IXORGANIC CHEJIISTRT. In o r g a n i c C h e m i s t r y. 207 Oxidation of Hydrogen Iodide by Oxy-aeids. By 0. BURCHARD (Zeit. physikab. Clzenz., 2, 796-8;39).-This paper contains the details of st large number of experiments on the action of chloric, bromic, and iodic acids on hydrogen iodide, with varying concentrations of the solutions, and also in the presence of other acids. Of the above three208 ARSTHXCTS OF CHEMICAL PAPERS. acids, iodic acid is the one which acts most readily on hydrogen iodide and chloric acid the least. With dilute solutions of iodic acid and hydrogen iodide, a certain period elapses before t,he reaction sets in, this period increasing, of course, with the dilution and also with the temperature. It was necessary in studying the reaction, as a time reaction, t o neutralise the mixture o€ both acids at a given moment, without affecting in any way the liberated iodine.This it was found could be done by the addition of acid sodium crtrbonate to the solution. The carbonate, which is thus introduced in excess, interferes, however, with the estimation of the iodine by means of thiosulphate, but for thc latter, sodium hydrogen sulphite may be substituted, the action of which on iodine in solution is represented by NaHS03 + H20 + I, = NaHS04 + 2H1, and is not affected by the preseiice of the carhonate. The action of iodic acid on hydrogen iodide is representcd by 5HI + HIO, = 3H20 + 31,. For the action of bromic acid, solutions of at least dDC normal concentration are required, and the action takes place in three stages : (1) HBrO, + 5HI = 3H,O + Br + 51; (2) Br + HI = HBr + I; ( 3 ) HBrO, + 5HBr = 3H20 + 3Br2.Hence, if there are less than six equivalents of hydrogen iodide to each equivalent of bromic acid present, bromine will be found in the solution a t the end of the reaction. For the action of chloric acid, the solutions miist be so concentrated, and the time required is so long, that a decomposition of the hydrogen iodide itself occurs, and thus the exact nature of the reactlion carinot be ascertained. None of the ordinary equations satisfy the conditions of the above changes studied as time reactions. The author finds, however, that the time required for the oxidation of a given quantity of hydrogen iodide depends in a similar manner on the concentration, as the action of sulphurous on iodic acid studied by Landolt (Abstr., 1886, 658).An excess of either acid above the equivalent quantities produces an acceleration of the reaction. The presence of other acids, both those which take part in the reaction and tliose which do not, also causes an acceleration, which in the case of the latter is in proportion to their avidities. H. c'. Action of Incandescent Platinum on Gases and Vapours. By W. R. HODGKINSON and F. K. S. LOWNDES (Chew,. News, 58, 223- 2'24 ; compare Abstr., 1889, 20).--It is now shown t,hat iodine mono- chloride or trichloride, or chlorine in the presence of iodine or iodine bromides, when vaporised and submitted to the action of incandescent platinum wire, in the manner already described, give rise to flames, to mixtures of platinous chloride or bromide and iodide, and in the presence of chlorine to the deposition of crystals of platinum on the hottest part of the wire. I n similar experiments with carbon tetra- chloride, there is no flame, chlorine is liberated and carbon and carbon sesquichloride are deposited ; with phosphorus pentachloride there is flame, but the wire soon becomes alloyed with liberat,ed phosphorus, and, consequently, melts ; with hydrochloric acid, platinons chloride is formed ; hydrogen fluoride yields a soluble platinum salt ; mercurous chloride gives platinons chloride and mercury ; phosphorus or arsenicINORGANIC CHEMISTRY.209 vapour destroys the wire at once, whilst there is no apparent action with mercury, sulphur, nilrogen oxides, or sulphurous anhydride. D.A. L. Analysis of Atmospheric Air. By UFFELMANN (Chen?. Cmt~., 1888, 1324-1325, from Archiv. f. Hygiene, 8, 262--350).-The carbonic anhydride was deterniined in a flask of about 24 to 4 litres capacity. After filling with water, the flask was completely filled with the air under investigation, by running out the water again and allowing it to drain for 10 minutes. 50 c c. of baryta-water (7 : 1000) was next added, the flask closed with a tight stopper with india- rubber cap, and after shaking for one minute, allowed to remain by itself for 24 hours. The stopper was now replaced by a double-bored one, 60 C.C. of freshly boiled water added to wash down the sides of the flask, and the excess of baryta titrated with oxalic acid from a burette with a very long nozzle, reaching through the stopper down to the liquid at the bottom of the flask. (The oxalic acid = 2.8636 grams per litre.) The determination of the organic matters of the atmosphere was made by passing a definite quaiitity of the air through a solution of potassium permanganate, of which 1 C.C.= 0.395 gram KMn04 (= 0.1 mgrm. = 0.07 C.C. 0 = 0 7875 mgrm. oxalic acid) and the excess of permanganate determined by oxalic acid. The dust was collected on an asbestos filter and titrated with permanganate accord- ing to the method above mentioned. The micro-organisms were determined, after collecting in sterilised water, according to Esniarch's method. The ammonia was determined by projecting a spray of water against a slanting glrtps plate at the distance of 1 metre, and titrating with Nessler's reagent.The principal results of a long series of examinations of the atmo- sphere in the neighbourhood of Rostock are as follows :-(1,) The carbonic anhydride amounted to 3.18 in 10,000 in the open field ; the amount increased with land-winds or fog. (2.) The organic matters were equivalent to 2.71 C.C. oxygen per 1,000,000 vols. of air in the open field ; this amount varied very greatly, i t being decidedly less after continued rain. (3.) The amount of organic matter in the air at the sea-coast was found to be but one-third of that found 12 kilo- metres inland. (4.) The air of the Rostock University yard contained one-tenth more carbonic anhydride and one-third more organic matter than the air of the open fields.