ii. 34 ABSTRACTS OF CHEMICAL PAPERS. Inorganic Chemistry Electrolytic Production of Active Hydrogen. Y. VENKA- TARAMAIAH and B. S. V. R. RAO ( J . Xci. Assoc. Vixianagaram 1923,1 45-46 ; cf. A. 1923 ii 235,482 ; Wendt and Landauer A. 1922 ii 369; Grubb A. 1923 ii 403).-By electrolysing dilute sul- phuric acid with a current of 3 to 15 amperes in a cell with platinum electrodes one being a perforated tube of length 30 mm. internal diameter 1.0 mm. and external diameter 1.5 mm while a current of nitrogen is passed through the perforated electrodeIN ORGANIC CHEMISTRY. ii. 35 into the solution ammonia is formed. In another form of apparatus the permeability of iron to nascent hydrogen a t the ordinary temperature is utilised; a small quantity of triatomic hydrogen present in the atomic hydrogen reacts with sulphur and forms hydrogen sulphide . Revision of the Atomic Weight of Bromine by the Complete Synthesis of Silver Bromide.0. HONIGSCHMID and E. ZINTL (Annulen 1923,433,201-230).-The stoicheiometrical relationship between silver and the halogens has hitherto always been found in modern work on atomic weights by measuring the ratio Ag AgHal. The results now communicated are theref ore novel since in addition to the ratio Br AgBr the ratio Br Ag is determined directly. An apparatus constructed entirely of glass is described by means of which specially purified bromine is distilled under reduced pressure into glass bulbs in which it is weighed. It is then reduced by means of ammoniacal ammonium arsenite solution an excess of which does not matter and the solution treated with a weighed quantity of silver dissolved in nitric acid.Equivalence of the bromine and silver is determined nephelometrically ; also the silver bromide is weighed. As the mean of ten determinations of the ratio Br Ag and nine of the ratio Br AgBr the atomic weight of bromine is found to be 79.916 on the basis Ag = 1074380. It is agreed with Baxter (A. 1922 ii 377) that contrary to the results of Guye and Germann (A 1914 ii 727) the purity of atomic weight silver is fully adequate. Hypobromous Acid and the Determination of Hypobromous Acid and Bromic Acid. E. BIILMANN and E. RIMBERT (Bull. Soc. chim. 1923 33 [iv] 1465-1473).-Hypobromous acid may be prepared by the action of bromine water on silver nitrate solution or on silver oxide the equations given being Br2+AgN03+H20 -+ HBrO +AgBr +HNO and 2Br2+Ag20 +H20 + 2HBrO + 2AgBr and the acid is then distilled under reduced pressure at 30-35".Various experimental difficulties and the methods of obviating them are discussed. Both reactions are reversible ; this may be shown by removal of free bromine on passing a current of air through the mixture when it becomes clear by solution of the precipitated silver bromide. Hypobromous acid even after distillation always contains bromic acid although the latter is not volatile under the conditions of distillation. The bromic acid may be determined by adding excess of phenol which removes the hypobromous acid then titrating with sodium thiosulphate after addition of dilute sulphuric acid and potassium iodide.If the titration is carried out without the addition of phenol the total of hypobromous and bromic acids is obtained and from this the former is estimated by difference. Mechanism of Oxidation Processes. VI. H. WIELAND (Annalen 1923 434 185-203).-I. [With A. WINGLER and H. RAU.] The Activation of Oxygen by Metallic Copper.-When A. A. E. W. S. N. H. J. E. 2-2ii. 36 ABSTRACTS OF CHEMICAL PAPERS. dilute aqueous solutions of carboxylic acids (acetic pyruvic succinic malic benzoic or better oxalic lact,ic malonic fumaric and citric) are shaken with copper powder in an atmosphere of oxygen the latter is rapidly absorbed (with oxalic acid 3Cu absorbs 20,) and the acids are converted into carbon dioxide; the metal passes into solution and thc action which is therefore not a catalytic oiie ceases when dissolution is complete.Neutral substances (alcohol dextrose etc.) are only oxidised when acid is present since the autoxidation of copper depends on its dissolution to form a salt and the metal cannot therefore alone activate molecular oxygen. The sodium salts of the above acids for example are not oxidised by copper and oxygen. As is shown by experiments conducted in an atmosphere of nitrogen the hydrogen acceptor molecular oxygen may be replaced by p-benzoquinone Methylene-blue or potassium persulphate. The copper dissolves and equivalent reduction (to quinol etc .) occurs but there is no oxidation when copper passes into solution in absence of molecular oxygen. An oxidation process similar to that observed with copper and oxygen occurs with cuprous salts and hydrogen peroxide.Similar COz:O relations are observed (potassium cyanide acts as a strong anticatalyst). An ice-cold suspension of cuprous chloride with hydrogen peroxide gives a strongly oxidising brownish-yellow precipitate probably identical with Moser’s copper peroxite (A. 1914 ii 467) which since it is formed by the direct combination of cupric oxide and hydrogen peroxide is probably OH*Cu*O*O*Cu*OH. With acids it affords a cupric salt and hydrogen peroxide and its oxidising properties are simply those of the latter substance. It plays no part in the autoxidations under discussion in which the active intermediary is probably produced (R = acid radical) thus CUR +HO*OH -+ R*CuH-O*OH + R*H +Cu*O*OH or O:Cu*OH (again possibly a basic salt R*Cu:O functions as the intermediary).Cuprous salts like copper also activate molecular oxygen. Cuprous chloride (2 mols.) effects similarly the reduction of 1 mol. of p-benzoquinone in presence of aqueous oxalic acid in absence of oxygen 2CuC1+2CLH +O:O (or O:C,H,:O) -+ 2CuC1 +KO-OH (or OH*C,H,*OH). Hydrogen peroxide so formed during an autoxidation cannot however be detected since it is a t once destroyed (cf. Traube A. 1882 795) but both with copper and with cuprous salts a cuprous-hydrogen peroxide intermediary complex is responsible for the autoxidations observed. [With A. wrn~LE~.]-The Oxidation of Phosphorous Acid.- Aqueous solutions of phosphorous acid do not absorb oxygen but do so in presence of palladium black phosphoric acid resulting.For oxygen may be substituted p-benzoquinone etc. oxygen being excluded. Bromine does not react in dry ethereal solutions with anhydrous phosphorous acid nor does dry palladium black cause absorption of oxygen in dry ether whereas on admitting traces of water both reactions proceed rapidly. Similarly palladium black p-benzo- quinone and phosphorous acid do not interact in absence of water 11. Charcoal does not act like palladium.IXORCAMC (XIEMISTRY. ii. 37 It is therefore concluded that the reducing action of phosphorous acid is due not to its absorption of oxygen but to the loss of hydro- gen from a hydrate which is the active form O:PH(OH) -+ PH(HO),-+ 2H+O:P(OH) (cf. Mitchell T. 1923,123 629 etc.). E. E. T. Diffusion of Sulphur Vapour in Air at the Ordinary Tem- perature. CHAVASTELON (Cmpt.rend. 1923 177 1040-1041 1217-1218).-1f small pieces of rhombic sulphur are kept in confact with or very near to sheets of silver copper or lead a circular sulphide film forms on the metal round the sulphur. Temperature time and presence or absence of light affect the result slightly. The effect is due to sulphur possessing an appreciable vapour pressure a t the ordinary temperature. Silver lead or copper wire was wound round a quartz tube closed a t one end and containing sulphur. The whole was placed (open end inwards) inside a similar but larger tube. With dry air inside the apparatus which was kept closed no metallic sulphide formed a t the ordinary temperature during nineteen months' exposure to light.With moist air a faint tarnishing was noticed. E. E. T. Pyrosulphates and Acid Sulphates. L. CAMBI and G. BOZZA (Ann. Chim. Applicata 1923 13 221-238).