : As,06 + 6S2C12= 4AsC1, + 3S0, + 9s.The reaction is complete in about an hour, and on cooling, nearly ail the sulphur formed crystallises out and the arsenic trichloride can be separated by decantation. Analogous reactions take place when antimony or bismuth sesquioxide is substituted for the arsenic compound. Hydrates of Sulphuric Acid. By EUGEN VON BIRON (Chem. Centr., 1899, ii, 467-468 ; from J. Russ. C'hem. Xoc., 31, 517-522).- The author has succeeded in crystallising the hydrate H,S04,2H20, predicted by Mendel6eff. A solution of the composition H,S0,,2H20, cooled with liquid air, solidifies to an amorphous mass. This mass, if rubbed a t rather a higher temperature with a glass rod, becomes crystalline, the thermometer rising a t the same time to - 35".The crystals thus obtained may be used to start crystallisation in a solution of the composition H2S0,,2H,0 cooled merely to - 75" with solid carbon dioxide and ether. Cooling with liquid air is detrimental t o the formation of the crystals. The freezing point of the hydrate was determined in a Beckmann's apparatus of small size, well protected by surrounding tubes, and cooled in a mixture of carbon dioxide and ether. During solidification, the thermometer remained steady for about 10 minutes, until practically no liquid was left, showing that the separation of the solid did not alter the freezing point of the remaining liquid. I n solutions which deviated from the composition H,S04,2H20, the thermometer was steady for only 1-2 minutes.The freezing point of the hydrate is - 38.9'; with the same apparatus, the hydrate N,S0,,4H2O solidified a t - 69". Polymerisation of Inorganic Chloro- Anhydrides. I. By GIUSEPPE ODDO and E. SERRA (Gazxetta, 1899, 29, ii, 318- -329. Corn- pare this vol., i, 92),-Determined by the boiling point method, the T. H. P. T. H. P. J. C. P.IN ORGAN I C C H E MIST R Y. 95 molecular weight of phosphorus oxychloride in various solvents is as follows : I n tetra,chloromethane, 325-362 ; benzene, 283-309 ; carbon disulphide, 214-244; chloroform, 159-164 ; ether, 147-157. The cryoscopic method gives a value 149-152 in benzene. The number corresponding with the formula POCI, is 153.5. The molecular weight of thionyl chloride in boiling chloroform is 229-235 and in freezing benzene 108-120; the formula SOC1, requires 11 9.Phosphorus oxybromide has a molecular weight of 287-310 in boiling tetrachloromethane and 334-352 in boiling benzene ; the cryoscopic method gives a value 293-308 in benzene. The calculated value for POBr, is 287. Phosphorus thiochloride gives the following molecular weights. I n boiling tetrachloromethane, 209 ; in boiling benzene, 236-247 ; in freezing benzene, 158-161. The calculated number is 169.5. For sulphur monochloride, the values obtained are, in boiling tetra- chloromethane, 169-173, and in boiling benzene, 180-193 ; the formula S2CI2 requires 135. Chromyl chloride in boiling tetrachloromethane gives the molecular weight 225-243 ; in boiling benzene, 441-528 ; and in freezing benzene, 165-175, The formula CrO,Cl, corresponds with st value 155.5.The cryoscopic method gives for the molecular weight of sulphuryl chloride in benzene the number 131, the value for S02C1, being 135. T. H. P. Properties of Solutions of Sodium Nitrite. By JOSSIF JURI BOGUSKI (Chem. Centr., 1899, ii, 470 ; from J. IZuss. Chem. Xoc., 1899, 31, 543 -551).-Sodium nitrite, prepared from the commercial article by crystallising from solutions saturated at 1 2 5 O , has a yellowish tinge which disappears, however, when the salt is dried over sulphuric acid or washed with alcohol, but even then the colourless salt still forms yellowish solutions. This coloration is dne to the presence of a foreign substance, The variation of the sp. gr. with the concentration of solutions of the nitrite is represented by a curve which is very nearly a straight line. The specific refractive index of solutions a t 20' may be calculated from the formula [nID = 1.33336 + 0*0011559 P, where P= percentage of nitrite dissolved. The original paper con- tains copious data and many tables.NOTE.-AS to the colour of sodium nitrite and its solutions, compare Divers (Trans., 1899, 75, 16), Groves (Proc., 1898, 222).-EDIToRs. By EDMUND GRAEFE and MORITZ ECKARDT (Zait. ccnoq. Chem., 1899, 22, 158-1 60).