(5.) The air in the open fields near Rostock contained, on the average, 250 micro-organisms per cubic metre, that of the university yard 450, whilst the air on the sea-coast contained but 100 per cubic metre ; these quantities being less after continued rain and greater during fog. (6.) The amount of carbonic anhydride in cellars depends greatly on the barometric pressure, and varies inversely as the height of the barometer. (7.) The air of cellars contained spores of fungi. (8.) The air of house sewers was found to be but little richer in organic matter than that of well ventilated rooms, and coutained but few germs. (9.) Atmospheric air may be considered impure when it contains so much210 ARSTRACTS OF CHEMICAL PAPERS. oxidisable organic matters that in 1,000,000 vols.12 or more vols. of oxygen are required in the permanganate test. J. W. L. Hydrogen Telluride. By BERTHELOT and FABRE (Ann. Chim. Pkys. [GI, 14, 103--106).-Pure hydrogen telluride (compare Bineau, Aim. Chin?. Phys. [2], 47, 232) can be prepared by treating maqne- fiium telluride with very dilute hydrochloric acid in an atmosphere of nitrogen. The gas thus obtained is completely and rapidly absorbed b r a1 kalis, yielding white or colourless, crystalline tellurides which dissolve in pure water, forming colourless solutions ; if, however, a trace of oxygen is present, violet or purple solutions are produced, and, with a large excess of oxygen, metallic tellurium is immediately precipitated. It is very unstablc ; when kept over dry mercury, it de- composes in a few hours, even in the dark, but in presence of moist air tlecomposition is instantaneous.The smell of hydrogen telluride differs considerably from that of hydrogen sulphide or selenide, and when the gas is inhaled the effects produced are far less disagreeable than i n the case of the latter. The magnesium telluride referred to aI)ove is prepared by the action of excess of tellurium vapour on heated magnesium in an atmosphere of pure, dry hydrogen. It is a white substance, which quickly darkens on exposure to the air, and dissolves in water, forming a blackish-purple solution, owing to the presence of oxygen ; it dissolves in water saturated with nitrogen, yielding an almost colourless solution. F. S. I(. Amides of Phosphorus and Sulphur.By A . MENTE (Annalen, 248, 232--869).-GGladstone (this Journal, 1864, 225 ; 1866, 1 and 290 ; 1868, 64 and 261, and 1869, 55) obtained a series of compounds by the action of gaseous ammonia on phosphorus oxychloride, which he regarded as amic acids of pyrophosphoric or tetraphosphoric acid. The author considers t h a t these compounds are imido-acids. Imidodi- ph~~phoric acid, NH<potoH)>O PO(0H) (Gladstone's ppophosphamic acid), \ I can be prepared by the action of ammonium cerbamate on phosphorus oxychloride. The product is dissolved in water containing hydro- chloric acid, and the acid is precipitated in the form of the barium or iron salt by the addition of barium or ferric chloride. The barium salt contains 1 mol. H20. Di-imidodbhosphoric acid, NH<pOIOH)>NH7 PO(0H) is best prepared by .I adding phosphorus oxychloride diluted with 10 times its volume of benzene to excess of ammonium carbarnate. The barium salt is sparingly soluble in water and is anhydrous. A basic sodium salt, NaN:P,O,(ONa)?:NH, is known. Di-in, idod ip ho'sphormonamk acid ( pyrophospho triamic acid), N"<pocoH) PO(NH2) >NH, is an insoluble, white powder, obtained by \ I saturating phosphorus oxychloride with ammonia a t loo", and was11- ing the product with water. The acid is decomposed by boiling with sodium hjdroxide, yielding the basic sodium salt of di-irnidodiphos-INdRGANIC CHEMISTRY. 21 1 phoric acid. Silver nitrate gives. with neutral solutions of the acid, the monobasic silver salt, P,O,(NH),(NH,)*OAg, with ammonincal solutions a salt of the formula NAg:P,O,(NAg),(NH,)-OAg.The salts of this acid have already been described by Gladstone (loc. cit.). ing imidodiphosphoric chloride, NH( POCl,),, at 290", yields an insoluble barium salt, NP307Ba + H20. The cnlcium and ferric salts are insoluble, even in strong acid. Most of the metallic salts are insoluble in water but dissolve in acids. Neither the acid itself nor the salts of the alkali metals have been obtained in a crystslline form. Am?nonium imidosulphonate, NH( S0,*ONH4),, is formed by +he action nf ammonium cai*bamate on sulphuryl chloride, pyrosulphuryl chloride, or sulphuric monochlorhydrin. Amnionium chloride is removed by digesting with alcohol, and tlie residue is recrystallised from water. The compound crystallises in the monoclinic system, :ind its sp.gr. is 1.965. Imi~oszcZphuryZnmide, NH(SO,NH,),, is tlie first product of the action of ammonium carbamate on pyrosulphuryl chloride. I t is decomposed by water, yielding ammonium imidosul- phonate. The amide forms beautiful crystals, and is freely soluble i n water. IV. c. w. Allotropic Arsenic. Reply to Geuther. By ENGEL (BUZZ. XOC. Chinz., 50, 194--1'37).-Several new determinations of the specific gravity of allotropic arsenic gave the mean 4.6, which is the same number as that previously obtained (compare Geuther, Abstr., 1887, 888). The ratio of the densities of crystalline and allotropic arsenic is 1.245, and is practically the same as the ratio of white and red phosphorus = 1.244. The molecular constitution of the two modi- fications of arsenic seems to be the same as that of the corresponding modifications of phosphorus.Action of Ammonia and Amines on Arsenious Bromide. By W. LANDAU (Cheltz. Centr., 1888, 135&1355).