GENERAL AND PHYSICAL CHEMISTRY. ii. 525 Inorganic Chemistry. Hydrogen from Formates and from Carbon Monoxide. M. G. LEVI and A. PIVA ( A n ? i . ChLirn. ,4ppJicata 1916 5 271-301).-Further experiments havel been made on the decom- position of the formates (compare A. 1914 i 480) and investiga- tions made on the conditions of formation of formates and inii. 526 ABSTRACTS OF CH:EMIC1A.L YBPERS. general on the conditions of reaction between carbon monoxide and alkali. The presence of calcium hydroxide lowers the temperature a t which calcium forma%e decomposes from 3 7 5 O to 2800 for an equi- rnoiecular mixture and to 260° for the mixture (HCO,)?Ca+ 2Ca(OH)z. As regards the gases generated the presence of lime causes (1) the proportion of carbon dioxide to diminish gradually to zero (2) that of carbon monoxide to decrease markedly aiid (3) that of hydrogen t o increase from a maximum o€ 45 C.C.for 1 gram of the pure formate to 220 C.C. for I gram of forinate in presence of the hydroxide the proportion of liydrogeln in the gas formed rising from 35% to 93%. Assuming that the reaction is expressible by the equation (HCO,),Ca + Ca(OH) = 2CaCO,+ 2H2 1 grain of the forniate should be capable of yielding 344 C.C. of hydrogen. The small proportion of methane formed disappears in presence of large amounts of lime. With referenoe t o the action of carbon monoxide on lime or sodium hydroxide or soda lime either dry o r in presence of water vapour the results of the author's experiments point to the follow- ing conclusions (1) With pure lime carbon monoxide gives formate ab 250-300° and carbonate and hydrogen above 300O; the forma- tion takes place moderately slowly and t o a much less extent than with sodium hydroxide since in the fundamental equilibrium between carbon monoxide and water CO+ H,O f H*CO,H the lime being a weak base exercises only a slight displacement.At 500-600° the formation of hydrogen is almost quantitative pro- vided always that the elements of water are prwent. (2) The addition of sodium hydroxide to the lime facilitates the formation of formate rendering it possible even a t 150-200° and lowers the temperature a t which hydrogen is formed the reaction proceled- ing in this direction t o the extent of 50% a t 300O; the presence of the two bases together increases the yield of hydrogen. T.H. P. Revision of the Atomic Weight of Bromine E. MOLES (Compt. rend. 1916 163 94-97. Compare this vol. ii 314).-The author has made further determinations of the density of hydrogen bromide gas prepared by the action of bromine on naphthalene in the cold and on paraffin heated a t ZOOo the resulting gas being carefully purified in each case. As a mean of thirty-three determinations he now gives the value 3*64442+0*00013 grams as the weight of a normal litre of hydrogen bromide. The weight of a litre of the gas under reduced pressures has been determined and these results give for the coefficient of compressibility of hydrogen bromide 1 + h = 1.00931 whence the molecular weight of the gas as compared with oxygen is 80.934; and the atomic weight of bromine taking hydro- gen = 1.008 is 79.926.E. WARBURG (Rer. Deut. physikal. Ges. 1915 17 194-197).-Small quantities of ozone are formed by the action of ultra-violet radiation on oxygen The Density of Hydrogen Bromide Gas W. G. Oeonisation of Liquid Oxygen by Radiation.INOROANI C CHEMISTRY . ii. 527 a t atmospheric pressure but since rays of wavelength exceeding 0 . 2 ~ are only slightly absorbed by oxygen under these conditions it was considered possible that larger quantities might be obtained by the use of liquid oxygen. Experiments made with liquid oxygen and air exposed t o the rays emitted by the spark discharge between zinc electrodm show that appreciable quantities of ozone are formed after a short exposure. Preparation of Free Hydroxylamine from Hydroxylamine Sulphate. OSKAR BAUDISCH and F.JENNER (Bcr. 1916 49 11 82-1 185) .-Finely powdered and carefully dried hydroxylamine sulphate is cautiously added to liquid ammonia in a quartz tube the ammonia is then removed by evacuation and the free hydroxyl- amine extracted by alcohol. The operation is attended by grave risks from explosions but details are given of the manipulation which the authors recommend. H. M. D. J. C. W. The Constitution of Coal. DAVID TREVOR JONES and RICHARD VERNON WHEELER (T. 1916 109 707-714).