-None of the methods proposed for the preparation of pyrosulphates yields the pure salts with the exception of the direct synthesis from anhydrous sulphates and sulphur trioxide. Dehydration of the hydrogen sulphates does not proceed to completion even under reduced pressure or in a current of sulphur trioxide. Sodium pyrosulphate solidifies in lustrous translucent white crystals m. p. 400.9" dy 2.668 begins to exhibit evident dissocia- tion a t about 460" and in moist air is rapidly converted into concentrated sodium hydrogen sulphate solution. The system Na,H,S,O,-Na,S,O with which supercooling is common forms an eutectic containing 6.8 l u 0 1 .~ ~ of the pyrosulphate a t 182.7". Sodium hydrogen sulphate has m. p. 185.7" (cf. ICendall and Landon A. 1921 ii 45). The pure pyrosulphate and also all the mixtures containing it in greater proportion than 10% are greenish-yellow in the liquid state. The dissociation of sodium hydrogen sulphate has been investigated by measuring the pressure of the water vapour emitted a t different temperatures; this pressure is very slight a t 180" and becomes equal to 1 atmosphere a t a little above 320". Potassium pyrosulphate crystallises in transparent colourless prisms m. p. 414.2" di5 2.512 and absorbs moisture from the air only very slowly. The solid undergoes a polymorphic transform- ation at 315" assuming an opaque porcelain-like appearance and a second transformation may be detected thermally at 225"; both changes are frequently accompanied by marked supercooling and usually proceed slowly.The eutectic temperature for the system K,H,S,08-K2S,0 is 201 -2" and the corresponding composition 14 mol. yo of the pyrosulphate. The polymorphic transformations of the pyrosulphate in the mixtures generally escape the thermalii. 38 ABSTRACTS OF CHEMICAL PAPERS analysis owing to the sluggishness with which they occur at low temperatures. The divergent results for the system Na,SO,-H,SO obtained by Pascal and Ero (A. 1919 ii 154) and by Kendall and Landon (loc. cit.) depend on the fact that the former authors fused the mixtures in open tubes and followed the crystallisation during the cooling whilst the latter melted the mixtures in sealed tubes.The curve of unstable crystallisation at 0-25 mol.yo of Na,SO is observed (cf. Kendall and Landon) and the transition point between the compounds Na2S0,,2H,SO and 2Na,S04,9H,S0 fixed a t 54.6". In the range 25-60 mol.% of Na,SO anomalies are manifested owing to dehydration of the sodium hydrogen sulphates occurring in the mixtures melting at the higher temperatures. The non-existence of any compound of sodium pyrosulphate with sodium hydrogen sulphate and the existence of the compounds 2Na,S04,9H,S0 Na2S0,,2H2S04 and Na,S0,,H,S04 are confirmed. The results obtained for the system K,#O,-H,SO differ little from those of Kendall and Landon (loc. cit.). Less dehydration occurs than with the sodium sulphate system but anomalies occur at temperatures above 120".The results indicate potassium hydrogen sulphate to be tetramorphous (cf. Bridgman Proc. Nat. A d . Sci. 1916 52 124). The double salts K2S0,,3H,S04 and &SO,,H,SO are formed. The results obtained with the system (NH,),SO,-H,SO coincide with those of Kendall and Landon. The Preparation of Pure Ammonia. L. MOSER and R. HERZNER (Monutsh. 1923 44 115-122).-Pure magnesium and calcium nitrides readily yield pure ammonia if allowed to fall gradually into gas-free water. On account of the highly exothermal nature of the reaction decomposition of the gas always occurs if water is allowed to fall on the nitride. The only impurity present in the gas obtained by allowing the nitride to fall slowly into water is hydrogen traces of which are formed from the traces of metal always present in the nitride. Commercial ammonium salts always contain organic matter the carbon dioxide obtained by oxidation corresponding with 0-39-0.62 yo of carbon in the samples examined. By oxidation with nitric acid or permanganate a t high temperatures the organic matter may be completely destroyed and pure ammonia obtained from the residues.Ammonia from technical salts may be completely freed from pyridine and other organic impurities by being passed over prepared wood charcoal. s. I. L. Decomposition of Nitric Oxide by Heating with Metals. E. MULLER and H. BARCK (2. anorg. Chem. 1923,129,309-320).- The action of metals on nitric oxide at high temperatures is studied. Copper a t 500" decomposes 99.7% of the gas but the action at lower temperatures depends on the purity of the metal.Silver is without action up to 700" ; iron if reduced in hydrogen is better than copper for decomposing nitric oxide whilst brass is without action a t 600" and decomposes only 29% at 700". Tin has no action up to 4W" T. H. P.INORGANIC CHEMISTRY. ii. 39 but then commences to act rapidly forming a nitride which may be completely decomposed a t 600". Zinc has no action below 350"; at; 600" the action is slow but the decomposition is complete. Bismuth decomposes 74% at 400" forming bismuth trioxide. Magnesium and calcium decompose the gas at 500" with the form- ation of metallic oxide and nitride. Aluminium has little action below 600". Manganese decomposes 68% of the ga,s a t 400" all of it a t 500" manganese nitride being formed.Chromium has no action below 700" and ferrochrome (60% Cr) only 19% at 650". Lead peroxide absorbs nitric oxide at laboratory temperature form- ing lead nitrite. On heating the gas is liberated but at 200" oxygen is liberated as well and nitrogen peroxide is formed. Litharge is practically without action on nitric oxide even a t 650". Red lead has no action on nitric oxide at laboratory temperature but at 200" absorbs it almost completely forming lead nitride ( '2 nitrite). Vanadium trioxide decomposes the gas completely at 500" forming vanadium tetroxide. H. H. The Intermittent Glow of Phosphorus. K. R. KRISHNA IYER (Chem. News 1923 127 321).-The author states that the pheno- mena of intermittent luminosity and the propagation of luminous pulses exhibited by phosphorus contained in an exhausted vessel into which air is allowed to leak (Lord Rayleigh A.1921 ii 546; 1923 ii 755) may be produced in an open vessel containing traces of inhibitors e.g. naphthalene carbon disulphide turpentine and light petroleum. A simple method of exhibiting the effect as a lecture experiment is described. P. PASCAL (Compt. rend. 1923 177 1298-1300) .