-A repetition of Winkler's experiments (Abstr., 1890, 331) on the reduction of alkaline carbonates proves Beketoff's assertion that the reduction to metal takes place more easily as the atomic weight increases. The preparation of ctmium is effected at a lower temperature than that of rubidium or potassium.A mixture in the proportion R,CO,: 3Mg is heated in an iron tube, free from rust, in a slow current of dry hydrogen and the metal, which distils E. W. W. Preparation of Cesium from the . Carbonate.76 BUSTRACTS OF CHEMICAL PAPEltS. collected under petroleum ; a theoretical yield is obtained. Metallic czsium is silver-white with B yellow tinge, retains its lustre under petroleum, oxidises with development of heat on exposure t o the air, then melts and explodes, swims on water, and burns with a reddish- violet flame with liberation of hydrogen. Calcium and its Compounds. By HENRI MOISSAN (Ann. Chim. Phys., 1899, [vii], 18, 2€?3-343).-A detailed account of work already published (compare Abstr., 1894, i, 313; 1898, ii, 116, 161, 578; 1899, ii, 25, 152, 153, 155, 219, 241, and 418).By KARL ZULKOWSKI (Chem. Centr., 1899, ii, 602-603; from Chem. Ind., 22, 349--352).-The setting of the ordinary roasted gypsum is due to the formation of the calcium salt, S(OH),:O,Ca, of hexabasic sulphuric acid by the action of water (2 mols.). Gypsum which has been roasted at a moderate red heat is also capable of taking up water, but ou account of its greater density i t only combines with 1 mol. of water, forming the calcium salt, SO(OH)2:0,Ca, of tetrabasic sulphuric acid. The water has a twofold action, first dissolving the more soluble calcium compound, and then combining with i t to form a less soluble compound which separates in a crystalline form and sets t o a compact mass.Preparation and Properties of Crystallised Barium and Strontium Phosphides. By A. JAROIN (Compt. rend., 1899, 129, 762-765).-Crystallised strontium phosphide is obtained by heating a mixture of lampblack and strontium phosphate for 3 or 4 minutes in an electric furnace with a current of 45 volts and 950 ampbres, and the barium compound is obtained in a similar way. StrontiulrL phosphide, Sr3P2, burns in fluorine a t the ordinary temperature, in chlorine a t about 30°, in bromine at 170-175', in iodine at it red heat, in oxygen above 300°, and in sulphur vapour at a higher temperature. It is decomposed by carbon at a high'temperature, but not by sodium a t a red heat ; by dilute acids and gaseous hydracids, but not by concentrated acids, nor by hydrogen sulphide or ammonia, nor by organic solvents.It alters rapidly i n moist air, is decomposed by water with liberation of hydrogen phosphide, and is violently attacked by oxidising agents. Its sp. gr. is 2.68 and it melts only in the electric furnace. Bccrium phosphide, Ba3P2, has similar properties, but is not so readily attacked; it burns in chlorine a t 90' and in bromine at 260-300". I t s sp. gr. is 3.183. Formation of Oceanic Salt Deposits, Particularly of the Stassfurt Beds. XIV. Influence of Pressure on the Forma- tion of Tachyhydrite. By JACOBUS H. VAN'T HOFF and H. M. DAWSON (Chem. Centr., 1899, ii, 401-402; from Sitxungsbey. Akccd. Viss. Uerlin, 1899, 557-562. Compare Abstr., 1899, 759).-The effect of increase of temperature and pressure on the evaporation of sea-water is to cause the separation of new compounds, for when the change of the solubilities of the various compounds affected by the alteration clf conditions exceeds certain limits, supersaturation with new compounds is rendered possible.Increase of temperature and pressure results in the formation of kieserite, liiweite, kainite, and lnngbeiuite, none of E. C. R. G. T. M. Setting of Gypsum. E. 177. W. C. H. B.I N 0 HG A N Ic' C H E M 1 YTR Y , 77 these compounds being formed when sea water is evaporated at 25' under the ordinary pressure. The salt basins of Besangon have a temperature of 62" a t a depth of 1.35 metres, and reckoning on this basis the variations of temperature and pressure in the case of the Stassfurt beds which have a maximum depth of 1500 metres amount t o 40 and 180 atmospheres respectively. The effect of changes of temperature and pressure has been investi- gated in the case of the formation of tachyhydrite (MgCl,),Ca,l 2H20, which easily separates from mixed solutions of the hexahgdrates oE magnesium and calcium chlorides a t 22".Experiments with the manocryometer and Beckmann's thermometer show that an extra pressure of one atmosphere only raises the temperature of the forma- tion of tachyhydrite by 0 . 0 1 7 O . This effect is comparable with that of pressure on melting points ; i t lies between the raising of the melting point of ice by 0.0073° and that of paraffin by 0.035" caused by a pressure of one atmosphere. Hence, since a n increase of pressure of 180 atmospheres would only raise the temperature of formation by 3', the effect 3f increase of temperature must be of much greater import- ance.E. w. w. By ALEXANDER P. LIDOFF (Chem. C'erdv., 1899, ii, 471 ; from J. Rum. C'hem. SOC., 1899, 31, 571--.572).