20 ABSTRACTS OF CEIEMICAL PAPERS. In o r g ani c C h e mi s t r y. Critical Constants of Hydrogen Chloride, Phosphide, and Sulphide. By ANATOLE LEDUC and P. SACERDOTE (Compt. 7vend., 1897, 125, 397--398).-The trihydrogen phosphide was prepared from a solution of cuprosodiphosphonium chloride, the hydrogen sulphide byIN ORU AN 1C C H Elf ISTRY. 21 the action of hydrochloric acid on antimony sulphide. results were obtained : The following Critical Critical temperature. pressure. Hydrogen chloride ........... 52' 83 atmos. Hydrogen phosphide ......... 52.8 64 9 9 Hydrogen sulphide ......... 100 90 ?? The results agree somewhat closely with those obtained by Dewar in the case of the chloride and sulphide, and also agree fairly well, SO far as temperature is concerned, with the results of Vincent and Chappuis for the chloride, and of Olzewski for the sulphide, although there are considerable differences in the pressures. C.H. B Hypoiodous Acid and Hypoiodites. Ry ROBERT L. TAYLOR (Clmn. News, 1897, 76, 17-20, 27--29).-Schonbein found that ammonia decolorised an aqueous solution of iodine, yielding a liquid that bleached indigo and gave a blue colour with starch ; the author confirms this statement and shows that similar solutions may be pro- duced with potash, soda, lime-water, and barium hydroxide. The bleaching strength of such solutions, ascertained by titration with standard indigo-carmine, corresponds with tohe amount of iodine present, which amounts to about 1 per 5000 ; stronger solutions may, however, be prepared by using some precipitated iodine with the iodine water.These solutions decompose in a few hours a t ordinary tem- peratures, and in a few minntes when boiled, with the -production of the iodide and iodate. Acids decompose them with the liberation first of hydriodic and hypoiodous or icdic mids, which at once react so that free iodine and water are obtained. With silver nitrate, the solutions yield a dark buff precipitate of the hypoiodite mixed with hydroxide and iodide; with a cobalt solution, a black precipitate on standing; with a manganous salt, a dark brown precipitate immediately ; with lead salts, a precipitate cont,aining brown lead peroxide, and with hy- drogen peroxide, an immediate and copious evolution of oxygen. The author has also obtained such solutions by Lunge and Schoch's method of triturating iodine, lime, and water together, and attributes their failure to delay in testing these solutions or to their having been heated.Moreover, solutions obtained by shaking mercuric oxide with iodine water are observed to have a feeble bleaching action, and on the addition of a drop of alkali, immediately become as active as the hypoiodite solutions just referred to, which they then resemble in all other respects. These solutions of free hypoiodous acid can be also made stronger if iodine is suspended in the iodine water; they can also be obtained by shaking iodine water with silver salts; in the latter case, however, they are very unstable, whilst when made by the iise of meicuric oxide they are more stable than the hypoiodite solu- tions. The free acid decomposes into hydriodic and iodic acids, which react and yield free iodine and water; it does not turn starch blue until after exposure to the air, but with silver nitrate it gives a pre- cipitate consisting of the iodide, and iodate after boiling.D. A. L.22 ABSTRACTS OF CHEMICAL PAPERS, Formation of Active Oxygen. By W. P. JORISSEN (Ber., 1897, 30, 1951-1953. Compare Engler and Wild, Abstr., 1897, ii, 402).-The author considers that the mechanism of oxidation by oxygen gas, in which a quantity of active oxygen equal to that con- cerned in the primary oxidation is produced, is not necessarily the same in all cases. The primary formation of a peroxide frequently affords the most simple explanation ; whilst, on the other hand, cases may arise accompanied by electrical phenomena or by anomalies in the velocity of change which are best explained by van't Hoff's assumption of the dissociation of the oxygen molecule into oppositely charged ions (corn- pare Jorissen Abstr., 1897, ii, 253 and 487).Reduction of Concentrated Sulphuric Acid by Copper. By CHARLES BASKERVILLE (J. Anzer. Chenz. Xoc., 1S96, 18, 942-947. Compare Abstr., 1896, ii, 474).-The author has repeated some of his former experiments, and now states that concentrated sulphuric acid (1.84) is reduced by copper when air is absent and at temperatures far below 86", in fact, at ordinary atmospheric temperatures, with the formation of copper sulphate, cnprous sulphide, and sulphurous anhy- dride, The author also adheres to his previous statement that copper reduces concentrated sulphuric acid a t 0" (compare Andrews, Abstr., 1897, ii, 22).Conversion of Nitrosohydroxylamines into Hyponitrous Acid, By ARTHUR R. HANTZSCH (Ber., 1897, 30, 2356-235S).