68 ABSTRACTS OF CHEMICAL PAPERS. Inorganic Chemistry. New Form of the Hydrogen Generator. By E. W. MAGRUDER (Amer. Chem., J., 1897, 19, 810--812).-The apparatus consists of a tube A, closed at the bottom, and connected by means of the inclined side tube C, with the tube B, furnished a t the -- bottom with the stop- cock D, and a t the top with the funnel E, into which fits, with a ground joint, the delivery-tube F. B also contains a platinum electrode G, suspended by a platinum wire fused through the tube at H. Zinc amalgam is put into A until it nearlyreachesthe tube C, A being then closed with a rubber stopper I, carry- ing a glass tube M, con- taining a-platinum wire, fused through the tube at L and dipping into the amalgam. The apparatus may be simplified by leaving out the stopcock D, and fusing F directly to the top of B.A --GINORGANIC CHEMISTRY. G9 AS the oxygen and hydrogen electrodes are in separate tubes, oxygen or any other escaping gas simply rises to the surface in A, and does not, therefore, mix with the hydrogen. By having the amalgam nearly up to the side tube, the zinc sulphate formed flows into B, from which it can be drawn off through the stop- cock. The generated gas is perfectly pure hydrogen. A. W. C. Nascent Hydrogen. By DONATO TOMMASI (J. Pl~ysical. Chent., 1897, 1, 555).-A priority claim in connection with Franchot's work on nascent hydrogen (Abstr., 1897, ii, 138). WILDER D. BANCROFT (ibid., 1897, 1, 556) replies by pointing out that Franchot's work had not reached the point a t which one would naturally make a reference t o Tommasi.H. C. Heat of Dissociation of Iodine. By JOACHIM SPERBER (Zed. anorg. Chem., 1897, 15, 281-282).-The heat of dissociation of iodine, c:tlculated according to the author's theory (Abstr., 1897, ii, 307j, is 13.132 Cal. ; he points out that this is in close agreement with the nnmber obtained by Roltzmann, namely, 14,265 Cal. E. C. R. Evolution of Oxygen during Reduction. By K. FRENZEL, S. FRITZ and VICTOR ~UEYER (Bey., 1897, 30, 2515-2519. Compare Abstr., 1897, ii, 19, 93; this vol. ii, 19).-Erdmann (Abstr., 1897, ii, 97) has attempted to explain the evolution of oxygen during the reduction of rubidium dioxide by hydrogen by the intermediate formation of hydrogen peroxide; in order to test this, experi- ments were made in some of which such formation was impossible.Potassium peroxide yields no oxygen when heated to the softening point of glass (in a silver vessel) in a current of air. In carbonic oxide, carbonic anhydride, and hydrogen, on the other hand, i t gives a considerable amount of oxygen at the boiling point of sulphur. I n the cases of carbonic oxide and anhydride, the intermediate forma- tion of hydrogen peroxide is excluded, and it is probable that the three reactions are quite analogous. K,O, + CO = K,CO, + 0,. 2K,O, + 2C0, = 2K2C0, + 30,. K,O, + H, = 2 KOH + 0,. Silver oxide decomposes in a current of air a t 250", but when heated at the boiling point of water in hydrogen, it yields free oxygen, and does this even at the ordinary temperature in carbonic oxide ; this is due to the fact that the heat evolved by the reduction of a portion of the oxide is so great as to raise the temperature of the rest above the point of decomposition.Similar results were obtained with potassium per- manganate, which decomposes in air a t 2 1 So, and in hydrogen to a very small extent at 155", freely at 182'. No evolution of oxygen was observed during the reduction of barium peroxide. A. H. Solubility of Ammonia in Water at Temperatures below 0". By JOHN W. XALLET (Ames.. Chern. J., lS97,19,804-809).-0n pass-70 ABSTRACTS OF CHEMICAL PAPERS. 