BBSTRACTS OF CHEMICN; PAPERS. 1210 Analytica1 6:hemistry. TIr:)Deemyosition of IiIimds by ,Yodilcrjl aiLd Potassiism. By 11. sCHO C NX.* -LLZiOST dl substances inso!uble in acids such as the silicates flnodi- CZ~CS cyanides titanates cliromie iron &c. are decomposed 11-lien igliitccl with soclinm or potassium. The recluetion is best efiectecl in a six~rtll cyliiidrical iron crucible about 4centirnetres in diameter and of eqxl height ; snfficiently thin to be capable of being brought rapidly to LZ red heat when ignited in a Bunsen flame. The fragments of sodium or potassium are arranged at the bottom of the crucible and covered ii-ith the mineral intended to be decom-posed the latter having been previously wll dried ; the crucible is then covered and quickly heated to briglit redness.TVheri the action is comphtc the crucible is allowed to cool and a little water added in tlie first instance drop by drop to the fused mass. The contents are f;n,zlly treated with boiling water and the insoluhie portion is. filtered off The solution contains the alkali in combination with the chlorons portion of tlie niiiiei-al while the metals or their oxides remain upon thc filter. The latter having been well mashed are dissolved in hydro- chloric or nitric acid mcl the analysis of both solutions is completed in the usual manner. When cai-bonates are decomposed by tlie method above described the insoluble portion will of course contain a certain qnantity of free carbon ; in the case of sulphates however YO corre-sponding reduction takes place ; nevertheless when the solution is ex.uuined the whole of the snlphuric acid will be found in combina-?ion with the sodium.The following notes relate to the author's experience :-Oxtde of CIiro7xium and C7msne Iroiz.-According to the qunntity of sodium enip1o-j-ccl the chromium is either reduced to the metallic state OT partially coiivcrted into chromate. FZzco~*idc of c(ilcivi,b is not attzckcd by sodium. To separate tit&c tic Id from its combinations (titanic iron for example) after ignition v-ith sodium the hydrochloric solution is neutralised with sodium cnlbonate and the iron having been peroxidised with nitric acid tl'e solution is boiled with sodium acetate. The whole of the titniiic acid is then precipitated with the iron.The calcined pre- X J.Phami. Chim. [4],xiv 190-192. ASXLPTICAL CHEMISTRY. 1211 cipitate is tlieii digmtec! :irith concentrated hydroch'oric aci~~, which slowly but completely remores the iron. R~i@7~ide XoZybde?~z~~~z of is attacked by sodium producing a brown alkaline solution which turns green when exposed to the air. The acid solution of the deposit which becomes blue on exposure is de-colorised when neutralised with ammonia and brought to ebullition and gives on the addition of barinni chloride a white precipitate of barium moly bdate. In a similar manner minerals containing timystic ad yield an alkaline solution which furnishes on addition of excess of h-j-dro- chloric acid a v;hitc or flesh-coloured precipitate rapidly chaiiging to lemon-yellow.The precipitate is pure tungstic acid. J. w. EstlirLafio~~ of Cream of Tartar Wine. By P. CARLES.* THEauthor having recently determined the percentage of cream of tartar in several varieties of wine both by the old process of incinera- tion and by the improved method lately introduced by Berthelot pnblishes the results of his analysis in order to show how n-idely the two processes sometimes differ in results and to urge upon chemists the necessity of rejecting the old process and of adopting instead the more exact iiiethocl of Berthelot and Fleurieu. In this latter process the acid tartrate of potassium is precipitated by addiiig to n known rolume of the wine an equal volume of a mixture of alc~holand ether.The crystalliiic precipitate is collected slightly washed with alcohol redissolved in water and the tartayic acid estimated by a standard alkaline solution. The following are some of the analyses above referred to. The results calculate to one litre respectively :-! I Red wine. 1 White nine. 1 White nine. Malags. (Eordeaux.) 1 I I A. B* ---I--!-[---I_-Alcohol.. ............../ 120-0 I 110'0 I 110.0 164 -0 Cream of tartar (by cal-2.86 1 .UP aa 1 -82 3 *4 cination) ..........} Cream of tartar (Ber-2 '32 1 -45 1 -10 0'05 thelot's methocl) .... } Extract ................ 24 -50 18 50 18 -00 171-40 It will be noticed that in the first two wines the percentages of cream of tartar obtained by the two processes are very nearly identical bnt the variation is much greater in the two latter instances especially * J.Pharm. Chim. {4] xiv 188-190 1212 ABSTRACTS OF CHEMICAL PAPERS. with the sample of Afalaga in which the proportion of cream of tartar indicated by calcination is actnally greater than the above mixture of alcohol and water is capable of dissolving. In the opinion of the author this anomalous result may be accounted for by the natural presence in the wine of organic potassium salts or perhaps from the addition of molasses or some similar product rich in salts of this nature ; at all events it is obvious that the old process is liable to fre- quent sources of erroi, especially mhen sweet mines are examined; it is better therefore to reject it altogether and to conduct the analysis by direct estimation of the cream of tartar.J. W. Dry Process tJLe Estiilzatiorb of Silicon in Cast hou 'IVyozcght I~*on CLiZCZ Xteel. By H. BOUSSINGAULT.