-Hexametaphosphate is the name usually given to the product of the action of heat (with subsequent rapid cooling) of an acid pyrophosphate or metal dihydrogen phosphate. If the rapid cooling is omitted trimetaphosphate is also formed and some colloidal products Hexametaphosphates are also supposed to result from the interaction of alkali sulphides and the metaphos- phates of heavy metals.The products however are mixtures of trimetaphosphates hexametaphosphates and colloidal meta- phosphates. To obtain a pure alkali hexametaphosphate a pure trimetaphos- phate is fused (at 700") in a platinum crucible and the latter then rapidly placed in cold water. The vitreous product is free from colloidal material and from trimetaphosphate. Tri- and hexa-metaphosphates are interconverted a t a definite temperature (607" & 2" for the sodium salts). The hexameta- phosphate is the form stable at high temperatures. Interconversion is almost instantaneous at one definite temperature being slower as the temperature falls until at the ordinary temperature no change occurs. The identity of the melting points of the two sodium salts (638") is thus explained as are also the details necessary for obtaining " Graham's soluble salt.'' After their formation hexametaphosphates undergo slight J.S. G. T. Hexametaphosphates.ii. 40 ABSTRACTS OF CHEMICAL PAPERS. changes (complete in sixteen hours at 648" or in six hours at 835"). The viscosity of t,heir solutions increases slightly owing to the formation of a little colloidal salt (1.8% and 1.2% at 648" and 835" respectively). The hexametaphosphates are distinguished by the formation of complex salts of the types M,[Fe(PO,),$ M,[Pe(PO,),] and M,[UO,(PO,) s]. Thus an excess of an alkali hexametaphosphate decolorises ferric thiocyanate and prevents the formation of a colour when uranyl salts and ferrocyanides are mixed.The Vapour Pressure of Arsenious Oxide. P. SMELLIE ( J . SOC Chm. Ind. 1923 42 466468~).-Experiments were made to extend the work of Schwers (A. 1920 ii 247) on the vapour pressure of arsenic in sulphuric acid solution Using the same method as Schwers but taking special precautions to prevent sulphuric acid being carried over mechanically much lower results were obtained than those recorded by Schwers. At 60-64" the vapour pressure found was 2.7 to 9 x mm. of mercury. Using a similar method the vapour pressure of pure arsenious oxide was found to be at 60-61" 2.4 x ; a t 81-86" 2-5 x ; a t 101-105" 4.6 x ; at 117-124" 1-9 x lop3 ; at 119-126" 2.2 x ; a t 149-152" 2.6 x 10-2 mm. To obtain hydrochloric acid free from arsenic for the Marsh test it is recommended to dilute the acid to d 1.10 and boil it gently with precipitated copper for several hours using a reflux condenser.The copper is prepared from pure granulated zinc and cuprous chloride dissolved in hydrochloric acid. The acid is then distilled over fresh precipitated copper and stored in Jena glass vessels. It is also recommended to use electrolytic hydrogen for the Marsh test. P. GAUBERT (Compt. rend. 1923 177 1123-1 125).-Very thin graphite laminae result when a lamina (obtained by cleavage) is rubbed between two microscope slides. The lamin= are trans- parent (if sufficiently thin) green in transmitt'ed light give complete extinction between crossed Nicols and in convergent light give a black cross without coloured rings. The crystals are optically negative and have a refractive index bctween 1.93 and 2.07.Graphitic oxide (green) obtained by treating graphite lamina? with fuming nitric acid and potassium chlorate has a crystalline structure (crystals optically negative ; black cross distinct). The oxide is stable while it contains a little nitric acid but after washing with water becomes brown grey or yellow when it also deflagrates on heating giving pyrographitic acid i.e. graphite. The dried oxide has a refractive index of 1-93-2. E. E. T. E. E. T. E. H. R. Optical Properties of Graphite and Graphitic Oxide. Azido-carbondisulphide. I. Formation Preparation and General Properties. A. W. BROWNE A. B. HOEL G. B. L. SMITH and F. H. SWEZEY [with Microscopical Studies by C. W. MASON] ( J . Amer. Chem. soc. 1923 45 2541-2550).-The action of oxidising agents on a 30/6 aqueous solution of potassium szido-INORGANIC CHEMISTRY.ii. 41 ditbiocarbonate (A. 1922 ii 847) has been investigated; in each case a 1% solution of the oxidising agent was used with 5 C.C. of the azido-salt (cf. Sommer A. 1916 ii 29). It is found that hydrogen peroxide after producing a green colour gives a precipitate of azido- carbondisulphide ( S*CS*N,) ; ozonised oxygen potassium chlorate potassium chromate potassium persulphate mercuric chloride ferric chloride (acidified) stannic chloride acidified potassium permanganate solid manganese dioxide cerium dioxide sodium nitrite chlorine bromine and iodine all produce the same substance in greater or lesser yields. It can also be prepared by the electrolysis of a 20 yo solution of potassium azido-dithiocarbonate between a rotating platinum anode of 30 sq.cm. total area and a stationary platinum cathode enclosed in a parchment thimble with an anode current density of 18.6 amperes/dcm.2 when relatively large yields a're obtained. The best method of preparation consists in taking 5 C.C. of the clear concentrated solution of potassium azido- dithiocarbonate obtained by filtering the liquid resulting from the interaction of 6 g. of potassium azide and 6 g. of carbon disulphide in 25 C.C. of water. This is diluted to 55 C.C. with water and treated drop by drop with a normal solution of iodine in potassium iodide and continually stirred until the precipitation of azido-carbon- disulphide is complete. The precipitate is washed wit,h water dried by suction and transferred to porous plates by a boiie spatula pressure and tapping being avoided and placed in a desiccator over phosphoric oxide a t 10" or lower.It is a white crystalline solid soluble to the extent of 3 parts in 10,000 of water at 25" ; it is very unstable and particularly sensitive to impact and heat more so than potassium azido-dithiocarbonate. It explodes with the liberation of heat and with more smoke and less flame than the azido-dithiocarbonate. It undergoes spontaneous decomposition at the ordinary temperature giving nitrogen sulphur and a polythiocyanogen ( S*CS*N3),=2N,+2S + (SCN),. Examin- ation of the rate of decomposition shows that the spontaneous decomposition probably takes place in the two stages (SCSN,),= N,+S+SCSN,*SCN ; SCSN,*SCN=S+N,+(SCN),.In the spon- taneous decomposition the colour changes to dark orange passing through various yellow shades. The crystals are tetragonal or orthorhombic and have a refractive index about 1.8. It is soluble in the commoner organic solvents but decomposes slowly in them. Dilute sulphuric acid (1 6) has little action on azido-carbondisulphide until the temperature is raised above 40° but more concentrated acid decomposes it at all temperatures ; hydrochloric and nitric acids both decompose it. It reacts with potassium hydroxide in much the same manner as a halogen to form the potassium salts of azido-dithiocarbonic acid and azido-oxydithiocarbonate (S*CS*N,) + 2KOH = KSCSN + KOSCSN,. It reduces potassium permanganate potassium iodate and potassium iodide.On the basis of the reactions of azido-carbondisulphide the authors adopt the formula NiN:N*CS.S*S*CS*N:NIN suggested by Sommer. J. F. S. An excess of iodine must be avoided. 2"ii. 42 ABSTRACTS OF CHEMICAL PAPERS. Azido-dithiocarbonic Acid. I. Formation Preparation and Properties. G. B . L. SMITH and F. WILCOXON [with A. W. BROWNE and with Microscopical Studies by C. W. MASON] ( J . Arner. Chem. Xoc. 1923 45 2604-2613).-Axido-dithiocarbolzic acid HS-CS-N a new halogenoid hydracid has been prepared by treat- ment of concentrated solutions of the sodium salt with concentrated hydrochloric acid. It is a white or very pale yellow crystalline solid; it crystallises in the monoclinic system has a strong double refraction and is readily soluble in non-aqueous solvents. It has the characteristic properties of it strong acid and its strength approaches that of hydrochloric acid.It is easily oxidised by various oxidising agents yielding the free halogenoid (S*CS*N,),. In the solid form the acid is very sensitive to both shock and to heat. It undergoes spontaneous decomposition at the ordinary temperature in keeping with the laws of unimolecular change. In the dry state this reaction is catalysed by an intermediate product or by the thiocyanic acid formed but not in aqueous solution. The decomposition may be represented by the equation HS*CS*N,=HSCN+S+N,. The solid product formed consists of polymerised thiocyanic acid and free sulphur. Diffusibility of Helium through Crystalline Septa. A.PIUTTI and E. BOOGIO-LERA (Rend. Accad. Xci. Fis. Mat. Napoli 1923 [iii] 29 111-115; cf. A. 1923 ii 20).