ii. 136 ABSTRACTS OF CHEMICAL PAPERS. Inorganic Chemistry. Preparation of Iodine Monochloride. Josh J. CERDEIRAS (Anal. Fis. Quim. 1915 13 460-462).-Iodine monochloride is best prepared by passing dry chlorine over excess of iodine. It forms a red liquid b. p. 102'. Oxidation of Hydrogen Sulphide by Means of Ozone in Steam at 120". U. BRESCIANI (Ann. China. Applicata 1915 4 343-346).-Under these conditions the oxidation of hydrogen sulphide t o sulphuric acid is not complete even when the ozonised air is in very large excess. Sulphites Thiosulphates and Polythionates. 111. Action of Mercuric Chloride on Sulphuroue Acid Sulphites Thio- sulphates and Polythionates. A. SANDER (Zeitsch nizgezo. Chew 1916 29 11-12 16).-In previons work on this subject (A. 1915 ii 161 629) the author has assumed that the statement of Feld (A.1911 ii 289 769) is correct namely that the' action of mer- curic chloride on thiosulphates and polythionates is an oxidising one) as expressed for example by the equation Na,S30 + 2HgC1 + 2H,O = N3,S04 + Hg,Cl,* 2HC1+ H,SO,+ S. This is apparently supported by the1 fact that the1 precipitate obtained is white. Further investigation however supports the older statements of Herschel Rose and Kessler (1819-1849) that the white precipitate obtained is the compound Hg3S,Cl formed by the interaction of mercuric sulphide and the excess of mercuric chloride the reactions taking place according to the equations 2Na,S,03 + SHgCl + 2H,O = Hg3S,Cl + 4NaCl+ 2H,S04 ; 2K,S30 + 3HgC1 + 4H,O = Hg3S,Cl + 4KC1+ 4H,S04 ; 2K,S40 + SHgCl + 4H20 = Hg3S,C1 + 4KC1+4H,S04+2S.The samel amount of acid is liberated as according t o Feld's equations so that previous estimations depend- ing on acidimetry are correct. The validity of the last equations was proved by weighing the quantity of precipitate obtained in various experiments the precipi- tate consisting of the compound Hg,S,Cl mixed in the case of the t'etrathionate with sulphur. The estimation of a mixture of trithionate and tetrathionate can be best carried out as follows The solution is oxidised with hydrogen peroxide in the presence of a known excess of sodium hydroxide the reactions being Na,S,O + 4H,O + 4NaOH = 3Na2S0 + 6H,O ; Na2S40p + 7H,02 + 6NaOH = 4Na2S0 + 10H,O. The unused alkali is then titrated. Another sample of the solution is heated with mercuric chloride (Zoc.c i t . ) and the acid liberated is estimated. From these two1 results i t is easy t o calculate the amounts of tri- and tetra-thionate present. G. CES'ZRO (Chem. Zeiztr. 1915 ii 874; from Bull. SOC. franc. Mintrd. 1915 38 38-74).-The A. J. W. T. H. P. T. S. P. Constitution of Thio-salts.INORGANIC CHEMISTRY. ii. 137 author discusses the structural formulz of thio-salts R:I1S3,rnR1IS where RIII repre’sents As Sb Bi Fe etc. and RII Cu2 Pb Hg etc.; m is termed the basicity of the thio-salt and varies from 4 t o 12. The various types of thio-salts are considered in relation t o natural minerals especially argentiferous pyrites pyrrhotite and berthierite. The natural thio-salts are divided into the following types poly-salts in which m is less than 1 ; meta-salts with m = 1 ; salts intermediate to meta- and pyro-salts m lying between 1 and 2 ; pyro-salts with rn = 2 ; salts intermediab t o pyro- and ortho-salts m being between 2 and 3 ; ostho-salts with m=3; argentiferous pyrites which represent molecular mixtures of a poly-salt having m={ with an ortho-salt; basic thio-salts with m between 3 and 4 ; and perbasic thio-salts with rn between 5 and 12.T. H. P. Baeyer’s Tension Theory and the Structure of the Diamond ERNST MOHR (Chem. Zentr. 1915 ii 1065; from Sitzungsber. d. Heidelberger Bknd. Tl’iss. Mathern. naturwiss. Kl. [Abt. A] 1915 7 18 pp.).-The author’s considerations are based on Saclise’s theory according t o which two different tension-free forms of cyclo- hexane are conceivable with the atoms of the ring not in one plane.Sachse’s models are not a t variance with observation if i t is assumed that impacts with other molecules are sufficient t o give to the cyclohexane ring all the possible forms in t u r n ; the number of isomerides with cyclohexane and its substitution products are then with tension-free arrangement of the carbon atoms the same as if the atoms of the ring were situated permanently in one and the same plane. The case of tension-free ring-combinations in which several carbon atoms of one cyclohexane ring belong also t o another ring is explained. Superposition of a number of such combinations leads