14 ABSTRACTS 011’ CHEMICAL PAPERS.In o r g a n i c C h e mi s try.On the Supposed Compound NH,. By COMBES (Cornpt. rend.,94, 1717).-Reference is made to a communication from Mnumene(Cornpt. re.nd., ISSS), in which he claims to have produced a new sub-stance, N&, by the reactmion of potassium permanganate on am-monium oxalate. As the existence of NH, or rather of N2H1, is theo-retically probable, the author has repeaked the experiments, with thefollowing results :-The carbonate mentioned by Maumene, treatedwith hydrochloric acid and p l a h u m chloride, givcs a crystallineprecipitate identical in form with ammonium platinochloride, andcontaining 1.9 per cent. of hydrogen (NH,Cl,PtCld = 1.8 per cent. H),whilst Maurnen6 obtained only 1.35. The aqueous solution of thesupposed new body, saturated with hydrochloric acid, Fields crystalsidentical in form with ammonium chloride, and this is confirmed bythe analytical results.The reaction, therefore, yields only ammoniaand carbonic acid. R. RINORGANIC CHERIISTRT. 15Silicon. By P. SCH~~TZENBERGER and A . COLSON (C’owpf. Tend., 94,1710--1713).--Platinum foil, heated to a reddish-white heat amidst amass of lamp-black, is found to contain silicon, which must have beencarried through the lamp-black from the crucible. Other experimentsdetailed in the paper show that the silicon reaches the platinum in thevaporous form, and that nitrogen, and probably also oxygen, play apart in the transference of the silicon, a s well as in the formation of thecarbo-silicic compounds which the authors have previously described.R.R.Compounds of Silicon with Sulphur. By P. SABATIER (BuZl.Xoc. Chirn. [2], 38, 153--154).-When dry hydrogen sulphide ispassed over crystalline silicon a t a red heat, a violent reaction occurs,and at t>he cooled part of the tube a ring of a reddish substance is ob-tained, in which fine white needles of silicon disulphide, SiS,, are found.Beyond the ring, the tube is covered with an orange-yellow powder,which is given off in fumes during the course of the reaction. TheVellow and brown substances seem to be identical with those obtainedby Colson (vide following Abstract). The latter has a variable composi-tion, and is probably a mixture of the disulphide with amorphous siliconor a subsulphide ; on treatment with water, if, gives off hydrogen sul-phide and leaves a brown residue.In the tube, there is alwayspresent a deposit of crystalline silicon, which the author explainsby supposing the formation a t the high temperature of a volatile sub-sulphide, which a t the lower temperature is decomposed into disulphideand silicon. The yellow substance seems to consist for the greaterpart of the disulphide contaminated with a certain quantity of thesubsulphide, to which the author attributes a probable formula Si,S4.V. H. V.Combination of Tetratomic Elements. By A. COLSON (Bu77.Xoc. Chirn. [2], 38, 56-60, and Conzpt. rend., 94, 1526--1528).-1fa current of ethylene or hydrogen saturated with benzene is passedover silicon contained in a porcelain tube surrounded by a clay jacket-ing tube which is heated in a reverberatory furnace, a carbosilicide ofthe formula SiCO, is obtained ; the necessary oxygen is derived fromthe silica of the tube.This compound is a whitish powder, and is unat-tacked by acids, chlorine, or oxygen a t a red heat. It is decomposedby fused potash, or a mixture of litharge and lead chromate. On sub-stituting carbonic anhydride for ethylene, a compound of the formulaSi,C,30 is formed, with liberation of carbonic oxide. The authorexplains the fact that an oxygen-containing compound like carbonicanhydride yields a less oxygenated product than ethylene, by suppos-ing a. simultaneous loss of oxygen of the silica and carbonic anhydride,analogous to the simultaneous loss of hydrogen when benzene andmethane are passed through a red-hot tube.On heating pulveriseilsilicon in a carbon crucible surrounded by a t.itaniferous (carbon andrutile) jacket, and heated to a white heat, a compound of formulzSi2C30, was obtained.If vapour of carbon bisulphide is passed over silicon a t a whiteheat, two componncls arc formed, the one a yellow volatile compoundof the formula SiS, the other a yellowish substance of probable conipo16 ABSTRACTS OF CHEMICAL PAPERS.sition SiSO. Both these substances are decomposed by water or dilutealkalis, with evolution of hydrogen. If the contents of the tnbz arefurther heated with a boiling soliition of potash to remove the excessof silicon and its sulphur compounds, and then digested for some timewith warm hydrofluoric acid, a greenish powder of the compositionSi4C4S is obtained: when heated in a current of oxygen this doesnot alter in weight, but is converted into an oxygen&gd compound,Si4C402.- I The author draws attention to the fact that the analogy of sulphurand oxygen does not hold good. a t high temperatures, for CO, yieldsSi4C402, but CS2 yields SiS and Si4C4S. V. H. V.Extraction of Selenium from a Waste Product. By P.KIENLEN (Bull. SOC. Chinz. [2], 3 7, 440-443).