-From the results of experiments oc the action of copper salts on alkaline eolutions of albumin, it appears that the biuret reaction really depends on the dissolution of copper, the copper salt being reduced t o soluble colloidal copper. An alkaline solution of gelatin in which copper gauze was placed had acquired a violet coloration after remaining a day, and after 48 days 3-54 per cent. of the copper was found to have been dissolved. By RICHARD Jos. MEYER and HANS BEST (Zeit. anorg. Chem., 1899, 22, 169-151).- The dark green solution of oxides of manganese in hydrochloric acid contains manganese trichloride. Xach of the oxides, Mn203, Mn304, or Mn02, when dissolved in absolute alcohol or ether saturated wlth hydrogen chloride, yields a solution of manganese trichloride which is decomposed by water and also by evaporation in a vacuum over sulphuric acid, and consequently the trichloride cannot be isolated.I t , however, yields very characteristic double salts with pyridine and quinoline hydrochlorides which crystallise in lustrous needles. The action of hydrogen bromide on the oxides of manganese results in the formation of the dibromide which yields a crystalline double salt with pyridine hydrobromide. When potassium permanganate is boiled with acetic acid, carbon dioxide is evolved and a brown solution obtained, which on cooling deposits the salt, 3Mn02,Mn2(C2H30,)G + 2C,H,02, and if a small quantity of water is added t o the mother liquor, Christensen's tri- acetate, Mn,(C2H30,), + 4H20, crystallises out.The solution obtained by reducing potassium permaaganate with acetic acid with the addition of 1 mol. of potassium acetate, when saturated with hydrogen chloride yields the double salt MnC13,BKC1 (Rice, Trans., 1898, 75, 258). If, however, the solution is only partially saturated with Solution of Copper in Gelatin Solutions. E. W. W. Manganese Trichloride and Tetrachloride. VOL. LXXVIII. ii. 678 ABSTRACTS OF CHEMICAL PAPERS, hydrogen chloride, the salt MnCl,,MnCI3,5KCl is obtained. When the acetic acid solution of potassium permanganate, without the addition of potassium acetate, is saturated with hydrogen chloride, a small qnantity of the salt MnCl4,2KC1 is obtained as a black, crystal- line precipitate, and the mother liquors contain large quantities of manganese trichloride.The corresponding ammonium salts could not be isolated, although the reaction proceeds apparently in a similar manner. Caesium permanganate, under similar conditions, yields the salt MnC13,2CsCl. Thallium permanganate does not yield a double salt . The double sulphate, Xn,(S0,),,K,S04, is precipitated quantitatively by adding sulphuric acid to the solution of potassium permanganate in acetic acid. The higher chlorides of iron, cobalt, and nickel are not obtained by the above method. Lead dioxide, when dissolved in cold alcohol saturated with hydrogen chloride, is converted into the tetrachloride, which yields crystalline double salts with pyridine, di- and tri-methyl- amine hydrochloride, and tetramethylammonium chloride.FhdZ&a permanganate, TlMnO,, jorms large, nearly black prisms. E. C. R. Atomic Weight of Cobalt. 111. Analysis of Cobaltous Chloride and Cobaltous Oxide. By THEODORE W. RICHARDS and GREGORY P. BAXTER (Zeit. anorg. Chem., 1899, 22,221-234. Compare Abstr., 1895, ii, 377, and 1899, ii, 753).-The authors have determined the weight of the cobalt obtained from a known weight of cobaltous chloride by reducing it in a current of hydrogen. The methods em- ployed are essentially the same as those previously described. The mean of the two experiments gives for the atomic weight Co = 59.045.The cobaltous chloride was prepared by decomposing purpureo-cobalt chloride at 200' and eliminating the remaining ammonium chloride by heating in a current of nitrogen and hydrogen chloride. The cobalt- ous chloride was found to contain small quantities of ammonium chloride, alkali salts, and silica. Cobaltous oxide is prepared by precipitating a solution of pure cobalt in nitric acid with ammonia, and after heating the precipitate over a spirit burner, decomposing the resulting black oxide at a red heat in a vacuum. The reduction of the cobaltous oxide by means of hydrogen gave for the atomic weight Co = 58.954 (mean of three experi- ments), An examination of the cobaltous oxide shows that it contains a small quantity of a higher oxide; also, by heating the oxide in a vacuum at 800°, a partial reduction takes place into metal and oxygen, and since a perfectly constant oxide could not be obtained, the authors have abandoned the experiments.The results of all the experiments show that the atomic weight of cobalt lies between 58-93 and 59-06 and the most probable value is Co = 58.995 when 0 = 16. Occlusion of Hydrogen by Cobalt and other Metals. By GREGORY P. BAXTER (Anzer. Chem. J., 1899, 22,351-364).