to large carbon atom systelms of simple regular st’ructure all the carbon valencies in the interior being satisfied without tension; only a t the surface of the system do valencies protrude unsatisfield into space.I n one form of this model the six atoms of each cyclohexane ring stand exactly above the six ring- atoms of the next lower layer (“geradex Bau”); in a second possible form the rings are so arranged that the centre of each lies immediately above a ring carbon atom of the adjacent layer (“ schiefer Bau ”). The latter carbon atom system is identical with the space-lattice of the diamond derived by Bragg from the reflection of X-rays by crystals (Proc. Roy. SOC. 1913 [,4] $9 277). The constitution of graphite may be regarded as correspond- ing either with the direct superposition of the cyclohexane rings or with a space-lattim in which the structure is alternately direct and oblique.T. H. P. The Electrolysis of an Aqueous Solution of Potassium Orthosulphoantimonite [Potassium Thioantimonite] and the Constitution of this Compound. J. A. MULLER ( B d . Soc. chim. 1916 [iv] 19 3-8).-1n the experiments the cathode was of antimony and the anode’ of platinum. I n the first series both electrodes were in the solution of the thioantimonite whilst in theii. 138 ABSTRACTS OF CHEMICAL PAPERS. second aeries only the cathode was in this solution the anode being in a 17% solution of sodium hydroxide the two solutions being seperated by a porous diaphragm which was impregnated with the solution of the thioantimonite. I n the first series of experiments a precipitate of antimony was obtained a t the cathode together wit.h a fee,ble evolution of gas whilst a t the anode a precipitate of antimony sulphide containing a little free sulphur and some potassium sulphide was obtained.The results of the second series of experiments show that during the electrolysis there do not pass from the cathode cell to the anode cell any negative ions contain- ing potassium but there does pass in this direction a certain weight of a grouping corresponding with the formula SbS,. The author considers that these results indicate that the original com- pound is really the combination of the radical SbS with three atoms of potassium and not the combination of three negative groups KS with an atom of antimony. W. G. Approximate Determinations of the Boiling Points of some Alkali Haloids. L. H. BORGSTROM (Jahrb. *&fin.1915 ii Ref. 298 ; from Tidskriften. Teknikern 1915 24).-'I'he following values are the means of several determinations which varied 10-2oo LiCl ...... 1360" NaCl ...... 1490" KCl ...... 1500" NaBr ...... 1455 KBr ...... 1435 NaI ...... 1350 KI ......... 1420 L. J. S. The Constitution of Brasses Containiog Small- Percentages of Tin. 0. F. HUDSON and R. M. JONES ( J . Irzst. Metals 1915 14 98-1 15).-That portion of the ternary system copper-zinc-tin has been examined which includes alloys containing from 50% t o 70% of copper and from 0% t o 5% of tin. An equilibrium diagram for this region has been determined. There are no distinct differences between the y constituent of the brasses and the 6 constituent of the bronzes and the two substances appear t o pass continuously into one another.The constituent which appears light blue under the microscope may thus be termed y when in contact with P I or 6 when in contact with a. C. H. D. The Copper-Rich Kalchoids. SAniuEL L. HOYT and PAUL H. M. P. BRINTON ( J . Inst. Metals 1915 14 178-188. Compare A. 1914 ii 366 ; Hudson and Jones preceding abstract).-The position of the eutectoid h e in the copper-zinc-tin alloys has been further determined and the syst7em is shown t o be truly ternary. Both therma.1 curves and microscopical examination are employed. C. H. D. The Micro-chemistry of Corrosion. IV. Gun Metal. CEcIr H. DESCH and HENRY HYMAN ( J . Inst. Metnls 1915 14 189-198. Compare A. 1914 ii 367 655; 1915 ii 689),-The corrosion ofINORGANIC CHEMISTRY. ii. 139 copper-tin alloys containing zinc has been examined by the method previously described.The total corrosion is much less than that of the brasses under similar conditions and the course varies in a marked degree with the applied E.M.F. When the current. is small only the a-phase is athacked but with larger currents the eutectoid is also corroded. The protective action of the layer of basic tin salts is strongly marked. The a-constituent develops a distinct crystalline structure during corrosion. I n quenched alloys contain- ing a and fl there is very little difference between the two con- stituents in their rate of corrosion. C. H. D. Polyiodides. 