-Coal is considered t o have been formed from decayed vegetable matter by the action of pressure and heat the latter not having exceeded a temperature of 300O. The coal is a conglomerate which can be resolved by sol- vents into “ cellulosic ” and “resinic ” portions (T. 1913 103 1706) ; thus on extracting coal with pyridine an insoluble cellulosic residue is obtained the extract containing a mixture of cellulosic material with resinic compounds which can be separated by treat- ment with chloroform in which the former is insoluble.It is the resinic component of coal which exerts photochemical action on a photographic plate the effect being due to some slow oxidation; the cellulosic component does not exert this action. When the cellulosic compounds are submitted to destructive dis- tillation phenols are produced (compare Wichelhaus A. 1910 i 868) the format~ion of these being due to the presence of the furan grouping *?*CH>O; there are also present in the csllulosic con- stituents compounds of which the molecules structurally resemble the carbon molecule (compare Cross and Bevan Phil.Mag. 1882 [v] 13 325 ; Diinroth and Kerkovius A. 1913 ii 774 ; Pictet and Ramseyer A. 1911 i 850) but there is no likelihood of the presence of free carbon. The resinic compounds are probably of more diverse nature than the cellulosic ; they contain alkyl naphthene and unsaturated hydroaromatic radicles linked to larger and more complex group- ings. Under the influence of pressure the resinic derivatives have become highly polymerised. The oxygenated derivatives pressnt among the resinic compounds are probably cyclic oxides. Free hydrocarbons exist to some extent in the resinic portion of coal (compare Pictet and Ramseyer Zoc. c i t . ; Pictet and Bouvier A. 1915 i 512) but paraffin hydrocarbons are present only in small quantity.The outstanding difference between the (‘ petroleum ” distilled *C=Cii. 528 ABSTRACTS OF CHEMICAL PAPERS. from coal and the natural oils is the presence of phenols in tlle former and the diff erencel is explicable when the) conglomerate character of coal is borne in niind especially in view of the fact that phenols are distilled only from the cellulosic portions. The absence of phenols from petroleums suggests that the origin of these is of a non-vegetable character. The Influence of Iron Pyrites on the Oxidation of Coal. THOMAS JAMES DRAKELEY (T. 1916 109 723-733).-A detailed account is given of experiments made' t o determine the influence of iron pyrites on the oxidation of coal to carbon dioxide and on the absorption of oxygen by the coal.F o r this purpose a quanti- tative comparison was made of the oxidation products obtained from (1) pyrites (2) coal (3) an artificial mixtura of coal and pyrites (4) ferrous sulphate (5) a mixture of coal and ferrous sul- phate (6) a mixture of coal and sulphuric acid. The rates at which oxygen was absorbed by these systems were also compared. The results obtained seem to show that pyrites has a measur- able influence in the sense that it increases the rate of oxidation of coal and its presence may therefore not be entirely disregarded in connexion with the question of the spontaneous ignition of coal. D. F. T. H. M. D. The Compoaition and Use of Greek Fire. C. ZENGHELIS (Compt. rend. 1916 163,125-127).-A brief survey of the modern and the ancient literature dealing with the composition and method of using Greek fire.The author considers that Greek fire was as suggested by Berthelot based on the addition of nitre to inflam- mable mixtures and that the gases produced by the explosive portion propelled the iiicendiary portion against the enemy. W. G. The Action of Oxygen on Metallic Oxides at High Tem- peraturea and Pressures. JAROSLAV MILBAUER (Chem. Z e d 1916 40 587).-When heated f o r an hour in closed tubes a t 480° in oxygen a t 12 atm. pressure the majority of metallic oxides remained unchanged. The following exceptions were observed Potassium oxide was partly converted into peroxide barium oxide gave barium peroxide lead oxide was converted into red lead antimony oxide into the tetroxide and manganous oxide into the sesquioxide nickelous and cobaltous oxides contained traces of the nickelic and cobaltic compounds whilst silver became covered with black crusts probably of a peroxide since on treatment with hydro- chloric acid chlorine was generated.Intimate mixtures of oxides or carbonates with chromic oxide heated under similar conditions were found in all cases but that of cerium to have been more o r less completely converted into the corresponding chromate. With oxides (or carbonates) of the following metals f o r example the product contained the annexed percentage of chromate! Silver 100% ; magnesium 82.7% ; calcium 56.9% ; zinc 72% ; lead 100% ; bismuth 6.3%; barium 52.8%. The reaction is apparently in some cases suitable for the technical production of chromic acid salts.G. F. M.INORGANIC CHEMISTRY. ii. 529 Pure Sodium Chloride. CLIFFORD LOHMANN (C'hem. Nezus 1916 114 53).