-The selenious anhy-dride produced by the combustion of seleniferous pyrites is reducedby sulphurous anhydride in the Glover tower to the state of selenium,which partly dissolves in the acid, partly remains in suspension.At aworks where pyrites from Sain Bel, near Lyons, are used, the amountof selenium present in the acid is often sufficient to impart to it adistinct blood-red tint.The amount of selenium in the sulphuric acid may be estimated bydiluting a considerable quantity of the acid with three times its bulkof water, and leaving it, in a warm place for a long time. The clearliquid is then decanted or siphoned off, the selenium collected on aweighed filter, washed, and dried a t 100". Glover tower acid of sp. gr.1.606 was found to cont'ain 28.3 mgrms. of selenium per litre, or 17.6mgrms. per 1000 grams, whilst chamber acid of sp. gr. 1-532 con-tained 34.2 mgrms. per litre, or 22.3 mgrms. per 1000 grams.When the sulphuric acid containing selenium is used for the manu-facture of salt-cake, the selenium volatilises along with the hydro-chloric acid, and is deposited in the first condensers, sometimes insuch quantity that it imparts a red fluorescence to the acid.It is thedeposit in these condensers which constitutes the new source of sele-nium. This deposit forms a brick-red mud, which becomes black ondrying. When dried a t 100" it contains from 41 to 45 per cent. ofselenium. The selenium is estimated by suspending 20 grams of thedried mud in water in a flask with a long neck, adding soda tofeeble alkaline reaction, and then adding bromine drop by drop withcontinual agitation. After some time the liquid is filtered, the filtratemixed with the washings, boiled with a little hydrochloric acid, andthe selenium precipitated by sulphurous acid.I n order to extract selenium from the deposit, it is suspended inwater, and treated with a current of chlorine in large Woolf's bottles.The selenium is converted into tetrachloride, and this is decomposedby the water, yielding selenious acid, which is partially oxidised toselenic acid.As fioon as the brick-red tint in the first vessel has dis-nppeared, the vessel is removed, and the second vessel put in its place,another vessel containing fresh mud being put on at the end. Thedark-coloured liquid thus obtained contains selenious, selenic, andhydrochloric acids. It is filtered through cloth, and boiled withexcess of hydrochloric acid, which reduces the selenic acid to seleINORGANIC CHEMISTRY.17nious acid, then diluted to its original volume, and the selenium pre-cipitated by adding sodium hydrogen sulphite until the liquid smellsstrongly of sulphurous anhydride. The selenium is deposited in largered flakes, which agglomerate t,o a pitchy mass with a bronze lustre.The liquid is boiled by passing in steam, when the precipitate rapidlyagglomerates and contracts, forming a spongy steel-grey mass, whichis then washed, dried, fused in a clay muffle, and cooled under wateror in glass moulds. By this method large quantitie8 of selenium canbe easily and rapidly obtained in a state of considerable purity.C. H. B.Boiling Point of Selenium. By L. TROOST (Compt. rend., 94,1508-1 510).-The author finds that the boiling point of seleniumnnder 760 mm.pressure is 665", and he suggests the use of boilingselenium as a means of maintaining a constant temperature for thedetermination of vapour-densities, &c. R. R.Coefficient of Expansion of Sodium Sulphate Solutions.By W. W. J. NICOL (Ber., 15, 1931--1932j.-On the assumption thatin sodium sulphate solutions the salt is in the anhydrous conditionabove 33-34', and hydrated below Chis temperature, the authorthought i t probable that a solution of this salt would show a suddenlyincreased or diminished coefficient of expansion a t about this tempera-ture. He has examined solutions of different strengths between 20"and 40', and has found that the coefficient of expansion gradually in-creases with rise of temperature up to 34-36', when it suddenlydiminishes.It increases again with further rise of temperature.A. K. M.Chloride of Lime '' and Chloride of Lithia." By K. KRAUT(AnnuZen, 214, 254-36O).-When chlorine is passed over moistlithia, a mixture of lithium chloride and hypochlorite is produced, buthalf of the lithia present takes no part in the reaction-4LiOH + 2C1 = LiOCl + LiCl + H20 + 2LiOH.When exposed to the action of carbonic acid, the hypochlorite is de-composed, and the hypochlorous acid which is set free acts on thechloride, and chlorine is evolved. A similar reaction takes placewhen a mixture of basic calcium chloride and calcium hypochlorite issubmitted t o the action of carbonic acid.As it mould not be possible for a monad metal, such as lithium, toform a compound having a composition analogous to ClCaOCl, theauthor concludes that Odling's formula, for bleaching powder is incor-rect.w. c. w.Calcium Hypoiodite. By 0. LUNGE and R. SCHOCH (Bw., 15,1883- 1888) .