111. The System CuI-I,. R. KREMANN and V. BORJANOVICS (Monatsh. 1915 36 923-927. Compare A. 1913 ii 15).-Cooling curves for mixtures of cuprous iodide and io.dine previously heated for several hours a t teniperatures varying from 180° t o 240° have shown no evidence of the formation of any compound.The only change indicated by these curves corresponds with the freezing point of iodine a t 1 1 4 O . This result has been confirmed by measurements of the vapour pressure of the iodine in equilibrium with the above mixtures a t 2 6 O . The values obtained were identical with the vapour pressure of pure iodine. H. M. D. Aluminium.-Aluminates. ERNST MA~ETIX (,Voaz. Sc? . 1 9 15 [v) 5 225-232).-Aluminium hydroxide produced by the hydrolysis of aluminates is quite different in properties from that produced by the precipitation of solutions of aluminium salts with ammonium hydroxide. It is of a sandy nature non-hygroscopic and when dry has a composition corresponding with the formula A1,o3,3H,0.On heating i t commences t o lose water a t about 1 6 0 O ; a t 225O it contains 2M,O and at 235O lH,O but it is only completely dehy- drated in the neighbourhood of 1000°; the loss of water during heating does not take place a t a uniform rate. The partly dehy- drated hydroxide takes up water again on exposure t o the air t o the extent of half that which has been lost but this water is readily removed in a desiccator or on heating a t looo. The author endeavours t o explain these results by means of constitutional f ormulz. The hydroxide obtained by precipitation with ammonium hydr- oxide when d-ried a t 80° has a composition approaching that of A1,0,,2H,O.The loss of water on heating takes place fairly uniformly and i t is impossible t o say that a t any particular temperature a definite hydrated oxide exists. When mixtures of alumina and barium carbonate no matter in what proportions are heated a t temperatures not exceeding 1500° the aluminate' Al,O,,BaO is always formed. A t the temperature of the electric arc trzbarium aluminate A1,0,,3BaO is obtained from a mixture of alumina and barium carbonate in the molecular p(ro- portions of 1 3. It is soluble in water and the solution gives with calcium chloride a precipitate of the formula (A1,03,3CaO)5,3CaC12. Solutions in water of the product obtained by heating mixtures ofii. 140 ABSTRACTS OF CHEMICAT4 PAPERS. alumina and barium carbonate a t temperatures up t o 1500O are not stable and slowly deposit aluminates which are less and less rich in barium. From a solution made from alumina and barium carbonate in the molecular proportions of 1 1 on concentration a barium alziminccte 10A1,0,,1lBa0,55H,O can be obtained as needles ; by precipitation with alcohol an nlicininate (doubtful whether pure) 5A1,03,4Ba0,4H,0 is obtained.Uiccrlciurn al?imiirnt? A1,0,,2CaO is obtained by heating a mix- ture of alumina and calcium carbonate in the molecular proportion of 1 2 a t 1500O. By appropriate variation of the composition of the mixture used the aluminates Al,O,,CaO A1,03,3Ca0. 2A1,03,3Ca0. 5A!,O,,GCaO and 10A120,,11Ca0 have been ob- tained but not properly investigated. Their solubility in water is very small. Double aluminates of the alkali metals and alkaline earth metals can be obtctined by heating mixtures of alumina alkaline earth carbonates and sulphates o r chlorides of the alkali metals. Prom mixtures of alumina barium sulphatel or calcium sulphate and carbon s~~lpl~ido-nl~~rni~~cctes are obtained of which the following have been isolated A1,03.2BaS ; 2A120,,3BaS.A mixture of silica alumina and calcium carbonate in th+ molecular proportion of 2 l 2 when heated t o 1400O gives the silico-aluminate 2Si0,,A1,03,2Ca0 which is insoluble in water. Under similar conditions using the molecular proportion 2 1 1 the silico-nlzc?niizate 2Si0,,A1,03,Ca0 is obtained. The Conservation in the Cold of Solutions of Sodium Aluminate. G. A. Ls ROY (Co?npt. wizcZ. 1916 162 74-7.5).- Concentrated solutions of sodium aluminate which slowly decom- pose when allowed to remain a t the ordinary temperature can be kept unaltered for an indefinite time a t -lo to - 2 O in an ice.chest. W. G. Constitution of Ultramarine. HISINRICH PUCHNER (Kolloid Zeitsch. 1915 17 119-123).