INORGANIC CHEMISTRY. I9 Ino r g an ic C h emi s t r y. Does Hydrogen find its Proper Place at the Head of Group I. or at the Head of Group VII. 3 By ORME MASSON (Chena. News, 1896, 72, 283-284).-As reasons for assigning a position t o hydrogen at the head of Group VII. in the periodic system, the author points to its monad valency, its diatomic molecule, its gaseous character and extremely low boiling-point. The atomic weight of hydrogen places it rather in Group VII. than in Group I., and arguments based on substitution cut both ways, as an inorganic chemist would be most struck with the analogy between hydrogen and tho alkali metals, an organic chemist with the analogy between hydrogen and the members of the chlorine group. JOHN A. R. NEWLANDS (ibid., 1896, 72, 305) points out that in 1872 he classed hydrogen and chlorine together on account of their mutual replaceability, and because the atomic weight of hydrogen indicates that it is really the lowest member of the chlorine group.Slow Oxidation of Hydrogen and of Carbonic Oxide. H. C. By VICTOR MEYER and MAX VON RECKLINGHAUSEN (Bey., 1896, 29, 2549 -2560).--It is known that hydrogen is slowly oxidised, when led through a solution of potassium permanganate, and a repetition of this experiment showed that the gas if left over an alkaline or neutral solution of the salt slowly decreased in volume until all had been absorbed. Carbonic oxide behaved similarly. 2-220 ABSTRACTS OF CHEMICAL PAPERS. On shaking the mixture on the shaking-machine lately devised by von Recklinghausen, a similar result was arrived at ; if, however, an acid solution of permanganate was employed, a large volume of gas remained unabsorbed, which on examination was found to be oxygen.39 C.C. of hydrogen after shaking for 15 hours with 29 C.C. of an acid solution of permanganate yielded 17-8 and 20.4 C.C. of oxygen. Since an acid solution of permanganate evolves oxygen by itself on standing, three experiments were conducted side by side, the hydrogen being replaced in one case by air and in the other by carbonic anhydride ; the quantities obtained were, however, very different, the carbonic anhydride experiment yielding only 2.5 C.C. of oxygen, whilst with the hydrogen 19.1 C.C. was obtained. The amount of oxygen evolved does not increase after a certain limit has been reached, the same volume being obtained whether the mixtures were shaken for 17 or for 87 hours.The quantity of oxygen evolved from an acid solution of perman- ganate on standing differs considerably from that given off on shaking. A closed tube completely filled with the solution burst after standing 24 months in the dark, whilst one which had been filled one-third with permanganate and two-thirds with air was found to contain 50 C.C. of oxygen at the end of the same period, that is, nearly 20 times as much as was given off in the majority of the experiments when the liquid was shaken. It was further found that the temperature had a con- siderable influence on the amounts evolved. The formation of oxygen in these experiments could be readily understood on the supposition that hydrogen peroxide was at first produced and that this subsequently reacted with the permanganate, evolving oxygen.Since, however, carbonic oxide behaves in a pre- cisely similar way (yielding 14.5-115 C.C. of oxygen when 39.5 C.C. of the gas is shaken with 20 C.C. of perrnanganate solution for 234 hours), this explanation does not hold good. The evolution of oxygen is irregular, and the numbers obtained are merely approximate, the quantities evolved in two similar experiments seldom agreeing. It might be supposed that the large quantity of oxygen evolved might be due to the diminution of pressure consequent on the absorp- tion of the hydrogen; that this is not the case, however, was shown by shaking carbonic anhydride and air with permanganate under diminished pressure, the volume of oxygen given off (2-3 c.c.) remaining unaltered.The authors hope t o throw light on these phenomena by means of further experiments conducted on other reducing gases, such as the hydrocarbons, &c. Meanwhile, they consider that the most probable explanation is that put forward by van’t Hoff in explanation of the simultaneous oxidation of triethylphosphine and indigo solution, namely, that in slow oxidations the oxygen molecule acts as two atoms ( 3 ) or ions (2). The reaction in this case proceeding thus 00 = 2H20+00, H2 H% 00 +INORGANIC CHEMISTRY. 21 Pressure - S3.7" Hydrogen iodide.. . . . . 739'8 - 34'1" Hydrogen bromide . . 738'2 - 64.9" the quantity of oxygen given off agreeing approximately with this equation.J. F. T. Behaviour of Halogen Hydrides at Low Temperatures. By TEIADDAEUS ESTREICHER (Zeds. physikal. Chem., 1896,20,605-6U9).- The author bas determined the boiling points and freezing points of hydrogen chloride, bromide, and iodide, as well as the critical tempera- tures of the last two compounds. The low temperatures were obtained by means of ethylene, and by carbonic anhydride and ether, and were measured by a helium thermometer. (Abstr., 1896, ii, 597). The results obtained were m. p. Crit. tenip. - 111.1" - - 87'9" + 91.3" -- 50.8" + 150.7" The liquid was in each case clear and colourless, and the solid an iron grey mass, but in the case of the iodine compounds the products were frequently coloured by dissolved iodine. L. M. J. Crystallography of Iodates and Periodates.By ARTHUR S. EAKLE (Zeit. Kryst. Min., 1896, 26, 558-588).--The author has investigated the conditions of growth most favourable to the cultiva- tion of well-developed crystals of a number of salts and dotermined their geometrical and optical properties. Well-developed transparent hydrated crystals of the trigonal pyramidal sodium periodate, NaIO, + 3H,O, can be obtained by very slow spontaneous evaporation of the pure aqueous solution, but in a much shorter time from solutions containing sodium nitrate ; a : c = 1 : 1.0942. The crystals deposited from a pure aqueous solution consist of dextrogyrate and hvogyrate individuals in approximately equal numbers, but in solutions containing sodium nitrate the remarkable fact is observed that the laevogyrate crystals outnumber the dextrogyrate ones by about four to one.The anhydrous ditetragonal dipyrnmidal crystals of sodium periodate, NaIO, deposited from aqueous solution a t above 30° have the axial ratio a : c = 1 : 1.5900. Aqueous solutions of ammonium periodate containing ammonia deposit, on spontaneous evaporation, trigonal rhombohedra1 crystals having the composition (NH4),T20,,3H20, and the axial ratio a : c = 1 : 1.6118 ; when the solution during evaporation has lost nearly all its free ammonia, crystals of the normal periodate, NH410,, isomorphous with those of anhydrous sodium periodate, separate ; a : c = 1 : 1.5228. Aluminium periodate, A1 (IO& + 3H20, crystallises in holohedral cubic forms from a solution containing nitric acid. Rhombic bipyramidal crystals of anhydrous sodium iodate, NaIO,, are deposited from an aqueous solution at 100" ; a : b : c = 0.9046 : 1 : 1.2815 ; at lower temperatures, long needles of the hydrated salt are22 AIWTRACTS OF CHEMICAL PAPERS.deposited. Ammonium iodate, NH,I03, crystallises in long prisms ; a : b : c = 0.9948 : 1 : 1.4335. Silver iodate, AglO,, in thin, colourless, transparent plates of adamantine lustre from solutions containing ammonia ; a : b : c = 0*8832 : 1 : 1.3072, These three salts are isomorphous. Potassium iodate is cubic and holohedral, but the habit of the crystals varies widely with changes in the temperature of crystallisation ; the crystals are isomorphous with those of rubidium iodate, which consist of opaque, white cubes. Calcium iodate, Ca(10,)2 -t GH,O, separates from an aqueous or nitric acid solution in rhombic prisms; a : 6 : c = 0,4357 : 1 : 0.5231.If, however, the solution contains a small quantity of calcium chromate, a second form of calcium iodate having the same composition, is deposited ; the new crystals are rhombic pyramidal prisms ; a : b : c = 0.64652 : 1 : 0.27682. The double compound of sodium iodate and iodide, ZNaIO,, 3Na1, + 20H,O, crystallises in di hexgonal bipyraniidal prisms ; a : c = 1 : 2.1892. Rubidium nitrate crystallises in ditrigonal scalenohedral prisms ; a : c = 1 : 1.2360. The crystallographic properties of the rhombic' aluminium nitrate, Al(NO,), + 9H,O, the monoclinic barium chlorate, Ba(C10,), + H,O, and the rhombic barium permanganate, BaMn,O,, were also determined.Calcium iodate is thus dimorphous. W. J. P. Reduction of Sulphuric Acid by Copper, aB a Function of the Temperature. By LAUNCELOT W. ANDREWS (J. Arner. Chem. Xoc., 1896, 18, 251-254).-By means of suitable apparatus, the author has been able to show-(1) That sulphuric acid (98.4 per cent.) begins to appreciably dissociate at a temperature somewhat below 70°, probably 67" ; (2) That sulphuric acid in the presence of air or oxygen mill dissolve copper a t the ordinary temperature, without the evolution of a trace of sulphurous anhydride; (3) That the reduction of sulphuric acid by copper does not begin below 86", that is, not until the acid has become partly dissociated into anhydride and water. Baskerville's statement (Abstr., 1896, ii, 474) that sulphuric acid is reduced by copper at 0" is therefore incorrect.J. J. S. Free Hydraaine. By CORNELIS A. LOBRY DE BRUYN (Eec. Trav. Chim., 1896, 15, 174-184; Compare Abstr., 1895, ii, 347).- Free hydrazine is best prepared by the action of barium oxide on hydrazine hydrate. The hydrate is added in small quantities a t a time to the barium oxide contained in a flask, which is cooled from time to time and should be provided with a neck some 50 cm. long, bent at right angles a t its upper end so that it can be directly attached to the condenser. The flask is first heated for several hours in a glycerol bath at 110-120°, and the contents are then distilled under a pressure of 150-100 mm. To prevent oxidation, the apparatus is filled with dry hydrogen. The base, which is collected in a cooled receiver, still contains 3-4 per cent.of water, and is, therefore, again distilled in a smaller flask, with a little barium oxide, the apparatus again being filled with dry hydrogen.INORGANIC CHEMISTRY. 