IN0 ROAN10 OEEM ISTRY. 9 I n o r g a n i c Chemistry. Ratio of the Atomic Weights of Hydrogen and Oxygen. By L. MEYER and I(. SEURERT (Ber., 1894, 27, 27;0-2773).-'l'homsen has recently (Abstr., 1894, ii, 277) calculated the ratio of the atomic weights of hydrogen and oxygen from the experimental equivalent- ratio of ammonia and hydrogen chloride. The authors point out that in the equation employed by Thomsen, H = -., in which r is the experiment,allg determined ratio of ammonia to hydrogen chloride, there is an uncertainty of 0.01 in the atomic weights of chlorine and nitrogen, which may produce an error of no less than 0.5 per ceut. in the irttio of hydrogen to oxygen, whereas, to be of value in deciding the question as to the true atomic ratio of these two elements, the possible error should not exceed 0-06.TC1 - N 3 - r A. H. A New Explosive Mixture. By A. AXGELI (Qazzetta., 1894, 24, ii, 52--58).--The action which occurs when a mixture of potassium nitrate, potassium carbonate and sulphur explodes is usually repre- sented by the equation 6KNO3 + 2K2CO3 + 5s = 6 N + 2C0, + 5K2SOA. On snbstitathg potassium nittitc for the nitratc, a mixture is ob- tained which explodes i n much the same way as the above. It is therefore probable that, during the fusion preceding thc explosion, potassium nitrite is formed by the reducing action of potassium snlphide, which, in turn, owes its formation to the interaction of the potassium carbonate and sulphur. A mixture of potassium nitrate with hypophosphite detonates in a similar manner, and potassium nitrite is found in the product obtained on cautiously melting potassium nitrate and hypophosphite together ; the same is true of mixtures of nitrates and thiocyanatcs.On heating R mixture of sodium thiocyanate with potassium nitrate, it f i ~ s t melts and then deflagrates slightly, whilst a mixture of the thiocyanate with potassium nitrite detonates violently. W. J . P. Hyposulphurous acid. By E. SCHAR (Ber., 1894,27,2714-2722). -Attention is drawn to the fact, hitherto overlooked, that Schoubein had prepared hyposulphurous (hydrosulphurous) acid and investigated10 ABSTRACTS OF CHEMIOAL PAPERS. many of its properties, several years bcfore Schiitzenbeiger occupied himself with the subject. Schonbein found that when an indigo solution has been bleached with hyposulphurous acid, the colour can bc restored, not oiily by oxidising, but also by reducing, agents, and even by heating in the absence of air.He thought, in consequence, that the original loss of colour must be due to the formation of an unstable compound of the colouring matter with the hyposulphurous acid. The author has repeated and exter_ded SchBnbein’s experi- ments, arid confirmed his results; be also is of opinion that the decolorisation of indigo by hyposulphurous acid cannot be explained satisfactorily a s a reduction of the colouring matter. Electrolysis of Nitrosylsulphuric acid in Sulphuric acid Solution. By A. GURC~IIAY (Zeit. nnorg. Chem., 1594, 7, 161-16ti). - The apparatus employed for the electroljvsis is so arranged that the elec- trodes arc separated by a narrow tube in order that diffusion may be prevented as much as possible. A current of 4 volts, and not more than 0.1 arnp&re, is used In the case of a solution containing 0*00316 gram of nitrous anhydride per 1 c.c., and using sulphurk acid of sp.gr. = 1.837, gas is evolved more quickly a t the positive than at the neg:itive pole, but the amount of gas evolved at tbc negative pole soon becomes grcnter than that a t the positive pole. The liquid then becomes violet, and, after 40 hours’ action, the evolution of gas at the negative pole ceases. As the electrolysis proceeds, the amount of nitric oxide evolved at the negative pole gradually diminishes ; even when the amount has become ve;:v small, the solution still contains nitrosylsulphuric mid.The positive ions of the nitrosylsulphuric acid, NO and H, are liberated a t the negatire pole, the negative ion, SOr, wanders to the positive pole and is split up into 0 and SO,; the latter then re-forms sulphuric acid. The oxygea