- Hyponitrous acid is formed by the direct action of nitrous acid on hgdroxylamine in solution in methylic alcohol. It is also produced when dimethylnitrosohydroxycarbamide, NMe,* CO*N(OH) *NO, which may also be regarded as isonitraminecarbnmide, NMe,. CO*N,O,H, is treated with alkalis a t 0". From this it would appear probable that nitrosohydroxylamine and hyponitrous acid are tautomeric. The parallel formation of nitramide, N,O,H,, from potassium nitrocarba- mate, KO*CO*N,O,K, points to the probability that nitramide and hyponitrous acid are in reality stereoisomeric diazohydrates, T.E. J. J. 8. HO-R HO-8 N-OH HO*N Action of Arsenious Acid on Metallic Oxides, Oxychlorides, and Amidochlorides. By C. REICHARD (Bey., 1897,30,1913-1916. Compare Abstr., 1894, ii, 350).-The author has studied the behaviour of the oxides of copper, mercury, silver, nickel, cobalt, tin, chromium, manganese, and bismuth,copper oxychlorides, and the mercury ammonio- chlorides towards solutions of arsenious oxide in water, soda, and ammonia. The observations, mainly qualitative, are embodied in a lengthy table, which does not admit of compression. By NICOLAE TECLU (J. pre Chern. 1897, S3, 178--18O).-Hirn (Ann. chim.phys., 30, 319) has stated his inability to demonstrate optically the presence of particles of carbon in an ordinary gas flame, and concluded that at a high temperature these particles are transparent.A. H. M. 0. F. Characteristics of Flames.INORGANIC CHEMISTRY. 23 The author finds that when the shadows of a yellow and blue gas flame, a candle flame, and the flame from a petroleum lamp are thrown on to a screen and photographed, no difference is noticed in the case of the two gas flames, but a slight darkening in the shadow of the candle flame is observed, and a very marked one in the case of the petroleum flame. The difficulty in obtaining the shadow of the particles in an ordinary luminous gas flame is due t o their exceeding fineness. If, however, twelve fish-tail burners are placed ona behind the other and lit up by an electric lamp, the darkening in their shadow is very pronounced.This cannot be caused by gases or vapours, because a layer of gas of t.he same thickness as the twelve fish-tail burner flame gave no shadow at all. A. W. C. NOTE BY EDITOR.-soret (Phil. Mag., 1875, p. 50), and Burch (Abstr., 1885, 466) have incontestably proved that the flame of a candle contains solid particles. Preparation and Properties of Potassium Percarbonate. By ARTHUR VON HANSEN (Zeit. EZektil.ochem., 1897, 3, 445-448. See Constam and von Hansen, Abstr., 1897, ii, 550).-The influence of small variations of temperature on the yield of potassium percarbonate is not very marked, if the specific gravityof the solution of potassium carbonate surrounding the anode is not allowed to fall below 1.52. Under these circumstances, the temperature may rise as high as 0' without sensibly affecting the yield.If the concentration falls, how- ever, even a little below saturation, the yield of solid percarbonate suffers considerably, owing to the great solubility of the salt in more dilute solutions of potassium carbonate. High current density a t the anode is also essential ; with 0.5 to 2 arnpBres per sq. dcm,, the product contains from 25 to 55 per cent. of percarbonate, whilst with 30 to 60 ampAres per sq. dcm. it contained 85 to 95 per cent. The best results are therefore obtained by allowing a saturated solution of potassium carbonate to flow slowly into the anode compartment of the electrolytic cell at the bottom ; the solution which has already undergone electro- lysis, and in which the solid percarbonate remains suspended, floats on this and is slowly expelled from the cell, carrying with it the percar- bonate; this is collected and dried on a porous plate in a current of dry air, which is finally warmed to about 40'.The yield of 87 to 93 per cent. salt is from 2.2 to 2.4 grams per ampere hour. The salt is more stable than was at first supposed. When dry, it is only slowly decomposed a t loo', a temperature of 200' to 300' being required for its rapid decomposition. An aqueous solution decomposes slowly at the ordinary temperature, but rapidly at 45'. Completely dry potassium percarbonate may be preserved without undergoing appreciable change, but when moist it suffers somewhat rapid decom- position. It is very little soluble in alcohol, but extremely soluble in water, from which it cannot be recrystallised.It may, however, be purified by digestion at - 5' to - 10' with a concentrated solution of caustic potash ; this dissolves the bicarbonate, and the remaining solid, after filtration and washing with alcohol, contains 95 to 99 per cent. of potassium percarbonate. T. E.24 ABSTRACTS OF CHEMICAL PAPERS. Influence exercised by Ferric Oxide on the Formation of Sodium Sulphate from Sulphurous Anhydride, Air, and Sodium Chloride. By JEAN KRUTWIG (Eec. Tyav. Chim., 1897, 16, 173-180).