23 The free base melts at 1 0 4 ~ ; it can, however, be cooled several degrees below this before it freezes. Its boiling points, determined in an atmosphere of dry hydrogen, are 56" at 71 mm., 113.5" at '7615 mm., and 134.6" at 1490 mm. pressure. Its sp. gr. =1*014 at 16"/15", that is, almost the same as that of the hydrate. The refractive index, as determined by Eijkman, is given as 1'47988 1.46953 1;:l;693 I 0.01035 0'33 Lorentz's formula 1 ':% I 8.62 0.17 ~~ ~~~ d = 1.0114 t = 17.8" MV = 31.64. Calculated for the group NH,, the dispersion = 0.16 according to Gladstone's formula and 0.09 according t o Lorentz.These numbers are practically. the same as those calculated from pentamethylene- diamine (Rec. Trav. Chinz., 1893, 12, 281). As regards its solubility, hydrazine resembles both water and hydroxylamine ; it is miscible in all proportions with methylic, ethylic, propylic, isobutylic, and amylic alcohols, but is only very sparingly soluble in other organic solvents. Many salts dissoIve in it, and determinations of their solubilities are given. Hydrazine apparently combines with sodium chloride, as a consider- able amount of heat is developed on mixing the two and the whole sets to a crystalline mass on cooling. Ammonium salts are decom- posed by the base, and lead nitrate is dissolved by it.The base itself may be heated t o 300-350" without undergoing perceptible decomposition ; even after heating for an hour at the boiling point of sulphur part of i t remains unaltered. The critical temperature, as determined in Altschul's apparatus (2eits.phgsikal. Chem., 1893'11,582) is 380". According to Boltwood, stability of the base at high temperatures depends on the pressure. The decomposition first proceeds according to the equation 2N,H4 = N, + H, + 2NH,, but the final decomposition is 3N,H, 7 N, + 4NH,. The base is a strong reducing agent, it readily bursts into flame in chlorine ; bromine and iodine also react violently. It is slowly oxidised by dry air free from carbonic anhydride, and also by oxygen ; in the latter case, the temperature gradually rises to loo", and ammonia can readily be recognised.Yellow phosphorus also acts on an aqueous solution of the base, forming, probably, hydrides of phosphorus. Sodium reacts with the base suspended in dry light petroleum, yielding hydrogen, ammonia, and a brown solid substance. J. J. S. Oxidation by means of Hydroxylamine. By FRITZ HABER (Bey., 1896, 29, 2444--2445).-Hydroxylamine readily converts24 ABSTRACTS OF CHEMICAL PAPERS. ferrous hydroxide into the ferric compound, and may be recognised by this reaction, which is not given by hydrazine. The hydroxylamine is almost completely converted into ammonia. An interesting lecture experiment to illustrato the double function of hydroxylamine as an oxidising and reducing agent may be carried out by adding hydroxylamine hydrochloride to a hot emulsion of ferrous hydroxide precipitated by means of caustic soda.The greenish mass rapidly becomes reddish brown, owing to the oxidation of the hydroxide. If now the liquid be acidified and a fresh quantity of hydroxylamine hydrochloride added, the latter acts as a reducing agent, and rapidly converts the ferric into a ferrous salt, the change being rendered evident by the disappearance of the yellow colour. A. H. Nitrohydroxylamine. By ANGELO ANGELI (Gaxxetta, 1896, 26, ii, 1 7-25).-0n treating hydroxylamine with acidic chlorides or anhydrides, or with an alkylic salt of an organic acid in alkaline solution, derivatives having the constitution R*CO*NH*OH or OH*CR:N*OH are obtained. It seemed of interest; t o endeavour to prepare a substituted hydroxylamine, in which the acidic group was an inorganic acidic radicle.With this object, a cold saturated solution of hydroxylamine hydrochloride (1 mol.) is poured into a concentrated alcoholic solution of sodium ethoxide (3 mols.), and to this, after filtration, ethylic nitrate (1 mol.) is added ; a white precipitate of sodionitrohydroxylamine, Na,N,O,, separates immediately, but cannot be further purified because of its great solubility in water. On adding barium chloride to its dilute aqueous solution, a heavy, white, minutely crystalline precipitate of the corresponding barium salt, BaN,O, + H,O, is deposited ; it loses its water at 115-120". On treating either the sodium or barium salt with a mineral acid, a violent evolution of nitric oxide occurs, almost quantitatively in accordance with the equation, H,N,O, = H,O + 2N0 ; the salts also decompose on beating, giving nitric oxide, and leaving the metallic oxide.The aqueous solution of the sodium salt is readily decomposed by boiling, and nitrites and nitrates formed from the nitric oxide evolved are found in solution; the barium salt is not decomposed by prolonged boiling with water, and is much more stable in the air than the sodium salt. The aqueous solution of the sodium salt gives a white precipitate with lead acetate or cadmium sulphate; a yellow solution, which rapidly deposits mercury, with mercuric chloride ; a deep brown-red coloration, which disappears on boiling, with uranium nitrate ; and a yellow precipitate, which rapidly blackens, with silver nitrate ; it reduces Fehling's solution, and gives a violet to brown coloration with ferric chloride.Mixtures of the sodium salt with sodium thiocyanate or hypo- phosphite, or with antimony trisulphide and sulphur or carbon, explode violently when struck ; when mixed with red phosphorus, the sodium salt also detonates if struck, but the mixture is less sensible t o shock than that of red phosphorus with potassium chlorate.INORGANIC CHEMISTRY. 25 According to its method of formation, nitrohydroxylamine should have the constitution NO,*NH*OH ; it: might readily undergo intramolecular change and assume the constitution OH*N:NO*OH or H0.N HO.N>O- It is interesting to note that nitrohydroxylamine is the last missing member of the series of acids corresponding with the five oxides of nitrogen; it takes the vacant place between hyponitrous and nitrous acids in the series and corresponds with a hydrate of nitrogen dioxide.W. J. P. The Electrolysis of solutions of Ammonia containing Salts and Bases. By SIMA M. LOSANITSCH and MILORAD 2. JOVITSCHITSCH (Bey., 1896, 29, 2436-2438). When a solution of ammonia and common salt is electrolysed in the usual Hofmann apparatus, there is always a deficiency in the nitrogen evolved. This is due to the occur- rence of secondary reactions. When a chloride is present, the liquid is found to contain hypochlorite, which is either a direct product of the action of chlorine on the ammonia, or an indirect product formed by the action of water and ammonia on previously formed chloride of nitrogen.A loud explosion occurred when the graphite electrode was heated, and the authors ascribe this to the presence of chloride of nitrogen. When the formation of hypochlorite is prevented by heating the liquid, the correct ratio of nitrogen to hydrogen is obtained. When potassium bromide is used, a hypobromite is formed, and iodide of nitrogen and a hypoiodite when potassium iodide is employed, When caustic potash or potassium sulphate is added, as well as when ammonia itself in dilute solution is submitted to electrolysis, a nitrite is produced and the volume of nitrogen obtained is too small. A. H. Hyponitrous Acid. By ARTHUR R. HANTZSCH and LUDWIG KAUFMANN (Alzrzalcn, 189 6, 292,3 1 7-340. Compare Abstr., 1896, ii, 520).-Silver hyponitrite is prepared by reducing sodium nitrite with sodium amalgam in strongly alkaline solution, and after removal of hydroxylamine with mercuric oxide, adding silver nitrate to the solu- tion acidified with nitric acid; treated in this way, 20 grams of nitrite yields 2-3 grams of silver salt, and the method compares favourably with Piloty's (Abstr., 1895, i, 556), which consists in hydrolysing benzosulphydroxainic acid with 40 per cent.aqueous potassium hydroxide. Hyponitrous acid is obtained in the solid state by adding the silver salt to ethereal hydrogen chloride until silver chloride ceases to be formed ; the filtered medium is then evaporated as rapidly as possible in the desiccator, when the acid crystallises in leaflets. It is deliques- cent and dissolves readily in water and alcohol; it is only sparingly soluble in petroleum, but more readily in ether, chloroform, and benzene; the crystals, when wet with ether, do not rapidly undergo change, but the dry substance explodes spontaneously, even at - 6".Determinationla of the molecular weight, effected by adding the silver26 ABSTRACTS OF CHEMICAL PAPERS. salt to water containing a known quantity of hydrogen chloride, yield results in accordance with the formula H2N202. Titration of hypo- nitrous acid with alkali, using phenolphthalein as indicator, shows that the hydrogen sodium salt, like hydrogen sodium carbonate, is neutral towards this indicator; if the solution is kept for a time at the ordinary temperature, sodium hydroxide and nitrous oxide are formed. Solutions of pure hyponitrous acid do not immediately libe- rate iodine from potassium iodide, but the amount of the free halogen gradually increases with time ; diphenylamine develops an intense blue coloration, and if ferrous sulphate is added to concentrated sulphuric acid mixed with silver hyponitrite, the brown ring charsc- teristic of nitrous and nitric acids is produced.Ammonium 1yponit&e, H,N*O-N:N*OH, is obtained by passing dry ammonia gas into a cooled solution of the acid in ether ; it melts and decomposes a t 64-65', and at theordinary temperature is resolved spon- taneously into ammonia, water, and nitrous oxide. The salt dissolves readily in water, the solution being alkaline, but the aqueous and alcoholic solutions leave no residue on evaporation ; normal ammonium hyponitrite has not been obtained.BenqZic hyponitrr.ite, CH,Ph*O-N: N* 0 CH,Ph, is obtained from benzylic iodide and silver hyponitrite in ether, and crystallises from light petroleum in leaflets ; it melts and evolves nitrogen a t 43-45', and explodes a t 60" when rapidly heated. The salt is volatile at the ordinary temperature. It has been generally accepted that hyponitrous acid decomposes into nitrous acid and water, but it is now established that nitrousacid and ammonia are among the products of change; it is found that hyponitrous acid is most stable in alkaline, least stable in acid, solution, an aqueous solution occupying tlie mean position. Determinations of the molecular conductivity of hyponitrous acid have been made, and the results show that it is very low in a freshly prepared solution, increasing, however, with the lapse of time; in the case of the sodium salt, it is higher, and gradually decreases with time.The conductivity constant shows that hyponi- trous acid is very feebly acidic, being of the same order as carbonic acid, and it is interesting t o notice that the value for nitramide is even lower than that for hyponitrous acid. M. 0. F. The Isomerism of Compounds N,O,H,. By ARTHUB R. HANTZSCH (AnmZen, 1896, 292, 340--358).