127" 188" 223-224.3" 261-263" 289-294" ing gaseous ammonia into an already strong aqueous solution a t tem- peratures below O", the following results mere obtained. At - 10" 1 gram of water dissolves 1.115 grams of ammonia 9 ) - 20 >, 9 , Y, 1.768 9 , 9 , j , - 30 9 , 9 7 7, 2.781 ,, 9 , 9 , - 40 ,, 9 , 9 , 2.946 9 , 9 , No visible change marks the presence of enough ammonia to represent the hydroxide of ammonium, assumed t o exist in ordinary solution, and the hydroxide, if it exists, continues to dissolve ammonia, or mixes with liquefied ammonia down to and beyond the normal boiling point of the latter.The proportion of ammonia dissolved by water is much greater than would be called for by a n extension of the curve representing the solubility a t temperatures above 0". Chloronitrides of Phosphorus. 11. By HENRY N. STOKES (Amer. Chem. J., 1897, 19, 782-796. Compare Abstr., 1895, ii, 217).- Continuing his researches on the chloronitrides of phosphorus, the author finds they are best prepared by heating phosphorus penta- chloride and ammonium chloride, in molecular proportion, in sealed tubes, PCl, + NH,Cl = PXCl, + 4HC1.The operation requires to be conducted with great caution on account of the large amount of hydrogen chloride generated. There is formed a mixture of chloro- nitrides, which is partly crystalline and soluble in light petroleum, and partly liquid and insoluble in this solvent ; the latter portion consists of about 50 per cent. P3N3C16 and 25 per cent.:P,N,Cl,, the remainder consisting of the higher homolognes. A separation is effected by means of fractional distillation, and, in the case of the solids, cry stallisation from benzene. A. W. C. The series, as at present known, consists of the following.256.5" 3285" plymerises polymerises polymerises Melting point. Triphosphonitrilicc chloride, (PNCI,), 1 114" Tetraphosphonitrilicc chloride, (PNCI,), 123'5" Pentaphosphonitrilicchloride, (PNC12)5~ 40'5-41" He~tcc.phosphonitril~c chloride, ( PNC12)7 i liquid 910 at - 18" Rezaphosphonitrilic chloride, (PNCI,), Polyphosphonitrilic chloride, (PNCI,),/ below red heat Boiling point. 13 mm. 1 760mm. I There were obtained, further, a liquid residue of the same empirical composition with a mean molecular weight of (PNCI,),, and a small amount of nit.i.ilo~exaphos~on~€r~l~c chloride, P6N7CI,, not belonging to the above series. It melts a t 23'7.5' (corr.) and boils a t 251-2GlO (13 mm.), cryststllises in transparent rhombic prisms, often united to acicular forms, and when phlverised becomes electrified.The complete absence of the lower members of the above seriesINORGANIC CHEMISTRY. 71 PNC1, and (PNCl,), is very remarlittble. Xach member of the series is converted by heat into the caoutchouc-like polyphosphonitrilic chloride, which ia highly elastic and insoluble in all neutral solvents. I t swells enormously in benzene, and on heating breaks down into a mixture of all tbe lower members mentioned above. By STEPHEN H. EMNENS and NEWTON W. EMMENS (Chem. News, 1897, 76, 37).--Minute variations in weight observed in little discs, one of silver and one of lead (free from silver), suspended in air in a bottle and both in contact with the same spiral of copper wire are attributed to migration of matter. On cupelIation, the lead now showed a minute amount of silver.D. A. L. By ARTHUR ROSENHEIM and PAUL WOGE, (Zeit. unovg. Chenb., 1897, 15, 283-318). Hydrated beryllium oxide, when treated with solutions of oxalic acid and hydrogen alkali oxalates, yields a series of double salts and not complex compounds as in the case of the hydrated oxides of chromium, iron, and aluminium. When treated with hydrogen alkali tartrates, it yields a series of complex compounds of the formula R,0,4Be0,2C,H40, + 8H,O, in which beryllium displaces the hydroxyl hydrogen atom of the tartaric acid. When treated with salts of paramolybdic acid, it yields beryllium molybdate, BeO,MoO, + 2H,O, and with hydrogen alkali sulphites it yields characteristic crystalline double salts. All these reactions point to the conclusion that beryllium is bivalent ; and the determination of the molecular weight of beryllium chloride by Beckmann’s boiling point method, using pyridine as the solvent, gives numbers which agree closely with the formula BeC1,.Thedouble salts,K20,2Be0,2C,0, + 2&H,O; Na,0,2Be0,2C20, + 5H,O, and (NH4),O,2BeO,2C,O, + 2+H,O, are obtained by treating a boiling solution of potassium, sodium, and ammonium oxalate respectively with hydrated beryllium oxide. The potassium salt separates in lustrous crystals, the sodium salt, which is very soluble, in ill-defined crystals, and the ammonium salt in slender needles ; the two former are stable on exposure to the air, but the ammonium salt is extremely hygroscopic. Oxalic acid and beryllium can be detected in solutions of these salts by the ordinary methods.The sults K2O,Be0,2C,O, + H,O ; Na20,Be0,2C,0, + H,O, and (NH4),0,Be0,2C,0, are obtained by saturating solutions of the cor- responding hydrogen alkali oxalates with hydrated beryllium oxide, and then adding an equal quantity of the hydrogen alkali oxalate. When treated with calcium or barium chloride, they are quantitatively decomposed into calcium or barium oxalate and beryllium chloride. When boiling oxalic acid is saturated with hydrated beryllium oxide, a non-crystallisable syrup is obtained containing 1 mol. of oxalic an- hydride to 2.5-3 mols. beryllium oxide; if to this solution an equal quantity of oxalic acid is added, beryZliunz oxcclute, BeOC,O, + 3H,O, separates in needles, which are obtained in quadratic tablets on recrystallisation from water.The ucid oxalate, 2Be0,3C2O, + 6H20, is obtained by dissolving 1 mol. of beryllium oxide in 2 mols. of oxdic acid. The compZex tartswtes, K,0,4Be0,2C,H,O5 + 8H,O, the sodium salt A. W. C. Migrant Matter. Valency of Beryllium. It crystallises in beautiful prisms.72 AESTRACTS OF CHEMICAL PAPERS. with 8H20,. and the ammonium salt with 8H20, are obtained by satu- rating a boiling aqueous solution of the corresponding hydrogen, alkali tartrate with hydrated beryllium oxide. The potassium salt, which is extremely soluble, crystallises in large, lustrous prisms, the sodium and ammonium salts in microcrystalline crusts ; the beryllium in these salts is a t once precipitated by ammonia. However, the complex character of the salts is shown in the abnormal behaviour of the tartaric acid.Potassium chloride does not cause a precipitation of potassium tartrate, even on adding dilute acids; no precipitate is ob- tained with calcium chloride, and barium chloride gives a precipitate only after some time. With neutral silver nitrate, a flocculent, voluminous precipitate is a t once formed. All the salts are neutral to ordinary indicators. These salts are the first compounds isolated in which a metal is sub- stituted for, not only the carboxyl hydrogen atom, but also for the hydroxpl hydrogen atom of the tartaric acid. The salt K,0,2Be0,2C4H,O, + 2H20, obtained by dissolving the theoretical quantity of hydrated beryllium oxide in a solution of hydrogen potassium tartrate, forms a glassy, crystalline mass, and has properties similar to those of the preceding beryllium tartrates ; it is strictly analogous to the copper compound of Fehling's solution.Beryllium ,molybdate, BeOMoO, + 2H20, obtained by boiling the theoretical quantity of hydrated beryllium oxide with molybdic acid suspended in water, is an oily liquid which solidifies to an aggregate of slender aeedles if kept for some time in the cold. The complex compounds corresponding with those obtained by saturating potassium or sodium paramolybdate with sesquioxides cannot be obtained in the case of beryllium oxide, beryllium para- molybdate being always formed. With ammonium paramoly bdate, beryllium molybdate is usually formed, but in some cases the double salt, 10(2BeO,MoO,) + 2(NH4),C),3MoO, + 1SH 2O is obtained, crystallis- iag in microcrystalline crusts The double salts, X,0,2Be0,3S02 + 9H,O and (NH,),0,2Be0,3S02 + 4H,O, are prepared by digesting freshly prepared solutions of the hydrogen alkali sulphites with hydrated beryllium oxide on the water bath ; the solutions, which must contain excess of sulphurous acid, yield crystals of the salts on cooling.They are extremely easily oxidised, and when exposed to the air rapidly evolve sulphurous anhydride, ammonium sulphite being also formed. E. C. R. Potassium Lead Iodide. By J. M. TALMADGE (J. Physical Chena., 1897,1,493-498).