oxidises a portion of the nitrosylsulphuric acid to nitric acid, the latter being reduced by the nitric oxide to nitrous acid, which then regenerates nitrosylsulphuric acid. Neither ammonia nor hydroxyl- amine are formed during the electrolysis. With a solution containing 0.00326 gram of nitrous anhydride per 1 c.c., and nsing sulphiiric acid of sp. gr. = 1.65, the electrolysis proceeds in a similar manner, but, after electrolysis, the residual acid contains no nitrous acid. J3. C. R. C. El. B. Metallic Double Salts of Diammonium and Diamide.By T. CITRTIUS and 3’. SCHRADER ( J . p . Chenz., 1894, [2], 50, 311-346).- Bgdrazine uniteswith either one or two equivalents of an acid to Porm salts. The compounds with two equivalents may be looked on as con- taining a bivalent radicle, diamrnonium, (N2H6)”, related to the metals of the alkaline earths in the same way as the ammonium radicle is related to the alkali metals. The salts containing one equivalent of acid are more stable than the others, and correspond with the v e r j stable hydrate, N2H4,H20. These salts form a number of crystalline double salts with the sulphates and chlorides of many of the metals, but do not appear to be capable of forming alums. The following salts hare been prepared.ISORGANIC CHEMISTRY. 11 1.Double snlphates. These are of the general formula (KZHj)2SOJ,RSOi, i n which R may be Cu, Co, Ni, Fe (ous), Bin, Zn, or Cd. These salts are all sparingly soluble i n water, and differ from the corresponding ammonium salts [such as (NHi),S0,NiS04 + tiH,O] by containing no water of crystallisntion. They are all obtained in the form of c.i*ystalline precipitates by niixing solutions of the two component salts. 2. Double chlorides. These are of the types N2HjCl,RCI, and Diammonium, i)zercuric ch7oride, 2N,H5CI,HgC12, is readily soluble, and may be recrystallised from water or alcohol. It forms well- ilereloped, six-sided pi-isms, which become opaque in the ah-, and melt at, 178" without decomposing. Dia?nnzoniuna cndrniiim clduride, N2H,C1,CdCI2, forms delicate needles, whilst the hi-salt, containing tn-o molecules of hydrazine hydrochloride, 2N2H5Cl,CdC12 + 4H20, separates in coarse prisms. Diummoniwn ziuc chloride, N2H6C1:ZaCl2, forms hygroscopic, six-sided prisms, melting between 180" aiid 185".The bi-sndt, 2N2H5CI,ZnCI,, er~-stallises in brittle, white, deliquescent needles, melting a t 135", and is difficult t o prepare pnrc. 'l'he two double salts with stannous chloride are difiicult to separate. The salt, N2H5C!,SnCI2. ciystallises in large, nacreous plates, melting a t 105"; the salt, 2N2H,Cl,SnC1,, is very hygroscopic, and iiielts at Hydrazine also forms a series of saIts analogous to tile ammonia compounds, as ammonionickel snlphate, NiS0,,6NH3, and aiiimonio- zinc chloride, Z n C 12,4NH3. When diammoniuni copper sulphate is dissolved in ammonia, thc cupric salt is reduced to a cuprous salt, and nitrogen is rapidly evolved.The double salts of iron, mangznese, mercury, and tin behave in a similar or analogous manner, x-hilst those of nickel, zinc, and cadmium form double salts, NiS0&lN2H,, ZnS04,0N2H,, ZnClt,2N2H4, and CdC12,2N2H, + H20, which may also be prepared by adding hydrazine bydrate to a solution of the metallic salt. It has hitherto been found impossible to prepare free hydrazine from any of these salts ; the double zinc chloride, for example, does not yield any reducing substance when heated, but evolves ammonia. Researches in this direction are being continued. Hydrazine hydrate reduces molybdenum trioxide to the dioxide, does not affect tungstic acid, and reduces chroniates and ferric salts to the corresponding lower oxides.Hydrazine hydrate in the pure state call be preserved without undergoing any alteration, whereas its dilute solution yapidly loses its characteristic properties, even when kept in a sealed tube. The cxact nature of the decomposition bas not yet been made out. Action of Nitric Oxide on Metals at High Temperatures. By F. EMICH (Monafsh., 1894, 15, 375-390 ; compare Abstr., 1892, 940 ; 18'33, ii, 66).