-This paper does not lend itself to condensation. An attempt is made to establish a stereo- isomeric relationship between the two compounds of the formula N202H,, nitramide and hyponitrous acid. The author claims to have refuted Bamberger's criticisms on his views regarding the structure of diazo-salts, M.0. F. Nitrogen Pentasulphide. By WILHELM MUTHMANN and A. CLEVER (Zeits. anorg. Chem., 1896, 13, 200--208).-Nitrogen penta- sulphide, N2S,, is obtained by heating nitrogen sulphide N4S4 (Abstr., 1896, ii, 29s) with pure carbon bisulphide at 100" for two hours under a pressure of five atmospheres. A deep red solution is obtainedINORGANIC CHEMISTRY. 21 together with a yellowish-brown, amorphous precipitate ; the filtrate is distilled until the greater part of the carbon bisulphide is separated, and the residue is allowed to evaporate at the ordinary temperature in a vacuum. The red oily product, consisting of a mixture of the penta- sulphide with sulphur, is triturated with perfectly dry ether as quickly as possible, when the greater portion of the sulphur remains undissolved.The residual sulphur is crystallised out by means of a freezing mixture ( - 25") and the filtrate allowed to evaporate in dry air ; finally, the last traces of ether are removed by allowing the product to remain over calcium chroride, and the residual red oil is filtered. If the product is pure and contains no sulphur, it crystallises from a well-cooled solution in ether in metallic tablets resembling iodine. A cryoscopic determination of the molocular weight, using benzene as the solvent, gave numbers agreeing with the formula N,S,. The amorphous bye-product mentioned above gave, on analysis, numbers agreeing with the formula C,N,Ss ; it is a very finely divided yellow powder, is extremely hygroscopic and obstinately retains traces of carbon bisulphide and nitrogen pentasulphide.It dissolves without alteration in concentrated sulphuric acid, and has properties similar to those of pseudocyanogen sulphide, C,N3S,H. The formation of the pentasulphide takes place according to the equation N4S4 + 2CS, = N,S, + S + 2CNS. The pentasulphide is also formed by heating a mixture of nitrogen sulphide and carbon tetrachloride, an amorphous black bye-product being also formed containing carbon, nitrogen, sulphur and chlorine ; this quickly decomposes on exposure to the air with evolution of sulphurous anhydride. It is also produced by reducing with zinc dust thiotrithiazyl chloride, N,S,Cl, suspended in methylic alcohol ; and generally by the decomposltlon of nitrogen sulphide and its derivatives ; for instance, when nitrogen sulphide is exploded by friction, when the compounds of nitrogen sulphide with the halogens, or with nitrous or nitric acid are boiled with water, and when nitrogen sulphide is heated cautiously with lead oxide.It is not, however, formed by heating nitrogen sulphide with ether or benzene a t 110"--125" in a sealed tube; in this case, the greater part of the sulphide is decomposed into nitrogen and sulphur. With ether, a small quantity of a compound having an odour resembling that of acetamide is formed, and with benzene a small quantity of a compound having the characteristic odour of "oleum absy nthii. '' Nitrogen pentasulphide is a deep red liquid, does not wet glass, has a sp. gr: = 1.901 a t 18", is partially decomposed when distilled even under diminished pressure,solidifies in a freezing mixture to a crystalline mass resembling iodine, and melts at 10-11".It is insoluble in water, but soluble in most organic solvents, and is stable in solution pro- vided it is not exposed to light; in the latter case, it decomposes into nitrogen sulphide and sulphur, The pure compound quickly decom- poses into nitrogen sulphide and sulphur. The absorption spectrum of the solution shows a broad band extending from the D-line into the blue. When boiled with water, a small quantity distils unchanged, but the remainder is converted into ammonia and sulphur. Alkalis decompose it completely in a similar way. When a very minute quantity28 ABSTRACTS OF CHEMICAL PAPERS. of alcoholic potash or soda is added to the alcoholic solution, an intense but transient violet-red coloration is produced, and this reaction is so characteristic that the smallest quantity of the pentasulphide can be detected. When it is added to alcoholic potash, a dark-coloured solution is obtained, from which ether precipitates a yellow, crystalline compound ; this is probably a sulphonitrate, and decomposes directly it is removed from the solution.With alcoholic alkali sulphides, it yields polysulphides and ammonia, and with hydrogen sulphide ammonium polysulphide and sulphur. It is violently oxidised by concentrated nitric acid with formation of sulphuric acid ; and with dilute hydrochloric or sulphuric acid, it yields ammonium salts and sulphur. E. C. R. Trimetaphosphimic Acid and its Decomposition Products.By HENRY N. STOKES (Amer. Chem. J;, 1896,18, 629-663).-A meta- phosphimic acid is defined as a metaphosphoric acid (HPO,),, in which one-third of the oxygen is replaced by an equivalent number of NH groups. TO prepare trimetaphosphimic acid (Abstr., 1895, ii, 2173, the sodium salt, P3N,0,H,Na, + 4R,O, is first isolated by shaking a solu- tion of 30 grams of triphosphonitrllic chloride (P3N3C16) in 150 C.C. of ether with a solution of 110 grams of sodium acetate crystals in 200 C.C. of water ; after agitation for 80 hours, the new salt will have crystallised. When formed at temperatures below 80°, the crystals are orthorhombic (measurements given), with 4H20 ; above 80" they are microscopic flat prisms, with 1H,O. The salt is neutral to litmus.When it is dissolved in an excess of caustic soda solution and the liquid is concentrated, the tetra-sodium salt, P3N,O7H4Na4 + H20, crystallises in long, brittle needles which are very unstable, being easily decomposed by carbonic anhydride. Salts with fewer than three or more than four sodium atoms could not be prepared. The potassium salt, P3N306H3K3, the ammonium salt, P,N,O,H,(NH,), + H20, the magnesium salt, the barium salt, and the baraum sodium salt, P,N,O,H,NaBa + 1&H,O (?), are described. The szlver salt, P,N,O,H,Ag,, is precipitated by adding silver nitrate to a solution of the sodium salt; in the presence of nitric acid, it separates slowly in the form of colourless, monoclinic plates (measure- ments given) ; it is soluble in ammonia.When the tetrasodium salt is used, and the silver nitrate is added in excess, the salt P,N,0GH2Ag, (or P,N,07H4Ag4) is formed as a white precipitate which changes into the yellow, crystalline salt, P,N,O,Ag, (or P,N,07H2Ag,), when left under pure water. A red, crystalline salt of the same formula was also obtained. Trimetaphosphimic acid is prepared by decomposing the silver salt with hydrogen sulphide; it is a gummy mass, distinguished from metaphosphoric acid in that it does not coagulate albumin. The final products of the hydrolysis of this acid are ammonia and orthophos- phoric acid, but the intermediate products, diiminotriphosphoric acid, iminodiphosphoric acid, and pyrophosphoric acid, have been isolated. To separate these, advantage is taken of the insolubility of magnesium iminodiphosphate, pyrophosphate, and orthophosphate, and the solubilityINORGANIC CHEIYIISTRY, 29 of magnesium diirninotriphosphate in a feebly ammoniacal solution.The filtrate from the magnesium precipitate may be neutralised and treated with silver nitrate, to obtain silver diiminotriphosphate, from which the sodium salt may be isolated by treatment with sodium chloride. The magnesium precipitate is dissolved in dilute nitric acid, ammonia is added to the solution until incipient precipi- tation occurs, and then an excess of silver nitrate, whereby silver iminodiphosphate is chiefly precipitated ; it must be purified by dis- solving it in ammonia, reprecipitating as the magnesium salt, and repeating the process. irrisodium diirni91 ot+phosphccte, P,N,O ,H4H a3, is precipitated in the form of microscopic, flat, pointed prisms or rhombic plates on adding alcohol to its aqueous solution ; it is neutral.The silver salt, P,N,O,K,Ag, forms colourless, monoclinic crystals (measurements given), which are easily soluble in ammonia, but only sparingly so in dilute nitric acid. The salt, P,N,O,H,Ag, is precipitated in a white, amorphous form when an nmmoniacal solution of the sodium salt is added to an excess of silver nitrate solution. (rrisodium iminodiphosphate, obtained by decomposing the silver salt, is precipitated by alcohol from its aqueous solution as a syrup which dries t o a gummy mass; it is alkaline. Silver inainodipphosphcbte, P,NO,H,Ag,, is a white precipitate characterised by its granular form; the salt, P,NO,Hhg,, exists