-At 480” helium is unable to traverse plates of laevo- or dextrorotatory quartz cut either parallel or perpendicular to the optic axis and having the thickness 0.3-1 mm. although the gas is able to penetrate ordinary or quartz glass a t this temperahre (cf. Williams and Ferguson A. 1922 ii 841). V. CHLOPINE (Bull. Xoc. chim. 1923 [iv] 33,1547-1551).-The values obtained by Engel (A 1894 ii 40) and by Masson (T. 1911 99 1132) for the solu- bility of barium chloride in hydrochloric acid of various concentra- tions were used as the basis of a method of separating the chlorides of radium and barium from one another. Gaseous hydrogen chloride is passed into a solution containing the mixed chlorides until the optimum concentration of the acid has been attained.This value depends on the “coefficient of enrichment,” which is the ratio of the percentage of total radium precipitated to the percentage of total barium precipitated. This coefficient is a regular function of the percentage of the barium salt precipitated and varies from 3.7 to 1.3 with a corresponding precipitation of 7-8% to 55-57% of t’he total barium present. The presence of other chlorides with the exception of that of lead does not interfere with the efficiency of the method which may therefore be carried out without pre- liminary purification of the mixed chlorides. Solubility of Potassium Perchlorate in Salt Solutions and the Corresponding Activity Relations.R. M. BOZORTH ( J . Amer. Chem. Xoc. 1923,45 2653-2657).-The solubility of potass- ium perchlorate has been determined in O-lN 0*3N and 0.6N J. F. S. T. H. P. The Separation of Radium and Barium. H. J. E.INORGANIC CHEMISTRY. ii. 43 solutions of potassium chloride nitrate and sulphate sodium chloride nitrate perchlorate and sulphate and barium chloride and nitrate. The individual effects of the separate ions on the activity coefficient product of the potassium- and perchlorate-ions have been compared and at the concentrations involved (0.25- 0.75N) have been found to be markedly specific and additive. J. F. S. The Third Form of the Amwonia-soda Process. P. P. FEDOTBEV and A. KOLOSSOV (2. anorg. Chem. 1923 130 3 9 4 5 ; cf. A. 1914 ii 268).-A third form of the ammonia-soda process involves the reaction between sodium sulphate and ammonium hydrogen carbonate which is complicated by the existence of the double salt NaNH,SO,,aq now shown to crystallise with 3H,O.The equilibrium conditions for this salt pair have been studied a t 35" under a pressure of 3 atm. of carbon dioxide. Measurements were made of the solubility of ammonium sulphate in sodium sulphate solution and of sodium sulphate in ammonium sulphate solution (these at 15" as well as 35") of ammonium hydrogen car- bonate in ammonium sulphate and vice versa of sodium hydrogen carbonate in ammonium hydrogen carbonate and of sodium hydrogen carbonate in sodium sulphate solution. Three solutions are possible saturated simultaneously with three salts namely (1) NaHCO NH,*HCO and (NH,),SO ; (2) NaHCO NaNH4S0,,3H,O (NH,),SO ; (3) NaHCO NaNH4S0,,3H,0 and Na,SO,.The solubility of sodium sulphate in ammonia solutions a t 35" was also determined. The disadvantage of this form of the ammonia-soda process is the low solubility of sodium sulphate in ammonia solutions. A Caesium Cupric Mercuric Chloride Cs,CuHgCl ; the Failure to Prepare Cs-Cu-Cd or Cs-Cu-Zn Chlorides ; and the varying Complexity of certain Triple Salts. H. L. WELLS (Amer. J. Sci. 1923 6 521-525).-The addition of strong hydro- chloric acid to mixed solutions of cesium cupric and mercuric chlorides with subsequent concentration and cooling causes the separation of the triple salt Cs,CuHgCl in the form of black cubes which when broken show a pale brownish-red colour and thus appear to be strongly dichroic or pleochroic.Triple chlorides of cmium and copper with cadmium or zinc could not be obtained. A comparison of the formulze of various triple salts of the metals of groups I and I1 shows that ammonium and the alkali metals of lower atomic weight than that of cesium tend to form more complex triple salts than cesium itself; this may be attributed to the more electro-positive character of cesium which gives it a E. J. WEEKS and W. V. LLOYD (Chem. News 1923,127,362) .-When stibine prepared as previously described (T. 1923 123 456) is washed dried and passed into a cold N/2-solution of silver nitrate pure silver antimonide is deposited. E. H. R. greater combining power. s. I. L. Preparation of Pure Silver htimonide. J. S. G.T. 2'-2ii. 44 ABSTRACTS OF CHEMICAL PAPERS. Solubility and Surface Energy of Calcium Sulphate. M. L. DUNDON and E. MACK jun. ( J . Amer. Chem. SOC. 1923 45 2479- 2485).-Some of Hulett's experiments (A. 1901 ii 493; 1904 ii 321) on the variation of solubility with the size of the particles have been repeated and further experiments have been carried out with calcium sulphate of different sizes of particles with the object of obtaining trustworthy values of the surface energy. The method of working consisted in adding finely powdered calcium sulphate to a saturated solution a t 25" in equilibrium with large crystals and measuring the increase in concentration and the return to the original concentration by conductivity measurements. On adding particles of average size 0 .3 ~ to a solution in equilibrium with particles 20-5Op long the electrical conductivity rose from 2208 x 10-6 to 2616 x 10-6 in a minute ; aftcr ~ W O days it had rcturned to 2276 x This represents an increased concentration of 24%. It was found that grinding gypsum and precipitated calcium sulphate causes them to lose water and since the particles added in the above experiment con- tained only 12% of watcr instead of 21-17( it is possible that some of the increased solubility is due to this cause. The concentration can be increased if the dehydrated salt is added to the saturated solution without any previous fine powdering. Thus a solution saturated as above rose in conductivity to 2450 x 10-6 in one day and to 2520~10-6 in two days on adding dehydrated crystals of the same sort as those with which the original solution was in equilibrium. Calculations from the results obtained with particles of calcium sulphate 0.2 and 0 .5 ~ in diameter give a value 370 ergs/ cm .z for the surface energy of the dihydrate of calcium sulphate. J. F. S. and a,fter six days to-2213 x Action of Silica and Alumina on Calcium Sulphate. (MLLE) G. MARCHAL (Compt. rend. 1923,177,1300-1302).-Using a method similar to that employed previously (cf. A. 1923 ii 139) the author has studied the reactions (1) CaS0,fSi02=CaSi0,+S02~+O~60 and (2) CaS0,+A1203=CaAl,0,tS0,+0~502 by measuring the equilibrium pressures at difterent temperatures (the systems being univariant). The first reaction begins a t 870" the second at 940-950" both being studied up to 1,250". Both are very rapid thus if gas is removed at 1,230" equilibrium is restored in a few seconds.Por the first reaction the total gas pressure becomes 760 mm. at 1,273" for the second at 1,363" (extrapolation value). E. E. T. New Reaction pmducing Strontium. C. MATIGNON (Cumpt. rend. 1923 177 1116-1118).