-The author generally confirms the results obtained by Sabatier and Senderens (Abstr., 1892,1151 and 1271), and concludes 2NzH5C1,RCIz. 55-60'. A. H.12 ABSTRACTS OF CHEMICAL PAPERS. that the action of nitric: oxide on the metals closely resembles, but is much slower than, that of oxygen, provided the temperature of the former gas i s kept below f h e point a t which its decomposition com- mences.Exlierimelits with copper, zinc, cadmium, mercury, tin, lead, chromium, rri0l-y bdcntim, tungsten, uranium, manganese, iron, nickel, cobalt,, titauium, aid vtinadium show that oiily three of tbese metals, namely, copper, lead, and vanadium, give products (Cu,O, PbO, and Vd,O, respectively) diit'iret~t from those obtained on heating them in a current of oxygen. I n the case of mercury, the metal is not affected by nitric oxide a t or below the temperature at which it boils. G. T. If. Nitrous Anhydride. By G. LUNGE and G. PORSCHNEW (Zeit. rmorg. Chem., 1894, 7, 209-250) -'l'hese investigations were carried out in order to settle the question of the existence or non-existence of nitrous nnhydrsde in the gaseous state.All conceivable precautions were taken to o t h i n pure materials tor the investigation and to avoid sources of error, and rrrany of the experimertts made by previous investigators we1.e repeated with additional precantions. The composition of the nitiogc n oxides was determined hy nbgorbing them by sulphuric acid, and eiiher, a, estimating tllc total weight, and total nitrogen, or, b, estimating the total nitrogen, and the atiiount of oxygen (in the form of permariganate) needed for vomplei e oxidation. 1. Arti119i o j Oxygen on excess of A7itric Oxide.-Special precautions were taken to ensure thorough mixiiig of the gases. I n no case did any oxygen remain uncomlined, the whole quantity employed being fourid iii the absorbed nitrogen oxides.a Ni.:.turc: oj* Nitric O.eide and Peroxide towards Sul- phzwic aczd.-Nitric peroxidr i n well known to be readily absorbcd by strong snlphuric ncid, whilst nitric oxide is scarcely absorbed at all by that acid. I n mixtures, ititric oxide W ~ I S found to be readily absorbed when riot present in excess of the proi~ortion NO : NO: : : 1 : 1. The absorption 1Iy su1phui.i~ acid gives, therefore, no indication whether a gas is nitrous atrhydi*ide or a mixture of nitric oxide and peroxide in molecular proporti, n. Action of excess of Oxygen on Volutilising Nitrous Anhydride.- The results showed that when excess of oxjgen was present the whole of the nitrogen oxides (nitrous auhydride ?) was converted into peroxide. The results obtained by Lung+ which seemed to show a resistant power of the N 2 0 B molecule to further addition of free oxygen, were probably due to incomplete mixing of the reacting gases.Action of Nitric Oxide on iVitric Pei*oxide.-Above 28", no contrac- tion, and consequently 110 combination, of the two gases could be ob- served when they were mixed. At -21 and below, ni ric oxide is freely ahsorbed by liquid nitric peroxide with formation of a, blue liquid having the composition N&; mixtures of the two oxides, when coaled to --'Ll", yield the same blue liquid. These results are in contradiction to those of Hasfabach (J.pr. C'hern., 1871, [2!, 4, 1) and of Rihmsay (Trans., 1890, 590). LCawsny states that by passing nitric oxide through liquid nitric peroxide ouly about 3.5 per ueut. of the latter is absorbed.He estimated this absorption, however, by the gain in 2. UehnrivurINURQANlC CHEMISTRY. 13 weight of the peroxide, and did not allow for the loss of weight dire to nitric peroxitle carried amag hay the eswpitig nitric oxide. The blue liquid lias the exact composition N20.c, and when exposed in sealed tubes to the ordinary tenipevatnre the coloiir changes to preen -the colour of liquid nitric peroxide, From this, the nuthors coriclude that nitrous auhydride exists as a well-c.harac:tei.iseri chemical com- pound a t -21", but that at ordiiiary temI)erattrm (that is, above its boiling point) partial decomposition sets ill, even under pressure, and whilst it is still in the liquid state. Vapour Density lht riui,intiom of VolatiliGed hTitrous A?IJiy&ide.