in a white and R yellow form.The n&gnesium, fern+ and cup& salts are also described. In discussing the constitution of the foregoing acids, the author expresses the opinion that Gladstone’s pyrophosphodiamic acid (this Journal, 1868, 67) is trimetaphosphimic acid, and his pyrophosphamic acid iminodiphosphoric acid. Trimetaphosphimic acid probably has the constitution expressed by the formula PO(OH)<NH.Po(oH. “PO(OH)>jyH* Imino-diphosphoric acid and di-iminotriphosphcric acid are regarded as PO(OH),*NH*PO(OH), and PO(OH),*NH*PO(OH)*NH*PO(OH), re- spectively. A. G. B. By WILHELRI MUTHMANN and A. CLEVER (Zeits. an0s.y. Chim., 1896, 13,191-199). -The compounds described are very similar to those of arsenic and selenium (Abstr., 1896, ii, 18).Phosphorus pentaselenide is prepared according to the method described by Bogen (Annalen, 1862, 124, 57), by melting together the theoretical proportions of red phosphorus and selenium in an atmosphere of carbonic anhydride. It is most important that the materials should be thoroughly dried, as otherwise large quanti- ties of hydrogen selenide are formed. Phosphorus triselenide is pre- pared in a similar way. Both compounds are unstable when exposed t o the atmosphere, and absorb moisture with the formation of hydrogen selenide and the cori esponding phosphorus acid. The compounds obtained by treating the phosphorus selenides with alkalis are more easily formed and crystallise better than the selenoarsenic salts ; the phosphorus selenides also, unlike the arsenic selenides, are capable of forming ammonium compounds with ammonia, ammonium carbonate, and ammonium sulphide: and double salts with magnesia similar to Compounds of Phosphorus with Selenium.30 ABSTRACTS OF CHEMICAL PAPERS.the compounds obtained from the snlphophosphat8es. The alkali carbonates decompose the phosphorus selenides whereas they do not decompose arsenic selenide. Potassium, triselenophosphite, K,HPSe,, 2&H20, is obtained by the action of potassium hydroxide on phosphorus pentaselenide. It is more easily prepared by gradually adding the powdered triselenide to a cold concentrated solution of potassium hydroxide, cautiously warm- ing the mixture a t 75"--78" until solution takes place, and allowing the filtered solution to crystallise in a freezing mixture; if the mixture is heated above SO", complete decomposition takes place and hydrogen selenide is evolved.It crystallises in highly lustrous octahedra-like forms, which do not, however, belong to the regular system as they show double refraction ; i t gradually decomposes in damp air with separation of selenium and hydrogen selenide : when heated, it gives off water vapour and hydrogen selenide, whilst selenium, and then selenious acid sublime, leaving a residue of phosphate. It is sparingly soluble in cold water, easily so in hot water, and can be re- crystallised from concentrated potassium hydroxide. The aqueous solution is neutral and pale yellow. With dilute acids, it yields hydrogen selenide, and selenium is gradually precipitated.Lead and silver salts give brownish, amorphous precipitates which gradually darken, and the supernatant solution then contains phosphoric acid. Barium salts give a yellow, amorphous precipitate which quickly turns red. Potassium thioselenophosphite, 2K2S, P,Se,, 5 H,O, is obtained by the action of potassium sulphide on the triselenide ; it forms small, colour- less, doubly-refractive crystals having an octahedral habit and an adamantine lustre, and is decomposed at about the same temperature as the preceding salt with evolution of hydrogen sulphide and selenide. The aqueous solution is colourless, and when treated with acids yields hydrogen sulphide and selenide, whilst a reddish-brown compound mixed with sulphur is precipitated.When treated with salts of the heavy metals, it yields dark, amorphous precipitates. Sodium oxyselenophosphate, Na,PSe,O + 10H20, is obbained by the action of sodium hydroxide on the triselenide in a similar way to the potassium salts just described. It crystallises from concentrated sodium hydroxide in pale green, doubly-refracting prisms. The aqueous solution quickly decomposes ; with acids, it yields hydrogen selenide and a brown precipitate containing selenium. When heated in a tube, it yields a sublimate of selenium, and a phosphate. All the above substances give well crystallised compounds with ammonia magnesia mixture. E. C. R. Boric Acid and its Salts. By Lours KAHLENBERG and OSWALD SCHREINER (Zeits. physikal. Cl~enz., 1896, 20, 547--568).-The deter- mination of the molecular depressions due to solutions of boric acid and its salts, lead to the conclusion that in aqueous solution only the one boric acid exists, namely H,BO,, and this is formed if the anhydride or R partial anhydride is dissolved.A solution of borax is identical with the solution containing equivalent quantities of boric acid and sodium hydroxide, By the determination of the conductivity and ofINORGANIC CHEMISTRY. 31 the E.M.F. of borate chains, the number of ions present in the solu- tions was determined, and the conclusions arrived at are that in con- centrated solutions of borax there are present sodium ions and anions containing 2 atoms of boron. These anions are, by dilution, dissociated so that in dilute solutions sodium ions and anions contain- ing 1 boron ion are present, and also undissociated boric acid. A solution of sodium metaborate is identical with a solution of boric acid with the equivalent quantity of sodium hydroxide, or to borax solution with sodium hydroxide ; the solution contains very few hydroxyl ions and is only slightly decomposed by water.By the addition of more sodium hydroxide, only small quantities of higher compounds are formed, since almost the calculated excess of hydroxyl ions are found in the solution. The nature of the reaction of bomtes with polyvalent alcohols was also investigated. The addition of 2 molecules of mannitol to 1 mole- cule of borax in solution has little effect on the freezing point depression, which is increased by further addition of mannitol ; it, moreover, causes the solution to become acidic.If the solution contains sodium hydrox- ide as well as borax, the addition of mannitol a t first decreases the depression, and the results are best explained on the assumption of the formation of a complex anion containing 2 boron atoms and 1 molecule of mannitol, which on adding more of the latter yields an anion containing 2 boron atoms and 2 mannitol molecules; no further combination, however, occurs, so that after this stage is reached the depression is increased. The effects of the addition of mannitol, erythritol, glycerol, and glycol on the electrical conductivity of solutions of borax were also determined, and diagrams representing the results are given. These effects further support the freezing point evidence of the forma- tion of complex ions.L. M. J. Attempt t o Lique* Helium. By KARL OLSZEWSKI ( A 9 2 1 2 . Phys. Chem., 1896 [2], 59, 184--192).-Helium from clheite of density 2.133 from which all gases that are absorbed by magnesium, copper oxide, phosphoric anhydride and caustic soda had been removed, mas cooled down by boiling ethylene and afterwards by boiling air first to - 150" and then to - 220". It was submitted a t the lower temperature to a pressure of 140 atmospheres and the pressure then suddenly reduced to 20 atmospheres, but the gas showed no sign of liquefaction. With the help of the well-known equation T'/Tl=(p/ply$ the author calculates the boiling point of helium as below - 2640, or at least 20" lower than that of hydrogen. The monatomic character of the helium molecule probably accounts for the fact that this gas although of greater density than hydrogen has the lower boiling point,.A comparison of the helium and hydrogen thermometers for temperatures between - 182" and - 210" shows a very fair agreement, and that therefore the coefficient of expansion of hydrogen remains the same even at these low temperatures. Combination of Argon with Water. By P. VILLARD (Conipt.rend., 1896, 123, 377--379).-When carefully purified argon is compressed at 150 atmospheres in presence of water cooled to about O", and the water is H. C.32 ABSTRACTS OF CHEMICAL PAPERS. further cooled so that it solidifies at one point, combination takes place between the gas and the water with production of a crystalline hydrate similar to those formed by many other gases.Mere compression of the gas in presence of water is not sufficient, but the presence of previously formed crystals of the hydrate has the same effect as the freezing of a small quantity of the water. The tension of dissociation of the hydrate is 105 atmospheres at 0" and 210 atmospheres a t 10". C. H. B. A form of Silver obtained on the Reduction of the Sulphide by Hydrogen. By FRANCIS C. PHILLIPS (J. Amey. Chem. Xoc., 1894, 16, 700--703).-1f silver chloride is heated in hydrogen to 300°, re- duction occurs, and the metallic silver appears as a compact, rounded mass having a moderate lustre. Precipitated silver snlphide when heated to temperatures above 450" in hydrogen, is reduced and the dull black powder gradually changes into a mass of tangled wires or threads which exhibit the lustre of highly polished silver.The wires vary in thickness from that of the finest hairs to that of coarse sewing- thread; and they are often so knotted and linked together as to be almost inextricable. Argentite behaves in a similar manner, but stephanite, Ag,SbS+ yields no silver even when heated in hydrogen to dull redness. Artihcial copper sulphide when slowly reduced at about 600" yields lustrous threads of copper resembling those obtained in the case of silver. Compare Opificus (Chem. Zeit., 1888, 649), Bischof (Annah, 1843, 289), and Hampe (Chem. Zeit., 1893, 1692). J. J S. Oxidation of Silver. By CHARLES E. WAIT (J. Amer. Chem. SOC., 1896, 18, 254--259).