-The method previously used for the preparation of barium (A. 1913 ii 504) has been successfully applied to the case of strontium which is obtained in a pure state by heating silicon (2.5 parts) with strontium oxide (20 parts) in an exhausted iron tube at 1,250" for three hours (2.1 grams of strontium were actually obtained). Ferro-silicon if sufficienkly rich in silicon may be used instead of silicon. E. E. T.INORGANIC C€IEMISTRY. ii.45 The Technical Electrolysis of Fused Carnallite. P. P. FEDOTBEV~W~~~ N. WORONIN] (2. anorg. Chem. 1923,130,25-38). -A study of the conditions favourable for the preparation of magnesium by the electrolysis of carnallite. [Cf. B. 19.1 E. H. R. The Preparation of Zinc by Electrolysis of Sulphate Solutions. P. P. FEDOT~EV and W. W. STENDER (2. anorg. Chem. 1923 130 51-62).-This paper records the results of a study of the influence of different factors (concentration of electrolyte concentration of acid current density temperature and impurities) on the efficiency of the electrolysis of zinc sulphate. [Cf. B. 18.1 E. H. R. Zinc Halide Ammines. W. BILTZ and C. MESSERKNECHT (2. nnorg. C'hem. 1923 129 161-175).-Compounds of the zinc halides with ammonia have been prepared and their vapour tensions determined. From these measurements the isothermals have been plotted.The halides all form monoammines diammines tetra-ammines and hexa-ammines whilst the chloride in addition forms a deca-ammine. By plotting the heats of formation of the ammines against the valency of the halide it is found that for the hexa-ammines and tetra-ammines the heats of formation are in the order C1< Br < I whilst for the diammines and monoammines the reverse is the case. Lead Acetato- (Oxalato-)complexes and Basic Lead Salts. R. WEINLANU and F. PAUL (2. anorg. Chem. 1923 129 243-262; cf. A. 1922 i 981 ; ii 767).-By means of replacement reactions carried out with the lead acetatoperchlorates previously prepared some other salts have been obtained notably the dithionate picrate nitrate and bromate. Further it has been found possible to prepare lead oxalato-salts in a similar way having kations of the structure Pb,Ox (I) and Pb,Ox (11).H. H. (Ox=C2O4.) Pb- ** Yb ! 1; ox 'Ox" ox (11.) ;I / \ I! -Pb (I ) [Pb -' o=(;-o\'.h]'* ! i I *-.. [ l i b \O"-O .*' 1- Basic lead salts have also been prepared these are considered to be aquo-salts of the types LPb<EE>Pb] (C10J2,H20 and [ P b ( gg>Pb) ] (ClO3),. Of these many examples including dithionates nickelicyanides and a nitroprusside have been prepared. Other basic salts have been prepared derived from the kation [Pb3(OH),]" by elimination of water to give [Pb,( OH),O]" or 1 Pb=O--Pb*$,H>Pb] OH . Salts with tervalent ions (111) 2 .. -J -~ L ~ < ~ ~ ~ ; * P b - O H - - - - P b ~ ~ ~ ~ g ~ > P b ] " ' or [Yb,(OH),]"' and (IV) [(Pb<gE>:.) 2 P b - O H - - - P b ~ ~ ~ ~ > P b ) ~ * * * or [Pb,( OH),]"' arcii.46 ABSTRACTS OF CHEMICAL PAPERS. also described. Finally are described salts containing the phos- phito-plumbo-kations (V) [Pb--O=PH<O-pbl and (VI) O-pb .*'. The following salts are new diacetatodiplumbo-dithionate [Pb,( OAc),]S,O irregular greasy platelets ; picrate lemon-yellow long tablets ; bromate irregular very thin platelets ; diacetutotri- plumbo-nitrate [Pb,(OAc),](NO,) granular crystalline masses. Omlatodiplumbo-perchlorate [Pb,Ox]( C10,),,3H20 long thin tablets ; nitrate thick irregular platelets (2H,O) ; trioxalatotetraplumbo- perchlorate [ Pb,Ox2](C10,),,5H,O prismatic six-sided crystals ; dioldiplumbo-chlorate [Pb,( OH),](C1O3),,H,O large greasy well- defined prisms dithionate (1 H,O) irregular pearly plates bromide (anhy.) whitish microcrystalline powder nitroprusside (2H,O) red cubes nickelicyanide (anhy.) microcrystalline pale yellow powder bromate (anhy.). Tetroltriplumbo-chlorate [Pb,( OH),](ClO,) sharp spiky crystals dithionate (0-5H2O) rods and truncated prisms bromide (anhy.) microcrystalline powder darkening in the light niclelicyanide (anhy.) irregular platelets bromate (anhy.) light orange needles basic bromate [Pb,( OH) ,]( Bra,) lustrous irregular six-sided platelets.Dioloxotriplumbo-bromate Pb(BrO3),,2PbO,H2O yellow microcrystalline powder trioxotetraplumbo-dichromate 4Pb0,2Cr03 fiery red powder dioxotriplumbo-dichromate 3Pb0,2Cr03 deep orange-red powder basic nitrite (06H,O) orange-yellow tufted prisms.Di(dioldip1umbo)chlorate ferricyanide exists in a brown and black form the bromateferricyanide (lH,O) forms reddish-brown irregular tablets the hypophosphite ferricyanzde (anhy.) bright yellowish-green powder. Di(dioldip1umbo)tetroltri- phmbo-tribromate f erricyanide [ Pb,( OH) J( BrO,),[Fe( CN) ,I bright yellow irregular platelets dioldiplumbotetroltriplumbo-nitrateferri- cyanide [Pb5(oH),]No,[Fe(CN),] minute cubes. Dioldiplumbo-p- oldioldiplumbo- f erric yanide olive- green and red forms. Tetroltriplumbo-~-oltetroltri~lumbo-ferricyanide dirty violet powder olacetatodiplumbo-chlorate dioldiacetatotri- plumbo-chlorate phosphitotriplumbo-bromide phosphitodiplumbo- chloride lead chlorate monohydrate and dileadthiocyanate mono- hydrate.H. H. [Pb,(OH),I(ClO,)[Fe(CN) 61 [Pb,( OH),]( H,PO,)[ E'e( CN) ,] Metallisation of Organisms. N. D. ZELINSKI ( C m p t . rend. 1923 177 1041-1043).-A bee placed in a platinum boat and lightly covered with cupric oxide was heated in a glass tube in a current of carbon dioxide. On removing the excess of cupric oxide a copper replica of the bee was obtained the minutest details of the structure being preserved. Beneath the metal was a coke- like mass containing nitrogen and hydrogen. The metallised bee contained 48.8470 of copper. Vegetable products e.g. leaves undergo similar metallisation. The effect is ascribed to the volatility of cupric oxide (cf. following abstract). E. E. T.INORGANIC CHEMISTRY. ii.47 Transportation of Copper in the Gaseous State and Copper Carbonyl. G. BERTRAND (Compt. rend. 1923 177 997-999).- Zelinski's results in connexion with the effect of heat on organisms which have previously been covered with cupric oxide (preceding abstract) are most readily explained on the assumption that copper curbonyE is formed being decomposed almost immediately but being sufliciently stable to account for the alleged volatility of copper oxide. Ignition of animal or vegetable matter not con- taining copper using copper or brass gas burners causes the introduction of traces of copper into the ash obtained this result being observed in absence of halogen compounds. If carbon dioxide is passed over heated cupric oxide in a glass tube no unusual effect is observed but if before passing over the cupric oxide the gas first passes over red hot carbon a red mirror-like deposit of copper is produced presumably owing to the inter- mediate formation of copper carbonyl.Formation of Sulphides Selenides and Tellurides of certain Metals. I. Copper Compounds. F. GARELLI ( A t t i R. Accud. Sci. Torino 1923 58 193-200).-The author confirms the observation made by Wicke (Annulen 1852 82 145) who found that a piece of sulphur wrapped in clean copper wire and immersed in saturated copper sulphate solution at the ordinary temperature gradually becomes coated with indigo-coloured crystalline copper sulphide. The copper sulphate solution which need not be saturated does not change in composition during the reaction. Very pure cupric sulphide may be prepared in con- siderable quantities by adding copper together with rather more than the atomic proportion of sulphur to neutral aqueous cupric sulphate or nitrate kept boiling by means of a current of ,steam.If the proportion of sulphur added is halved pure cuprous sulphide is obtained. I n a similar manner by the interaction of copper and selenium in copper sulphate solution the selenides Cu,Se CuSe and Cu3Se2 may be obtained in pure condition. The copper telluride Cu4Te may be prepared similarly. Basic Copper Sulphate. F. S . WILLIAMSON ( J . Physicul Chem. 1923 2'7 789-797) .