- Very careful detei*minations a t ordiiiary t c myteratures and in vacuous ressels failed to show any excess in density over diat of a mixture of nitric oxide and peroxide.The density was determined 111 the weight arid pressure it8 registered hy a manornt*ter formed of a Hempel burette partly filled with carbonic anhydride. Formation of Hydrogen Phosphide. By J. W. RETGERS (Zeif. umrg. Chrwz., 1894, 7, 26%---266).-Ba-etJ on some very old experi- ments of Fourraroy and Vauquelin, the text-books stnte that free hydrogen cannot I)e made to ct~rnt)ine directly wiih phosphorus. The author finds that i f a strmm of hydrogen is passed over heated amorphous phosphorns, gaseous hydi-opn phosphide is formed. The liquid aiid solid compounds are, a t the same time, produced in small quantities.L. T. T. L. T. T. Thiohypophosphates. By C. FRIEDEL (Compt. 1 . e d , 1894, 119, 260--264).-FVllt1n v:irious metds are heate.1 with the calculated pro- portion of phosphorus pentasulphidc), or, better, auiorphous phosphorus arid sulphur, thiohypc>phosphates arc obtained pure and in crystals. In some cases any excess of phosphorus sulphido can be removed by boiling with sodium hj-drox de solution, or by heating out of contact wit,h air, when the sulphide sublimes; butl piiritic*ation is always difficult, and the best results are only obtziined by using the calculated quantities of ma terinls. Iron t hioh ypoplz osphti te, Fr,P2S6, forms b ril lian t , grey i sh-bl a&, hexagon d l n ~ u e l l i ~ resemhliiig those of graphite or speculnr tieinatite.The latnellt~, when very thin, arc brown by traiisiritted light, and have no action on para1 lel polarised light. 'I'he cornpouiid is attacked bj- nitric acid, but more easily by it mixture of the acid with potassium chlorat J. The alurniniuin compound forms elongated, white larnellae, which act on poliirised light ; i t alters rapidly when exposed to air, anit is -decornl,osed IIY water with evoliition of hydrogen sulphide. The z n c s a l t is pale yellow, and is difficult to obtain in a state of purity. The copper salt, Cu2P2S6, forms small, acicri lar crystals, which act stronely on polarised light. The lead salt is purified by boiling for some time with water, arid then forms m i orauge-yellow, crystalline powder, which acts strongly on polarised light, and is not decomposed by water.The silcer salt is snlphur-yellow ; a t a, dull red heat, it burils with a flame like that of' phosphorus, and leaves a brown, brittle residue. The mercury salt, Hg2P,Ss, forms sulphur-jellow14 ABSTRACTS OF CHEXICAL PAPERS. lamella?, which act strongly on polarised light, and with rz convel-gent bean1 shows rings. This compound is slowly decomposed by boiling water, and more rapidly by potassium hydroxide solntion, a residue of mercuric sulphide bein? left. I n presence of an excess of phos- phorus sulphide, the compound can be sublimed, but it decomposes if strongly heated. Tin yields two compounds, Sn2P& and SnP2Ss, according to the proportion of tin employed ; the first is orange-yellow, is decomposed by boiling water with evolution of hydrogen sulphide, and disso 1ve.s completely in concentrated potassium hydroxide solution.The second compouud forms yellowish-brown radiating crystals, which alter when exposed to air, becoming yellow aid opaque, and giving off hydrogen sulphide. C. H. B. Sodium Pyrophosphates. By T. SALZER ( A ~ c h . Pharna., 1894, 232, 3 6 5 - 4 7 5 ) ,-‘li-isodium hydroyeiL pyrophosphatP, Na,HY20, + H20, is prepared by evaporating a solution of the tetrasodium salt (13.5 grams) and the disodium salt (10 grams) in water (50 grams) ; it forms small crjstals, is soluble in three parts of water a t the ordinary temperature, is neutral to litmus, and with silrer nitrate gives silver pyrophosphate, AgaP20,. A salt of similar composition, b u t containiug i H 2 0 , and crystallking in prisms, is formed when the disodium salt is emploged in slight excess, and the solution gently warmed.It mas only obtained in small quantity. Sodimt trihydrogm p.1/mphosphate, NaH3P207, mixed with