* THEprocess generally adopted for the determination of silicon in iron and steel is not sufficiently trustworthy in its results when the pro- portion of silicon contained in the iron is extremely small and when consequently tlie weiglit of silica obtained cannot possibly exceed a fern milligrams. In fact the question not nnfrequeiitly arises whether the silica obtained really existed in the iron as silicoii or wliether it was derived fi-on1 the vessels 01%reagents employed in the operation. It became necessarr therefore in order to resolve definitely the questions which the author hacl undertaken that some more satisfactory process for the estimation of silicon should be devised and after inany trials the follon-ing clry metliod was decided upon experience having shown that it mas cntircly free fi-orn those particular sources of error which in the old pi-ocess are so very liable to falsify the results.The iron at red heat is submitted successively to the action of two gases. 1,Atmospheric air to oxiciise tlie iron carbon and silicon. 2. Dq hydrochloric acid gas to convert andl volatilize the metal in the state of chloride. The operation is effected consequently in two stages ; aiicl this is the more necessary inasniuch as it was found that when the iron is oxidised aid coiiverttd into chloride simultanconslg- by passing over the heated metal a current of hydrochloric acid gas mixed with air the chlorination of tlie iron takes place nioi*e rapidly than the oxida- tion and the silicon in combination passes into the state of chloride before it is oxidised aid the whole or at least a considerable portion is in this way lost.The analysis is conductecl in the following manner :-About one gram of the metal in a tolerably fine state of division is placed in a small platinum boat and carefnlly weighed. The whole is * &in. Chim. Phys. [4] xxii 45'7-472. ANALYTICAL CHEJIISTRY. 1213 then introduced into a muffle and heated in a furnace to the tempera- ture usually employed in cupellation. In two or three hours the iron will be found to be completely oxidised and to have increased about 35 or 36 per cent.in weight; this indicates the composition 4Fe0 Fe,O, in which 1 gram of metal would require 0.53 gram of oxygen. The boat and its contents are next placed in a porcelain tube heated to redness and a slow and continuous current of dry hydrochloric acid gas passed over them. The iron is thus converted into chloride and volatilised towards the cooler portion of the tube while the silica not being affected by the gas remains untouched in the platinum vesscl. When the operation is judged to be complete the latter is withdrawn placed in a glass tube and re-weighed ; the increase in weight is due to silica. The silica obtained in this manner is perfectly white in an ex- tremely fine state of division geiierally preserving the form of the oxide as withclrawn from the muffle.If this silica for example was derived from an iron turning coiled spirally the silica will represent this spiral ; while so great is its tenuity that it may be compared to the ash of a linen thread which has been carefully burnt in a still atmosphere. It is obvious therefore that it is important to moderate the rapidity of the current of gas during the operation lest the silica or any portion of it be mechanically carried away. The purity of the silica is ascertained by pouring into the boat a few drops of hydrofluoric acid and a drop of sulphuric acid ; on warming the whole should volatilize. In every estimation made by the author upon malleable iron and steel the silica disappeared entirely by this treatment but cast-iron not unfrequently yielded a silica containing a small quantity of earthy matter resulting probably from an admix- ture of scoria with the melted metal ; it was certainly not due to the presence of alumina inasmuch as aluminium is never found in cast- iron the author having even failed to detect it in Wootz steel not-withstanding that Faraday’s analyses tended in the opposite direction.Considering that a certain proportion of phosphorus was contained in many of the samples of iron and steel examined by the author it mas somewhat surprising that no indications of ferric phosphate or phosphoric acid should be found in any of the silica collected by this process ; on the contrary it was perfectly pure volatilizing entirely under the influence of hydrofluoric