-The formation of basic sulphates of copper has been investigated by adding various definite molecular qyan- tities of sodium hydroxide to copper sulphate solution and analysing the precipitates. The results of Pickering (A.1883 853) have been confirmed namely that the precipitate obtained by adding varying amounts of sodium hydroxide not exceeding 1.5 mols. per mol. of copper sulphate is practically constant in composition. This precipitate has the composition CuS04,3Cu0,4H20 which may be written (Cu0,H2)3CuS0,,H,0. By precipitating with alkali as in the above case the compound CuS0,,2Cu0,2H20 is not obtained as postulated by Stearn and Young (Phil. Mag. 1843 [3] 23 501). Since copper sulphate will peptise hydrated copper oxide and be carried down by it there is no certainty that the basic salts may not contain adsorbed copper sulphate. E. E. T. T. H. P. J. F. S.ii. 48 ABSTRACTS OF CBEMXCAL PdPERS. Absorption of Halogens by Mercurous Salts.K. G. NAIK and M. D. AVASARE (J. Arner. Chem. Xoc. 1923,45,2769-2770).- Mercurous chloride sulphate and nitrate are treated with an alcoholic solution of iodine or bromine in excess. Mercurous chloride yields mercuric iodochloride HgIC1 and mercuric bromo- chloride HgBrCl. The sulphate yields di-iodomercuric sulphate (IHg),SO and a perbromide (BrHg),X04,Br2. The latter reacts with dry chlorine to give dichloromercuric sulphate (ClHg),SO,. The nitrate gives a periodide (IHgNO,),,I and a perbromide (BrHgNO,),,Br,. The periodide when treated with chlorine gives a perchloride (ClHgNO,),,Cl,. Separation of Rare Earths by Basic Precipitation. VI. W PRANDTL and J. RAUCHENBERGER (2. anorg. Chem. 1923 129 176-180 ; cf. A. 1922 ii 769).-The influence of mercuric cyanide and of nickel nitrate on the precipitation of the rare earth oxides from solutions of their nitrates by means of ammonia has been studied.The authors conclude that although these and other salts tend to keep t'he rare earths especially lanthanum in solution in the presence of ammonia the most effective for this purpose is cadmium nitrate. H. H. Action of High Temperatures on some Refractory Sub- stances. C. MATICNON (Cmpt. rend. 1923 177 1290-1293).- Various substances have been heated in the form of pellets in an electric furnace in an atmosphere of nitrogen. Sodium aluminate Pu'a,O,Al,O (m. p. 1,660') is prepared by heating the appropriate mixture of aluminium oxide and sodium carbonate a t 1,100-1,200" a pure product also being obtained using a slight excess of the oxide.Aluminates of the type A1 O,,(n+l)Na,O dissociate below 1,650'; sodium oxide being lost until the simple aluminate is formed. Zircon used in the natural crystalline form does not melt a t 2,126'. At 1,800' some dissociation occurs dense white fumes of silicon dioxide being formed a t 1,900". The residue melts a t about 2,600'. Zirkite (that is zircon containing excess of silica and some ferric oxide) does not melt a t 1,950'. Tungstic oxide does not melt a t 2,130'. In presence of carbon it is converted into tungsten and a carbide. Aluminium nitride (AN) does not melt at 2,200". At a higher temperature it dissociates into its elements but is under these conditions only slowly affected by oxygen. E. E. T. Composition of the Precipitate from Partially Alkalinised Alum Solutions. L.B. MILLER (U.S. Pub. Health Repts. 1923 38 1995-2004 ; cf. Theriault and Clark ibid. 181 ; Williamson A. 1923 ii 324).-Varying amounts of sodium hydroxide were added to 0.005 and 0-02M- (in aluminium) alum solutions a t room temperature and to the latter at 100". After precipitates had settled for half an hour the pH value of the liquid was determined colorimetrically . After using a centrifuge and decanting the precipitate was treated in a centrifuge with successive 200 C.C. portions of water until nearly free from sulphate-ion. At this \V. S. N.INORGANIC CHEBfISTRY. ii. 49 point dispersion of the precipitate commenced. Bringing the pE of the wash-water to that of the solution had little effect on the composition of the precipitate.For additions of sodium hydroxide up to 2.5 mols. per mol. of aluminium at room temperature the composition of the precipitate was constant and approximate to 5A1,0,,3S03. Increasing concentration of sulphate-ion over a wide range (by addition of potassium or ammonium sulphate) or increasing concentration of aluminium up to 0.1M had no effect on the composition of a precipitate formed at a definite p E value. For pH 4-0-5.5 the ratio of aluminium to sulphate in the pre- cipitate is constant; at higher p values the sulphate rapidly disappears. When three or more mols. of sodium hydroxide are added for each mol. of aluminium the precipitate can be washed free from sulphate. The precipitate appears to consist of two components of nearly equal solubility.For 0.005M-aluminium the greatest insolubility of the precipitate was found at pH 6.7-7*0 at which point 2.75 mols. of sodium hydroxide have been added. On both sides of this however pH 54-8.5 are zones of great insolubility. Theriault and Clark's (Eoc. cit.) point of greatest flocculation pa 5.5 (2.4 mols. of sodium hydroxide added) is the point where precipitation of aluminium first approaches completion and is in the region where the greatest amount of sulphate is found in the precipitate. In Blum's method for the determination of aluminium it is essential that sulphate should be absent or present only in small amount; otherwise a second precipitation from hydrochloric acid solution is necessary. Chloride is satisfactorily removed by ignition over a MBker burner for ten minutes.Experi- ments with solutions of aluminium chloride were abandoned on account of the formation of colloidal suspensions not flocculated by the prolonged use of a centrifuge. The Production of Manganese by Electrolysis of Aqueous Solutions of its Salts. P. I?. FEDOT~EV (2. anorg. Chem. 1923 130 18-24) .-Attempts to obtain pure manganese by electrolysis of manganous chloride or sulphate under dserent conditions were unsuccessful. Under the most favourable conditions using a neutral or weakly acid 6.5N solution of the chloride at 5" with a current density of 20 amperes per sq. dcm. the deposit on the copper cathode contained only about 65% of metallic manganese the remainder being hydroxide. [a. B. 19.1 E. H. R. Formation of Manganese Carbide from Carbon Dioxide and Manganese.E. MULLER and H. BARCK (2. anorg. Chem. 1923 129 321-322) .-When a mixture of carbon dioxide and hydrogen is passed over manganese at SOO" some of the manganese is con- verted into manganous oxide and some into manganese carbide; the latter yielded methane when treated with water. The gaseous mixture was found to contain carbon monoxide. H. H. Polymorphic Transformation of Iron at 370" and the Possiblity of Dissolution of Cementite in a1-Iron. G. SIROVICH (Gaxxettn 1923 53 674-688) .