-1ron pyrites was finely divided and thoroughly dried, and then mixed with finely divided common salt which had previously been heated to redness. The mixture, placed in a porcelain boat and in- troduced into a combustion tube, was heated in a Glaser furnace, while dried air was passed slowly through the tube; the air was then made to traverse tubes filled with glass wool, and finally through a tube containing a solution of potassium iodide.The first experiments were made with a mixture of pyrites and sodium chloride only; in later experiments, ferric oxide was introduced. The conclusions arrived at are : (1) ferric oxide acts as an oxygen carrier ; (2) the conversion of sodium chloride into sodium sulphate depends on the amount of ferric oxide present ; (3) the temperature a t which the operation is carried on has a great influence on the results. By GILBERT J. FOWLER and PHILIP J. HARTOG (J. 5'oc. Chem. Ind., 1895,14,Z43--245).-The authors have attempted to prepare silver alloys which, whilst possessing the whiteness of silver, should not be liable to tarnish.The following alloys were obtained by fusing their constituents and then quickly cooling. Silver-zinc alloys :- 1-Silver 95, zinc 5 per cent, ; 2-silver 93, zinc 7 per cent. ; 3-silver 90 and zinc 10 per cent. Silver-nickel alloys :--1-Silver 95 and nickel 4 per cent. ; 2-silver 90 and nickel 10 per cent. Silver-nickel-zinc alloy:-silver 90, nickel 5, and zinc 5 per cent. Silver-copper-zinc alloys : -1-Silver 75, copper 15, and zinc 10 per cent.; 2-silver 67.87, copper 5.1'7, and zinc 27.47 per cent. Silver-aluminium alloy :-Silver 90 and aluminium 10 per cent. Silver-tin alloy :-Silver 95 and tin 5 per cent. All these alloys tarnished readily, but it was found that the stain was more easily removed by rubbing with a chamois leather than in the case of pure silver.The authors also tried to obtain suitable alloys by electrolytic deposition from solutions containing silver and zinc ; silver, zinc, and copper ; or silver and aluminium, but the results were not satisiactory. (Compare S. P. Thompson, Proc. Roy. Soc., 1387, 42, 387, and C. Winkler, this Journal, 1872, 1134). Solubility of Calcium Sulphite in Water and in Sugar Solu- tions. By JULIUS WEISBERG (Bull. Xoc. Chinz., 1896, [iii], 15, 1247--1250).-The calcium sulphite used was prepared by the action of sulphurous anhydride on milk of lime, and throughout the experiments care was taken to avoid contact with air as far as possible. At 18", a litre of water dissolves 0.043 gram of calcium sulphite only; a litre of a 10 per cent.sugar solution dissolves 0.0825 gram, and a litre of a 30 per cent. sugar solution dissolves practically the same amount. The sulphite in solution oxidises rapidly, especially if the solutions are heated. C. H. B. By Josh R. MOURELO (Compt. rend., lS97, 125, 462-464).-The various forms of phosphorescent strontium sulphide, when exposed to sunlight in presence of air, give off some hydrogen sulphide, and are more or less completely J. J. S. Silver Alloys. J. J. S. Stability of Phosphorescent Strontium Sulphide.INORGANIC CHEMISTRY. 25 oxidised. The least stable variety is that prepared by the action of hydrogen sulphide on the carbonate, and the most stable is that pre- pared by the author’s modification of Verneuil’s method (Abstr., 1897, ii, 450-469) ; the latter, in fact, changes very slowly.Impurities, more especially .alkali compounds, increase the stability in all cases, most probably because they cause incipient fusion, and the sulphide thus becomes covered with a kind of protective glaze. I n all cases, the rate of oxidation is more rapid if the sulphide is powdered, but the phosphorescence is not affected by the oxidation, except in so far as the sulphide is converted into non-phosphorescent compounds. C. H. B, Compounds of Metallic Hydroxides with Iodine. By THEODORE RETTIE (J. Arne?*. Clmz. Soc., 1897, 19, 333--.339).