metaphosphoric acid, was obtained from soda and pyrophosphoric acid. Attempts to prepare sodium tetraphosphate, Na6P401,, by heating trisodium hydrogen pyrophosphate were unsuccessful. A comparison of Fieitmann and Henneberg’s and Uelsmann’s sodium tetraphosphates shows that they are different ; the former is probably a mixture, i t crystallises in thin plates, and decomposes into disodium hydrogen phosphate and sodium dihydrogen phosphate in presence of water ; the latter can be readily reciytallised, is deposited in small needles, and is not decom- posed by the prolonged action of water.By G. KBSSNER (Arch. Phtxma., 1894, 232, 375--387).-Calcinm plumbate, ChPbO,, combines with water a t t(he ordinary temperature ; the mixture soli- difies, the colour changes from yellowish-red to pale yellow, and almost colourlees, microscopic, transpiiretit crystals of a hydrate containing 4H,C) are formed. With dilute nitric acid, the crystals turn brown, nncl finally become pulverulent. By the avtion of water at 150°, a yellowish-green, voluminous powder is formed; with acids at the ordinary temperature, it darkens, becoming finally black ; boiling dilute acetic acid rapidly produces the same change. Its coniposition agrees with that of a hydrate with 2H20, but it appears t o be a mixture of calcium hy- droxide (3 mols.) and a hydrogen calcium diplumbate, H2CnPb20G, resulting from the hydroljsis of the orthoplumbate ; the calcium hydroxide is extracted by washirig with water free from carbonic anhjdricle ; the residual diplumbate is yellow, and relatively When exposed to light, it gradually blackens.J. B. T. Orthoplumbates of the Alkaline Earths.INORGANIC CHEXISTRY. 15 stable to.cvarcls acids. On onc occasion, by the action of dilute nitric acid on the diplumbate, hydrogen calcium triplumbate, H,CaPb,O, was formed as a greyish-brown powder resembling the diplumbste in general properties. Both compounds decompose sud- clenly a t ,z definite temperature into oxygen and lead dioxide. Atten- tion is called to tbe analogy betxecn the highest oxidation products of lead, silicon, and carbon, which is emphasised by the discocery of these comples acids of leacl.Volatility of Mercuric Chloride. By H. ARCTOWSKI (Zeit. anorg. Cheni., 1894, 7, 167--175).-The author shows that mercuric chloride is decidedly vo1,ztdle at the ordinary temperature, and has also determined the volatility at various temperatures from 56" t o l%', keeping the other conditions constant. Taking the temperatures as abscissa, and the amounts of chloride volatiliaed as ordinates, a curve is obtained for the relatire volatility which is asymptotic to the temperature-axis. E. C. R. By L. 31. DENXIS and W. H. MAGEE (Zeit. mzorg. Clzern., 1894, 7, 250-264) .-The authors, having obtained con- siderable quantities of allanite, intend to caref ully irivestigzte the derivatives of cerium.For the separation of cerium from the other earths of this group, the best method is a modification of Debray's fusion with potassium nitrate (Abstr , 1883, 713). When employing potassium nitrate, the fusion point of the mixed nitrates is about 325", and at tbis temperature traces of diclymium nitrate are always decom- posed, so that the insoluble cerium oxide left after treatment of thc melt with water always contains didymium. The authors substitute a mixture of potassium and socliurn nitrates in molecular proportion for the potassium nitrate. Such R mixture melts a t 231' (Carnelly and Thornson, Trans., 1888, 792). By this means the melting point ~f the mixed nitrates was reduced below 230°, and the decomposition could be carried out a t 300".No didymium nitrate was then decom- posed, and the cerium oxide obtained after each fusion was free from impurity. After trial of the various qualitative tests for cerium, the authors recommend Cleve's (Abstr., 1885, 635) and Bois baudran's (ibid.) method of adding excess of ammonia, and then hydrogen peroxide, to the solution. The orange precipitate or coloration can be detccted with as little as 0.005 milligram of cerium oxide. In attempting to obtain a tetrachloride