--By passing dry ammonia into a solution of arsenious bromide and benzene, the compound 2AsBr3,7NH3 was formed. Ethylamine forms the com- pound AsBrj,4NH,Et + H,O, melting at 152". DipvopyZavuine forms with it the componnd AsBr3,4NHPr2 + H20, melting a t 258". The trimethylamine-derivative, AsBr3,3NMe3, melts at 235". 'The com- pound from triethylamine, AsBr,,3NEt3, melts a t 242". The aniliiLe compound, AsBr,,SNH,Ph + H,O, becomes changed into AsBrJ,4NH,Ph + H,O, by treatment with absolute alcohol or glacid acetic acid. Diphenylnmine arsenious bromide, AsBr3,3NHPh, + H,O, melts at 140°, and becomes changed, like the aniline compound, by the actiou of alcohol or acetic acid into a compound melting a t 230".The quinoline corn pound, As Rr3, C9NH7, HBr, melts at 13 7". The tyiethy Zp hosp hine- derivative, AsBr,,PEt,,HBr, melts a t 6Fj". J. W. L. N. H. M. Preparation of Boron and Silicon. By S . G. RAWSON (Chem. News, 58, 283).-A mixture of 3.5 grams of boric anhydride and212 ABSTRACTS OF OHEMTCAL PAPERS. 11 grams of calcium fluoride, is gently heated with concentrated sul- phuric acid ; the boron fluoride evolved is passed over heated potis- sium contained in a series of bulbs. Potassium fluoride and boroii arc formed, and are easily separated by washing with water. Amorphous silicon may be prepared in a similar manner, D.A. L. Preparation of Silicon. By H. N. WARREN (Chem. News, 58, 215-216) .-The author prepares amorphous silicoii by passing silicon fluoride over metallic magnesium heated in a combustion tube. Among the other products of the reaction is a magnesium silicide, and by the action of concentrated acids on this, a gas is evolved, which takes fire spontaneously in the air with explosive violence ; with feeble acids the gas is not spontaneously inflammable. Owing to the difflculty of eliminating free hydrogen from this gas, its exact composition could not be ascertained : it, however, appears t o he a mixture of solid, liquid, and gaseous silicon hydride. D. A. L. Effect of High Temperature and Pressure on Carbon. By C. A. PARSONS (Proc. Roy. Soc., 44, 320--323).-Carbon rods were snrrounded by benzene, paraffin, treacle, chloride o r bisulphide of car.hon, and submitted to great pressure in a hydraulic press, the rods being meantime heated by passing an electric current through them.In some cases a considerable amount of gas was evolved, and a soft friable deposit of carbon produced. In no case was the density of the carbon increased. When the rod was surrounded with silica, the latter fused, and the rod was largely conrerted into graphite ; the Rame occurred with hydrated alumina in lime or magnesia, the rod being rapidly destroyed with evolution of gas. With layers of coke, lime, and silica, the rod was rapidly corroded, and was found after the experiment to be coated with a coke-like layer of great hardness, sufh’cient to scratch rock-crystal and ruby, and to wear down the cut facets of a diamond.It resists the action of a mixture of hpdrofluoric and nitric acids. The conditions of temperature and pressure with presence of moisture, lime, and silica, resemble those which appear to have existed in the craters of the Cape diamond mines. The part played by the lime and silica is not clear. Formation of Carbon Oxysulphide by the Action of Carbon Bisulphide on Clay. By A. GAUTTER (Comyt. rend., 107, 911- 913).-K;lolin previously heated to incipient redness is packed into a ia1-g.e porcelain tube, which is then heated to bright redness, whilst car+on bisulphide vapour is passed through it. The gas which issues from the tube contains 60-64 per cent. of carbon oxysulphide, 35-39 per cent.of carbonic oxide, about 1 per cent. of carbonic anhydride, find traces of hydrogen sulphide, mixed of course with excess of carbon bisulphide vapour. The proportion of carbonic oxide is lower, and the proportion of oxysulphide higher the higher the temperature. The products are passed into a flask half filled with ice-cold water, which condenses the greater part of the bisulphide ; then through potaBsium hydroxide, which absorbs hydrogen sulphide and carbonic anhjdride ; then though acidified cuprous chloride, which absorbs car- H. K. T.INORGANIC CHEMISTRY. 213 bonic oxide, and finally through a 12 per cent. solution of aniline in alcohol, and over pumice and sulphuric acid. Alcoholic aniline has no action on carbon oxysulphide, but readily absorbs carbon bisul- phide.The properties of pure carbon oxysulphide agree mainly with the ordinary description. It has a very faint, somewhat et'hereal alliaceous odour. Carbon oxysulphide, as Berthelot observed, combines slowly with ammonia, forming yellow crystals of ammonium oxythiocarba- mate, which is decomposed by water. Sodium hydroxide slowly absorbs the gas, and a 35 per cent. solution in contact with excess of the gas yields pale-yellow tabular and acicular crystals of a thiocar- bonate, which is decomposed by water, as indicated by the equation 2NaHCS0, + H,O = NaHC03 + NaRS + CO, + H2S. When carbon oxysulphide is prepared by the action of carbon bisulphide 011 an oxide, alumina gives the best results, but it becomes too finely divided, and is difficult to manage.It mnst be heated to a white heat, since even a t a cherry-red heat the yield is small. If, however, sulphur vapour is passed over a mixture of alumina and carbon heated to bright, redness, almost pure carbon oxysulphide is obtained. Fremy observed that when carbon bisulphide is passed over heated oxides of lead, zinc, iron, and copper, crystalline sulphides are formed. The author finds that only in the case of zinc oxide is any notable quantity of carbon oxysulphide obtained. After carbon bisulphide has been passed over kaolin, the tube contains brilliant, acicular crystals of silicon sulphide, and somewhat large, hard, lozenge-shaped crystals, which evolve hydrogen sulphide when moistened, and are slowly decomposed by water into alumina and gelatinous silica,.They consist of a kind of thiosilicate, some of the oxygen in the kaolin having been replaced by sulphur, whilst some of the silicon has been removed in the form of sulphide. Alkaline Aluminates. By K. J. BAYER (Chem. %it., 12, 1209- 1210).