-Although cobalt in the forffi of ingots, which therefore presents a minimum amount of surface, is known to occlude practically no hydrogen, it is E. C . R.INORGANIC CHEMISTRY. 79 found that when reduced from the oxide, and therefore in a very finely divided condition, it occludes relatively large amounts of this gas.Electrolytic foil, which although somewhat porous lies between the two modifications in relative surface, falls between them also in its occluding power. The volume of occluded hydrogen varies, in the case of metal reduced from the oxide, with its purity and the temperature of reduction ; it is remarkable that the metal reduced from cobalt bromide occludes practically no hydrogen ; this appears to bedue to its being deposited in a more compact form than that reduced from the oxide, as the presence of sodium bromide has no perceptible effect on the amount of occluded gas. Since the occlusion of hydrogen progresses very slowly at the ordinary temperature and is practically negligible at the temperature of reduction (400 -500°), it must be a maximum a t some intermediate temperature; the time during which the metal is in contact with hydrogen determines largely the amount of gas taken up.Although practically none of the hydrogen occluded is given off in a vacuum at the ordinary temper- ature, yet. on heating in a vacuum nearly the whole is evolved. The occlusion of hydrogen by nickel appears to be governed by similar conditions to those dealt with in t,he case of cobalt ; with pure copper and silver, the occlusion is practically nil. Chromyl Chloride, Chlorochromic Acid, and Aminochromic Acid. By RICHARD Jos. MEYER and HANS BEST (Zeit. anwg. Chem., 1899, 22, 192--199).-Chromyl chloride is obtained by the action of hydrogen chloride on a solution of chromic acid in acetic acid, but cannot be separated from the acetic acid solution.When pyridine, dissolved in acetic acid, is added to the solution, chlorine is evolved, and the pyridine salt of hexachlorotrichrompl chloride, Cr3O,C1,,3PyHC1, is obtained, which crystallises in brownish-gold leaflets. The same salt is also obtained by adding pyridine hydrochloride to a solution of pure chromyl chloride in acetic acid. Pyridine and quinoline, when added to a solution of potassium chlorochromate in acetic acid, yield pyridine and quinvline chloro- chromates respectively, which separate in yellowish-red crystals, are stable, and can be recrystallised from dilute hydrochloric acid. No evidence of the formation of aminochromic acid was obtained by the action of ammonia on a solution of potassium chlorochromate in dry acetone.E. C. R. Recovery of Chromic Acid from Chromium Residues. By FRIEDRICH REGELSBERGER (Zeit. angew. Chew,., 1899, 1123-1 la€!).- Various methods have been suggested for recovering chromic acid (compare Haussermann, DingZ., 288,163 ; Lorenz, Abstr., 1896,ii, 265 ; Heibling, French Pat. 275274 ; Fitzgerald, Eng. Pat. 1896, 5542 ; Dercum, Eng. Pat. 18'38, 3801 ; Meister, Lucius, and Briining, German Pat. 103860). Two different electrolytic methods are de- scribed in the paper. The one consists in oxidation in alkaline solution: a current is passed through a saturated solution of an alkaline chloride containing an amount of chromic oxide or chromium salt equivalent to the current in unit time. When potassium chloride is employed, potassium dichromate crystallises from the hot solution W.A. D. 6-280 ABSTBACTS OF CIIEMICAJ, PAPERS. after some time, and chlorine is liberated. The metal vessel in which the electrolysis was conducted served as the cathode, and platinum gauze as the anode, and, t o ensure complete admixture, a i r was blown through. The method may be of practical use when the chromium liquors contain considerable amounts of organic matter, or when solid chromium residues hare t o be dealt with. The second method consists in the use of lead anodes in acid solu- tion, either with or without a diaphragm ; the lead is first converted into peroxide, which then oxidises the chromium compounds. The electrolysis proceeds best in hot solution, and almost any metal may be employed as cathode. Iron salts must not be present in the liquid.Preparation of Molybdenum and Uranium with the Aid of Liquid Air. By ALFRED STAIENHAGEN (Bey., 1899, 32, 3065. Com- pare next abstract but one).