-The author has previously drawn attention to the fact that a certain bismuth litharge from a Western smelting company contained a large amount of silver (2.94 per cent.).It mas also shown that the silver was not in the metallic state but was probably present as the oxide, since it readily dissolved in warm acetic acid. In order t o account for the formation of the silver oxide, the author has studied the action of various metallic oxides on silver a t fairly high temperatures. He finds that manganese dioxide readily oxidises metallic silver to the oxide. Experiments have shown that, other conditions being equal, the longer the duration of the heating the less silver oxide is formed. 0.5 gram of silver, and 2 grams of manganese dioxide, when heated for 2.5 mins., gave 34.28 per cent. of the silver in the form of oxide, but when heated for 20 mins.gave only 11-78 per cent. Oxides of iron, bismuth, copper and zinc cannot bring about this oxidation, but lead monoxide and dioxide and barium peroxide cm. The lead monoxide no doubt acts as an oxygen carrier, and the presence of silver oxide in certain litharges is to be attributed to the oxidising influence of the litharge. By ORAZIO REBUFFAT (Gaxxetta, 1896, 26, ii, 242--244).--The author replies to Oddo's criticisms (Abstr., 1896, ii, 246). W. J. P. 5. J. S. Hydraulic Cements.INORGANIC CHEMISTRY. 33 Atomic Weight of Magnesium. By THEODORE W. RICHARDS and H. G. PARKER (Zeits. anorq. Cherrt., 1896, 13, 81--100).-1’he methods employed consist in determining the weight of silver chloride obtained by precipitating a known weight of magnesium chloride with excess of silver nitrate: also by precipitating a known weight of mag- nesium chloride with the calculated quantity of pure silver dissolved in nitric acid and then by means of two equivalent solutions of silver and ammonium chloride (the silver solution = 1 milligram per c.c.) deter- mining the point at which equal addition of the two equivalent solutions produces an equal turbidity in the clear supernatant liquid.This point is determined by means of a nephelometer and the method is delicate enough to indicate the difference between 0.002 and 0,003 milligram of silver chloride. All the operations of precipitation, he., are performed in orange-coloured light. Pure magnesium chloride free from water, oxychloride or ammonium chloride is obtained by heating magnesium ammonium chloride con- tained in a platinum boat in a current of perfectly dry hydrogen chloride, at first cautiously, so as to remove any water that may be present : the temperature is then raised and the ammonium chloride volatilised, finally the magnesium chloride is heated to redness until it melts to a clear, colourless liquid.It is then cooled in a current of dry nitrogen, and when cold, the nitrogen is displaced by a current of dry air; the pure dry magnesium chloride is transferred to a weighing bottle without removing it from the drying apparatus. A full description, with drawings of the apparatus, is given in the original paper. The greatest care must be taken to exclude every trace of moisture, as otherwise oxychloride is formed, The following results were obtained.Series I. (5 experiments) the ratio 2AgCl : MgCI, gave Mg = 24.369. Series 111. (6 experiments) the ratio 2Ag : MgCl, gave Mg = 24.365, Series IV. (6 experiments) the ratio 2Ag : MgCl, gave Mg = 24.362. The individual results of Series IV. agree extremely well with one another, the greatest difference amounting to only one-tenth of a milligram of magnesium chloride. Hence the authors accept the last value as the probable atomic weight of magnesium; this compared with the atomic weight of oxygen gives 0 = 16. Mg = 24.362 : 0 = 15.96, Mg = 24.301 : 0 = 15.88, Mg = 24.179. E. C. R. Magnesium Nitride. By A. SMITS (Bec. Trav. Chim., 1896, 15, 135-137. Compare Abstr., 1894, ii, 16).-When an intimate mixture of magnesium nitride, Mg,N,, and anhydrous nickel chloride is heated gently in a sealed tube, il brisk reaction takes place; the product after washing with warm water forms a black mass which, when dissolved in mineral acids, yields a green solution, containing ammonia; the black compound is therefore, probably, a nitride of nickel.Anhydrous ferrous, ferric, and cobalt chlorides behave in a somewhat similar manner, except that the black compounds formed are free from nitrogen. Chromic chloride yields a nitride, CrN, which is not acted on by acids at the ordinary temperature ; this forms small black plates. VOL. LXXII. ii. 334 ABSTRACTS OF CHEMICAL PAPERS. Mercuric chloride yields a green nitride of mercury which is decom- posed by acids. Silver nitrate gives a yellow silver nitride which is decomposed by water, yielding silver oxide and ammonia.Platinic chloride when heated with magnesium nitride, even a t a low tempera- ture, yields metallic platinum. Cupric oxide or sulphate yields a greenish yellow nitride or at higher temperatures metallic copper. Lead peroxide and the oxides of iron also react vigorously with magnesium nitride. J. J. S. Corrosion Phenomena of Zinc Plates. By FRANZ MYLIUS and ROBERT FUNK (Zeits. anorg. Chena., 1896, 13, 151--156).-Cast zinc and cadmium anodes which are contaminated with a layer of oil or grease, when subjected to a current of 1 ampere per square decimetre in a bath of zinc sulphate, are pitted and the skin is undermined by the action of the electric current in such a way that a thin skin is separated.Rolled zinc plates suffer a like corrosion, which is probably due to the mixing of the impurities contained in the zinc by the process of rolling. The corrosion of zinc containing lead is hindered by sudden cooling when cast and by the mechanical treatment of rolling. The mechanical treatment of pure zinc has, however, a very slight influence on the ease with which corrosion takes place. An analysis of the outer skin and the inner layer of a commercial zinc plate gave practically the same percentage composition, so that the corrosion phenomena cannot be due to a difference in the composition of the outer skin. E. C. R. Zinc Carbonate. By KARL KRAUT (Zeits. anorg. Chem., 1896, 13, l-l5).-The author has determined the composition of the products obtained by precipitating zinc salts with alkali carbonates, and hydrogen carbonates under varying conditions.The first product obtainbd when solutions of these salts are poured into solutions of zinc sulphate is normal amorphous zinc carbonate. This, according to circum- stances, is then either converted into the stable crystalline zinc car- bonate, ZnC03,H20, or int?the basic carbonate, 5Zn0,2C02,4H,0, with evolution of carbonic anhydride. The basic zinc carbonate, 2Zn0,C02, described by Boussingault, Wackenroder, and H. Rose, and probably all the basic carbonates having compositions between the compounds ZnO,CO, and 5Zn0,2C02, are mixtures formed by the partial conversion of the primary amorphous zinc carbonate into crystalline carbonate and into the basic carbonate, 5Zn0,2C0,,4H20.If the solution of the zinc salt is added to a solution of alkali carbonate, the normal car- bonate is not formed, but only the basic carbonate, and this is due to the presence of an excess of alkali carbonate during the precipitation. Finally, the basic carbonate, when boiled with excess of alkali car- bonate, is converted into anhydrous zinc oxide, Basic zinc carbonate, 5Zn0,2C02,4H,0, can be obtained pure by dis- solving zinc, hydrated zinc oxide, or zinc carbonate in an excess of aqueous carbonic acid and then boiling the solution, when it is pre- cipitated. It is also obtained very nearly pure by precipitating zinc sulphate with an alkali carbonate from cold dilute solutions or from boiling solutions, if equivalent quantities of both salts are em-INORGANIC CHEMISTRY.35 ployed. Concentration of the solutions determines the formation of the normal carbonate, and dilution, the formation of the basic oar- bonate ; above a certain concentration, excess of potassium or sodium hydrogen carbonate combines with the zinc carbonate to form double salts. A solution of zinc sulphate in water was mixed with a solution of sodium hydrogen carbonate in water, in the proportion ZnSO, : 4NaHC0, ; the precipitate was allowed to remain under the mother liquor until it had become crystalline, then collected, and washed with a solution of sodium hydrogen carbonate. It consisted of small quad- ratic optically active crystals of zinc carbonate, ZnCO,,H,O, and of larger octahedral skeleton forms which were inactive, and gave on analysis numbers agreeing with the formula Na20,3Zn0,4C02,3H,0.The two compounds can easily be separated mechanically. The author was unable to prepare a basic carbonate of cadmium. Only a very small quantity of cadmium carbonate is dissolved by aqueous carbonic acid, even under a pressure of several atmospheres. The precipitate obtained by adding cadmium sulphate to a solution of sodium carbonate consists of cadmium carbonate, CdCO,, mixed with a small quantity of the basic sulphate, 2CdO,SO, ; whilst by adding cad- mium nitrate to an excess of ammonium carbonate, the carbonate CdCO, + &H,O is obtained. Electrolytic Refining of Cadmium. By FRANZ MYLIUS and ROBERT FUKK (Zeits. ccnom~. Chem., 1896, 13, 157--160).