-The bypoiodites are generally stated to be highly un-sfable compounds, of which, however, little is known, as they have notbeen isolated.By the action of iodine on lime suspended in water (several hoursbeing allowed to complete the reaction), the authors have obtained acolourless solution, which has an odour of iodoform, and gives theVOL. XLIV. 1s ABSTRACTS OF CHEMICAL PAPERS.following reactions :-Addition of acid produces immediate separationof iodine ; solution of starch gives no coloration ; hydrogen peroxidein acid solution produces turbidity and abundant evolution of oxygen ;cobaltous nitrate gives a green-coloured precipitate ; coal-tar coloursare not affected, whilst cochineal, logwood, litmus, &c..are bleached,From these results, and more especially from the bleaching power ofthe solution, the authors conclude that by the action of iodine on limeat the ordinary temperature the compound CaOI, = Ca(OT), + CaIzis produced, besides calcium iodide and iodate. From quantitativeexperiments on the bleaching power of this iodide of lime, they showthab it is much more stable than the alkaline hypoiodites are generallysupposed to be. It decomposes slowlyin the dark, more rapidly whenexposed to sunlight, and by boiling for many hours is decomposed onlyto the extent of one-half.A. K. M.Didymium. By P. T. C L ~ E (Compt. rend., 94, 1528-1530)-The author has for several years suspected the presence of a newelement accompanying didy mium, and he has recently by fractionalprecipitation and decomposition separated a portion, the spectrum ofwhich, besides the known lines of didymium and lanthanum, gave newlines, and amongst these a very strong one of wave-length = 4333.5.This line was previously observed by Thalhn in 1868, in a mixture oflanthanum and didymium, but was absent from the spectra of samplesof lanthanum and of didymium prepared by the author in 1874.The atomic weight of the first fraction precipitated by potassiumsulphate was 146; that of the last fraction, 142. Without namingtbe new element, the author proposes to designate it by the symbolDip’.The author intends to continue his researches on didymium.R.RDidymium. By B. BRAUNER (Compt. rend., 94,1718-1719).-1nthis paper the author does not claim priority over ClBve (precedingAbstract), but merely asserts that his observations are independent,and were announced in the Anzeiger dei. Acnd. Wissenschaft in W i e n of6th October, 1881, and 9th June, 1882. He found that lanthanumsulphnte may by repeated crystallisations be divided into two frac-tions, the more basic having an atomic weight = 138.3, and the lessbasic = 140.2. By repeatedly treating didymium free from oxide oflanthanum with ammonium nitrate, the author obhined an earthhaving the atomic wclight, of 140.6, the atomic weight of the remainingdidymium being 142.5 ; but by repeated precipitations n product wasobtained with an atomic weight of 146.6.In the spark spectra of thedifferent fractions, raye were found belonging to none of the knowncerite metals. These phenomena are doubtless due to the fourthelement designated Dip by ClBve. The author succeeded in sepa-rating another earhh of a higher atomic weight than 145.4. Ordinarydidymium appears to be a mixture of a t least three elements. Oneis true didymiuni (Di = 145.4) ; another (the Dip of ClBve) is a morebasic metal, and has an atomic weight of about 141; the tbird, of ELhigher atomic weight, is less basic than didymium. R. RISORGANIC CHEMISTRY. 19Explosive Alloys of Zinc with Certain Platinum Metals.By H.SAINTE-CLAIRE DEVTLLE and H. DEBRAY (Comnpt. rend., 94,1557-1560) .-Oxide of iridium is projected into fused zinc, themass is kept in fusion for six hours, and the cooled ingot treated withhydrochloric acid to remove the excess of zinc, &c. When thegraphite-like residue, washed and dried at lOU", is heated to 300" itinstantly takes fire, almost explosively, giving off fumes of zinc and ofosmic acid. This deflagration occurs also in a vacuum, but naturally,without production of zinc oxide or of osmic acid. At 300", there istherefore a change of state attended by great development of heat,which in the air occasions combustion. This phenomenon is somarked that by its means 1 or 2 per cent. of iridium may be detectedin platinum.Ruthenium and rhodium produce similar effects.R. R.Action of Aluminium on Cupric Chloride. By D. TOIIIIASI(BUZZ. SOC. C'him. [a], 37, 4&3-445).-Aluminium acts rapidly, evena t ordinary temperatures, on a solution of cupric chloride, with libera-tion of hydrogen and copper, and formation of an aluminium oxychlo-ride, the composition of which depends on the concentration of thecopper solution. With a 31.25 per cent. solution of cupric chloride,the aluminium oxychloride had the composition 2A12H606,3A12C16,and with a 7-81 per cent. solution, the composition Al~H,0s,4Al,Cl,.These oxychlorides are easily- decomposed and will not crystallise :Tbey are not true compounds, but variable mixtures of aluminiumchloride and oxychloride.The action of metallic aluminium on theseoxychlorides yields as a final product the compoundTo obtain this compound, a 31.25 per cent. solution of cupric chlorideis treated with alnminium until all the copper is precipitated ; theliquid is filtered, the filtrate heated, and aluminium added in succes-sive small quantities unbil it ceases to dissolve, water being added fromtime to time to make up for loss by evaporation; The clear liquid isthen evaporated to a syrup, and finally dried a t 40-50". In this waythe oxychloride is obtained in white flakes resembling those of potas-sium boro-tartrate. A solution of this oxychloride, like that of ferricoxychloride, is precipitated by sulphuric acid and by certain salts, suchas the sulphates of sodium, ammonium, potassium, magnesium, zinc,copper, and iron ; but it is not precipitated, evec on boiling, by thechlorides of potassium, ammonium, sodium, copper, or barium, bypotassium iodide, potassium bromide, ammonium nitrate, o r potassiumnitrate.The aluminium hydroxide thrown down is but slightlysoluble in sulphuric acid, and appears to be an isomeric modification,AIZCl,,GAI~H,O6 + I2Hz0.probably the modification 6, described by the author (Conzpt. yenti.,1880). C. H. F.Stability of Cupric Hydroxide. By D. TOMMASI (UZCZZ. SOC.Chim. [el, 37, 197-202).