-The yield of molybdenum from a mixture of molybdic acid, aluminium, and liquid air is poor, owing to the vola- tility of molybdic oxide. A mixture of uranic acid and aluminium explodes with great difficulty, but when liquid a i r is added, the reaction is very violent, and a thoroughly fused uranium regulus is obtained. G. T. M. J. J. S. Molybdenum Dioxide. By MARCEL GUICHARD (Compt. rend., 1899, 129, 722-785).--Various oxides of molybdenum, intermediate in composition between MOO, and MOO,, have been described by former observers as being produced by heating molybdenum trioxide with ammonium molybdate and by the electrolysis of fused molybdenum trioxide.Both these reactions have bean studied by the author, who finds that the sole product, after excess of molybdenum trioxide has been removed by successive washing with soda and hydrochloric acid, is in each case molybdenum dioxide, which was obtained in a pure, By ALFRED STAVENHAGEN (Bey., 1899,32, 3064-3065).--The addition of liquid air t o the mixture of aluminium and tungstic acid employed in the preparation of tungsten (Abstr., 1899, ii, 489), produces, on explosion, so great a rise of temperature that a completelyfusedregulus of tungsten is obtained which contains only traces of aluminium. The author was unable t o obtain t h e element by Hallopeau's electrolytic method (Abstr., 1899, ii, 158). When molten lithium paratungstnte is electro- lysed with a current of 3.5 amperes and a n E.M.P. of 12 volts, bluish- black crystals of lithium-tungsten bronze are produced. By EDGAR F.SMITH (J. Amer. Chern. Xoc., 1899,21, 1007--1008).-An introductory paper, referring t o those treated in the following abstracts and in this vol., i, 76, 89. Atomic Weight of Tungsten. By EDGAR F. SMITH and WILLETT L. HARDIN (J. Amer. Chem. Soc., 1899, 21, 1017-1037. Compare Abstr., 1898, ii, 336).-Recent experiments prove t h a t practically no hydrogen is occluded when metallic tungsten is allowed to cool in a n atmosphere of this gas (compare Waddell, Abstr., 1887,112 j Derenbach, crystal1 ine state. N. L. Preparation of Tungsten with the Aid of Liquid Air. G. T. M. Tungsten. J. J. S.INORGANIC CHEMISTRY.81 fnaug. Diss. FVu~xbu~y, 1892). It has been proved that tungsten tri- oxide exists in two varieties, crystalline and amorphous. These differ in specific gravity and also in their solubility in sulphur mono- chloride; they may be converted into one another. If the insolu- ble oxide is converted i r t o ammonium tungstate and then ignited, the oxide which is formed dissolves in sulphur chloride a t 145O, and when the metal obtained from the oxychloride is heated in a current of oxygen, the oxide formed is insoluble in sulphur chloride a t 145'. The authors conclude that, so far, there is no trustwortby method for the determinationof the atomic weight of tungsten. The methods they have employed are (1) heating pure tungsten in air or in pure oxygen ; (2) precipitating metallic silver from silver nitrate solution by the aid of metallic tungsten (compare Smith, hbstr., 1893, ii, 170) ; (3) estimating the water of crystallisation in barium metatungstate (compare Scheibler, J.prakt. Chem., 1861, 83, 324). I n no case were concordant results obtained. Action of Sulphur Monochloride on Tungsten Trioxide. By EDGAR F. SMITH and HERMAKN FLECK ( J . Amer. Chem. Xoc., 1899, 21, 1008--1013).-When tungsten tricixide, the dioxide, or the mineral wolframite or scheelite is heated with sulphur monochloride, a red solution of tungsten oxychloride, WOCl,, is obtained, together with a small amount of a brown, insoluble substance. The trioxide obtained by heating ammonium tungstate, or the trioxide which has been heated for some time in the aiy, is not completely acted on by sulphur chloride, whereas the trioxide obtained by beating the oxychloride is practically all dissolved by it.This difference is not due to the presence of a nitride or oxynitride. Tungsten itself is not acted on by pure sulphur monochloride, but if free chlorine is present, tungsten hexachloride is formed. By SKLODOWSKA CURIE (Compt. rend., 1898, 126, 1101-1 10S).