acid.It became necessary there- fore to institute a more careful examination and to ascertain exactly the manner in which the phosphorus was eliminated. To this end some ferric phosphate was prepared in the ordinary way mixed with a proportion of oxide and placed in the same platinum vessel in which the previous analyses had been conducted; the tube having been brought to a bright-red heat dry hydrochloric acid gas was passed VOL XXIV. 4 M 1214 ABSTRBCTS OF CHElIICAL PAPERS. over it for two hours. At the expiration of this time the boat ap- peared empty ; it contained nevertheless a few vitreous globules which proved on examination to consist of sodium phosphate mixed with a little silica the alkaline phosphate being doubtless a portion of that with which the iron was precipitated and which had escaped washing.As far as regards the phosplioric acid combined with iron the whole was volatilized and found in the acid solution of iron con- tained in the flask at the extremity of the porcelain tube. In order to avoid the error introduced by the employmeiit of artificial ferric phos- phate some finely-powdered vivinnite TVRS operated upon in tlie manner already described. In an hour the vivianite had disappeared with the exception of a little earthy gangue which contained neither iron nor phosphoric acid. As a deduction from the foregoing experiments it may be confidently stated that when oxide of iron containing a small quantity of phos-phoric acid is submitted to a current of liydrochloric acid gas at a temperature sufficiently high to isolate thc silica tlie whole of tlie phosphoric acid is eliminated at the same time as tlie metal.The following table shows the amount of silicon contained in various samples of iron and steel as estimated by the dry method. The quantity of iron operated upon was one gram. Swedish iron ............................ 0.00164 Puddled iron from Unieux ................ 0*00093 Bar iron ................................ 0' 00190 Iron wire ................................ 0.00230 Cast steel.. .............................. O*OOOiO Cast and cemented steel .................. 0.00440 Chincse steel ............................0.000'70 Soft steel.. .............................. 0.00093 Cast steel for carriage springs. ............. 0.00094 , for watch springs ............ 0.00044 Cast steel Krupp ........................ 0.00440 Tungstcn steel .......................... 0.00093 Wootz steel.. ............................ 0.000GB Grey cast iron from Ria .................. 0.0 140 White cast iron from Ria ................ 0-0034 J. W. A New i12ethod of Estimatixg Ziizc. By HUGO TAMJI.* WHEN a solution of zinc in any mineral or volatile organic :Lcid is supersaturated with ammonia then slightly acidulated with 1,ydro-* Chem. News xxiv 148-150. AShLYTlChL CHEMISTKY. 1215 chloric acid a double chloride of zinc and ammonium is produced.The addition of hydro-disodic phosphate to the solution thus obtained occa- sions a precipitate of zinc phosphate which by subsequent heating in the liquid combines with ammonium phosphate thereby forming ammonio-zinc phosphat e. The author recommends this reaction for the separation and estima- tion of zinc instead of the ordinary but very troublesome method of precipitating the zinc as sulphide. If great accuracy is required the liquor should be left in a m-arm place for ten or twelve hours to ensure the complete precipitation of the zinc ; but for ordinary assay purposes it is sufficient to allow the precipitate to rest for an hour. The ammonio-zinc phosphate dried at 100" has the constant formula Zn(NH1)PO1. It is slightly soluble in water and is converted by ignition into zinc phosphate with loss of zinc.When formed as described it retains sodium phosphate and ammonium chloride with some force and several successii-e washings are required to free it from these salts. Directions are given for estimating zinc in syelter blendes ccclnnzines and bmss. The general plan is as follows :-First remove from the respective solutions all substances precipitable by sodium phosphate in a neutral or alkaline solution ; saturate with ammonia ;acidulate slightly with hydrochloric acid ; add sodium phosphate in excess ;heat near to boiling point ; allow the liquid to remain in a warm place some time ; collect ; wash well with water ; and dry at 100" C. Most amlvoiiiacal solutions of oxides mixed with an excess of ammonia are also precipitated by sodium phosphate and when the mixture is boiled the corresponding ammonio-phosphate crystallises.Ainmoizio-iiia?zga.lLeseyl~osphateforms beautiful flesh-coloured crystal- line silky scales. Amnzonio-cobalt phosphate forms crystals of an indifferent pink colour which when calcined at a low temperature change to a deep blue colour without losing the shape of the original crystals. Am,no?Lio-fewozus phosphate forms silky crystalline scales resembling in colour certain greenish micas. An examination of the ammonio-zinc phosphate has led the author to conclude that ths equivalent of zinc is 32-75. A. T. On the Volumeti-icnl DeteyinincLtion of Zinc by Soclic Xulp7~icle. By 0. SCHOTT.