--Cupric hydroxide, perfectly free fromoxide, can be obtained only by using very dilute solutions of coppersulphate and sodium hydroxide, the precipitation being effected a t 0".The author has determined the influence of the presence of variousc 20 ABSTRACTS OF CHEMICAL PA4PERS.salts on the dehydration O F the cnpric hydroxide. In contact withdistilled water at 6-8", cu pric hydroxide undergoes sensible dehydra-tion after 120 hours.A sensible amount of dehydration takes placein 24 hours in presence of sodium hydroxide, and is greater the moredilute the soda solution. It is most marked with a 0.2 per cent. solu-tion, but with a 10 per cent. solution the hydroxide remains blue, evenafter 48 hours; it becomes black, however, after 96 hours. Inpresence of sodium acetate, carbonate, or sulphate, dehydration takesplace more slowly, and, in presence of calcium chloride, sugar, man-ganese sulphate, or potassium chlorate, no dehydration is perceptible,even after a long time.Halojid salts of the alkalis appear to formsmall quantities of oxyhdoid copper compounds. The presence ofsmall quantities of certain substances a1 together prevents the dehydra-tion of the copper hydroxide; the presence of 0.3 per cent. man-ganese snlphate, for example, prevents dehydration, even at 100".Cupric hydroxide added to a solution of nickel sulphate is convertedinto an apple-green precipitate which contains both copper and nickel,probably in the form of a double basic sulphate. No copper passesinto solution. When the hydroxide is added to a solution of leadnitrate, the copper displaces a portion of the lead, which is precipitatedas hydroxide, whilst the copper passes into solution.The influence of different salts on the temperature of dehydrationis shown by the following table :-Salt.Na2C03........KC1.. ........NaHO ........NaC2H302 ....NazSOa ......NaHO ........NaHO. .......KBr ..........KClO s . . . . . . . .KI ..........HZO ..........Strength ofsolution.5 per cent.10 ,,10 9,10 9,ITemperature ofdehydration.50"71747778798384858586 ........ .. .. No dehydration, ...... even at 100".C. H. B.Transformations of Cuprosocupric Sulphites. By A. ~ T A R D(Compt. rend., 94, 1475--1477).-The formula for the precipitat,e,obtained by the addition of an insufficient quantity of sulphurous acidor sodium hydrogen sulphite to a solution of cupric acetate, was givenby Pkan as SO~Cuz,S03Cu,5Hz0. The author's analyses of this saltlead him to assign to it the composition S8027Cu"lo(Cu2) + 26H,O,wit.h the following rational formula :-after a type already adopted by him, and he proposes to call the sub-stance acid cuproso-cupric octosulphite........... CaCl, 10MnS04.. 10Sugar 10S8032C~'Z,~u''z~C~''~(HB)H1 + 21H20INORGANIC CEEMISTRY. 2.1By the action of sulphuroun acid, the abwe salt is transformed intoChevreul’s salt, SO3Cu2SO3,’LH20, and by the action of sodiumhydrogen sulphite into a yellow salt already described by the authoras acid octosulphite of cz~proswz, cupricum, and sodium (ibid., 94,1422).The reaction is represented thus :-Ss03~Cu’zCu‘’2Cu’’8( He) He,21H20 + S80zrNasH8 = S20sH8 +8H2 0 + SSeparation of Gallium.By L. DE BOISBATJDRAN (Compt. rend.,94, 1439-1442 ; 1625-1629) .-Separ ation from Gluciwm.-T hegallium is precipitated by potassium fersocyanide from a solutioncontaining hydrochloric .acid in large excess, or it may be throwndown along with arsenious sulphide.Xeparatioiz from Cerium, Lant haw urn, Didy mium, Xamar him, Ytt r izcm,Holmium, and Thu/ium.-These earths may be precipitated by potas-sium hydroxide i n considerable excess a t the boiling temperature, andthe gallium separated from the alkaline solution by means of cuprichydroxide or by addition of ammonia and long boiling after previousneutralisattion with hydrochloric acid. Gallium may also be separatedfrom the above-named metals by precipitating it with potassium ferro-cyanide from solutions containing excess of hydrochloric acid.Gal-lium is carried down when arsenious acid is precipitated by hydrogensulphide.Separation, from Iron.-This is effected by a boiling solution of potas-sium hydroxide, but as the iron oxide carries down with it a littlegallium, it must be re-dissolved and re-precipitated four or five times.When the quantity of iron present is relatively large, it is preferableto reduce the ferric salt with metallic copper, add a small excess ofcuprous oxide, and, after repeating this operation three or four times,to pass hydrogen sulphide through the last strongly acid hydrochloricacid solution. The remainder of the iron is then eliminated by twoor three treatments with boiling potassium hydroxide.Separation f ronz Thorium.-The methods with potassium hydroxide,with potassinm ferrocyanide, and with arsenious sulphide are allapplicable in this case.h’eparutioiz f rona Zirconium.-This may be effected either by boilingwith potassium hydroxide or by arsenious sulphide, but not by potas-sium ferrocyanide, because the latter precipitates very acid and dilutesolutions of zirconium.Separation from 1Manganese.