by saturating with dry chlorine, at a low temperature, a concentrated hydrochloric solution of cerium oxide, a chloride, CeCI3,7H2O, was obtained ; the same salt was formed when a stream of dry hydrogen chloyide was substituted for the chlorine. The salt forms orthorhombic crystals showing the axial ratios a : b : c = 0.80834 : 1 : 1.44187.It loses part of its water of crystallisation in a vacuum over dehydrating agents, but, is stable in the air. It appears to be different from the salt, 2CeC1,,15H20, already known. Oerous hjdroxide when pure is white, but when moist readily absorbs oxygen, becoming first purple and then yellow, the latter J. B. T. Cerium Compounds.16 ABSTRACTS OF OHEMICAL PAPERI'. coloured substance yielding, on ignition, ceric oxide. hydrate is undoubtedly in an intermediate stage of oxidation. The purple L. T. T. Nitrogen Compounds of Manganese. By 0. PRELINGEK (Monatsh., 1894, 15, 391-401) .-Pentanaangzizesc nitride, Mn5N2, and trimanganese nitride, &ln,N,, are obtained on passing nitrogen and ammonia respectively over finely-divided manganese, heated t o redness in a hard glass tube.The former has a dull, metallic lustre, and, when finely divided, is somewhat darker than powdered man- ganese. On heating i n a stream of hydrogen, ammonia is formed, whilst if hydrogen sulphide is substituted for the element, ammonium sulphide results. On treatment with ammonium chloride solution, ammonia, hydrogen, and the double chloride of manganese and ammonium are obtained, and on fusion with caustic alkali much ammonia is liberated. The compound, Mn,N,, forms a darker powder, but otherwise closely resembles the substance previously described. Ttie author points out that the constitution of the two compounds may be explained by tissuming the nitrogen to be pentavalent in the case of the pentnmanganese nitride, and trivalent in that of the trimanganese nitride. G.T. M. Manganese Steel. By H. LE CHATELIEE (Con@. ?*end., 1894, 119, 272--874).-T he manganese steel discovered by Hadfield is practi- cally non-magnetic, and has a higher resistance than any other alloy of iron. Moreover, i t is more malleable, the more highly it has been tempered. Hadfield has found that an allotropic modification, whicli is magnetic, can be obtained by heating the ordinary manganese steel at a high temperature for several days. The author finds that the change from non-magnetic to magnetic metal takes place between 500" and 650", and, at the most favourable temperature, 550°, the change is complete in an hour or two. I n order to convert the magnetic metal to the non-magnetic, it mast be heated a t a temperature not lower than 800", and must be cooled s9mewhat rapidly in order t o prevent the reverse change between 500" and 600'.Since the rate of this latter change is very IOIT-, ordinary cooling in air is usually sufficient. The elecrrical resistance of a wire, 1 min. long and 1.4 mni. di- ameter, is as follows. Temperature.. . . . . , . 15' 90" 300" 500" 635" 730" 850' 965" 1020" non-magnetic 1-06 1-19 1-44 1.65 - - 1.88 - 1.9i magnetic.. .. 0.88 099 1.2'7 1'50 1.7 1.79 - 1.93 1.97 Resistance { The two curves touch a t 740", and coincide above this temperature, and hence this is the temperature of transformation of the two varieties of metal. This temperature is, however, alsc! that at which soft iron becomes non-magnetic.It would seem, therefore, that manganese steel i s a mixture of iron with a dcfinite compound of iron and manganese, the latter being non-magnetic under any con- ditions. The iron undergoes the normal changes, hut they take place more slowly, in con5equence of thc presence of the manganese compound.INORQANIO CEEMISTR Y. 17 The expansion of the two varieties of manganese steel is the same, and hence it would follow that there is no change of dimensions a t the point of transformation. The following table gives the expansion of a rod 100 mm. long. Temperature.. 230’ 500’ 680’’ 830” 990” 1060’. Expmsion.. . . 