-The electro- lytic refining of cadmium is carried out in a very similar way to that of zinc.The cadmium is deposited from a concentrated solution of cadmium sulphste, slightly acidified, on cathodes of platinum or cadmium foil, a current density of 0-5-1 ampere per square decimetre being employed. The deposited cadmium is easily separated from the cathodes, and when melted and sublimed in a vacuum is obtained in flat needles or silvery six-sided tablets. No impurities can be detected by chemical means in the metal so prepared, The lead in the impure metal is completely separated by electrolysis of the sulphate solution with a very weak current, and is deposited partly at the cathode as metal and partly at the anode as peroxide. Iron is most easily detected by means of potassium thiocyanate.The presence of zinc is detected by melting a sample of the metal in a por- celain crucibleand breaking the skin of oxide with a glass rod, so that the bright molten metal is exposed; if the cadmium is free from zinc, the bright surface in a few seconds becomes covered with a many- coloured oxidation skin ; in the presence of zinc, however, the oxidation of the cadmium is hindered, and zinc oxide is first formed. This re- action is so delicate that less than 0.01 per cent. of zinc can be detected. The quantitative estimation of the zinc in cadmium is effected by transferring the cadmium electrolytically through a very dilute acid or salt solution, whereby the whole of the zinc remains in solution together with a small quantity of cadmium, and can now be easily separated from the latter.E. C. R. E. C. R. Crystallised Slags from Copper-smelting. By ALFRED C. 3-236 ABSTRACTS OF CHENICAL PAPERS, LANE (Bull. Geol. Xoc. Amer., 1895, 6, 469--470).-At the smelting works at Dollar Bay and on Torch Lake in Upper Michigan, the copper which has been oxidised during the melting down of the ore is after- wards reduced. The slag produced in this reduction has a strong tendency to crystallise, and contains very large melilite crystals in the form of square tablets with rounded and reticulated faces. The crystals are optically uniaxial and negative; they are merely shells, with a rectangular network of enclosed matter, consisting of a greenish bi- refringent substance, iron oxides, and globules of copper ; a partial analysis, by R.L. Packard, gave SiO,. Fe20s. A1,0,. CaO,MgO, &c. [cliff.]. 34.84 16.78 13.26 [35-12] In cavities in the slag are small scales of hzmatite. L. J. S. Solubility of Lead and Bismuth in Zinc. By WALTH~RE SPRING and L. ROMANOFF (Zeits. cmov*g. Cltem., 1896, 13, 29-36).- A crucible is prepared with a hole bored in the side at a convenient distance from the bottom and stopped with a plug during the prepara- tion of the solution. The crucible is filled with the heavier metal (lead or bismuth) so that the level of the molten metal is just above the hole, the zinc is then added, covered with a protecting layer of fused salt or charcoal, and the crucible maintained at the temperature at which the solubility is to be determined. The contents are stirred about erery 10 minutes.The sample of the top layer is taken by bailing out a sufficient portion. The plug is then knocked out of the side of the crucible, and after the top layer has run out, a sample of the bottom layer is bailed out, A table of the composition of the two layers at temperatures from 266-900" is given. When the results are stated graphically, taking the temperature for abscisszle and the percentage of the two layers as ordinates, two points are obtained for each temperature : one the solubility of bismuth in zinc, the other the solubility of zinc in bismuth, and by prolonging these two until they meet, the temperature is determined a t which the two metals will mix in all proportions. The critical temperature for bismuth and zinc lies between 800-850", that for lead and zinc between 900-950".The curves are analogous to those obtained by Alexkeff for non-miscible liquids. E. C. It. Metaplurnbates. By M. HOEHNEL (A~cla. Plaurm., 1896, 234, 397--400).-Calcium metaplumbate (Abstr., 1896, ii, 248) is much more stable than the metaplumbates of sodium and potassium ; by digesting it with a solution of the acetate of a heavy metal, the meta- plumbate of that metal can be prepared. The following were obtained in this way : metaplunabate of xime, ZnPbO, + 2H,O, reddish-brown, crystalline ; of copper, CuPbO,, black, amorphous ; of mungunese, grey- black, crystalline ; of lend, PbPhO,, reddish-brown, amorphous, and identical with lead sesquioxide. By UBALDO ANTONY AND T. BENELLI (Guaxettcc, 1896, 26, ii, 97-107, 362--362).-The authors have examined the action of water of various C.P. B. Potable Waters which have flowed through Lead Pipes.INORGANIC CHEMISTRY. 37 4 '09 2'32 4-09 1.36 1.84 ~ 1-77 degrees of purity on lead, and have obtained analytical results which are summarised in the accompanying table. ~~ ~ Parts of lend per 100,000. Water alone, unfiltered ....................... Water alone, filtered ............................. Water containing chlorides, unfiltered ..... Water containing chlorides, filtered . . . . . . Water containing sulphates, unfiltered ..... Water containing sulphates, filtered ........ 1 8'19 3 -00 1-36 0 -68 3.41 2.05 2 12.98 4 -09 2.73 1 *50 6'83 3 '41 3 __ 8.19 2'07 0'68 0.67 2'05 1 '64 Water was allowed to remain in contact with lead mire for five days, each 150 C.C.of water being in contact with 285 square cm, of metal. The amount of lead in the unfiltered and filtered water was then determined. The column headed 1 gives tthe numbers for distilled water free from air,-first for the pure water, then for water cont,aining 0.049 gram of sodium chloride per litre, and finally for water con- taining 0.49 gram of sodium sulphate per litre; column 2 gives the corresponding data for distilled water aerated by agitation with air ; column 3 shows the results obtained with water continuously aerated by passing 1 litre of air through it per hour, whilst column 4 gives the data for distilled water through which 1 litre of a i r and 400 C.C. of carbonic anhydride were passed per hour throughout the experiment.The lead was determined by the process previously described by the authors (Abstr., 1895, ii, 549). The authors next experimented on the solubility of lead in water containing various dissolved substances, with the-result,s summarised in the accompanying table. ~ Parts of lead per 100,000. Water used. ~ 1 ' 2 Distilled water alone .............................. 8'19 Distilled water with NaCl ........................ 1.36 3 '41 CaHCO, solution alone ........................... ~ 2'45 CnHCO, solution with NaCl ............. ...I 2.05 CaHCO, solution with Na2S04 .............. , ' 2.18 CaS04 solution alone ............................ 6 '83 &SO4 solution with NaCl ....................... I 5.46 CaSO, solution with Na,SO, ..................... 1 4.78 Distilled water with Na,SO,..................... I 12-98 2'75 6-83 3'14 3'41 3 -32 6.83 6.57 5.87 3 - _ _ - 8 '19 0 '68 2.05 2'63 2 -35 2-05 3-41 3 '51 3-69 4 4 -09 4.09 1 '84 5.70 3.40 3'16 1 -35 1.50 1-77 The experiments were conducted under conditions similar to those prevailing in the earlier experiments. Columns 1, 2, 3, and 4 refer t o the solutions named in the first column (1) free from air ; (2) saturated38 ABSTRACTS OF CHEMICAL PAPERS. with a i r ; (3) continuously aerated by a current of a i r ; and (4) continuously aerated by a current of air and carbonic anhydride; the quantities of sodium chloride and sulphate present in the various samples of water were the same as before. I n the experiments with solutions of calcium hydrogen carbonate, each litre contained 0.04 gram of CaO as carbonate, whilst in the tests with calcium sulphate, each litre of water contained 0.095 gram of CaO as sulphate.W. J. P. Double Chlorides. By RAOUL VARET (Cowpt. vend., 1896, 123, 421--423).-The double chlorides formed by mercuric chloride 2HgCI,,MCI2 + nH20, and HgCI,,MCI, + mH20 are all more or less dissociated when their solutions are submitted to dialysis. The author finds that when dilute solutions of mercuric chloride and various met'allic chlorides are mixed, there is a development of heat which varies in different cases from 0.80 to 1.36 Cal., but generally approximates somewhat closely to 1.0 Cal. The double salts may be regarded as derived from two complex unstable acids, H,HgCI, and H2Hg2C16. Solutions of cupric or cadmium chloride give practically no thermal disturbance when mixed with solutions of other metallic chlorides, and when solutions of the double chlorides of copper or cadmium are dialysed, the proximate constituents can be completely separated.It follows that these double salts exist as such in the solid state only. C. H. B. The Components of Monazite. By G. PAUL DROSSRACH (Bey., 1896, 29, 2452-2455).-The monazite examined came from the Blue Mountains. It was finely powdered, and then decomposed with sul- phuric acid ; from the solution, the thorium was separated by fractiona- tion, and then a large excess of sulphuric acid was added. The erbium metals remain in solution, whilst the cerium metals are precipitated ; their precipitation is rendered more complete by partially neutralising the liquid with soda, for with sodium sulphate, they form double salts which are quite insoluble in a saturated solution of that salt.The solution of the cerium metals is then treated with potassium hydroxide and permanganate, which precipitates the cerium as the dioxide, together with some of the didymium ; the latter is extracted, with dilute nitric acid : strong nitric acid then extracts the cerium, leaving the manganese dioxide. The lanthanum and didymium in solution are converted into nitrates, and these are fractionally precipitated with sodium hydroxide ; didymium is first precipitated, as the basic nitrate, whilst the lanthanum remains in solution. Perhaps the two didymium precipitates obtained contain different elements.The erbium elements in solution are precipitated with oxalic acid ; the oxalates are converted into the hydroxides, and these into the nitrates, which are then fractionally precipitated with magnesia. Yttrium remains in solution ; from the precipitate, ytterbium can first be separated by the basic nitrate method, and erbium next fractionally precipitated with dilute sodium hydroxide, when the solution is foundINORGANIC CHEMISTRY. 39 to contain what is possibly a new element, with an atomic weight in the neighbourhood of 100. C. F. B. Analyses of Steel,-By GIOVANNI GIORGIS and UGO ALVISI. (Gaxxettcb, 1896, 28, ii, 167-1 78).-The authors have analysed a number of specimens of soft steel, the mechanical properties of which have been studied by Biadego (Rivista Tecniccc L'l?zdzcst?ia, 1895, 47, 52; 1896, 2, 6, 7).W. J. P. Steel.