-For this nine processes are giacn.That by potassium hydroxide is applicable, but it must be severaltimes repeated, and has no advantages in the presence of much man-ganese.Barium carbonate or calcium carbonate separates gallium i nthe cold after some hours, leaving manganous chloride in solution.Very good separations may be obtained by arsenious sulphide, also bycupric hydroxide used hot.The reaction with potassium ferrocyanidemay be used, but with special modifications, of which it long anddetailed account is given in the paper.Separation fq-orn Zinc.-The method with copper hydroxide com-pletely separates gallium from zinc. Barium or calcium carbonate32 C u’* C: u’’~ CU”~ ( Naa ) Hz , S 0 ,4Hz4, 5 H20.R. R22 ABSTRACTS OF CHEMICAL PAPERS.precipitates gallium, but considerable quantities of zinc are carrieddown with it. R. R.Action of Ammonium Sulphide on Stannous Sulphide. ByH. BAUBIGNY (Compt. rerid., 94, 1473--1475).-Stannous sulphide isquite insoluble in pure normal ammonium sulphide. If air has access,however, the oxygen decomposes the ammonium sulphide with forma-tion of sulphur ; this unites with a portion of the stannous sdphideand transforms it into stannic sulphide which is soluble in the liquid.Sulphide of ammonium or of the alkaline metals is employed inanalysis to dissolve and separate stannous sulphide, but these reagentsact as solvents only when they contain sulphur in excess, and arewithout action when reduced to the state of normal sulphides.Thissource of uncertainty would be avoided if the ammonium sulphideused in analysis were always fully sulphurised by previously dissolvingin it a sufficient quantity of sulphur.By M. PRUDHOMME and F.BINDER (BUZZ. SOL Chim. [2], 37, 194--196).-When barium chlorideis added to a solution of potassium dichromate, normal bariumchromate is precipitated, and potassium chloride and chromic acidremain in solution, thus : K2Cr207 + BaCI, = BaCrOa + 2KC1 + GO3.'l'his reaction furnishes additional evidence in favour of the view thatpotassium dichromate is a molecular combination of the normalchromate with an easily displaceable molecule of chromic anhydride,,z view also supported by the fact that many dichromates (NH4, K, Ca,kc.), can be prepared by tho direct action of chromic anhydride on amolecule of the corresponding normal chromate.By treating di-chromates with alkalis, alkaline earths, or the corresponding car-bonates, double chromates are freqaently formed. Zinc, aluminium,cupric, and chromic hydroxides, when heated with potassium dichro-mate, form normal potassium chromate and a chromate of the parti-cular base.I n this way, certain chromates, e.g., ZnCr04, can be pre-pared, which were formerly obtained only by the action of chromicacid on the carbonate or oxide. This method of preparation explainsthe formation of chromium chromate when potassium dichromate is'treated with hydrogen sulphide or sodium thiosulphate. Chromichydroxide is first formed, and is then acted on by the excesR of di-chromate. When a strong solution of potassium dichromate is addedt o a solution of sodium hydrogen sulphite of 30" B., a green solution isobtained, which rapidly solidifies, owing t o the formation of greenchromic oxide. If the dichromate is in excess, brown chromiumchromate is formed.R. R.Chmmic Acid and Chromates.C. H.B.Chromous Sulphate. By H. MOISSAN (BUZZ. Xoc. Chim. [3], 37,2136--298).-The greater part of this paper has already appeared intlie Conzpt. rend. (Abstr., 1881, p. 684). Chromous sulphate does notdecompose water a t 100". 12.35 grams of the salt dissolve in 100 C.C.of water at 0", but it is only slightly soluble in alcohol. With potas-sium or sodium hydroxide, a solution of chromous sulphate gives ablack precipitate, insoluble in excess ; with ammonia, a black precipiINORGANIC CHEMISTRY. 23tate, soluble in excess, forming a blue solution; with alkaline carbo-nates, a reddish precipitate ; with potassium chromate, a maroonprecipitate ; cupric salts, a brick-red precipitate ; ammonium molyb-date, a dark maroon precipitate ; gold chloride, a deposit of metallicgold ; hydrogen sulphide, no precipitate ; alkaline hydrosulphides, ablack precipitate.When moist chromous carbonate or acetate is treated with a largeexcess of concentrated sulphuric acid, the hydrate CrSO, + 3H,O isobtained in white crystals, more stable when exposed to air than thehydrate CrS04 + 7H,O.I n contact with a small quantity of water,it passes into the normal hydrate Cr,S04 + 7H20. C. H. B.New Class of Borotungstates. By D. KLEIN (Bull. HOG. Chim.[2], 3 7, 202--208).