-When the product of the ignition of alumina and soda in the molecular proportions 1 : 1 is treated with water, pure hydrated alumina, A1,o3,3H,0, is spontaneously and continuously deposited until the solution contains alumina and sodium hydroxide in t'he proportions of 1 mol. of the former to 6 mols. of the latter; the decomposition t,hen proceeds no further, and the solution remains clear. If instead of water a solution of sodium hydroxide, containing as much sodium as is already present in the ignited mixture is employed, then the solution remains quite clear, whether warm or cold, or a t rest or in motion, provided it is protected from carbonic anhydride ; but in the presence of this gas, or of spontaneously depo- sited hydrated alumina, this solution behaves in the manner already described.Neither pulverised glass, nor sand, nor granite powder, nor even ordinary gelatinous a,lumina can initiate this decomposition. If the Eolntions contain sodium salts with alkaline reaction, the podium determined by titration is somewhat higher, but the preci- pitated alumina is always pure. Potassium aluminate behaves in a similar manner. In explanation of these results, the author suggests that the A1,0,,2Na20 and A120,,6Na20 are really chemical compounds. C. H. B. Y VOL. LVI.214 ABSTRACTS OF CHEMICAL PAPERS.and therefore with the compounds A1203,Na20 and A1,03,3Na20 would indicate the existence of four different alkaline aluminates. D. A. L. Porcelain Glazes. By C. LAUTH and G. DUTAILLY (Bull. Soc. Chiw., 50, 221--251).-The first part of the paper contains the re5ults of experiments made with colourless glazes, which are divided into three classes :-(1.) Silicates of a single base associated with silica and alumina. (2.) Silicates of two bases, associated with silica and alumina. ( 3 . ) Silicates of several bases together, associated with silica and alumina. In the second part of the paper, experiments wit’h various coloured glazes are described. N. H. M. Cause of Rails Rusting less Quickly when in Use than when not. By W. SPRING (BUZZ.SOC. Chirrh., 50, 215-218).--The preservation of rails, when in use, from rust is not the result of vibratory movement, or of an electric action due to the passage of trains, but is due t o the superficial fovmation of the magnetic oxide produced by the compressign of the rust on the metal. - N. H. M. A Crystalline Subsulphide of Iron and Nickel. By J. B. MACKINTOSH (Chem. Neu-s, 58, 200).-A compound approximating in composition to the formula Fe4Ni2S, has been found in fern-like aggregations of srnall cnbical crystals in the cavities in the concre- tions in the hearth of a shaft furllace used for smelting roasted nickeliferous pyrrhotite. The author’s analytical numbers are : Cu 2.20 ; Ni,Co 26.16 ; Fe 61.685 ; S 8.305 ; SiO, 0.56 ; total, 98.91 ; deficiency and silica being due to adherent slag.D. A. L. New Fluorine Compounds of Vanadium. By A. PIccIx-r and G. GIORGIS (Gazzetta, 18, 186--194).-On dissolving ammonium metavanadate in aqueous hydrofluoric acid and treating the hot liquid with sulphurous anhydride, a blue liquid is obtained, which on neutralising with ammonia and adding ammonium fluoride, yields a bluish, crystalline precipitate This is collected, washed with a little water, and recrystallised from water; a t first monoclinic prisms having the composition of Baker’s salt (Trans., 1879, 760) separate, and then srnall, blue, lustrous crystals; these are mono- metric octahedra, which dissolve easily in water, yielding a blue solu- tion, from which the salt separates again on the addition of ammonium fluoride.On analysis, it gave results correbponding with the formula VOF,,SNH,F. The author calls it octahedral ammonium hydrqjluorozy- varbadats. When dissolved in concentrated hy drofluoric acid and allowed to remain for some days, it deposits crystals of Baker’s salt, VOF,,2NH4F + H,O. The new salt is also formed a t the negative pole, when a solution of ammonium metavanadate, acidified with hydrofluoric acid and mixed with ammonium fluoride, is submitted to electrolysis ; or when the mixture is reduced by alcohol. I n the latter case, how- ever, the mother-liquors, if exposed to the air after the hydrofluoroxy- vanadate has been deposited, yield yellow, octahedral crystals of t h e composition V02F,YNH4F. These can be obtained more easily byMINERALOGCICAL CHEMISTRY.215 dissolring vanadic acid in excess of hydrofluoric acid, and carefully neutralising w i t h ammonia while the liquid is still warm. On cooling, an abundant deposit of the crystals is obtained ; care must be taken not to add any great excess of ammonia, otherwise colourless crystals of ammonium vanadate alone will separate. By adding potassium hydrogen fluoride to a solution of ammonium metavanndate reduced by sulphurous anhydride, and evaporating the blue solution, the author obtained sky-blue, crystalline crusts, of the composition VOF2,2KC1. The remainder of the paper is devoted to theoretical consideration of the crystallographic connection of various isomorphous fluorides, the author drawing especial attention to the fact that the four com- pounds V0,F,3NH4F’, VOF,,SNH,F, FeF,,3NH4F, and CrF3,3NH4F, all crpstallise in the monometric system in cubes or octahedra, and all contain the same number of atoms.C. E. G.IXORGANIC CHEJIISTRT.In o r g a n i c C h e m i s t r y.207Oxidation of Hydrogen Iodide by Oxy-aeids. By 0. BURCHARD(Zeit. physikab. Clzenz., 2, 796-8;39).