-By FRANZ MYLIUS, FRITZ FOERSTER and GEORG SCHOENE (Zeits. nnoiy. Chena., 1896, 13, 38--58).-Iron Caybide in Steel.-The iron carbide was at first extracted from tool-steel by electrolysis. Anodes of steel plates or rods are subjected to a current of 1 ampere per quarter-decimetre in a bath of concentrated zinc sulphate containing 0.1 per cent. of free acid. The electrolysis proceeds without any appreciable evolution of hydrogen, but the anode becomes coated with a grey skin which must be occasionally rubbed off. The residue ob- tained consists of a grey powder which contains from 7 to 9 per cent. of carbon and about 90 per cent. of iron. It is not, however, pure carbide ; for whilst the steel dissolves completely in hot hydrochloric acid, this residue contains a small quantity of insoluble carbonaceous material, The carbide was also extracted with twice normal, and with normal sulphuric acid a t the ordinary temperature, washed as quickly as possible with water, alcohol, and finally with ether, and dried at 120" in a current of hydrogen, care being taken to exclude air as much as possible during the operations.Four experiments with the same steel gave products which contained from 6.5 to 7.2 per cent. of carbon, and were not completely soluble in hot hydrochloric acid, but yielded a small carbonaceoas residue ; this residue is formed by the oxidation of the damp carbide on exposure to air. I n order to obtain an iron carbide completely soluble in hydrochloric acid, the steel is extracted on a filter bed of asbestos with sulphuric acid in an apparatus so arranged that the extraction is performed in an atmosphere of hydro- gen.The residue is washed in the same apparatus with water, alcohol, and ether in an atmosphere of carbonic anhydride, and dried at 100" in a current of carbonic anhydride. This product is almost completely soluble in hydrochloric acid ; the residue, which amounts to 0.1 per cent., consists of silica. It gave on analysis-6.5 per cent. C, 91.96 per cent. Fe, 1.1 per cent. l\ln, 0.23 per cent. Cu, 0.02 per cent. Si, and traces of sulphur, phosphorus, arsenic and nickel. When heated to redness in a current of nitrogen or hydrogen, it lost only 0.4 per cent. in weight. I T O ~ carr*bide, when dry, is not altered by exposure to air.When dried at 100" in a current of hydrogen, and then exposed t o air, it glows and yields iron oxide ; the same phenomenon occurs occasionally when it is dried at a low temperature in carbonic anhydride or nitrogen. When heated to redness in hydrogen, i t is not pyrophoric; but the hydrogen must be completely displaced by carbonic anhydride or nitrogen to render it stable on exposure to air. When damp, it is rapidly oxidised on exposure to air, yielding a brown powder, which consists of a mix- ture of oxide of iron and the carbon compound C,,H,O, described by40 ABSTRACTS OF CHEMICBL PAPERS. Bourgeois and Schiitzenberger (this Journal, 18'75,7SS) and by Zabudzky (Abstr., 1882, 427, 660). This compound is most easily obtained by treating the carbide with a solution of ammonium sulphate in dilute sulphuric acid.The carbide is only slightly decomposed when heated with water a t 145", but when heated a t 400" to 430" i n a current of steam and nitrogen for half an hour, i t increases 13 per cent, in weight and yields a black residue containing oxides of iron and free carbon together with a combustible gas consisting for the most part of hydrogen. When heated at a red heat in a current of steam for half a n hour, it increases 29 per cent, in weight, yields a black powder of ferrosoferric oxide and a gaseous mixture of hydrogen, carbonic oxide and carbonic anhydride. It is not appreciably attacked by very dilute acids ; when treated with one-tenth normal hydrochloric acid:at SO", ft distinct evolution of hydrogen is observed.It is very gradually dis- solved by normal hydrochloric acid, and 1 gram after remaining sealed up in a vacuum tube for 10 days with normal hydrochloric acid gave only 2 5 C.C. of a combustible gas. I n concentrated hydrochloric acid, it is easily soluble and leaves only a minute residue which is insoluble in water, but easily soluble in alcohol or ether. In no case did the authors observe the formation of amorphous intermediate carbonaceous pro- ducts during the dissolution. An analysis of the gas evolved on dis- solving the carbide in hydrochloric acid gave 92.3 per cent. H, 6.3 per cent. hydrocarbons, 1.4 per cent. N. The ratio of the combined hydrogen to the carbon very nearly approaches that required for a member of the series CnH,,l+, and the density of the hydrocarbon is about the same as that of pentane.Although the iron carbide is not altered by heating to redness, it melts at a white heat in an atmo- sphere of nitrogen, and carbon is deposited. The cast iron regulus thus obtained contains 4.36 per cent. of caybon ; it is very brittle, shows a radial fracture like that of '' spiegeleisen," and contains no graphite. In order to prove that the iron carbide is present in the steel as a definite chemical compound, the authors extracted samples of the steel with various acids of different strengths. Normal acetic acid yielded the best results, over 90 per cent. of the carbon present in the steel being obtained as carbide ; whilst with N x 4 hydrochloric acid only 25 per cent.of the carbide mas obtained. The product, in all cases, contained the same percentage of carbon, Samples prepared from different sorts of steel gave the same composition onanalysis as did also a sample prepared from carbonised iron, made in the laboratory from pure electrolytic iron which was carbonised by means of pure acetylene. It is evident from the above results that the carbide is a definite chemical compound of the formula Fe,C. Only in one reaction does it show a behaviour different from that which one would expect from a compound of the formula Fe,C, which should decompose when treated with acids according to the equation Fe,C + 6HC1= 3FeC1, + CH4 + H2. E. C. R. Crystallised Martin-slag. By A. HARPF ( J c ~ ~ L Him., 1896, ii, Ref. 37 ; from astew. Zeits.f. Berg- umd Huttenwesen, 1895, No. 7).-INORGANIC CHEMISTRY. 41 Crystallised slag from the Martin-furnace at the iron and steel works a t Donawitz, gave the following results on analysis :- SiO,. FeO. h1,0,. MnO. CaO. MgO. Fe. Total. sp. gr. 30.75 60.23 2.07 5.10 1.30 1-10 0.G7 100.62 4.280. This shows the material to be fayalite. Some of the crystals are tabular, others are columnar; they show the forms b (OlO}, c ( O O l ) , tl { l l O ) , s {102), and x (104). By NAGENDRA CH NAG (Zeh. ccz20vg. Clma., 1896, 13, 16-18. Compare Durrant, Proc., 1896, 96 and 244).-When precipitated, cobalt carbonate is treated with bromine in presence of potassium hydrogen carbonate, carbonic anhydride is evolved, and a green solution is formed, from which a green compound, which does not contain bromine, is precipitated by shaking the solu- tion with alcohol and ether.The new compound is very unstable and has not yet been isolated, but it appears to be an oxidation product probably corresponding t o the ferates. A nickel compound was not obtained under like conditions. When either cobalt or nickel carbonate is treated with sodium acetate and bromine, a deep orange-red solution is obtained. The nickel solution when boiled gives a violet precipitate and a neutral apple-green solution. The cobalt solution gives no precipitate when boiled. These orange-red compounds appear to be lower oxidation products, for when potassium hydrogen carbonate is added to the brown cobalt salt solution containing an excess of bromine, the green solution is obtained.E. C. R. L. J. S. New Cobalt and Nickel Salts. Constitution of Cobalt, Chromium and Rhodium Bases. By SOFUS M. JORGENSEN (Zeds. anovg. Clt,em., 1896, 13, 172-i90 ; see also Abstr., 1895, ii, 47; 1896, ii, 424).-The author has determined the relative quantities of the compounds formed in the preparation of Erdmann’s cobalt triamine nitrite (J. pr. Clwnt., 1866, 97, 412), and gives a table of the results. Cobalt carbonate (10 grams) dissolved in hydrochloric acid (50 c.c.), was treated with varying quantities of sodium nitrite and 30 per cent. ammonia and the mixture oxidised in various ways ; the dark brown liquor obtained, was filtered from the insoluble matter and evaporated to dryness in a draught. The residue was dissolved in 50-70 C.C. of cold water, whereby a second residue was obtained, and a dark brown solution which, when treated with dilute nitric acid, yielded the xantho-nitrate. The insoluble residues were washed with cold water until free from chlorine, whereby the croceo- and xantho-chlorides were removed, and the residue free from chlorine was extracted with hot water containing some acetic acid, when the triamine nitrite was dissolved.This was separated from the croceo-diamine nitrite and other sparingly soluble impurities by frac- tional crystallisation from very dilute acetic acid ; the bulk of the croceo-diamine nitrite was obtained in the last extraction of the residues. The last two salts are more soluble the more concentrated the acetic acid used. Of the salts of the composition Co(NH,),(NO,),, only the ordinary triamine nitrite and the croceo-diamine nitrite were obtained,42 ABSTRACTS OF CHEMICAL PAPERS.The best yield of these salts, amounting to one-fourth of the cobalt used, is obtained by employing 27 grams of sodium nitrite and 200 C.C. of 20 per cent. ammonia and passing air through the mixture for two hours and then allowing the mixture to remain for 3 days in an open dish. The relative quantities of the materials employed, the dilution, and especially the method in which the oxidation is effected, have a great influence on the composition of the products obtained. B t y e r e n t f owns of the triccmine n i t y i t e ccnd the d i n i t i - o t w h n i n e chh%de.