-The disodium salt previously described (Bull.SOC. Chiin., 35, 14) may be a boroduodecitungstate, or a boro-quatuordecitungstate, or a boroyuindecitungstate. The analyticalresults agree equally well with all three formule,The bariunz salt is obtained in white crystals by adding a boilingsaturated solution of barium chloride in excess to a warm saturatedsolution of the sodium salt.If the mixed solutions are allowed toboil, the small quantity of hydrochloric acid which is set free preci-pitates tungstic acid. Too frequent crystallisation from water alsodecomposes the salt, pTobably with separation of metatungstic acidand formation of a basic salt. The addition of a few drops of hydro-chloric acid appears to prevent t'his decomposition. The amount ofwater of crystallisation in the salt appears to be very variable, andthe salt is in all probability efflorescent. When dried a t 160°, thecomposition of the salt agrees more closely with the formula14WO,,R2O3,3Ba0,5K2O than with 15WO3,B2O3,3Ba0,5Hz0,The potassium salt is obtained in slender needles, closely resem-bling dipotassium tungstoborate, by decomposing the barium salt withpotassium sulphate.It has the composition 14WOy,Bz0,,3K,0,Hz0 +The silver saZt is obtained by adding silver sulphate to a solution ofthe barium salt. It is a white crystalline powder, almost insoluble incold, and very slightly soluble in hot water. It cannot be completdxdried without partial decomposition, but appears to have the compo-sition 14W03,B203,3Agz0,Hz0 + 7H20.When a limited quantity of barium chloride is added to the solutionwhich yields the sodium salt on acidification, and the precipitate fil-tered off, the filtrate deposits small granular crystals, very slightlysoluble in cold, more soluble in hot water.They have the composition14W03,B,0a, (3&Ba0,1&Na20),6H,0 + 29H20. This complicated doublesalt resembles the double paratungstates obtained by Marignac. Thecorresponding strontiumcompound, 14Wo3,B2O~, (3$3r0.lhN~0),6HzO + 29.8,0, is obtained in a similar manner by mixing saturated solu-tions of strontium chloride and the sodium salt. All these salts forma new group of boroquatiiordecituiz~states. The barium-sodium andstrontium-sodium compounds are possibly not true double salts, butmolecular combinations of the two salts. The author was unable toobtain the tetrapotassium or pentapotassium salts. When potassium21HZO23 ABSTRACTS OF CHEMICAL PAPERS.carbonate is added to tripotassium boroquatuordecitungstate, potas-sium tungstoborate and a precipitate of potassium parat ungs tate areformed.C. H. B.Change which Ferric Hydrate undergoes after a Time. ByD. TONMASI and G. PELLIZZARI (Bull. Soc. Chim. [2], 37, 196--197).-Ferric hydrate kept under water for a year loses its gelatinousstructure, and changes in colour from brown to yellowish-red. About30 per cent. passes into a modification insoluble in dilute acids, andabout 0.3 per cent. is converted into a soluble modification identicalwith Graham's colloidal hydrate. The change is very slightly, if atall, affected by light. C. H. B.Ferric Hydrates. By n. TOMMASI (Bull. XOC. Chim. [Zj, 38, 152-153).-The author separates the ferric hydrates into two isomericclasses, the a or red series, and the /3 or yellow series, the main pointsof difference between which are given in the table below :-Yellow o r @-series.I Red or a-series.Obtained by precipitating aferric salt with alkalis.a-Fe203,2H20 begins to be de-a-Fe20s,H,0 dehydrated at 92".a-Fe203 is brown.Sp. gr. of Fe203 = 5.11.The hydrates dissolve in diluteThe hydrates are dehydratedhydrated at 50".acids.on boiling with water.Obtained by oxidation of ferroushydrate, ferroso-ferric hydrate,or ferrous carbonate./3-Fe203,2H20 begins to be dehy-drated at 105".p-Fe,03,H20 dehydrated at 150".p-Pe,03 is red or yellowish-red.Sp. gr. of %'c?203 = 3-95.The hydrates are sparingly solu-ble in concentrated acids.The hydrates, even on long boil-ing, retain a molecule ofwater, which can easily be re-moved by a concentrated soh-tion of calcium chloride.The hydrates of the a-series may not only be distinguished, butseparated from the hydrates of the @-series ; for the former are solublein ferric chloride and are reprecipitated by the addition of sodium sul-phate or sulphuric acid, whereas the latter are quite insoluble in theAction of Hydrogen Sulphide on Solutions of NormalNickel Sulphate.By H. BAUBIGNY (Compt. rend., 94, 1473-1475).-The experiments described in this paper show that the pre-cipitation of nickel from a solution of the normal sulphate by hydro-gen sulphide _depends on the tension of the gas. The quantity ofsulphide thrown down in a given time is greater as the liquid is richerin hydrogen sulphide, and the effect of heating at 100" in closedvessels is that the same limits of precipitation are obtained in a fewhours which at the ordinary temperature would require as manyweeks.same reagent.v. H. vINORGANIC CHEMISTRY. 25The pi*ecipitation of the nickel ia complete when the solution doesnot contain more than 1 gram of sulphate in the litre.The limit of precipitation does not depend entirely on the degree ofacidity acquired by the liquid, but varies according to other circum-stances.The action of heat on a solution of neutral nickel sulphate in thepresence of hydrogen sulphate furnishes an exact method of separatingnickel from manganese, aluminium, &c., whose salts are not decom-posed by hydrogen sulphide, but not from iron.R. R.Action of Hydrogen Sulphide on Nickel Sulphate in AceticAcid Solution. By H. BAUBIGNY (Compt. rend., 94, 1715-1717).-The action of hydrogen sulphide on nickel sulphate in solution is re-tarded or entirely prevented if a sufficient quantity of acetic acid isadded in proportion to the quantity of nickel salt present. At thetemperature of loo", however, and in a closed vessel, acetic acidhas no power to retard the action of hydrogen sulphide on dissolvednickel sulphate, the reaction taking place as with an aqueous solutionof the neutral sulphate. It. R.Action of Heat on an Acid Solution of Nickel Sulphate inPresence of Hydrogen Sulphide. By H. BAUBIGNY (Comnpt. rend.,94, 1595--1598).