-This paper contains the detailsof st large number of experiments on the action of chloric, bromic, andiodic acids on hydrogen iodide, with varying concentrations of thesolutions, and also in the presence of other acids. Of the above thre208 ARSTHXCTS OF CHEMICAL PAPERS.acids, iodic acid is the one which acts most readily on hydrogen iodideand chloric acid the least. With dilute solutions of iodic acid andhydrogen iodide, a certain period elapses before t,he reaction sets in,this period increasing, of course, with the dilution and also with thetemperature.It was necessary in studying the reaction, as a time reaction, t oneutralise the mixture o€ both acids at a given moment, withoutaffecting in any way the liberated iodine.This it was found could bedone by the addition of acid sodium crtrbonate to the solution. Thecarbonate, which is thus introduced in excess, interferes, however,with the estimation of the iodine by means of thiosulphate, but forthc latter, sodium hydrogen sulphite may be substituted, the action ofwhich on iodine in solution is represented by NaHS03 + H20 + I, =NaHS04 + 2H1, and is not affected by the preseiice of the carhonate.The action of iodic acid on hydrogen iodide is representcd by5HI + HIO, = 3H20 + 31,.For the action of bromic acid, solutionsof at least dDC normal concentration are required, and the actiontakes place in three stages : (1) HBrO, + 5HI = 3H,O + Br + 51;(2) Br + HI = HBr + I; ( 3 ) HBrO, + 5HBr = 3H20 + 3Br2.Hence, if there are less than six equivalents of hydrogen iodide toeach equivalent of bromic acid present, bromine will be found in thesolution a t the end of the reaction. For the action of chloric acid, thesolutions miist be so concentrated, and the time required is so long,that a decomposition of the hydrogen iodide itself occurs, and thus theexact nature of the reactlion carinot be ascertained.None of the ordinary equations satisfy the conditions of the abovechanges studied as time reactions. The author finds, however, thatthe time required for the oxidation of a given quantity of hydrogeniodide depends in a similar manner on the concentration, as the actionof sulphurous on iodic acid studied by Landolt (Abstr., 1886, 658).An excess of either acid above the equivalent quantities produces anacceleration of the reaction.The presence of other acids, both thosewhich take part in the reaction and tliose which do not, also causes anacceleration, which in the case of the latter is in proportion to theiravidities. H. c'.Action of Incandescent Platinum on Gases and Vapours.By W. R. HODGKINSON and F. K. S. LOWNDES (Chew,. News, 58, 223-2'24 ; compare Abstr., 1889, 20).--It is now shown t,hat iodine mono-chloride or trichloride, or chlorine in the presence of iodine or iodinebromides, when vaporised and submitted to the action of incandescentplatinum wire, in the manner already described, give rise to flames,to mixtures of platinous chloride or bromide and iodide, and in thepresence of chlorine to the deposition of crystals of platinum on thehottest part of the wire.I n similar experiments with carbon tetra-chloride, there is no flame, chlorine is liberated and carbon and carbonsesquichloride are deposited ; with phosphorus pentachloride there isflame, but the wire soon becomes alloyed with liberat,ed phosphorus,and, consequently, melts ; with hydrochloric acid, platinons chlorideis formed ; hydrogen fluoride yields a soluble platinum salt ; mercurouschloride gives platinons chloride and mercury ; phosphorus or arseniINORGANIC CHEMISTRY.209vapour destroys the wire at once, whilst there is no apparent actionwith mercury, sulphur, nilrogen oxides, or sulphurous anhydride.D. A. L.Analysis of Atmospheric Air. By UFFELMANN (Chen?. Cmt~.,1888, 1324-1325, from Archiv. f. Hygiene, 8, 262--350).-Thecarbonic anhydride was deterniined in a flask of about 24 to 4 litrescapacity. After filling with water, the flask was completely filled withthe air under investigation, by running out the water again andallowing it to drain for 10 minutes. 50 c c. of baryta-water (7 : 1000)was next added, the flask closed with a tight stopper with india-rubber cap, and after shaking for one minute, allowed to remain byitself for 24 hours.The stopper was now replaced by a double-boredone, 60 C.C. of freshly boiled water added to wash down the sides ofthe flask, and the excess of baryta titrated with oxalic acid from aburette with a very long nozzle, reaching through the stopper downto the liquid at the bottom of the flask. (The oxalic acid =2.8636 grams per litre.)The determination of the organic matters of the atmosphere wasmade by passing a definite quaiitity of the air through a solution ofpotassium permanganate, of which 1 C.C. = 0.395 gram KMn04(= 0.1 mgrm. = 0.07 C.C. 0 = 0 7875 mgrm. oxalic acid) and theexcess of permanganate determined by oxalic acid. The dust wascollected on an asbestos filter and titrated with permanganate accord-ing to the method above mentioned.The micro-organisms weredetermined, after collecting in sterilised water, according to Esniarch'smethod.The ammonia was determined by projecting a spray of water againsta slanting glrtps plate at the distance of 1 metre, and titrating withNessler's reagent.The principal results of a long series of examinations of the atmo-sphere in the neighbourhood of Rostock are as follows :-(1,) Thecarbonic anhydride amounted to 3.18 in 10,000 in the open field ; theamount increased with land-winds or fog. (2.) The organic matterswere equivalent to 2.71 C.C. oxygen per 1,000,000 vols. of air in theopen field ; this amount varied very greatly, i t being decidedly less aftercontinued rain.