-The electrical conductivity of air, when induced by the Becquerel rays emitted by uranium compounds, varies directly with the amount of this element present in the active substance. All uranium compounds are active, and the metal itself more so than any of its derivatives, except pitchblende and native chalcolite (copper uranylphosphate) ; the latter substance, however, when prepared artificially, is less active than the metal; these results seem to indicate that the two minerals con- tain an element far more active than uranium. Thorium compounds are very active, the action of thoria being more pronounced than that of metallic uranium ; cerium, niobium, and tantalum compounds are slightly active.Yellow phosphorus is extremely active, but its action is probably of a nature different from that of uranium and thorium ; iE the allotropic form and in the phosphates, it is quite inert ; the com- pounds of all other elements do not appreciably influence the electrical conductivity of air. The effects produced by the rays vary directly with the thickness of the layer of active substance ; the rays traverse thin sheets of glass, ebonite, paper, and the metals.The rays emitted by thoria are more penetrating than those from uranium, and the penetrative power is augmented by increasing the layer of the oxide. J. J. S. J. J. S. Rays Emitted by Uranium and Thorium Compounds.82 ABSTRACTS OF CHEMICAL PAPERS. Distinct photographic impressions are obtained in the case of uranium, uranous oxide, pitchblende, chalcolite, and thoria, but those produced by thorium sulphate and potassium fluoroxytantalate are very faint. When uranium and thorium compounds are subjected t o the influence of Rontgen rays, they emit secondary rays, which produce a more intense effect than those emitted by lead under similar conditions. New Radio-active Substance contained in Pitchblende, By P.CURIE and SKLODOWSKA CUR~E (Compt. rend., 1898, 127, 175-178. Compare preceding a bstract).-A specimen of pitchblende, having 2 i times the emissive power of uranium, was examined chemically wlth a view of isolating the radio-active principle which produces the abnormal activity. The mineral was dissolved in acids and treated with hydrogen sulphide, the thorium and uranium remain in solution, whilst the active substance is precipitated with the sulphides insoluble in ammonium sulphide; after separating these in the usual manner, i t is found that the substance in question remains with the bismuth. When the sulphides are treated with nitric acid, the less active por- tion dissolves more readily ; and when the solution of the nitrates is diluted with water the more active portion is first precipitated; the progress of the separation is controlled by determining the electrical conductivity of air induced by the various fractions.An extremely active product can be isolated from pitchblende by sublimation, and when the sulphides of bismuth and the active substance are heated in a vacuum a t 700°, a sublimate is obtained, the activity of which is 400 times t h a t of uranium. It is believed that the extremely active substance obtained from pitchblende contains an unknown metal to which the name polonium is given. Spectroscopic examination of the substance, however, has not revealed the existence of any characteristic lines indicating the presence of a new element. An extremely Radio-active Substance contained in Pitch- blende.By P. CURIE, SKLODOWSKA CURIE, and GUSTAVE B~MONT (Compt. rend., 1898,127, 1215-1217. Compare preceding abstracts). -In the course of their researches on radio-active substances, the authors have obtained a product having all the properties of barium chloride, and, in fact, consisting mainly of this compound, but differ- ing from the ordinary chloride in being extremely active. By repeated fractional precipitation of the active chloride from its aqueous solu- tion by alcohol, a product is obtained which is 900 times more active than uranium. Ordinary barium salts are never radio-active, and, moreover, spectroscopic examination of the active substance has revealed the presence of a well-defined line not belonging to any known element (compare following abstract) ; the distinctness of the line increases with the radio-activity of the fraction under inspection.For these reasons, it is supposed that the active barium chloride con- tains another radio-active element for which the name radium is proposed. The atomic weight of barium in the active salt is not markedly different from that of the element in its inactive compounds. The compounds of uranium, thorium, polonium, and radium all give photographic effects on sensitive plates, and in this respect G. T. M. G. T. M.INOHGANIC CHEMISTRY. 83 polonium and radium are far more active than the other two; the rays emitted by the new elements render barium platinocganide fluorescent, but the effect is less marked than with Rontgen rays. G. T. M. Spectrum of a Radio-active Substance [in Barium Chloride].By EUQBNE DEMAR~AY (Compt. rend,, 1898, 127, 1218. Compare pre- ceding abstract).