~ THE author made some comparative experiments (1)with lead paper prepared according to Fresenius's directions by coating paper with Jk Zeitschr.Anal. Chem. x 209. 121G ABSTRACTS OF CHEMICAL PAPERS. plumbic carbonate ; (2) with cobalt-paper prepared according to the directions of A. Deuss ;* and (3) with a paper known in comnierce as polka-paper i.e. ordinary glazed paper used for visiting cards covered with white lead; he gives tlie preference to the latter. The ammg- niacal solution of zinc is placed in a beaker diluted to half a litre and soclic sulphide is run in as usual. The end of the reaction is ascertained by dipping into the liquid a glass tube 25 centim. in length and 7 mm. in diameter wherewith a column of the liquid can be taken up by placing the thumb on the upper end of the tube.A piece of the glazed lead-paper is next pressed geiitly against the lower end of the tube held in a slightly slanting position over the beaker and the liquid is allowed to run out slowly OVC~the inclined card. The end of the reaction is indicated by the formation of a brown ring whicli appears where the liquid was brought in contact with the lead paper. The glass tube the end of which should be first groand and then smoothed in the Atmic should not be pessed too had against the paper in order to avoid any action of the siispencled zinc su1ph;de upon the white led. The ai~thor c1:iiiiis slight advantage for his indicator over that propoFcc1 by EIrCFeiLus and a st ill more decided advantage over Deuss’s iridicator.w. v. ~:l.oi:iiiie-2i.citi~~, 11s CL Test -for PlLemd mcl Allied Xzcbstamas. By 11. LANDOLT.? INanalysing the water of a well situated in tlie neighbourhood of a gas work the author mas induced to search for phenol. The methods already kiiown for cletectifig thk bdy are only of a moderate degree of delicacy. The bluish-violet coloration with chloride of iron is but slightly sensitive and is not always to be relied upon since minute quantities of free acids and even neutral salts as potassium sulphate sodium sulphate &c. interfere with the reaction. The most satisfactory indications are afforded by bromine-water M liich gives with a dilute solution of phenol an immediate yello-\.vish bulky precipitate of tribromopheiiol.An excess of bromhe-water must be employed. By this reaction 1part pf phenol in 43,700 parts of water may be easiig detected. The limit to tlie chloride of iron reactioii is 1in 2,100 of water. The smell of phenol cannot be detected in cz solution coiitaining less than 1part in 2,800 of water. The author shows that the piwipitation by bromine-\\ ater may be employed for the quantitative determination of phenol. The same reaction will show the phenol present in urine. 500 C.C. wine serve for its detection. T. E. T. ANALYTICAL CHEMISTRY. 12 7 On the Methods for Detemnining the Amount of Allcaloiids in Peruvian Bark. By C. SCHACHT.* THEauthor discusses at some length the respective advantages and disadvantages of the several methods for determining the alkalo'ids in Peruvian bark which have been recommended in various periodicals as well as in the German and English pharmacopmas published during the last few years.The results which he obtained by these various methods differed widely (from -95 p. c. to 8-67 p. c. of cinchona bases). Schacht fixed at last upon Hager's method,? which con-sists in converting the bases into picrates decomposing with potassic hydrate and extracting the free bases from the dried residue bymeans of absolute alcohol. The pulverised bark is first exhausted with water containing sulphuric acid and the acid extract is precipitated by means of picric acid. 10 grms. of bark when so treated by the author gave -6270 grm. of picrates ; that is less than what Hager obtained.Instead of weighing the picrates as such they may be decomposed by means of potassic hydrate. The mixture of potassic picrate and alkalo'ids is then evaporated to dryness and exhausted while warm with chloroform (free from alcohol) whereby a colourless filtrate is obtained ; or it may be shaken up repeatedly with chloroform the clear layer removed with a pipette the chloroform distilled off the residue taken up with water containing a little hydrochloric acid and the bases pre- cipitated with sodic hydrate. 