-The triamine nitrite crystallises,. according to the conditions, in needles or in rhombic tablets.The dinitrotriamine chloride crystallises in quadratic tablets, in octahedra, in elongated prisms and in other forms ; all these modifications when precipitated with hydrochloric acid are converted into quadratic tablets. The dinitrotriamine chloride is dissociated when dissolved in water, but only to a small degree a t the ordinary temperature : when heated with water containing a few drops of acetic acid on the water bath, it yields the triamine nitrite according to the equation Z(NO,),Co(NH,),Cl= (N0,),Co(N€€,),N0,+Co(NH,),(N02)CI,. The author was unable to obtain an isomeric triamine nitrite by adding sodium nitrite to the solution, and the only compound isolated was ammonium cobalt dinitrite. A c t i o n of d i l u t e ucids o n croceocobult-diuminecobcLZt nitrite.-Croceo- cobalt-diaminecobalt nitrite is converted by prolonged treatment with hydrochloric acid at the ordinary temperature into a dark brown crystalline powder which consists of %he chloronitro-tetramine chloride and c?~Zo~onit~otetrc~minecobccZt-dictn~i~.zecobcLlt nit?*ochlos*icle, NO,CoCl( NH,),CI, (NH3),Co( NO,),.The latter salt is separated by washing the precipitate with water, in which it is insoluble: it is then washed with alcohol (95 per cent.), dried over sulphuric acid, and purified by dissolving it in water and precipitating with hydrochloric acid. When heated with sodium nitrite, it is quantitatively reconverted into the croceodiamine nitrite. The above constitution is proved by the following reactions. When heated with hydrochloric acid, it yields the chloronitrotetramine chloride, and with dilute nitric acid, the aquoxantho-nitrate.When treated with water containing acetic acid, it yields aquoxantho- salts, Silver nitrate precipitates all the chlorine as silver chloride. When heated with water, ammonium nitrate and ammonia, it yields xanthonitrate and flavonitrate. It is also obtained by mixing a solution of the chloronitrotet'ramine chloride in water containing a small quantity of acetic acid with a solution of the ammonium-diamine nitrite in water and hydrochloric acid. E. C. R. Action of Chromic Acid on Thiosulphuric Acid. By ANTONIO LONGI (Gaxxettcc, 1896, 26, ii, 119--141).-1n view of the fact that Longi and Bonavia (Abstr., 1896, ii, 625) have shown Diehl's inter- pretation of the reaction between a chromate and a thiosulphate to be erroneous, the author has examined the reaction in detail.On adding a mineral acid and pot'assium dichromate t o a dilute sodium thiosulphateINORGANIC CHEMISTRY. 43 solution, the reaction which takes place is in accordance with the equs- tion 2H2Cr0, + 6H,S20, = 3H,S,0G + Cr,(OH), + 2H,O, small quanti- ties of sulphuric acid being simultaneously formed in accordance with a secondary reaction expressed by the following equation, 3H,S,oG + 14H,CrO4 + 16H,O = 13H,SO, + 7Cr,(OH)G. Hydrogen sulphide is also produced, and it is shown that both sodium thiosulphate and tetrathionic acid give this gns when small quantities of an acid or a chromic salt are added to their solutions ; more hydrogen sulphide is obtained at high than at low temperatures, and tetrathionic acid is more stable than the thiosulphate, for sodium thiosulphate gives hydrogen sulphide when treated with hydrogen peroxide or acetic acid, or when carbonic anhydride is passed through its hot solution.W. J. P. Formation of Antimony Cinnabar. By JOHN H. LONG (J. Amel.. Chem. SOC., 1896, 18, 342-347. Compare Abstr., 1895, ii, 399).-It has been previously shown that antimony cinnabar, obtained by boiling a solution of antimony chloride or tartrate with sodium or calcium thiosulphate, has the composition Sb,S,. On boiling solutions of antimony salts with hydrogen sulphide or an alkali sulphide and sulphurous anhydride, the amorphous sulphide mixed with free sulphur is always formed.A thiosulphata precipi- tates the cinnabar red product from either acid or neutral solutions of antimony salts, the addition of a few drops of ammonia, however, pre- vents the precipitation. Antimonious oxide remains unaltered when boiled with a thiosulphate, but on adding a small quantity of hydro- chloric acid, the red sulphide is slowly formed; only a small portion, however, of the oxide can be converted into sulphide by this method, notwithstanding the amount of acid added. Antimony oxychloride readily gives the cinnabar-red precipitate on boiling with a thiosulphate solution, evan without the addition of acid. J. J. 8. Action of Alkali Sulphides on Auric Sulphide. By UBALDO ANTONY and ADOLFO LUCCHESI (Gacxxettcc, 1896, 26, ii, 350-353).- Auric sulphide, Au, S,, dissolves in sodium monosulphide solution a t 3-4" giving an odourless, reddish liquid which, however, rapidly becomes yellow and acquires the characteristic odour of the polysulphides. On pouring the odourless solution into absolute alcohol, an oily heavy mass is deposited which rapidly changes into a mass of white needles ; the latter consist of sodium sulphcmwite Na,AuS2, which is a compara- tively stable substance and apparently results from the reduction of the oily sulphaurate, Na,AuS3. Sodium sulphaurite is soluble in water, and the solution gives precipitates with solutions of metallic salts ; siluev sulphawite, Ag,AuB, is obtained as a heavy reddish precipitate.Similar results are obtained with potassium monosulphide. W. J. P. Purple of Cassius. By UBALDO ANTONY and ADOLFO LUCCHESI (Gaxxetta, 1896, 26, ii, 195--196).-0n adding a smaller quantity of mercurous chloride t o a solution of auric chloride than is required by44 ABSTRACTS OF CHEMICAL PAPERS.the equation 3HgCl + AuCI, = SHgCI, + Au, metallic gold is deposited ; but if excess of mercurous chloride be added, that which remains unchanged assumes the characteristic colour of purple of Cassius ; similar results are obtained with cuprous chloride. Further, if barium sulphate and mercurous chloride are suspended in water and excess of auric chloride then added, the barium sulphate takes up the gold and becomes the colour of purple of Cassius. Since these several salts can assume the colour of purple of Cassius, it follows that the true purple of Cassius is not a compound but merely stannic acid mechanically coloured with metallic gold.W. J. P. New Double Salt of Platosemidiamine. By DAGMAR SCHOU (Zeits. anoyg. CJLem., 1896, 13, 36 -37)-A pZatosenLiclianai.lze clilode ccwbonate 2Pt(NH,),C12, Pt,(NH,)4C12C0,, is obtained by mixing a solution of potassium ylatinochloride in water at 40" with a mixture of ammonium hydrogen carbonate cnd water, and then passing carbonic anhydride through the solution until i t acquires an indigo blue colour and some quantity of a blackish-blue precipitate is deposited. The solution is precipitated with alcohol, and the precipitate washed successively with water and alcohol, and dried by exposure to the air. It forms small crystals and aggregates, is sparingly soluble in water, insoluble in alcohol and ether, is gradually decomposed by cold hydrochloric acid, and, by boiling with the acid, is converted into platosemidiamine chloride. When boiled with ammonia, a small quantity remains undissolved, and the solution when saturated with hydrogen chloride and treated with potassium platinochloride gives a precipitate of Alagnus's green salt. When treated with a slight excess of silver nitrate, a yellow solution of platosemidiamine nitrate is obtained. E. C. R. Iridio-ammonium Compounds. VI. By WILHELM PALMAER (Zeits. ccnoy. Chem., 1896,13,2 1 1-228; see also Abstr., 1889,35 2 ; 189 I, 402 and 11 65 ; and 1896, ii, 179).--l.1.idiotet.1.ccnaine diclJorochloYide, Ir(NH3)4C13, H,O, has been obtained in very small quantities by the action of ammonia on iridium chloride (Abstr., 1896, ii, 179). It is extracted from the mixture thus obtained, together with ammonium chloride, by means of cold water, and is separated from the am- monium chloride by fractional crystallisation from water, or by crystallisation from 40 per cent. alcohol. It cry stallises in aggregates of thin, yellowish-white needles, or, when slowly crystallised, in bright yellow prisms belongin.. to the monoclinic system (6 : b : c = 0-72078 : 1 : 0.65354 ; /3 = 53" l?. The water of crystallisation is not entirely removed by heating a t 100". It is soluble in 15 parts of water at the ordinary temperature, and has the electric conductivity p= 104.5 at 25" (V=lOOO). When heated, it yields ammonia and ammonium chloride and a residue of metal remains. With concentrated sulphuric acid, one-third of the chlorine is evolved as hydrogen chloride. The dichZorosuZp~~cte, [Ir(NH3),C12],S0,,2H20, obtained by treating the preceding compound with concentrated sulphuric acid, crystallises in thin, lustrous, bright yellow scales, gives off ammonia when heated,MINERALOGICAL CHEltIISTRY. 45 and is sparingly soluble in water ; the solution has a neutral reaction. The clici~Zo~oci~Zo~~oi~iclite, 3Ir(NH,),Cl, + 3IrCl,, is a leather-coloured precipitate, insoluble in cold water, and is attacked by concentrated sulphuric acid at 100". The insoluble double salt of the empirical formula Ir(NH,),Cl,, obtained by the action of ammonia on iridium chloride (Zoc. cit.), is partly grey and partly leather-coloured. The grey compound is attacked by concentrated sulphuric acid at 1 OO", the leather-coloured compound partly at 110" and partly at 100" ; hence the grey com- pound is the double salt of aquopentamine chloride, and the leather- coloured compound the double salt of pentamine or tetramine chloride and iridium trichloride. The reactions of the tetramine salts are described in detail. The bromide is precipitated in bright yellow, rhombic crystals, the CLZCYO- chloride in orange-red, four-sided tablets, the platinochloride in bright yellow aggregates. The solutions of the tetramine salts give a violet colouration with chlorine water. Finally, the author gives a complete YesumE of the chemical and physical properties of the pentamine, aquopentamine, hexamine, and tetramine compounds of iridium, and compares them with the corres- ponding compounds of chromium, cobalt, and rhodium. 8. C. R.