-The experiments detailed in this paper lead to theconclusions that-1.In acid solutious of nickel sulphate, as in neutral solutions,when the ratio of the weights of the acid and the metal remainsconstant, the precipitation of the nickel by hydrogen sulphide is moreoomplete as the solution is more dilute.2. Whatever the ratio of tlie volumes of gas and of liquid, theamount of nickel precipitated increases with the time. R. R.Cobalt Sulphate. By G. VORTMANN (Ber., 15, 1888-1889).-On adding concentrated sulphuric acid to an aqueous solution of ncobadt salt, and then evaporating, cobalt sulphate containing 1 mol. ofwater of crystallisation is produced. The same compound is alsoformed on treating purpureocobaltic chloride with a small quantity ofwater and strong sulphuric acid until dissolved, and then heating to220".It forms a crystalline peach-coloured powder, sparingly solublein cold water, being less soluble than the anhydrous salt. A low redheat is required to drive off the water. In contact with moist air, itabsorbs water very slowly.Cobaltarnine Compounds. (Part 111.) By G. VORTMANN (Ber.,15, 1890-1903) .--Octcwtine Compouwds.-IInstead of preparing thesecompounds from tthe carbonate, as described in his first paper (Ber., 10,154), the author heats the decamine-purpureo chloride for some timeon a water-bath with dilute ammonia and solid ammonium carbonate.On evaporating the solution it assumes the dark cherry-red colour ofthe octamine carbonate. If any luteo-chloride should be present, itwill separate out on cooling, and can be filtered off.Octamine-roseo-oobaltic chloride, Co,(NH&( 2Hz0) Cls + 4H20, generally crystallisesA. K. M26 ABSTRACTS OF CHEMICAL PAPERS.(as previously shown by the author) with 2 mols. H20, but by addingconcentrated hydrochloric acid to the cold solution above-mentioned,it is obtained with 4H20. With mercuric chloride, the compoundCoz(NH3),(2H,O) C1,,6HgCI2 + ~ H z O is precipitated of a pale-redcolour. Heated to 100" it loses 3H20. Octamine-purpureocobalticchloride also forms a double salt with mercuric chloride ; this whentreated with concentrated hydrochloric acid and evaporated on a water-bath yields greyish-violet crystals of Co,(NH,),C1,,SHgCl2 + H,O, buton further concentration, green crystals of the praseo double salt,CO,(NH,),C&,H~CI,, are obtained.This compound is sparinglysoluble in cold water ; hot water con-ierts it into octamine-purpureo-cobdtic chloride. On adding praseocobaltic chlornitrate to acidu-lated mercuric chloride solution, a green precipitate ofis produced. Octamine-cobaltic nitrate, prepared by adding nitricacid to a solution of the corresponding carbonate, forms a crys-talline precipitate of the formula Co~(NH3),(NO3),2HzO. It hasalso been obtained with 1 and with 6 mols. H20, also anhydrous.Octamine-cobaZtic chromate, Co2(NH3),(2H,0) (CrO& + 2H20, is pre-pared by the action of potassium dichromate on the octamine-pur-pureo chloride, sulphate, or nitrate, and may be purified by crystal-lisation from a weak acetic acid solution : it forms bronze-colouredplates.If normal potassium chromate is employed, the compound ob-tained is olive-green, and contains 8 mols. H20. Solutions of bothcompounds in strong acetic acid when evaporated yield an orange-red body of the formula Co,(NH3),(2HzO)(Cr04j2,Crz07 + H20. Aplatinnchloride has been prepared, but its formula is not established.The acid carbonate, to which the author previously gave the formula,Coz03(NH3),,4C0, + 2H20, contains 3H20, 2 mols. of which aredriven off a t 100". Praseocobaltic chlornitrate, COz( NH3),(N03),C1, +2H20, is precipitated on adding dilute nitric acid or potassium nitratesolution to solution of praseocobaltic chloride. With potassium dichro-mate the latter gives a yellowish-green precipitate of the chlorochro-mate, CO,(NH,),(C~,O~)C~~ + H,O.Hexamine Compounds.-To an ammoniacal solution of octamine-cobnltic chloride, ammoni urn carbonate is added, and the solutionevaporated to dryness.On redissolving and repeating the same treat-ment, the residue consists of cobalt hydroxide and hexaminecobalticcarbonate, and from this 0the.r hexamine salts are obtained by theaction of acids. Hexamine-cobaltic chloride is a green compound, ofthe formula CO,(NH~)~C~, + 2H20. It dissolves in cold water to abluish-violet solution, which changes to violet-red on warming ; fromthis solution concentrated hydrochloric acid precipitates octamine-purpureocobaltic chloride. Hexamine-cobaltic sulphate was obtainedas an oil, which after frequent treatment with alcohol gradually be-came crystalline.It forms a red powder, easily soluble i n water. Itsformula is CO,(NH,)~(SO~), + 6H20. The nitrate, CO,(NH,),(NO~)~ + 8H20, is a dark cherry-red deliquescent body. The carbonate,Co,( NH&( OH),( GO3), + 3H20, is formed in preparing the octnrnine-cmbonate, and is precipitabed by alcohol as an oil, which afterCog (NH,)sC16,2HgC1INORGANIC CHEMISTRY. 27repeated solution, precipitation, and treatment with alcohol, becomescrystalline.Heptaqnine Compounds.-The author has in a previous paper given amethod for the preparation of melanocobaltic chloride. In order toconfirm the formula which he gave for Rose's " black salt,"COz(NH,)s(NH,Cl) C Lhe has prepared other derivatives, which he finds to be of analogouscomposition.ikIelanocobaltic ch.Zorockromate,Co,(NH,),(NHzC1) C&,Cr207 + HzO,by precipitating the melanochloride with potassium dicliromate.With platinum chloride, the melanochloride gives a brownish-blackprecipitate of Go2( NH3)6 (NH,C1) C I,, P t C1,. 