10 grms. gave -220 grm. of bases equal to 2.2 per cent. Although this method was found to be pre- eminently quick and neat the author was never able to obtain the quantity of picrates which Hager obtained. By substituting amylic alcohol for absolute alcohol whilst proceeding in every other respect in the same manner he obtained more favourable results- 10 grms.of bark yielded 2.54 per cent. of bases. In a second experi- ment in which the picrates were washed by decantation decomposed by means of excess of potash-solution and shaken up at once with amylic alcohol he obtained 2.4 per cent. of bases. By modifying the process so as to precipitate the bark-extract (obtained by boiling the finely macerated bark with solution of potash and sulphuric acid) at once with excess of solution of potash without previously precipitating the bases as picrates and extracting with amylic alcohol he obtained 3.16 per cent. of bases. In his subsequent experiments Schacht treated an aliquot part (50 c.c.) of the filtrate directly with 2.5 C.C.of potash of sp. gr. 1.3 and then with 25 C.C. of amylic alcohol of boiling point 132" C. The layer containing the alkaloids was drawn off with a pipette and the extraction with amylic alcohol repeated three times over. In this manner he obtained in two experiments 3.16 and 3.12 * Arch. Phaxm. [2] cxlvii 97-106. t Pharm. CentralhaUe 1869 p. 146. 4~2 ABSTRACTS OF CHEMICAL PAPERS. per cent. of bases. By allowing the mixture of 50 C.C. of bark extract to which 2.5 C.C. of potash-solution and 25 C.C.of amylic alcohol had been added to stand for twelve hours before pipetting off and by shaking up three times more with amylic alcohol the author obtained in three experiments 3-50 3.52 and 3.82 p.c. of bases. In the last experiment the alkalo'ids were found to be contaminated with a little colouring matter carried down by precipitating with too strong a soda ley. Schacht usually prepared an extract from 10 grms. of bark by boiling with solution of potash and sulphuric acid made it up in a narrow cylinder to 110 c.c. and allowed it to subside sufficiently to obtain 50 C.C. of clear supernatant fluid which could be filtered rapidly into a flask. Caustic potash is added and without any delay amylic alcohol and the whole well shaken up. The liquid separates speedily into two layers. The upper one being very slightly coloured is trtms- ferred by means of a pipette to a porcelain dish wherein the por- tion of reddish-brown aqueous liquid subsides before it is filtered through a small filter moistened with amylic alcohol into the distilling flask.The washing with amylic alcohol is effected in like manner three times over. The amylic alcohol is next distilled off; the residue transferred from the flask to a small porcelain dish by rinsing with a little amylic alcohol (absolute alcohol has a tendency to creep over the dish and to cause a loss) and then evaporated to dryness. It is next compZeteZy taken up 11-ithwater and a few drops of dilute sulphuric acid of sp. gr. 1.16 and filtered cold through a small filter moistened with water. The precipitation of the bases is effected by means of uerp dilute caustic scda-solution (1part of soda of sp. gr. 1.3 to 9 parts of wator) and continuous stirring to prevent the baliing of the precipi- tate.After standing quietly for some hours it may be filtered through a weighed filter and washed cautiously with very little water Addition of glycerin to the sulphuric acid extract of the bark did not yield favourable results. By dispensing however with the pri-mary operation of treating the macerated bark with potash in order to disintegrate the fibrous tissues employing a mixture of 2 grms. of hydrochloric acid of sp. gr. 1.12 together with 50 C.C.of glycerin and 100 grms. of distilled wTter and digesting the bark for 12 hours in the cold a clear filtr.de may be obtained which not only passes freely through tlie filter but leaves also a residue which can be wnsllcd out very readily. On evaporating the whole the liquid remains invariably clear and when treated with caustic potash and amylic :~lcc;iiol,sepamtes after some time into two lajers which can readily be icmnved from each other and washed.3.41 per cent. and 3.72 per cent. of cinchona bases were obtained in this manner. It is however inipossibk to precipitate the bases as picrates on account of their solnliiliiy in h~diwchloric acid. w. v. TECIINICAL CHEMISTRY. 1219 072 a Metld of Determining the Gases resulting fsom the Explosion of Nitroylycerin. By L. L.HOTE.* TOexplode nitroglycerin the author introduces a small quantity of it from 5 to 6 milligrams contained in small bulbs of thin glass into a Mitscherlich's eudiometer and then passes into the eudiometer a small quantity of electrolytic gas ; on passing the electric spark the explosion of this gas cames the detonation of the nitroglycerin.The gases formed cause the mercury to turn brown on the surface. One gram. of r~itroglycerin yields 284 cubic centimetres of gas at 0" C. and 760 mm. which contains 45.72 p.c. COa,20.36 NO and 33.92 N. This method of explosion may be applied to the study of the gases furnished by other explosive compounds. Potassium picrate detonates easily under the conditions and yields an inflammable gas containing an appreciable proprtion of cyanogen. A. P.