-1n the course of experiments on the effect of sodium chloride solutions on various types of soil it has been found that when calcareous sand containing humus and gypsum obtained from the1 neighbourhood of Miiiiich is subjected t o the action of sodium chloride solution in such a way t h a t the solution is drawn up through the soil by capillary action the aqueous extracts from the middles portion of the soil yield a residue which on ignition exhibits the dark blue colour characteristic of ultramarine.Analyses of such residues show that the1 development of the blue colour is connected with the proportion of sulphate in the residue. The general behaviour of the residues seems t o show conclusively that the blue colour is really due t o the formation of ultramarine. I f the sodium chloride is replaced by potassium chloride the1 effect is not obtained. The observations are discuss'ed in reference to whether ultra- marine is a definite chemical compound or whether the blue colour is connected with adsorption phenomena but no definite conclusion is drawn. H. M. D. T. S. P.INORGANIC CHEMISTRY. ii. 141 Anhydrous Sulphates. VIII. Manganous Sulphate with Lithium Sodium and Potassium Sulphates. G. CALCAGNI and 1). MAROTTA (Gaszptta 1915 45 ii 368-376.Compare A. 1914 ii 52 205).-The fusion diagrams of the systems MnS0,-Li,SO MnS0,-nay so and MnS0,-K2S0 have been investigated. The first yields no compound between its constituents but the others give 3MnS04,Na2S0 MnS0,,3Na2S04 and 2&!tnSO,,I<,SO,. The Influence of Oxygen on Some Properties of Pure Iron. WESLEY AUSTIN (J. Iron Steel I ~ s f . 1915 92 157-163).-Alloys of iron apd oxygen are prepared by melting together iron and ferric oxide in a small arc furnace lined with magnesia. The product does not in any case contain more than 0.288% of oxygen. The A3 point is lowered by oxygen but the A2 point is un- changed. Pearlite is absent. Microscopical examination shows that the oxide is mechanically mixed appearing in globules. The alloys may be worked above 900° or between 750° and 850° but are brittle between those temperatures.The alloys are very resistant to atmospheric corrosion but are more readily attacked by a i d s than mild steel. ' Blue-brittleness ' occurs as high as TOOo C. H. D Sulphur in Malleable Cast Iron. R. H. SMITH ( J . Iron Steel Inst. 1915 92 141-156).-Sulphur is not removed from white iron in tlie ordinary annealing process whether the surrounding mixture be reducing oxidising or neutral. When oxidation takes place the sulphur tends t o diffuse into the unoxidised portions of the iron. Sulphur does not produce an injurious effect until its quantity exceeds 0.15%. The Occurrence and Influence of Nitrogen in Iron and Steel. N. TSCHISCHEVSKI (J. Iron Step1 Inst. 1915 92 47-105).-For the estimation of nitrogen in iron o r st'eel the metal is dissolved in hydrochloric acid the solution made alkaline with lime and the ammonia in the distillate estimated by means of Nessler's reagent or by titration with N/lOO-sulphuric acid.The addition of zinc which is sometimes recommended is t o be avoided. None of the nitrogen of tlie metal passes off as nitrogen oxides. A tin tube should be used f o r the distillation on account of the alkalinity of Iron in powder or filings reacts with dry ammonia a t 200° the maximum sate of absorption being a t 450° the compound formed being Fe,N. Carbon is practically without influence on the absorp- tion of nitrogen. Manganese reacts with ammonia a t 600-700° with considerable development of heat forming the compound Mn3N2 and finally MnN.Manganese also reacts with dry nitrogen but more slowly. Manganese nitride is soluble in molten iron. Silicon nitrides also dissolves in iron and as this compound is very stable i t remains in the residue when the steel is dissolved in acid f o r the purpose of analysis and so introduces an error into the C. H. D. glass.ii. 142 ABSTRACTS OF CHEMICAL PAPERS. estimation of nitrogen. Aluminium nitride passes into solid solution in iron. The cementation of iron by ammonia leads t o the formation of a brittle layer of nitride whilst the interior of the metal contains crystals of the same substance. Nitrogen hardens steel and decreases its ductility. Phosphorus in Iron and Steel. W. H. HATFIELD ( J . ITOIZ Steel Znst. 1915 92 122-140. Compare Stead A.1915 ii 778). -White irons containing about 2.9% of carbon are little changed in properties by the presence of phosphorus in quantities up to 0.20%. Above 0.25% free phosphide makes its appearance. When such iron is etched by means of Stead's copper reagent the varia- tions of deposition which are obtained may be due t o the irregular distribution of other elements besides phosphorus. C. H. D. C. H. D. The Carburiaation of Iron a t Low Temperatures. T. H. BYROM ( J . Iron Steel Inst. 1915 92 106-121).