-1f a solution of a magnesium salt is mixed with a supersaturated solution of iodine in aqueous potash, or if it is first mixed with a solution of iodine in potassium iodide, and then with a small quantity of aqueous potash, a dark, reddish-brown precipitate is fGrmed containing magnesium hydroxide and iodine.The author now states t h a t this compound must be considered rather as a mixture, since the ratio between the iodine and magnesium varies from 2.1 to 6.3. On substitubing zinc acetate for a magnesium salt, an iodiscd zinc hydroxide is obtained in which the ratio between iodine and zinc is fairly constant, varying from 2.4 to 2.9. Cadmium nitrate also gives a fairly stable precipitate. The iodine is, however, only in a weak state of combination, and is gradually removed by washing or drying. (Compare Walker and Kay, Abstr., 1897, ii, 261.) L. DE K. New Mode of Combination between M e t a l s : Alloys of Cadmium with Silver and with Copper.By JEAN B. SENDERENS (Bull. Xoc. Chinz., 2896, [iii], l5,1241--1247).--When a bar of cadmium is placed in a neutral solution of silver sulphate, the quantity of cadmium sulphate that is formed is exactly equivalent to the silver precipitated, but the loss of weight of the bar is much greater than the weight of the cadmium dissolved. Further, if the precipitated silver is allowed to remain in contact with the cadmium, the loss of weight of the latter increases during a period of five months or more, and in one set of experiments the loss was more than six times as great as the calculated amomt. There is no decomposition of the water, and no appreciable formation of cadmium oxide, prorided that air is excluded. Cadmium sulphate solution has no action on cadmium.If freshly precipitated silver is brought in contact with cadmium in presence of cadmium sulphate solution, similar results are obtained, and this is true also when pure water is substituted for the solution of the salt. The cad- mium removed from the mass of the metal seems to combine with the finely divided silver, forming an alloy, and since the action practically ceases when the alloy has a composition corresponding with the formula AgCd,, it would seem that this is the limiting compound. The important point is, that the formation of this alloy is subsequent to, and not simultaneous with, the precipitation of the silver. Silver26 ABSTRACTS OF CHEMICAL PAPERS. acetate behaves like the sulphate, but with silver nitrate, the pre- cipitated silver adheres so firmly to the cadmium that investigation becomes almost impossible, Copper sulphate, chloride, and acetate yield similar results, with the exception that the formation of the cadmium alloy is simultaneous with the precipitation of the copper.The velocity of the reaction varies with the different salts, but pre- cipitation is completed more quickly the more dilute the solution. No hydrogen is liberated. The ratio of the excess of cadmium removed to the calculated quantity is independent of the quantity of copper solution used, but varies with its concentration; it is greater the more dilute the solution. These facts may be due to the existence of several definite alloys of copper and cadmium, or to an influence of the concen- tration of the solution on the proportion of the copper that is converted into alloy; the latter seems more probable.Copper nitrate, like silver nitrate, does not lend itself to experiments of this kind. Cadmium precipitates all the lead from dilute solutions of lead acetate, but there is no evidence of any formation of an alloy. With solutions of lead nitrate, or concentrated solutions of the acetate, the precipitated metal adheres strongly to the cadmium. C. H. B. Improvements in the Preparation of Metallic Alloys by Electrolysis. By JOHANN WALTER (Zeit. Elektrochem., 1897, 3, 385--388).-In the electrolytic preparation of alloys, it is difficult to obtain them in a homogeneous condition, To overcome this difficulty, one of the metals may be employed as cathode in the molten condition, and during the electrolysis, it should be stirred continuously; the same result may be obtained by allowing it to flow through the electro- lyte in a thin stream.Secondly, the one metal may be suspended in the bath from which the other is deposited on it, the temperature of the bath being regulated so that the alloy formed fuses and falls t o the bottom, exposing a fresh surface. A third process consists in periodi- cally adding small quantities of the more readily separable metal to the bath ; the two metals are then deposited in thin, alternate layers. A number of possible modifications of each of these methods is men- tioned, and as examples of their application, the following processes are described. An alloy of sodium ccnd lend may be obtained by using lead as the cathode in a bath of fnsed sodium chloride.The molten lead