0 n heating melanocobalticchloride solution, i t becomes red ; with platinum chloride this redsolution gives a reddish-brown precipitate, which (air-dried) has theformula Coz (NH,),( NH2C 1 j CI, (OH),,PtCI,. With mercuric chloride,pale-red needles of Coz(NH,),(NHzCl) CIz( OH),,3HgCIZ + HzO, areprecipi tnted. With picric acid, melanocobaltic chloride gives a brownprecipitate, which explodes violently on heating. A. K. M.Electrolysis of Ammonium Carbamate and Carbonate withAlternating Currents and Platinum Electrodes. By B.GERDES(J. pr. Chem. [2], 26, 257-276).-Dnrechsel has observed (J. pr.Chem. [2], 22, 476) that on electrolysing solutions of ammoniumcarbamate and carbonate with a1 ternating currents, the platinumelectrodes become strongly corroded, with formation of soluble andinsoluble platinum bases, of urea, and of an oily substance soluble inammonia. The author has isolated and studied the platinum basesreferred to. He used platinum electrodes two inches by one inch, abattery of 6 6 Grove's cells, and alternated the current about, 10times in each second, the duration of each experiment being 10-12hours, and the solution being kept cool. After that time a thickyellowish or white precipitate had formed, the solution being colour-less, and the electrodes much attacked.Besides ammonium nitrite and nitrate, urea, and a fatty substance,the filtrate was found to contain a soluble platinum salt thrown downas a blue or green precipitate by hydrochloric acid, and crystallising inneedles.It was not, however, obtained in quantity suficient to allowof detailed investigation.The white precipitate contains most of the platinurn dissolved offthe electrodes. It is a carbonate insoluble in cold water, but on heat-ing dissolves sparingly, forming an a1 kaline solution. After havingbeen dried over sulphuric acid, it does not lose weight a t 110", but athigher temperatures it first becomes yellow and then suddenly decom-poses, with evolution of ammonia and water, leaving metallic platinumin a very fine state of division.Its composition corresponds wellwith the formula PtN,H&,O,. The carbonate dissolves in dilutesoda, and is precipitated without alteration from its solution by car-bonic anhydride. When dissolved in hydrochloric acid and precipi-tated by sodium carbonate in sufficiently dilute solutions sniall wellformed octohedrons separate28 ABSTRACTS OF CHEMICAL PAPERS.The chZoride, P t ( NH,) &I4, is easily obtained in small rhombohedronsor in needles; i t dissolves readily in hot water, and with gold orplatinum chloride gives precipitates resembling ammonium platino-chloride. The nitrate, Pt (NH,)6(NOs)4, is readily soluble iu water,.arid crystallises in needles. The sulphate is practically insoluble inwater, even calcium sulphate solution giving an immediate precipitatewith solutions of the soluble salts. It could only be obtained in theamorphous condition.Other compounds have been prepared, buthave not been analysed. C hromates produce a yellow, hydrofluosilicicacid a white precipitate. The free base has not yet been isolated in astate of purity.If during the action of the electric current the solution of carbonateor carbamate is not artificially cooled, the temperature of the fluidrises to 40-50", and no precipitate of the carbonate just describedis obtained, but on cooling, long prismatic highly refractive crystalsseparate from the solution. These likewise consist of a carbonate,PhN,C4OI4H,,. Their solution gives with nitric acid a colourless pre-cipitate, which gradually changes into bright blue octohedrons ofplatodianzntoniunz nit rate, Pt (NH,), (NO,),.The author believes the formation of the several platinum bases totake place as follows :-Electrolysis a t first takes place in the ordinarymanner, (NHJ2C03, splitting up into 2NH4 and GO,, the former com-bining in the nascent state with the negative electrode, formingPtH,(NH3),. The current alternating, the negative pole becomespositive, and C03 is added to the compound formed, yielding H, +Pt { ",$>CO, that is to say, platosammonium carbonate, which saltdirectly combines with NH3, yielding plntodiammonium carbonate,P t { ~ ~ : ~ ~ " , > C O . To the latter, on further alternation of cnrreut,~ -I -HNH3 NH NH.0 2NH4 are added, giving H-H,>Pt<NH::NH:o >CO, from which in- -a similar manner the insoluble carbonate-Ammonioplatinum Diammonium Compounds. By E. D RECH-SEL (J. pr. Chem. [2], 26, 277--281).--In an appendix to the previouspaper, the author points out the strikitig similarity of the reactions ofthe platinum base described by Gerdes with those of barium com-pounds, the solutions being precipitated by sodium carbonate, potas-sium dichromate, sodium phosphate, sulphuric acid, calcium sulphate,hydrofluosilicic acid, and alkaline oxalates, hyposulphates and ferro-cyanides. By the reagents named in fact the platinum compoundcannot be distinguished from baryta. Hydric sulphide and ammoniumsulphide produce precipitates only after some time. The ammonio-platinum diammonium compounds stand therefore in the same relationto the alkaline earths as ammonium does to the alkalis. 0. H