-The existence at 370" CHEMICAL ABSTRACTS.ii. 50 ABSTRACTS OF CHEMICAL PAPERS. of the point A corresponding with the polymorphic transformation of al- into ct,-iron (La Metall. Ital. 1922 16 3) is confirmed by the results of dilatometric measurements which exhibit anomalous behaviour a t that point.Both the al- and the a,-modifications of iron possess characteristic coefficients of expansion so that the conclusions of Le Chatelier of Charpy and Grenet and of Driesen with regard to the way in which the coefficient of expansion of iron varies below 700" require revision. The possibility of effecting dissolution of cementite in ccl-iron by means of maturation in the field of stability of this iron on pearlitic steels must be admitted and such maturation certainly induces in the metal abnormal hardness which indicates the occurrence of phenomena not yet observed. T. H. P. Ferric Salt as the " Solution Link " in the Stability of Ferric Oxide Hydrosol. A. W. THOMAS and A. FRIEDEN ( J . Amer.Chem. Xoc. 1923,45,2522-2532) .-Ferric oxide hydrosols prepared from ferric chloride have been investigated with the object of ascertaining the quantitative relationships between the substances constituting the colloid and of determining the amount of electro- lyte required to keep the colloidal particles dispersed. It is found that precipitation of the sol is imminent when the ratio Fe,O,,/FeCl is about 21 and that this value is unaffected by dilution. It is shown that the stability of ferric oxide sol stabilised by ferric chloride is due to the solubility or the solution forces of the adsorbed ferric chloride in the dispersion medium rather than to the mutual repulsive forces of the particles presumed to reside in their electrical charges of like sign. The so-called " meta-iron '' sol of Pkan de St.Gilles ( J . pr. Chem. 1855 [l] 66 137) is one in which the particles of the dispersed phase are less hydrated than those in Graham's ferric oxide hydrosol. J. I?. S. Mechanism of the Mutual Precipitation of certain Hydrosols. A. W. TIIonus and L. JOHNSON ( J . Amer. Chem. SOC. 1923 45 2532-2541) .-The mutual precipitation of ferric oxide and silicic acid sols and ferric oxide and arsenic trisulphide sols has been investigated. It is shown that the precipitating ratios depend on the peptising agent. There is chemical equivalence between the peptising agents of ferric oxide hydrosol peptised by ferric chloride and silicic acid sols peptised by sodium silicate provided the ratio of peptising agent to the dispersed phase falls within a certain range. Outside this range the precipitation is erratic.Ferric oxide-silicic acid sol precipitations showing chemical equivalence between the peptising agents at maximum precipitation exhibit little variance in precipitation ratios with dilution whilst those showing a divergence from chemical equivalence approach the chemical equivalence on dilution. The mutual precipitation of ferric oxide-silicic acid sols is due to the removal of the peptising agents by chemical action between them. Qualitative experiment shows that the mutual precipitation of arsenious sulphide hydrosol and ferric oxide hydrosol may be due to the chemical reactionINORGANIC CHEMISTRY. ii. 51 S"+2Fe"* -+ S+2Fe". This however has not been confirmed quantitatively. J. F.S. Solubility of Nickel Sulphate by the Floating Equilibrium Method. F. C. VILBRANDT and J. A. BENDER (Ind. Eng. Chem. 1923 15 967-969) .-The method for determining solubility described by Dundon and Henderson (A. 1922 ii 552) was used for measuring the solubility at different temperatures of hydrated nickel sulphates. The new method gives results in good agreement with those obtained by the gravimetric method and is much more rapid. The following results in g. of anhydrous nickel sulphate per 100 g. of water were obtained NiS0,,7H20 -4.25' 27.335; 40.594 ; 30" 43368 ; NiSO,,GH,O (blue) 31-71' 45-299 ; 40" 47.528 ; 50° 52.166 ; 53-26" 54.041 ; NiSO,,GH,O (green) 58-21" 55.389 ; 60*ll" 55.557 ; 79.75" 64.476 ; 94*22" 72.597. -2.O" 23.366; 0" 26.189; 6" 30.282; 15*65" 35.491; 25" E.H. R. System Chromium-Carbon. 0. RUFF (2. Elektrochem. 1923 29 469474).-A criticism of Nischk's paper (A. 1923 ii 762) in which it is stated that the carbide of chromium richest in chromium has the formula Cr,C. In an earlier paper (A. 1918 ii 399) Ruff and Foehr have shown that a carbide Cr5C2 exists and can be prepared in quantity. After a discussion of the data furnished in both papers the author maintains his assertion as to the existence of Cr5C,. Iso- and Hetero-poly-acids. XIX. Molybdi-phosphites and -pyrophosphates. The Structure of Phosphorous Acid. A ROSENHEIM and M. SCHAPIRO (2. anorg. Chem. 1923 129 196-205 ; cf. this vol. ii 54 ; Rosenheim Weinberg and Pinsker A. 1914 ii 58) .-Heteropolymolybdates similar to the hetero- polyvanadates have been obtained of the general formulae 2R20,P,03,12Mo03,xH20 (ammonium potassium sodium and lit hzum) and 2R,O P,O 5Mo0 ,xH20 (potassium ammonium and guanidinium) for the phosphites and 2R,0,P,0,,12M00,,xH20 (sodium and lithium) for the pyrophosphates.[With A. ITALIENER.]-By measurements of depression of freezing point and elevation of boiling points of aqueous solutions of phos- phorous acid it is concluded that phosphorous acid may be repre- sented as H,(P,O,H,) in complete analogy with H4(P,0,) for pyrophosphoric acid. There is thus a complete series of ions containing phosphorus (PO,) + + + (HPO,) + (H,PO,) + H,PO and (pH,)-. J. P. S. phosphate phosphite hypophosphite phosphine oxide H. H. phosphonium Complex Chlorotungstates. 0. COUENBERG and K.SANDVED (2. anorg. Chem. 1923 130 l-l7).-The only derivatives of tervalent tungsten known are the potassium salt K,W,Cl and the corresponding ammonium rubidium cesium and thallium salts prepared from it by double decomposition (A. 1913 ii 328; 1914 i 944). It is now shown that in these salts the tungsten is present as part of a complex anion and a number of new com-ii. 52 ABSTRACTS OF CHEMICAL PUERS. pounds have been prepared from the potassium salt by double decomposition with cliff ereiit chlorides. The complex character of the potassium salt was proved qualitatively by ion migration experiments. The freezing point depression in aqueous solutions of the salt K3WgCl indicated that not more than four ions are formed by dissociation of the salt and conductivity experiments were confirmatory.The complex chlorotungstates are strong reducing agents ; solutions of gold silver mercury and copper salts are readily reduced and powerful oxidising agents oxidise the tervalent tungsten to tungstic acid. With ferric salts this reaction is quantitative and can be used for the analysis of the complex salts of tervalent tungsten. The stability of the complex ion W2C$"' is shown by the fact that the potassium salt interacts with cadmium copper and silver salts in presence of strong ammonia to form ammines without precipitation of any tungsten hydroxide. The following new coinpounds are generally sparingly soluble in water; their most concentrated aqueous solutions have an intense green colour becoming yellower on dilution. At lower temperatures the solutions are fairly stable more so in presence of hydrochloric acid.The figures in brackets indicate solubility a t 20" per 100 C.C. of solution. [Cr(NH,),]W,Cl,,2H20 dark green thin microscopic rectangular prisms ; [ Co(NH,) ,]W,Cl,,GH,O (0-96) bright green microscrystalline powder. [Ag(NH,),],W2C1 is very unstable towards water and air losing ammonia and deposit- ing metallic silver. Cu(NH,),KW,CI,,H,O (1.85) forms a finely crystalline powder. Cu(NH,) NH4W2Cb,H20 (2.4) a green crystal- line powder. Cd(NH,),NH,&,CI (2.17). Cd(NH,)4KW2C1 (1.9). The pyridine compound (C,H,N),W,CI (4.7) forms brownish- green metallic tabular crystals ; the hexumethylenetctrumine com- pound (C,H,,N,),W,Cl,,H,O (0-2) is a yellowish-green crystalline powder ; tetramethylammonium compound (XMe,),W,Cl (1 .IS) yellowish-brown microscopic crystals ; (NEt,),W,Cl (relatively high solubility) ; trimethylamine compound (NHMe,)3W,CI dark green thin pyramidal prisms ; dimethylamine compound (NH,Me,),W,Cl (8.25) dark green rhombohedra ; aniline compound (NH3Ph),W2Cl (0.37) greenish-yellow leaflets; phenyltrimethyl- ammonium compound (Nl\le,Ph),W2C19 (0-21) a voluminous greenish-yellow crystalline powder ; p-tolyltrimethylammonium com- pound (C,H4Me*NiUe3),W,C1 (0-48) yellowish-green striated prisms ; trimethylsulphine compound (SMe3),FV2Cl9 (0-24) a bright green crystalline powder. A solution of the free acid H,W,Cl was obtained by interaction of the thallium salt and hydriodic acid but the acid could not be isolated.The solubility of the potassium salt K,W,Cl at 20" is 15.4 g.per 100 C.C. of solution. E. H. R. Solubility of Titanic Acid in Alkali Hydroxides and in Alkali Carbonates. Crystalline Titanium Oxychloride. V. AUGER (Compt. rend. 1923 177 1302-1304).