-The following method was employed for separating sodium and potassium chlorides About 0.5 gram of the salt was dissolved in a little water and treated with 20 drops of a 10% solu- tion of platinuni chloride. A few drops of water were added after which the precipitate was collected and washed five or six times with a mixture of water (2 vols.) and alcohol (1 vol.) and then about six times with a mixture of alcohol and ether. When dry the precipi- tate was washed into a weighed platinum crucible with boiling water. Three samples of sodium chloride supposed to be chemically pure were found t o contain from 0.45 t o 0.57% of pot'assium chloride.N. H. J. M. Ammonium Iodide its Solubilities and the Absence of a Transition Point. ALEXANDER SMITH and HERBERT E. EASTLACK (27. A mer. C h ~ m . Soc. 1916 38 1500-1502).-Amrnonium iodide crystallises in cubes whilst the bromide and chloride form penta- gonal icositetrahedra. Since i t is probable that both these are transformed into cubic forms above their respective transition tem- peratures there is some ground f o r the supposition that ammonium iodide will exist in two forms connected by a transition point. Measurements of the solubility of ammonium iodide in water between - 190 and 1 3 6 O afford no evidence of the existence of such a transition temperature. W. STAHL (MetalE uizd Erx. 1915 12 501-504).-Crude silver containing iron sulphate is liable t o ' spit' a t the moment of solidifying.This effect has been attributed to sulphur dioxide. This gas is however insoluble in fused silver (Sieverts and Bergner A. 1913 ii 321). Silver is known to react with sulphur dioxide according to the equation 4 A r ~ 2'W- Ag,SO,+Ag,S or in presence of oxygen 4Ag+2S02+202 ZAg,SO,. As the dissociation temperature of silver sulphate is 1030° i t can exist in contact with molten silver and forms a slag on the surface. If more strongly heated dissociation takes place Ag,SO = 2Aq + SO + 0,. The oxygen then dissolves in the silver and is liberated on solidification so causing the violent effect which has been observed. Action of Hydrogen Sulphide on Mixtures of the Alkaline Earths with the Alkalis and with Oxides of the Heavy Metals.11. ICILIO GUARESCHI (Atti R. Accad. Sci. Torino 1915-1916 51 951-962. Compare this vol. ii 324 325).-The effect of strontium and lithium oxides on the action of hydrogen sulphide on sodium hydroxide o r potassium hydroxide is consider- ably less energetic than that of calcium oxide or barium oxide no incandescence being observed under the conditions used. The incan- descence obtained with mixtures of sodium (or potassium) hydroxide and calcium (or barium) oxide cannot be attributed to the presence or formation of peroxides. The alkali peroxides (not however of recent preparation) act almost like the hydroxides whilst barium magnesium and lead peroxides have no action Mixtures of calcium H. M. D. The 'Spitting' of Silver.C. IR. D.i i . 530 ABSTRACTS OF CHEMICAL PAPERS. (or barium) oxide with mercuric or nickel oxide also react vigor- ously with hydrogen sulphide and when the constituent oxides are in definite proportions vivid incandescence or even explosion takes place. When incandescence occurs with mercuric oxide sulphur dioxide is formed and this reacts with the hydrogen sulphide giving colloidal sulphur which is precipitated by water in a special blue form. No incandescence is observed when hydrogen sulphide acts on pumice and mercuric oxide or on a mixture of calcium (or \ barium) oxide with ferric oxide cuprqus o r cupric oxide o r litharge. T. H. P. Composition and Solubility of Calcium Hydrogen Carbonate. ALFREDO CAVAZZI (Gazzetta 1916 46 ii 122-135).