is stirred by passing through it some indifferent gas such as nitrogen, or, better, some gas which combines with the chloriiie evolved a t the anode, for example, hydrogen, carbonic oxide, or methane, An cclurninium-tin ulloy can be prepared by allowing tin to flow over a channelled carbon cathode forming the bottom of a bath of fused aluminium sodium chloride. Silicon bronze is obtained by the electrolysis of fused sodium silicate, the melting point of which is preferably depressed by admixture of other suitable salts, a suspended rod of copper being employed as the cathode. The sodium formed reduces the silicate to silicon, which alloys with the copper; this alloy then fuses and drops down, exposing a fresh surface of copper.INORGANIC CHEMISTRY.27 Alunziniurn bronze can be prepared by adding cryolite and copper oxide or chloride alternately t o a bath consisting of a mixture of sodium and potassium chlorides in fusion. The anode is of carbon, and the cathode of carbon or copper. Sodium amalgam is obtained by the electrolysis of a solution of caustic soda, the mercury which is employed as cathode being allowed t o flow slowly over a series of open gutters arranged in a terrace. I f acetone, for example, be added t o the liquid in the cathode compart- ment, it is reduced by the sodium amalgam to propylic alcohol. T. E. Compounds of Thallium. By LOUIS M. DENNIS and MARTHA DOAN (J. Amer. Chem. Soc., 1896, 18, 970--977~.-~hal~0zcs trinitride, TlN,, is obtained when a concentrated solution of potassium diazoate, containing a small quantity of the free acid, is added to a solution of thallous sulphate. It crystallises in fine needles, is strongly doubly refractive, and, in an atmosphere of carbonic anhydride, melts a t 334' ; i t is readily soluble in boiling water.When reduced in a current of dry hydrogen, i t yields about 30 per cent. of its nitrogen in the form of ammonia. Thallous thullic tyinitride, TlN,,TlN,, is most conveniently obtained by dissolving thallic hydroxide in diazoic acid (1 *6 per cent. solution), and allowing the solution t o stand at 0' in an exhausted Hempel desiccator ; it crystallises in yellow o r brownish crystals, which are highly explosive. Hot water decomposes the compound, part of the thallium being precipitated as thallic hydroxide, whilst part remains in solution and may be precipitated as thallous iodide.Z'l~allous tellumte, Tl,TeO,, is obtained as a white, flocculent precipi- tate, sparingly soluble in water, on adding a solution of pure telluric acid to a solution of thallous hydroxide. Thallous platinocyanide, Tl,Pt(CN), (compare Carstanjen, J. Chem., 1867, 102, 144), crystallises in nearly colourlees plates which are strongly doubly refractive. Nom.-It has long ago been shown that the salt described by Carstanjen was a double carbonate and platinocyanide of thallium, and that thallous platinocyanide, when pure, is perfectly colonrless. (Friswell, this Journal, 1871, 461 ; Friswell and Greenaway, ibid., 1877, ii, 251).J. J. S. Decomposition of Mercuric Sulphate by Water: Law of Thermochemical Moduli. By JOSEPH GUINCHANT (Bull. Xoc. Chim., 1896, [iii], 15, 1185-1 191).--Varet's conclusion that mercuric sulphate dissolved in dilute sulphuric acid exists in the solution in the form of a hydrogen sulphate strictly comparable with the alkali hydrogen sul- phates (Abstr., 1896, ii, 648), is not only contrary t o the results of Ditte, Cameron, Le Chatelier, and other observers, but is contrary t o the known chemical properties of the mercuric salts. The author points out that Favre and Silbermann's laws of thermochemical moduli are only approximate, and are not trustworthy criteria when dealing with questions of equilibrium, Moreover, Berthelot has shown t h a t the laws only hold good in the case of strong bases, and are not28 ABSTRACTS OF CHEMICAL PAPERS.applicable at all t o mercuric salts. All attempts to isolate a mercuric hydrogen sulphate have failed, and the fact that, when strong sulphuric acid is added to a solution of mercuric sulphate in dilute sulphuric acid, normal mercuric sulphate is precipitated, seems to show clearly that no hydrogen sulphate exists in the solution. Varet’s observation, that the heat of dissolution of the sulphate in sulphuric acid is inde- pendent of the relative proportions of acid and salt and of the con- centration of the acid, is analogous to Berthelot’s in the case of potassium chloride and hydrochloric acid. Further, the author’s own cryoscopic observations indicate that no mercuric hydrogen sulphate is formed in the solutions. Action of Fused Sodium Hydroxide under Pressure on Wrought Iron and Cast Iron.By AUGUSTE SCHEURER-KESTNER (Bull. Xoc. Chim., 1896, [ iii], 15, 1250--1252).