-The two com- pounds Na,Ti0,,4H20 and K,Ti0,,4H20 described by Demoly in 1849 are shown iiot to exist. The approximate solubility ofINORGANIC CHEMISTRY. ii. 53 t'itanic oxide in alkali hydroxides etc. has been sbudied using either the hydrated oxide or titanic chloride. The results are as follow (figures refer to mg. of TiO dissolved in 100 C.C. of solution) Sodium hydroxide loyo 2-2.5 ; 36% 6-10. Potassium hydr- oxide lo% 3 0 4 5 ; 40% 70-90. Sodium carbonate (saturated) does not dissolve titanic oxide. Sodium hydrogen carbonate (10 yo) 35.Potassium carbonate 30% 2 ; saturated solution 30. (All the titanic oxide present in the last three solutions is precipitated on diluting and boiling.) Potassium hydrogen carbona4te saturated sohtion 700. When 1 part of titanic oxide is fused with 40 parts of sodium hydroxide and the cooled melt extracted with water the only crystalline material obtained is a hydrate of sodium hydroxide. A little titanic oxide remains in solution the majority being prc- cipitated even from concentrated solutions. Fusion of titanic oxide with potassium hydroxide etc. affords a metastable eolu- tion of a titanate (containing up to 1,800 mg. of oxide per 100 c.c.) but this deposits most of its titanium (leaving about 100 mg. in solution) in a few hours. Fusion of titanic oxide with sodium carbonate also gives metastable solutions containing 25-10 mg. of oxide per 100 c.c.all of this being precipitated on keeping. When titanic chloride is added t o a mixture of potassium hydrogen carbonate and its aqueous solution as much as 2,000 mg. per 100 C.C. of titanic oxide remain in solution 1,300 mg. however being precipitated in a few hours. A double carbonate is probably present in the more concentrated solutions. If a solution of titanic chloride in concentrated hydrochloric acid is evaporated in the cold over sulphuric acid an oxychloride Ti02,HC1,3H20 or Ti( OH),C1,2H20 is deposited as large colourless rhombohedra1 plates decomposed in moist air. Reduction of Thorium Zirconimn and Titanium Dioxides. 0. RUFF and H. BRINTZIRGER (Z.anorg. Chem. 1923 129 267-275) .-The reduction of these oxides by means of metallic calcium and sodium a t high temperatures was studied. At 900- 950° sodium alone will not reduce the oxides to any great extent although calcium gives yields of from OO-l00./,. The bcst results were given by a mixture of calcium and sodium containing 30% of the latter. H. H. T. L. WALKER (Nature 1023 112 831).-A suggestion that Sorby's " jargonium " (1869) may have priority over either hafnium or celtium. G . HEVESY and V. T. JANTZEN (Chem. News 1923 127 353-355).-Detailed accounts are given of the preparation of ammonium zirconium fluoride and ammonium hafnium fluoride from alvite and of the separation of hafnium from this mineral by the double fluoride method (A. 1923 ii 570) employing the ammonium double fluorides in place of the potassium double fluorides as previously described. J.S. G. T. E. E. T. Hainiwn or Jargonium. A. A. E. Separation of Hafnium froin Zirconium.ii. 54 ABSTRACTS OF CHEMICAL PAPERS. Iso- and Hetero-poly-acids. XVIII. Vanadioiodates Vanadioperiodates and a few Vanadiophosphates. The Alkalimetric Determination of Vanadic Acid. A. ROSENHEIM and K. H. YANG (2. anorg. Chem. 1923 129 181-195; cf. A. 1922 ii 47).-By introducing vanadium pentoxide into a boiling aqueous solution of iodic acid the acids V205,1205,4H20 and V205,21205,10H20 were obtained according to the amount of iodic acid used. The potassium ammonium and guanidinium salts of these acids are described. Similarly by using alkali periodates and vanadates the sodium potassium and ammonium salts of the type 3R20,2V,05,120,,xH20 were obtained.By using phosphates and vanadates if care be taken that the phosphates are sufficiently dilute (2-3 normal) it is possible to obtain salts of the types 2R20,V205,P205,xH20 and R,0,2V205,P205,xH20. Conductivity measurements were made to ascertain the structures of the valrious anions. Vanadic acid is best determined in solution by the addition of excess of sodium hydroxide followed by back titration with sulphuric acid at loo" using a-naphtholphthalein as indicator. Iodic acid is also estimated by this method so that a mixture of the two may be analysed by adopting the above procedure and then reducing iodate to iodide and determining this with silver. The Gravimetric Ratio of Antimony to Antimony Tetroxide. J.KNOP (2. anal. Chem. 1923,63 181-188).-Pure antimony was treated with nitric acid and subsequently converted into the tetroxide by ignition at 800-900". The purity of the product was determined by the iodine-thiosulphate method. The results agree with a mean value of 122.04 for the atomic weight of antimony. A. G. P. Gold-Chromium Alloys. R. VOGEL and E. TRILLING (2. anorg. Chem. 1923,129 276-292) .-The temperature-composition diagram for gold-chromium alloys is given in full. Three kinds of mixed crystals are distinguished one rich in chromium and two rich in gold. No compounds of definite composition are formed. Chemistry of the Platinum Metals. IV. Alkali-Ruthenium Double Sulphites. 11. H. REMY and C. BREIMEYER (2. anorg. Chem. 1923 129 215-242; cf. A. 1922 ii 857).-The prepar- ation of the following new double sulphites of ruthenium and the alkali metals is described potassium trisulphitoruthenate dark green needles ; sodium undemsulphitodiruthenute yellowish-white crystals ; sodium ruthenoruthenisulphite 2RuS03,Ru2(S03),,4Na,S0 dark blue amorphous powder. In addition a method of analysis of these salts involving the determination of the ruthenium as tetroxide and the sulphur as barium sulphate is described. H. H. H. H. K,IIRu(SO3)31 Ru2(S03)3,8Na2S03,3H20 H. H.ANALYTICAL CHEMISTRY. Miner a1 o g i c a1 C h 8 mist r y. ii. 55 Bitumen of Judaea. The Sensitivity of Bitumen to Light as a Function of its Degree of Dispersity. J. ERRERA (Bull. SOC. chim. 1923 33 [iv] 1409-1414).-Bitumen which is sensitive to light appears to contain three distinct substances a- p- and y-bitumen of which only the last is sensitive. The author regards bitumen as a " polydispersoid," the sensitive portion being that which is in the form of a disperse colloid. It is the portion with the greatest sulphur content and this appears to be correlated with polymerisation molecular and colloidal bitumen being the extrema between which are found -intermediates of various degrees of asso- ciation. Isolation of any one of these probably has a coagulating action. The conclusions are confirmed by evidence from measure- ments of viscosity and of molecular weight dialysis and sensitive- ness to light. Uka-filtration is suggested as a method of separating the light-sensitive colloidal bitumen from the molecular portion. H. J. E.