(3.) The amount of organic matter in the air atthe sea-coast was found to be but one-third of that found 12 kilo-metres inland. (4.) The air of the Rostock University yard containedone-tenth more carbonic anhydride and one-third more organic matterthan the air of the open fields. (5.) The air in the open fields nearRostock contained, on the average, 250 micro-organisms per cubicmetre, that of the university yard 450, whilst the air on the sea-coastcontained but 100 per cubic metre ; these quantities being less aftercontinued rain and greater during fog. (6.) The amount ofcarbonic anhydride in cellars depends greatly on the barometricpressure, and varies inversely as the height of the barometer.(7.) The air of cellars contained spores of fungi.(8.) The air ofhouse sewers was found to be but little richer in organic matter thanthat of well ventilated rooms, and coutained but few germs. (9.)Atmospheric air may be considered impure when it contains so muc210 ARSTRACTS OF CHEMICAL PAPERS.oxidisable organic matters that in 1,000,000 vols. 12 or more vols. ofoxygen are required in the permanganate test. J. W. L.Hydrogen Telluride. By BERTHELOT and FABRE (Ann. Chim.Pkys. [GI, 14, 103--106).-Pure hydrogen telluride (compare Bineau,Aim. Chin?. Phys. [2], 47, 232) can be prepared by treating maqne-fiium telluride with very dilute hydrochloric acid in an atmosphere ofnitrogen. The gas thus obtained is completely and rapidly absorbedb r a1 kalis, yielding white or colourless, crystalline tellurides whichdissolve in pure water, forming colourless solutions ; if, however, atrace of oxygen is present, violet or purple solutions are produced,and, with a large excess of oxygen, metallic tellurium is immediatelyprecipitated.It is very unstablc ; when kept over dry mercury, it de-composes in a few hours, even in the dark, but in presence of moist airtlecomposition is instantaneous. The smell of hydrogen telluridediffers considerably from that of hydrogen sulphide or selenide, andwhen the gas is inhaled the effects produced are far less disagreeablethan i n the case of the latter. The magnesium telluride referred toaI)ove is prepared by the action of excess of tellurium vapour onheated magnesium in an atmosphere of pure, dry hydrogen.It is awhite substance, which quickly darkens on exposure to the air, anddissolves in water, forming a blackish-purple solution, owing to thepresence of oxygen ; it dissolves in water saturated with nitrogen,yielding an almost colourless solution. F. S. I(.Amides of Phosphorus and Sulphur. By A . MENTE (Annalen,248, 232--869).-GGladstone (this Journal, 1864, 225 ; 1866, 1 and290 ; 1868, 64 and 261, and 1869, 55) obtained a series of compoundsby the action of gaseous ammonia on phosphorus oxychloride, whichhe regarded as amic acids of pyrophosphoric or tetraphosphoric acid.The author considers t h a t these compounds are imido-acids. Imidodi-ph~~phoric acid, NH<potoH)>O PO(0H) (Gladstone's ppophosphamic acid),\ I can be prepared by the action of ammonium cerbamate on phosphorusoxychloride. The product is dissolved in water containing hydro-chloric acid, and the acid is precipitated in the form of the barium oriron salt by the addition of barium or ferric chloride.The bariumsalt contains 1 mol. H20.Di-imidodbhosphoric acid, NH<pOIOH)>NH7 PO(0H) is best prepared by. I adding phosphorus oxychloride diluted with 10 times its volume ofbenzene to excess of ammonium carbarnate. The barium salt issparingly soluble in water and is anhydrous. A basic sodium salt,NaN:P,O,(ONa)?:NH, is known.Di-in, idod ip ho'sphormonamk acid ( pyrophospho triamic acid),N"<pocoH) PO(NH2) >NH, is an insoluble, white powder, obtained by\ I saturating phosphorus oxychloride with ammonia a t loo", and was11-ing the product with water. The acid is decomposed by boiling withsodium hjdroxide, yielding the basic sodium salt of di-irnidodiphosINdRGANIC CHEMISTRY.21 1phoric acid. Silver nitrate gives. with neutral solutions of the acid,the monobasic silver salt, P,O,(NH),(NH,)*OAg, with ammonincalsolutions a salt of the formula NAg:P,O,(NAg),(NH,)-OAg. Thesalts of this acid have already been described by Gladstone (loc. cit.).ing imidodiphosphoric chloride, NH( POCl,),, at 290", yields aninsoluble barium salt, NP307Ba + H20. The cnlcium and ferric saltsare insoluble, even in strong acid. Most of the metallic salts areinsoluble in water but dissolve in acids. Neither the acid itself northe salts of the alkali metals have been obtained in a crystslline form.Am?nonium imidosulphonate, NH( S0,*ONH4),, is formed by +he actionnf ammonium cai*bamate on sulphuryl chloride, pyrosulphurylchloride, or sulphuric monochlorhydrin.Amnionium chloride isremoved by digesting with alcohol, and tlie residue is recrystallisedfrom water. The compound crystallises in the monoclinic system,:ind its sp. gr. is 1.965. Imi~oszcZphuryZnmide, NH(SO,NH,),, is tliefirst product of the action of ammonium carbamate on pyrosulphurylchloride. I t is decomposed by water, yielding ammonium imidosul-phonate. The amide forms beautiful crystals, and is freely soluble i nwater. IV. c. w.Allotropic Arsenic. Reply to Geuther. By ENGEL (BUZZ. XOC.Chinz., 50, 194--1'37).-Several new determinations of the specificgravity of allotropic arsenic gave the mean 4.6, which is the samenumber as that previously obtained (compare Geuther, Abstr., 1887,888).The ratio of the densities of crystalline and allotropic arsenicis 1.245, and is practically the same as the ratio of white and redphosphorus = 1.244. The molecular constitution of the two modi-fications of arsenic seems to be the same as that of the correspondingmodifications of phosphorus.Action of Ammonia and Amines on Arsenious Bromide.By W. LANDAU (Cheltz. Centr., 1888, 135&1355).