-Schreinemakers (Abstr., 1893, ii, 260) has shown that concentration a t ordinary temperatures of a solution from which the solid double salt PbI,,KI + 2H20 has separated is always attended by solution of the double salt.The author shows that, a t looo, the behaviour is reversed and there is continued precipitation of the double salt on evaporation. H. C. By J. K. HAYWOOD (J. Physical Chem., lS97, 1, 41 l--413).-Lescaeur has shown that when cuprous chloride is Cuprous Chloride.INORGANIC CHEMISTRY. 73 treated with water, cuprous oxide and hydrochloric acid are formed, and states that when cuprous chloride is washed with water, the amount of acid in the wash water remains practically constant until all the cuprous chloride has been converted into oxide. The author shows that this last result must have been obtained by accident, as the action is dependent on time and the mass of water present. H. C. Possible New Element or Elements in Cast Iron and Blast- furnace Boiler-Dust. By GETHEN G.BOUCHER (Chenz. News., 1897,7S, 99-100 ; 182).-The suspected new metal has been obtained from iron to the extent of from 0-0019 to 0.006 p r cent., in the following manner. The iron is treated with dilute 1 : 5 sulphuric acid and the cold solution saturated with hydrogen sul phide. The mixed precipitate and undissolved matter, after being washed until free from iron, is boiled with hydro- chloric acid and potassium chlorate until the evolution of chlorine censes,and the filtered solution is saturated with hydrogen sulphide. The precipitate is again submitted to the treatment with hydrochloric acid and potassium chlorate ; the arsenic precipitated by magnesia mixture and filtered off; the copper, by treatment with hydrogen sulphide, pre- cipitated as sulphide, which is removed, and ,the solution slightly acidified with hydrochloric acid.The precipitate is repeatedly boiled with hydrochloric acid to remove antimony sulphide and a heavy dark-brown sulpbide remains. This is dissolved in potash, the scllution treated with hydrogen sulphide, filtered, and reprecipitated with hydrochloric acid, these operations being repeated until no further precipitate is obtained from the alkaline solution on treatment with hydrogen sul- phide. The sulphide is dissolved in hot (1 : 2) nitric acid, the solutions filtered, made slight1y:alkaline with ammonia, any precipitate removed, and the metal again precipitated as sulphide. The metal is obtained as a black powder by reducing the oxide in a current of hydrogen, or by fusion with potassium cyanide; it is in- soluble in dilute hydrochloric and sulphuric acids, sparingly soluble in these acids when strong, is soluble in dilute and strong nitric acid, and very readily soluble in aqua regia.When beated in a current of air, the metal glows and forms a yellow, volatile oxide. The oxide, which is also formed on roasting the sulphide, melts at a low temperature and sub- limes at a full red heat, yielding transparent, colourless needles; i t is sparingly soluble in hydrochloric acid, almost insoluble in sulphuric acid, and insoluble in nitric acid. The borax bead is clear and colour- less in the outer, and pale pink in the inner, flame, the microcosmic salt beads are chrome-green, that in the inner flame being darker.When fused with sodium carbonate, the oxide yields a colourless mass soluble in water. The sulphide, precipitated from slightly acid solutions by hydrogen sulphido, is dark-brown, and soluble in the sulphides of ammonia and sodium and in the alkali hydroxides, also in nitric acid and aqua regia, but insoluble in dilute hydrochloric and sulphuric acids. The solution of the chloride gives the following reactions : with sodium thiosulphate, a violet coloration, turning brown on heating with hydrochloric *acid and yielding a precipitst,e of the sulphide ; with potassium ferricyanide, a dark