-The author demonstrates the accuracy of the common supposition that the salt formed by the action of carbon dioxide on lime water or by that of a water more or less rich in carbon dioxide on the normal carbonate is calcium hydrogen carbonate of the composition Ca(HCO&. The maximum quantity of calcium carbonate which after prolonged shaking (not less than ten hours) dissolves a t Oo in 1 litre of water saturated with carbon dioxide and maintained so in presence of the gas a t atmospheric pressure is 1.56 grams that of calcium hydrogen carbonate being theref ore 2.5272 grams. Under similar conditions and after shaking extending over some days 1 litre of water a t 1 5 O dissolves 1.1752 grams of calcium carbonate and consequently contains 1.9038 grams of calcium hydrogen carbonate.When a very violent current of carbon dioxide is passed into lime water saturated a t 15O the latter becomes at first very turbid but after about a minute almost clear and highly supersaturated with carbon dioxide ; this solution which only per- sists for a short time contains 2.29 grams of calcium carbonate or 3.71 grams of calcium hydrogen carbonate per litre. Action of Hydrogen Sulphide on Mercuric Iodide. GIAMBATTISTA FRANCESCHI (Boll. chim. farm. 1916 55 481-483). -Study of this reaction in alcoholic solution shows that i t takes place in three phases with formation of three different conipounds according t o the conditions. These phases may be represented 2(2HgS,Hg12) ; and (3) 2HgS,Hg12 + H2S = 3HgS + 2HI. Thus the action of hydrogen sulphide on mercuric and mercurous salts always yields mercuric sulphide the mercuric salts being previously reduced t o mercurous salts and an intermediate compound formed which may be regarded as a thiobasic mercury compound of the constitu- tion Hg(*S*HgR),.The compound Hg(*S*HgI) forms a tobacco-coloured precipi- tate. T. H. P. The Influence of Tungsten on Nickel. R. IRMA" (Metall ulzd Em. 1915 12 358-364).-Tungsten has very little tendency t o dissolve in molten nickel a t 1800° but the alloys may be prepared by melting the nickel in an electric crucible furnace with carbon electrode adding nickel oxide t o remove carbon then adding the T. H. P. thus (1) H,S + 2Hg1 = 2HgI + 2HI + S ; (2) 6HgI + H,S + 4s =NINERALOGICAL CHEMISTRY. ii. 531 tungsten and after all is melted deoxidising by means of mag- nesium.A 50% alloy made in this way may be melted with excess of nickel in the electric furnace without difficulty. For analysis alloys containing less than 30% of tungsten may be dissolved in aqua regia with tlie addition of bromine; richer alloys are fused with potassium hydrogen sulphate. Cooling curves are taken by melting 100 grams of each alloy in a carbon crucible lined with magnesia. The arrests are small prob- ably on account of the rapid ratje of cooling. The freezing-point curve up t o 50% TV’ shows maxima a t 1480O and 1495* correspond- ing with the compositions Ni,,W and Ni,W respectively. The former is perhaps not a definite compound. Eutectic points are observed a t 1426O 1435O and 1456O a t the compositions 2 9.2 and 17 atomic % TV respectively.Solid solutions are probably formed up to 2 atomic %. The solubility of nickel in sulphuric acid and its liability t o corrosion are greatly reduced by alloying with tungsten. The tensile strength shows a minimum a t 15% W by weight beyond which i t increases rapidly. The alloy with 19% W may be rolled hot into thin sheets. C. H. D. The Annealing of Nickel Brase (German Silver). F. C. THOMPSON ( J . I77st. Jf etals 1916 15 230-263).-The specific volume of alloys containing 60% of copper and from 7 to 28% of nickel the remainder being zinc diminishes as the nickel content increases. A t the same time the electrical resistance increases and in the annealed state the specific resistance is independent of the copper-zinc ratio and may be expressed by the formula (T= 12 + Ni% microhms per C.C. An abrupt change takes place in the resistance between 300° and 400° the temperature being independent of t h e composition within the limits given. It is connected with the lcnown instability of these alloys when used as resistance wires a t a teinprrature above 300° and may be due t o a transformation of tlie compound NiZn,. The heating and cooling curves plotted by the inverse-rate method do not exhibit any discontinuity b u t it is 1~7ell marked in the thermo-electric properties. When the hard- rolled alloys are annealed a t different temperatures and tested for hardness by the Brine11 method a clear increase of hardness is f o m d to take place a t 300-330°. This temperature is slightly lowered by long annealing but otherwise the influence of time is negligible. C. H. D.