0-35 0.67 1 0 5 1.43 1.97 2.09 mm. Manganese steel tempered in water shows, at once, when annealed, C. H. B.a definite contraction of 0.4 mm. on 100 mm. Chemical Behaviour of Arsenical Pyrites. By T. M. LIGHTFOOT ( J . Amer. Chewz. Soc., 1894,16, 624--633).-The author has tried the action of solutions of potassium permanganate, of different styengths and at different temperatures, on arsenical pyrites, aud has tabulated the results. Potassium permanganate partially oxides the sulphur, b u t even a t the boiling heat nothing like a complete oxidation takes place. The sulphur oxidised was estimated in the filtrate by the usual process. Hydrogen chloride a t a hiqh temperature does not expel large quantities of sulphur, but removes considerable amounts of arsenic. I n these experiments, the sulphur left in the boat was oxidised in the ordinarv way, and deducted from the total sulphur, the differenca being the sulphur volatilised.So!ution of copper snlpliate under pressure dissolved but little ferrous iron, which goes to prove that the mineral contains the iron in the ferric state. A little green copper arsenate was also formed in. the reaction. L. DE K. Action of Ferric Sulphate on Potassium Iodide and Hy- driodic acid, By K. SEUBERT and R. ROHRER (Zeit. anorg. Chew,., 1894, 7,137-153).-The action between ferric sulphate and potassium and hydrogen iodides takes place in a manner very similar to that de- scribed by Seubert and Dorrer for ferric chloride (Abstdr., 1894, ii, 191). The reaction with one equivaIent of ferric sulphate, +Fe,(SOJ,, and varying proportions of potassium iodide, approaches the theoretical when 20 mols. of potassium iodide are present.The yield is then 97 per cent., and any further increase in the proportion of potassium iodide pro- duces no increase in the amount of free iodine. When the results R I e compared with those obtained with ferric chloride, it is evident that the action is much slower, and f o r small excesses of potassium iodide is nerer so complete. The results obtained with the propor- tion FeC13 : K I correspond with those obtained with the proportioil Fe2(S04)3 : 4KI. When the action is allowed to procesd for 18 hours, using an equivalent of ferric sulphate, and varying proportions ot potassium iodide, the amount of iodine liberated is a t first much less than the amount liberated when ferric chloride is used, but the difference becomes small for 6 mols.KI, and disappears for 10 mols. With 1 mol. of potassium iodide and varying proportions of ferric sulphste, the action is also much slower than with ferric chloride ; and the numbers obtained with the proportions 2KI : 4Fe,(S04j, VOL. Lxvm. ii. 318 ABSTRACTS OF OHEMICAL PBPERS, and 2KT : 10Fe(S0J3 agree with those obtained for the proportions KI : FeC1, and KI : SFeCI, respectirely. The action between ferric sulphate and hydrogen iodide is also slower and less complete than that between ferric chloride aucl hydrogen iodide. It is evident that tlhe reaction Pez(SOa)3 + 2KI = 2FeS04 + K,(SO,) + I, is a reversible one wlicn none of the products are removed from the sphere of action. And experimcnts with mixtures containing various inolecular proportions of ferrous sulphate potas- sium iodide, iodine, and potassium snlphate, after rernaininz 144 hours, a.lways contained an amount of iodine corresponding with that liberated by an equivalent mixture of ferric snlphate and potassium iodide. E.C. R. Attempt to prepare Nlolybenum Hexachloride. By E. F. SMITH and H. C. BURR (J. Amer. Chem. SOC., 1894, 16, 577--578).- The authors thought that if a partially chlorinated liiolybdic acid were exposed to chlorinating agents, the residual oxygen might be re- moved, and that then perhaps the hexachloridc might be obtained as a final product. To this end, molybdenyl chloride, MoO~CI,~ was mixed with an equivalent quantity of phosphorus pentachloride, End heated in a sealed tube, in the rresence of chlorine, at 170". Oii cooling, it was found that the tube contained a, mass of greenish-black crystals ; these, when freed from phosphorus oxychloride by distillation in a cnrrent of chlorine, yielded it product which, on analysis, gave figures approximating more to the pentachloride than to