--By passing dryammonia into a solution of arsenious bromide and benzene, thecompound 2AsBr3,7NH3 was formed. Ethylamine forms the com-pound AsBrj,4NH,Et + H,O, melting at 152". DipvopyZavuine formswith it the componnd AsBr3,4NHPr2 + H20, melting a t 258".Thetrimethylamine-derivative, AsBr3,3NMe3, melts at 235". 'The com-pound from triethylamine, AsBr,,3NEt3, melts a t 242". The aniliiLecompound, AsBr,,SNH,Ph + H,O, becomes changed into AsBrJ,4NH,Ph + H,O, by treatment with absolute alcohol or glacid acetic acid.Diphenylnmine arsenious bromide, AsBr3,3NHPh, + H,O, melts at 140°,and becomes changed, like the aniline compound, by the actiou ofalcohol or acetic acid into a compound melting a t 230". The quinolinecorn pound, As Rr3, C9NH7, HBr, melts at 13 7". The tyiethy Zp hosp hine-derivative, AsBr,,PEt,,HBr, melts a t 6Fj". J. W. L.N. H. M.Preparation of Boron and Silicon. By S . G. RAWSON (Chem.News, 58, 283).-A mixture of 3.5 grams of boric anhydride an212 ABSTRACTS OF OHEMTCAL PAPERS.11 grams of calcium fluoride, is gently heated with concentrated sul-phuric acid ; the boron fluoride evolved is passed over heated potis-sium contained in a series of bulbs.Potassium fluoride and boroii arcformed, and are easily separated by washing with water. Amorphoussilicon may be prepared in a similar manner, D. A. L.Preparation of Silicon. By H. N. WARREN (Chem. News, 58,215-216) .-The author prepares amorphous silicoii by passing siliconfluoride over metallic magnesium heated in a combustion tube.Among the other products of the reaction is a magnesium silicide,and by the action of concentrated acids on this, a gas is evolved,which takes fire spontaneously in the air with explosive violence ;with feeble acids the gas is not spontaneously inflammable.Owing tothe difflculty of eliminating free hydrogen from this gas, its exactcomposition could not be ascertained : it, however, appears t o he amixture of solid, liquid, and gaseous silicon hydride. D. A. L.Effect of High Temperature and Pressure on Carbon. ByC. A. PARSONS (Proc. Roy. Soc., 44, 320--323).-Carbon rods weresnrrounded by benzene, paraffin, treacle, chloride o r bisulphide ofcar.hon, and submitted to great pressure in a hydraulic press, the rodsbeing meantime heated by passing an electric current through them.In some cases a considerable amount of gas was evolved, and a softfriable deposit of carbon produced. In no case was the density ofthe carbon increased.When the rod was surrounded with silica, thelatter fused, and the rod was largely conrerted into graphite ; theRame occurred with hydrated alumina in lime or magnesia, the rodbeing rapidly destroyed with evolution of gas. With layers of coke,lime, and silica, the rod was rapidly corroded, and was found afterthe experiment to be coated with a coke-like layer of great hardness,sufh’cient to scratch rock-crystal and ruby, and to wear down the cutfacets of a diamond. It resists the action of a mixture of hpdrofluoricand nitric acids. The conditions of temperature and pressure withpresence of moisture, lime, and silica, resemble those which appearto have existed in the craters of the Cape diamond mines. The partplayed by the lime and silica is not clear.Formation of Carbon Oxysulphide by the Action of CarbonBisulphide on Clay.By A. GAUTTER (Comyt. rend., 107, 911-913).-K;lolin previously heated to incipient redness is packed into aia1-g.e porcelain tube, which is then heated to bright redness, whilstcar+on bisulphide vapour is passed through it. The gas which issuesfrom the tube contains 60-64 per cent. of carbon oxysulphide, 35-39per cent. of carbonic oxide, about 1 per cent. of carbonic anhydride,find traces of hydrogen sulphide, mixed of course with excess of carbonbisulphide vapour. The proportion of carbonic oxide is lower, andthe proportion of oxysulphide higher the higher the temperature.The products are passed into a flask half filled with ice-cold water,which condenses the greater part of the bisulphide ; then throughpotaBsium hydroxide, which absorbs hydrogen sulphide and carbonicanhjdride ; then though acidified cuprous chloride, which absorbs car-H. K.TINORGANIC CHEMISTRY. 213bonic oxide, and finally through a 12 per cent. solution of aniline inalcohol, and over pumice and sulphuric acid. Alcoholic aniline hasno action on carbon oxysulphide, but readily absorbs carbon bisul-phide.The properties of pure carbon oxysulphide agree mainly with theordinary description. It has a very faint, somewhat et'hereal alliaceousodour. Carbon oxysulphide, as Berthelot observed, combines slowlywith ammonia, forming yellow crystals of ammonium oxythiocarba-mate, which is decomposed by water. Sodium hydroxide slowlyabsorbs the gas, and a 35 per cent. solution in contact with excess ofthe gas yields pale-yellow tabular and acicular crystals of a thiocar-bonate, which is decomposed by water, as indicated by the equation2NaHCS0, + H,O = NaHC03 + NaRS + CO, + H2S.When carbon oxysulphide is prepared by the action of carbonbisulphide 011 an oxide, alumina gives the best results, but it becomestoo finely divided, and is difficult to manage.It mnst be heated to awhite heat, since even a t a cherry-red heat the yield is small. If,however, sulphur vapour is passed over a mixture of alumina and carbonheated to bright, redness, almost pure carbon oxysulphide is obtained.Fremy observed that when carbon bisulphide is passed over heatedoxides of lead, zinc, iron, and copper, crystalline sulphides are formed.The author finds that only in the case of zinc oxide is any notablequantity of carbon oxysulphide obtained.After carbon bisulphide has been passed over kaolin, the tubecontains brilliant, acicular crystals of silicon sulphide, and somewhatlarge, hard, lozenge-shaped crystals, which evolve hydrogen sulphidewhen moistened, and are slowly decomposed by water into aluminaand gelatinous silica,.They consist of a kind of thiosilicate, some ofthe oxygen in the kaolin having been replaced by sulphur, whilstsome of the silicon has been removed in the form of sulphide.Alkaline Aluminates. By K. J. BAYER (Chem. %it., 12, 1209-1210).