-MiId steel exposed for several years t o the action of blast-furnace gases a t about 50O0 becomes convezted almost completely into carbides analysis indicating a mixture of Fe,C and Fe,C. Electrolytic iron strips are readily carburised by carbon monoxide a t 550-600°.A layer of iron sulphide is also formed on the specimens exposed t o blast-furnace gas. I n the interior of the metal the carbide is formed at the boundaries of the crystal grains. The Influence of Heat Treatment on the Specific Resistance and Chemical Constitution of Carbon Steels. EDWARD D. CAMPBELL ( J . Iron Steel Inst. 1915 92 164-180).-The specific resistance of carbon steels increases with the temperature from which they are quenched. This is in accordance with the view t h a t the dissolved carbides dissociate with rising temperature ( J . Iron Steel Inst. 1899 ii 223). A steel containing 0.35% of carbon quenched from different temperatures shows a close parallel- ism bet'ween the specific resistance and the colour produced by the solution of the1 same specimens in cold nitric acid.When steels are regarded as solid solutions sufficient attention is not usually given t o the solute. The study of aqueous solutions is not regarded as that of the influences of various solutes on the properties of water but t'he properties of steel are commonly treated as modifica- tions of those of iron by the presence' of dissolved molecules. The term '' ionoids " is used t o denote the products of dissociation of the solutes in metallic solid solutions. The effect of the ionoids is to increase the specific resistance of the solvent metal. C. H. D. 0. LEHMANN (Jahrb. Min. 1915. ii 109-132).-A detailed summary of the C. H. D. Mixed Crystals of Iron Ammonium Chloride. iiterature and of the views previchsly expressed by the aithor.L. 5. S. Synthesis of Smaltite and Lollingite. A. BEUTELL and FR. LORENZ (Centr. Min. 1916 10-22).-Smaltite and lollingite (A,,INORGANIC CHEMISTRY. ii. 1413 1915 ii 633) after being heated in a vacuum until no further loss of arsenic occurred were the materials experimented on the ratio of (Co Ni Fe) As and of Fe :As being thereby reduced t o 1 0.95 and 1 0.88 respectively. Thesel and also pure cobalt were heated in an atmosphere of arsenic a t various temperatures f o r prolonged periods until equilibrium was attained. Using pure cobalt the following arsenides were obtained the range of temperature through which they are stable in arsenic vapour being indicated in brackets (415-430°) and CoAs3 (450-618O). The iron arsenides obtained under these conditions were FeAs (335-385O) Fe,As3 (395-415O) and FeAs (430-618O).The varying compositions of the minerals smaltite chloanthite skutterudite and lollingite are explained by the mixture of these various arsenides. L. J. S. COAS (275-335') CO~AS (345-365') COAS~ (385-405') CO~AS The Explosive Property of Uranyl Nitrate. ARNO WLLER (Chem. Zeit. 1916 40 38-39).-The author was unable) t o confirm the detonation or decrepitation of preparations of uranyl nitrate crystallised from ether as specific properties of t h a t substance. In only three out of twenty experiments was any detonation observed on mechanical handling of the preparations after moistening with water and in no case with the violence described by Ivanov (A. 1912 ii 455) o r Andrews (A. 1913 ii 60).The phenomenon was never observed in the case of preparations recrystallised from water or prepared in the absence of free nitric acid and the opinion is advanced t h a t i t may be due t o the decomposition of an unstable compound of a lower oxide of nitrogen with a uranium-ether additive compound. G. F. M. Densities of Solutions of Uranium Nitrate in Water i n Alcohol and in Some Acids. W. CECIISNER DE COXINCK ( B d I . SOC. chim. 1915 [iv] 17 422-424).-The8 author has determined the densities of solutions of uranium nitrate in water alcohol (85%) commercial methyl alcohol acetic acid (D 1-055) nitric acid (D 1.153) sulphuric acid (D 1.138) and hydrobromic acid (D 1-21> containing varying percentages of the salt. The results are tabu- lated in detail in the original.W. G. The Preparation of Metallic Vanadium. R. EDSON and D. MCINTOSH (Trana. Roy. SOC. Canada 1915 [iii] 9 81-83).