-Under pressure, the action of fused sodium hydroxide on iron is distinctly greater than under ordinary conditions. Whether under atmospheric or higher pressure, the amount of corrosion increases with the temperature, and wrought iron is attacked much more readily than cast iron. These observations are of considerable practical importance now that fusions with sodium hydroxide under pressure have frequently to be carried out, and it was, in fact, a serious accident clue to rapid corrosion of an iron tube that directed the author’s attention to the question. C.H. B. C. H. B. Chromium Tetroxide, and Salts of Perchromic Acid. By 0. FRITZ WIEDE (Bey., 1897, 30, 2178-2189).-1f the blue solution obtained by extracting an aqueous solution of chromic anhydride and hydrogen peroxide with ether is cautiously treated with aqueous am- monia, the colour gradually disappears, and if the solution has been kept sufficiently cool, the under aqueous layer will have acquired a deep brown colour, and mill deposit a greenish-brown precipitate on standing. This dissolves in warm 10 per cent. ammonia solution, and, on cooling, separates in the form of pale brown needles of a composi- tion corresponding with the formula Cr04,3NH, This compound dissolves in water, undergoing partial decomposition, explodes when heated, and evolves oxygen on treatment with strong acids: When treated with alkalis, or if the ethereal solution itself is treated with a concentrated solution of a fixed alkali instead of ammonia, chro- mates of the alkali metals alone are formed. Other salts of perchromic acid can be prepared by employing substituted ammonias in place of ammonia.Thus, if the ethereal sollition is treated with pyridine, and the ether evaporated in a brisk current of air, blue scales are left, and if one of these be introduced into the ethereal solution after the addition of the pyridine, long, dark blue, glistening prisms gradually separat e having a composi tion corresponding with the formula CrO,H,C,NH,. They are extremely unstable, exploding violently even a t the temperature of a hot summer’s day ; when dry, however, they can be kept i n the cold for weeks, but i n the presence of moisture rapidly decompose.The salt is soluble in almost all the neutral organic solvents, and is only gradually acted on by potassium per- manganate in acid solution.MINERALOGICAL CHEMISTRY. 29 On adding aniline to the ethereal solution of perchromic acid, and diluting with an equal volume of light petroleum, dark red crystals, CrO5H,NK2Ph, separate, resembling potassium permanganate in ap- pearance ; this is even more explosive than the pyridine compound. J. F. T. Parastannyl Chloride. By RODOLPHE C. ENGEL (Conzpt. rend., 1897, 125, 464-466).-When a solution of metastannic chloride (compare Abstr., 1897, ii, 376) in dilute hydrochloric acid is heated at about 100' for some time, i t rapidly acquires the property of giving a precipitate with dilute sulphuric acid which is characteristic of the p-stannic chloride of Berzelius. No compound of definite com- position could be isolated from the solution, but the product approxi- mated in composition somewhat closely to metnstannic chloride. When metastannic acid is boiled with water, it is converted into the compound Sn50,,H2+ 7H,O, and when this is dried in a vacuum it yields the hydrate Sn,O,,H, + 2H20. Neither of these hydrates dissolves in hydrochloric acid, and therefore they differ from meta- stannic acid ; they combine with it, however, and when the product has been dried on porcelain, it dissolves in water, and the opalescent solu- tion gives a precipitate with sulphuric acid, but is very slowly pre- cipitated by excess of hydrogen sulphide. The dried chloride, which has the composition Sn,09CI, + 2H20, is decomposed by excess of water, and the product, when dried, has the composition Sn,O,,H, + 2H,O, its potassium salt crystallises with 3H20, and from it the correspond- ing chloride can be prepared by the action of hydrochloric acid. The author distinguishes these compounds as parastannic com- pounds, and their relation to the rnetastannic compounds is shown in the following table. Dried in Air. Dried at 100". Chloride. Potassium Salt. Metastannic . . . Sn50,,H,,9H,0 Sn5011H,,4H,0 Sn50gClp, 4H20 SnjOl,K2,4H20 Parastannic ... Sn5011H2, 7H,O Sn,01,H2,211,0 Sn50,C1,,2€I,0 8n,0,11<,,3H,0 The contradictory statements of earlier investigators are probably attributable to the fact that they were dealing with mixtures of the different modifications of the stannic compounds. C. H. B.