the hemchloride.The experiment was repeated, silicon tetrachloride being substitiited for the phosphorus compound. The crystalline product was quickly removed to a boat, nncl distilled in an atmosphere of carbonic anhy- dride. The brown vaponrs condensed in t.hc colder portions of the tube, and on analysis gave results varying between those required by the tetrachloride and pentachloride. The authors have not, however, given up all hope of preparing the hexachloridc. They h a w observed that metallic molybclennm rapidly acts on ferric chloride solution with reduction of the ferric salt, and dissolves as a hexad.Further esperiiiients will bc made t o pet it in the solid state. L. DE K. Salts of Sulphomolybdic acid. By A. R,OSESIIEIX ( Z p i f . ~w-,J. C72ena., 1894, 7, l'iCi-l84).-The salts described in this communication have been previously prepared by P6chard (Abstr., 1893, ii, 530). The author confirms the results previoiisly obtained as to their pro- perties ; but his analytical results d~ not agree with the composition of the salts assigned to them by Pkchard. The ammoiiiuiii salt has the composition :3(NH,)20,2S02,8MO;$ + 5H20, the potassium salt is 4K,O,4S0,,9MO3 + 5H20, whilst the sodium salt has the composi- tion 9Na,O,8SO2,20MO3 + 37HzO. E. C. R. Complex Inorganic Acids.By V. ALTISI (Gnzcetia, 1894, 24, i, 523).-oii warming phosphomolybdic 3cid with hydrofluoric acid solu- tion, the yellow colour disappears, arid the solution 110 longer gives a,INORGANIC C HEMISTRY. 19 precipitate with ammonium nitrate ; on concentration, the yellow colour reappears, and ammonium nitrate then precipitates the salt Mo0,,2NH,F. Ammonium phosphomolybdate, when warmed with concentrated hydrofluoric acid, yields nionemmonium fluoroxymolybdate (compare l\lllauro, Abstr., 1891, IS). Neutral potassium fluoroxymolybdate is deposited on mixing solu- tions of potassium hydrogen fluoride and pbosphomol ybdic acid. Ammonium fluoride acts on ammonium phosphomolybdate with for- mation of the salt i\lo02,NH4F, and on phosphomolybdic acid, giving the salt MoOs,3NH4F and normal ammonium fluoroxymolybdate I\/lo0,F,,2NHiP.W. J. P. A New Oxysulphide of Tin. By F. W. SCHMIDT (Be?.., 1894,27, 27:39-2i43).-When stannic sulphide, obtained in the usual way by the action of hydrogen sulphide on a solution of stannic chloride, is allowed to remain in contact with ammonia, i t dissolves, forming a solution, wliich, on acidification, yields a wry voluminous white pre- cipitate. The same substance may be obtained by digesting stannic snlphide with ammonium carbonate solution, filtering, and acidifying ; about 10 per cent. of the stannic sulphide is dissolved. This compound when freed from sulphur by treatment with carbon bisulphide, has the composition Sn,S,O + llHE,O, is easily and com- pletely soluble in ammonium carbonate, and is also slowly soluble in water, When dried, the white mass gradually loses its property of dissolring in ammonium carbonate solution, and becomes amber- yellow, but, even after preservation for a year, this -pllow mass dissolves readily in ammonia, and the solution thus obtained deposits the white compound when excess of acid is added. The partial solubility of staniiic sulphide in ammonium carbonate solution is of great analytical importance, as this reagent is often used to separate the sulphides of tin and arsenic. A. H. New Method of Extracting Gold from Auriferous Ores by means of Bromine. By C. LOSSES ( B e y . , 1S94, 27, 2726-2727).- Hitherto bromine has not been used for this purpose, as i t could not be readily recovered, and the cost of the process was consequently too great. The author electrolgses a solution of potassium bromide, and obtains an alkaline solution which contains hypobromite and bromste, and which is capable of dissolving gold. The ore is treated with excess of this solution in rotating cylinders, the solution is then filtciwd, the gold is precipitated by passage over a mixture of iron and coal, and the solution, which now contains mainly potassium bromide, is clectrolysed and again used for extraction. c. F. €3. 3-2