brown, flocculent precipitate soluble in acids and alkalis ; with alkali hydroxides, a slight blue coloration, with zinc or iron, a black deposit of metal, some of which escapes with7s ABSTRACTS OF CUEMICAI, PAPERS.the evolved hydrogen; and when deposited in the usual way from the flame, is almost black with little lustre, and neither soluble in hydro- chloric acid nor calcium hypochlorite. No reactions, however, are obtained with carbonates, chlorides, sodium sulphate, phosphate or acetate, magnesia mixture, potassium cyanide, ferrocyanide or chro- mate, ferrous sulphate or stannous chloride. The chloride or nitrat,e when evaporated nearly to dryness with hydrochloric or, better still, sulphuric acid, gives a blue coloration destroyed by water. It does not give a precipitate when warmed with sodium phosphate and nitric acid, Heavy, dark-colourqd boiler-dust yields a metal that is similar in all respects except that, with stannous chloride, it produces a dark blue colour, turning brown when boiled with hydrochloric acid.The strong aqueous extract of the dust is treated with hydrogen sulphide and hydrochloric acid, the precipitate, when sufficient has been accumulated, is dissolved in aqua regia and evaporated nearly to dryness with sub phuric acid; the solution is diluted, filtered, made alkaline with am- monia, treated with hydrogen snlphide, filtered, and acidified with hydrochloric acid. The. precipitated sulphide is dissolved in potash and treated as in the preceding case. A ton of dust yielded 0.3 gram of this metal. Lucium. By WALDRON SHAPLEIGH (Chem. News, 1897, ”76, 41)- Following Barri6re’s methods, but making each separation thoroughly, the author did not obtain lucium from monazite, but, instead, less than 1 per cent.of mixed oxides of the yttria group. H e remarks that con- centrated solutions of yttrium potassium sulphate give a precipitate when heated, and suggests that this may be the origin of Barrikre’s lucium, inasmuch as Barrihe used concentrated solutions, and then heated (compare Crookes, Abstr., 1897, ii, 144). For- mation of Basic Aluminium Carbonate. By WILLIAM C. DAY (Amer. Chem. J., 1897, 19, 707-728).-When carbonic anhydride is passed into a solution of sodium aluminste (prepared from ‘( Connetable” phosphate rock by the action of quicklime and sodium carbonate, and subsequent washing), a precipitate is formed which differs from alu- minium hydroxide precipitated from an aluminium salt by means of ammonium hydroxide in being of an opaque white, and more granular. If this precipitate is washed with cold water until the wash- ings show no alkaline reaction, and the air-dried residue is analysed, it is found to contain more carbonic anhydride (from 5-5-9 per cent.) than is capable of uniting with the small amount of sodium present. The author concludes that this substance must be regarded as a basic aluminium carbonate. The literature bearing on this subject is die- cussed a t some length. A. W. C. [Reactions of Titanic Acid with Organic Acids.] By GEORG BERU (Zeit. anorg. Chem., 1897, 15, 328--330).-See tbis vol., i, 66. Platino-platinoso-additive Compounds, By MAURICE VEZES (Zeit. ccnoyg. Chem.; 1897, 15, 278--28O).--.The author points out that acid potassium triplatinosohexanitrite, Pti+O(N02)6K2H4, which he obtained by the cautious oxidation of potassium platinosonitrite, can- not have a constitution analogous t o the platinum compounds described D. A. L. D. A. L. Action of Carbonic Anhydride on Sodium Aluminate.iUINERALOGICAL CIIEMISTSY. 75 by Werner (Abstr., 1896, i, 465) and Hadow (this Journal, 1860, 13, 106), although it exhibits the same phenomenon, namely, it crystallises in copper-coloured crystals and gives an almost colourless solution in water. The same behaviour is characteristic of the compound ob- tained by Miolati by heating at 150' the compounds Pt(NOz)4K2,Nz0, and Pt(NO,),K,,HCl. Miolati states that it is identical with, or analogous to, acid potassium triplatinosohexanitrite. E. C. R.