-The method describemd whilst inapplicable for the preparation of vana- dium on a large scale may be used with ,advantage in the labora- tory 0.r f o r demonstration in the1 lectureroom. A flask containing some vanadyl chloride is fitted with a platinum wire filament attached t o two heavy copper leads and also with inlet and exit tubes for the passage of a current of dry hydrogen. The filame8ntl is made to glow by passing a current through it and the vanadium is deposited smoothly on the platinum as a silvery-grey coating. The experiment may be carried out either in a vacuum or in an atmosphere of hydrogen a t low pressure. As soon as the filamentii.144 ABSTRACTS OF CHEMICAL PAPERS. reaches a white heat the deposition begins and can be continued until the wire burns out. The vanadyl chloride is made by heating a mixture of the pent- oxide arid charcoal in a stream of chlorine and need not be purified from any vanadium tetrachloride i t contains. T. S. P. Antimony Pentachloride and Iodine. OTTO RUFF [with J. ZE~DNER and LEOPOLD HECHT] (Bey. 1915 48 2068-2076).-1n an earlier paper (A. 1909 ii 1023) it was meiitioned that antimony pentachloride reacts with iodine but not with bromine. The author hoped to make a complete study of this reaction but has had t o desist because of the effect of the vapours on the' respiratory organs. Recently however Moles (A. 1914 ii 812) has studied the depres- sion of the freezing point of antimony pentachloride when iodine iodine monochloride or iodine trichloride are dissolved in it and his results help t o elucidate the above reaction.Three main reactions proceed as indicated by the equations (I) SbC1 + 2 1 = SbC1 + 2IC1; (11) 2SbCl + 2 1 = SbC1,,2ICl+ SbCl ; (111) 3SbC1 41 = SbCl,,3IC1+ 2SbC1 + IC1. Equation I repre- sents the reaction which takes place when less than 1.5% of iodine is dissolved in the antimony pentachloride. Some iodine trichloride antimony tri-iodide and possibly a chloro-iodide of quinquevalent antimony are formed in addition but no evidence can be adduced of the formation of a compound SbClJ analogous t o SbF,I. The compounds SbC1,,2ICl and SbC15,31C1 may be isolated by sublimation from a mixture of antimony pentachloride (10 grams) and io'dine (4.3 o r 8.6 grams) under 15 mm.pressure from a bath a t 30-35O. They form bluish-black crystals stout prisms o r needles m. E. 62-63O which fume in the air and dissolve readily in carbon tetrachloridc o r chloroform but sparingly in antimony pentachloride. Temperature-concentration diagrams of mixtures of antimony pentachloride and iodine monochloride were also made in order t o confirm the existence of the above double compounds. I n the solutions they are strongly dissociated. J. C. W. Presence of Platinum in Spain. DOMINGO DE ORUETA and S. PIRA DE RUBIES (Gompt. rend. 1916 162 45-46. Compare this vol. ii 106).-The Ronda chain occurring in Andalusia between Malaga and Gibraltar is constituted by a series of rocks very similar in constitution and distribution t o those forming the Ural platiniferous beds.The central zone of peridotites is 72 kilometres long by 20 kilometres broad and is surrounded by other smaller ones. llilicroscopic examination further confirms the analogy between these rocks and those of the Urals particularly in the case of the dunite. The mean of fifty analyses gives the platinum content as 3 grams per cub. metre of platiniferous sand the varia- tion being from traces up t o 28 grams. IDE KOIFMANN. (Arch. Sci. phys. nut. 1915 [ivl 40 509-513).-The action of nitric acid on silver-platinum alloys was investigated ; the alloys W. G . Silver-Platinum Alloys and their Analysis.MI N ERALOGZCAL CHEMISTRY. ii. 145 used contained from 0,219 t o 5.162% of platinum and were p8re- pared by heating together mixtures of silver and platinum for twenty minutes a t a temperature above the fusing point of silver the mixtures being covered with a layer of pure sodium chloride. The preparation of the alloys confirmed the slight solubility of platinum in silver and i t was impossible t o make an alloy contain- ing much more than 5% of platinum. When these alloys were boiled wiZh nitric acid of varying concentrations a brown-coloured solu- tion was obtained together with a more or less voluminous black sediment and the latter always yielded a brown-coloured solution when treated with water. The author considers that silver and platinum in an alloy cannot be separated by means of nitric acid; part of the platinum forms a colloidal solution in the nitric acid and the remainder also passes into solution when treated with a sufficient quantity of water. w. P. s.