-The spectrum of the radio-active barium chloride, together with distinct indications of barium and faint lines due to lead, calcium, and platinum (from electrodes), contains a well-defined line of wave-length 3814.8 (Rowland's scale) which appears bet ween the platinum lines 3819.9 and 3801.5; this line has not been noticed in the spectra of any of the known elements. Spectrum of Radium. By E U ~ N E DEMARCAY (Compt. rend., 1899, 129, 716-71 7).-The following lines characteristic of radium were obtained from the photographed spectrum of a specimen of barium chloride containing that element. The numbers 1 to 16 indicate the order of intensity. A, 4826.3, (10) ; 4726.9, (5) ; 4699.8, (3) ; 4692.1, (7) ; 4683.0, (14) ; 4641.9, (4) ; 4627.4, (4), the centre of a nebulous band ; 4600.3, (3) ; 4533.5, (9) ; 4458.0, (3), the centre of a nebulous band ; 4436.1, (8); 4364.2, (8) ; 4340.6, (12); 3814-7, (16); 3649.6, (12).A number of feeble lines of uncertain origin were also observed. Atomic Weight of the Metal in Radio-active Barium Chloride. By SKLODOWSKA CURIE (Compt. rend., 1899,129,760-762). -Radio-active barium chloride obtained from a large quantity of uranium residues was fractionally crystallised, and the radio-active constituent was found t o accumulate in the less soluble portions; when the latter were dissolved in water and fractionally precipitated by alcohol, the active substance was concentrated in the first precipi- tates.The atomic weight of the radio-active barium increases with the intensity of the radiation ; with an intensity 3000 times as great as that of uranium, the atomic weight is 140.0; 4700 times as great, 140.9, and 7500 as great, 145.8. The radio-activity of the crystallised or precipitated compounds increases in a marked manner for several weeks after their preparation, but eventually attains a limiting value which may be five or six times as great as the value imfnediately after their preparation (compare Giesel, Ann. Phys. Chem., 1899, [ii], 69, 91). C. H. B. By ERNST COHEN and C. VAN EYCK (PYOC. K. Akad. Wetensch. Amsterdum, 1899,2, 77).-It is well known that at low temperatures tin becomes converted into a grey powder, but the change has been but little investigated, and various reasons have been assigned for it, The authors find that at - 83' the change occurred in about 24 hours, and the reverse transition could not be observed below 30°.At temperatures between these limits, the velocity of the change becomes so small as to be incapable of measurement. By the addition of a few drops of a 10 per cent. solution of ammonium stannic chloride, however, the reaction is considerably accelerated, and dilatometric observations employing this liquid for measurement purposes indicated G. T. M. N. L. Enantiotropy of Tin.84 ALTSTKhCIT8 OF CHEMICAL PAPERS. a temperature of between 10' and 20' for the transition. Detormina- tions of the E.M.F. of the cell : white tin/l0 per cent ammonium stannic chloride/grey tin, gave 20' for the temperature of transition.It hence appears that all tin exists, save in exceptionally warm weather, in the metastable condition. L. M. J. Thio- and Seleno-antimonites. By ISIDORE POUGET (Ann. Chim. Phys., 1899, [vii], 18, 508-571. Compare Abstr., 1897, ii, 499).- The following salts are described for the first time. Lithium orbhothio- antimonite, Li,SbS, + 3H20, forms white, deliquescent crystals ; the para-salt Li2Sb,S7 + 3H,O, is a dark red, gelatinous precipitate which has not been obtained crystalline. Ammonium orthothioantimonite, (NH,),SbS,, is precipitated as a white, crystalline powder on adding alcohol to the mother liquor from the meta-salt ; it readily decomposes and is only stable in the presence of ammonium sulpbide. Barium metathioccntipmonite, Ba(SbS,& + 44H,O, is an amorphous, brown precipitate.Triargentic thioantimonite, Ag,SbS,, is obtained as an amorphous, black precipitate by the addition of silver nitrate to a dilute solution of potassium orthothioantimonite ; the double salt, Ag,KSbS,, is a yellow, crystalline precipitate produced by the action of silver nitrate on concentrated solutions of the potassium ortho- o r pyro-salt (Somrnerlad, Abstr., 1897, ii, 500). A series of double salts of the formula RAg,SbS, is obtained from sodium, lithium, and ammonium thio- antimonites; these compounds are all decomposed by water into triargentic thioantimonite and the corresponding alkali salt. Zinc orthothioantimonite, Zn,( SbS,),, forms a yellow, crystalline precipitate, produced by the action of zinc salts on dilute solutions of potassium thioantimonite ; the double salt, ZnKSbS,, is obtained when concentrated