-When the product of the ignition of alumina and sodain the molecular proportions 1 : 1 is treated with water, purehydrated alumina, A1,o3,3H,0, is spontaneously and continuouslydeposited until the solution contains alumina and sodium hydroxidein t'he proportions of 1 mol.of the former to 6 mols. of the latter;the decomposition t,hen proceeds no further, and the solution remainsclear. If instead of water a solution of sodium hydroxide, containingas much sodium as is already present in the ignited mixture isemployed, then the solution remains quite clear, whether warm orcold, or a t rest or in motion, provided it is protected from carbonicanhydride ; but in the presence of this gas, or of spontaneously depo-sited hydrated alumina, this solution behaves in the manner alreadydescribed. Neither pulverised glass, nor sand, nor granite powder,nor even ordinary gelatinous a,lumina can initiate this decomposition.If the Eolntions contain sodium salts with alkaline reaction, thepodium determined by titration is somewhat higher, but the preci-pitated alumina is always pure.Potassium aluminate behaves in asimilar manner. In explanation of these results, the author suggeststhat the A1,0,,2Na20 and A120,,6Na20 are really chemical compounds.C. H. B.Y VOL. LVI214 ABSTRACTS OF CHEMICAL PAPERS.and therefore with the compounds A1203,Na20 and A1,03,3Na20 wouldindicate the existence of four different alkaline aluminates.D. A. L.Porcelain Glazes. By C. LAUTH and G. DUTAILLY (Bull. Soc.Chiw., 50, 221--251).-The first part of the paper contains there5ults of experiments made with colourless glazes, which are dividedinto three classes :-(1.) Silicates of a single base associated withsilica and alumina. (2.) Silicates of two bases, associated with silicaand alumina.( 3 . ) Silicates of several bases together, associated withsilica and alumina. In the second part of the paper, experimentswit’h various coloured glazes are described. N. H. M.Cause of Rails Rusting less Quickly when in Use thanwhen not. By W. SPRING (BUZZ. SOC. Chirrh., 50, 215-218).--Thepreservation of rails, when in use, from rust is not the result ofvibratory movement, or of an electric action due to the passage oftrains, but is due t o the superficial fovmation of the magnetic oxideproduced by the compressign of the rust on the metal.-N. H. M.A Crystalline Subsulphide of Iron and Nickel. By J. B.MACKINTOSH (Chem. Neu-s, 58, 200).-A compound approximating incomposition to the formula Fe4Ni2S, has been found in fern-likeaggregations of srnall cnbical crystals in the cavities in the concre-tions in the hearth of a shaft furllace used for smelting roastednickeliferous pyrrhotite. The author’s analytical numbers are :Cu 2.20 ; Ni,Co 26.16 ; Fe 61.685 ; S 8.305 ; SiO, 0.56 ; total, 98.91 ;deficiency and silica being due to adherent slag. D. A. L.New Fluorine Compounds of Vanadium. By A. PIccIx-r andG. GIORGIS (Gazzetta, 18, 186--194).-On dissolving ammoniummetavanadate in aqueous hydrofluoric acid and treating the hotliquid with sulphurous anhydride, a blue liquid is obtained, whichon neutralising with ammonia and adding ammonium fluoride, yieldsa bluish, crystalline precipitate This is collected, washed witha little water, and recrystallised from water; a t first monoclinicprisms having the composition of Baker’s salt (Trans., 1879, 760)separate, and then srnall, blue, lustrous crystals; these are mono-metric octahedra, which dissolve easily in water, yielding a blue solu-tion, from which the salt separates again on the addition of ammoniumfluoride. On analysis, it gave results correbponding with the formulaVOF,,SNH,F. The author calls it octahedral ammonium hydrqjluorozy-varbadats. When dissolved in concentrated hy drofluoric acid and allowedto remain for some days, it deposits crystals of Baker’s salt, VOF,,2NH4F + H,O. The new salt is also formed a t the negative pole, whena solution of ammonium metavanadate, acidified with hydrofluoricacid and mixed with ammonium fluoride, is submitted to electrolysis ;or when the mixture is reduced by alcohol. I n the latter case, how-ever, the mother-liquors, if exposed to the air after the hydrofluoroxy-vanadate has been deposited, yield yellow, octahedral crystals of t h ecomposition V02F,YNH4F. These can be obtained more easily bMINERALOGCICAL CHEMISTRY. 215dissolring vanadic acid in excess of hydrofluoric acid, and carefullyneutralising w i t h ammonia while the liquid is still warm. On cooling,an abundant deposit of the crystals is obtained ; care must be takennot to add any great excess of ammonia, otherwise colourless crystalsof ammonium vanadate alone will separate.By adding potassium hydrogen fluoride to a solution of ammoniummetavanndate reduced by sulphurous anhydride, and evaporating theblue solution, the author obtained sky-blue, crystalline crusts, of thecomposition VOF2,2KC1.The remainder of the paper is devoted to theoretical considerationof the crystallographic connection of various isomorphous fluorides,the author drawing especial attention to the fact that the four com-pounds V0,F,3NH4F’, VOF,,SNH,F, FeF,,3NH4F, and CrF3,3NH4F,all crpstallise in the monometric system in cubes or octahedra, and allcontain the same number of atoms. C. E. G