solutions are employed.The correspocding manganese salts, Mn,(SbS,), and MnKSbS,, are produced in a similar manner ; they form pale red, crystalline pre- cipitates. The Zeud salts, Pb,(SbS,), and PbKSbS,, are brown precipitates. The reaction between potassium orthothioantimonite and cadmium, nickel, cobalt, and ferrous salts, follows the same course as in the preceding examples, but the products are very unstable. Cuprouspotassium od't,othioantirnontite, Cu,KSbS, + 3H,O, is a yellow, crystalline precipitate obtained by treating a cupric salt with excess of a concentrated solution of potassium orthothioantimonite ; cuprous orthoantimonite, Cu3SbS,, is produced by treating the precediug double salt with water ; it is a red, crystalline precipitate. A black precipi- tate of cupric thioantimonite is formed when cupric salts are treated with a dilute solution of potassium orthothioantimonite ; if, however, the latter reagent is in excess, i b gradually reduces the cupric salt and the precipitate then consists of a mixture of this compound with the cuprous salt.When mercurous salts are added t o solutions of the alkali ortho- thioantimonites, the yellow precipitate at first produced is rapidly converted into a black deposit of mercury ; mercurous salts react in aINORGANIC CHEMISTRY. 8 5 similar manner. Gold chloride, when mixed with a concentrated solution of thioantimonite, produces a brown precipitato ; t h i s redis- solves on agitation, and the solution, when warmed, deposits metallic gold.Potassium ort?~oselen.oa?ztimonite, K,SbSe,, is obtained in the form of orange crystals on evaporating a solution of antimony selenide in potassium selenide in a current of hydrogen ; the salt is extremely unstable and its solution rapidly deposits selenium. The para-salt K,Sb4Se7 + 3H,O, separates as a gelatinous, brown precipitate on cooling a saturated solution of its generators. Sodium orthoselenoantirnonite, Na,SbSe, + 9H,O, crystallises in yel- low needles from a solution of antimony selenide in sodium selenide ; it is even more oxidisable than its potassium analogue, and its solution deposits red, tetrahedric crystals of sodium selenoantimonate, Na,SbSe, By the joint action OF selenium and antimony dissolved in potassium sulphide, the mixed salt, K,,Sb,S,Se, + 4H,O, is produced ; it separates from the concentrated solution in small, yellow crystals. When a solution of sodium selenide is employed as the medium, two compounds are produced ; the less soluble is a complex salt having the composition Na,SbS1.,Se2.5 + 9H,O, which corresponds with the thioantimonate, Na,SbS4 + 9H20 ; the more soluble salt, Na,SbS,.,Se,., + 9H,O, forms yellow needles and is analogous to the orthoantimonite, Na,SbS, + 9H,O.Tellurium derivatives, analogous to the preceding compounds, could not be prepared; antimony telluride does not dissolve in a hot solu- tion of potassium telluride, K,Te, or potassium hydrotelluride, KHTe, and tellurium itself is insoluble in alkali sulphides. 4- 9H20. G. T. M. Derivatives and Atomic Weight of Palladium. By WILLETT LEPLEY HARDIN (J. Amer. Chem. Xoc., 1899, 21, 943-955. Compare Rosenheim and Maass, Abstr., 1899, i, 163). -Palladobis-phenyl- ammonium chloride, Pd(NH,PhCl),, is obtained as a voluminous yellow precipitate when a slight excess of aniline is added to a hydro- chloric acid solution of palladious chloride, i t is insoluble in hydro- chloric acid, but dissolves in ammonium hydroxide solution ; the bromide, Pd(NH,PhBr),, resembles the chloride. Palladodiquinolin- ium cldoi*ide, Pd(C',NH,Cl),, the corresponding bromide and pallado- clipiljeridium cl~Zoi*ide, Pd(C5NHi,Cl),, all form pale yellow precipi- tates. They are not acted on by hydrogen at the ordinary tempera- ture, but are readily reduced to metallic palladium when heated in hydrogen and the product allowed to cool in a current of air. A com- pound, PdC1,,2NH2Et,2HCl, crystallising in brownish-red scales has also been prepared. The atomic weight of palladium has been determined by various authorities (Berzeliw, 1828 ; Keiser, Abstr., 1890, 17 ; Keller and Smith, ibid., 1893, ii, 73; Bailey and Lamb, Trans., 1892, 61, 745 ; Joly and Leidie, Abstr., 1893, ii, 284 ; Keiser and Breed, ibid., 1894, ii, 141) and has now been determined by analysing diphenyl- palladodiamnionium chloride (seven experiments) and bromide (five experiments), and also ammonium palladium bromide (four experi-86 ABSTRACTS OF CHEMICAL PAPERS. ments), all of which were prepared from carefully purified palladium. The mean result obtained was 107.014 (0= 16), which is somewhat above that given by other authorities. J. J. S.