44 ABSTRACTS OF CHEMICAL PAPERS. Analytical Chemistry. Standardising Acids. By ANTON SEYDA (Chem. Centr., 1899, i, 1164 ; from Zed. ofentl. Chem., 5, 141--151).-Sodium carbonate, used for stsndardising acids, is likely to retain excess of carbon dioxide, or if overheated it may contain sodium hydroxide. A suitable article is, however, obtained by heating sodium hydrogen carbonate for a n hour a t 220” in an air-bath. The results then perfectly agree with those obtained by standardis- ing the acid with the aid of potassium hydrogen tartrate or ammonium By STEPHEN I?. PECKIIAM and H. E. PECKHAM (J Amer. Cihem. Xoc., 1899, 21, 772--776).-1n reply to Hodgaon, the authors state that in order to get accurate estimations of sulphuric acid, any iron or aluminium oxides should be first removed.The following method has been finally adopted for the estima- tion of sulphur in bitumen. A quantity of the substance representing about 0.5 gram of real bitumen is mixed with 15 grams of pure dry sodium carbonate and 15 grams of potassium nitrate and the mixture then fused by degrees in a platinum crucible. The product is dissolved i n dilute hydrochloric acid, evaporated to dryness, and gently ignited to render any silica insoluble ; the residue is then treated with dilute hydrochloric acid and the solution precipitated, while boiling, with a slight excess of ammonia t o remove iron and aluminium. If desired, any calcium may be precipitated as oxalate. The filtrate is now acidified with hydrochloric acid and the boiling liquid precipitated with barium chloride solution added carefully from a pipette.chloride. L. DE K. Estimation of Sulphur in Bitumens. L. DE K. Estimation of Hyposulphurous Acid. By N. FRADISS (Chent. Centr., 1899, i, 1223 ; from Bull. Assoc. Chimistes, 16, 453).-The solution is neutralised and added from a burette to an ammoniacal standard solution of copper sulphate until this is quite decolorised. The presence of sulphites does not interfere. 1 mol. of hypo- sulphurous acid reduces 2 mols. of copper oxide to the cuprous state. L. DE K.AN A L Y TI C AL C H EM IST R T . 45 Volumetric Estimation of Sulphuric Acid. By FRANZ LITTERSCHEID and KARL PEIST (Arch. Pharrn., 1899,237, 521-525).- The authors have developed independently the method described by Griitzner (this vol., ii, 530).The solution of the sulphate is acidified with hydrochloric acid, heated to boiling, treated with a measured excess of N/4 barium chloride solution, stirred well, and allowed to remain in a warm place for half-an-hour; the liquid is then made strongly alkaline with ammonia, excess of ammonium carbonate solution added, and the whole stirred and allowed to remain a t 50-60' for ten minutes. The precipitate is then collected on a small filter, washed three or four times with hot water, and transferred to a conical flask, when the barium carbonate is titrated with N/10 hydrochloric acid. It may be titrated directly when methyl-orange is employed as an indicator ; if phenolphthalein be used, excess of N/10 acid must be added, the carbon dioxide driven off by boiling, and the excess of acid titrated with N/10 caustic potash.On multiplying by 0.4 the number of c c. of N/10 acid neutralised and subtracting the product from the number of C.C. of N/4 barium solution taken, the remainder multiplied by 0.01 gives the weight of SO, present, or multiplied by 0.004 the weight of S. The solution of the sulphate must not contain any acid of which the barium salt is insoluble, such as phosphoric and oxalic acids; neither must it contain substances which are precipitated by ammonium carbonate, or, like 'tartaric and citric acids, hinder the precipitation of the barium. Alkaloids, if insoluble in ammoniacal ammonium carbonate solution, must be removed by extraction with a suitable solvent. The sulphates of potassium, sodium, ammonium, rubidium, czsium, copper, zinc, cadmium, nickel, and cobalt, may be estimated directly ; so also that of lithium! provided the solution be dilute enough t o retain the lithium carbonate in solution (solubility : 1 in 75 parts of water a t 50') Iron sulphate, and the solution obtained by oxidising copper pyrites with nitric acid and potassium chlorate, may also be estimated, if the iron is first removed as hydroxide. With magnesium sulphate, good results have not yet been obtained.Titration of Persulphates. By MAX LE RLANC and M. ECKXRDT (Zeit. Elektrochern., 1899, 5, 355-357).-The reaction between a persulphate and a ferrous salt in solution is comparatively slow at the ordinary temperature ; erroneous results may therefore be obtained in the analysis of persulphates by treating the solution with ferrous sulphate and titrating the excess with permanganate.These errors are very easily avoided by warming the acidified mixture of persulphate and ferrous ammonium sulphate to 60-80' before titrat- ing with permanganate T. E. Estimation of Tellurous Acid in Presence of Haloid Salts. By FRANK A. GOOCH and C. A. YETERS (Arner. J. Xci., 1899, 8, 122-1 26). -Tellurous acid is generally estimated by adding to its alkaline solution a slight excess of standard potassium permangannte. A definite volume of standardised ammonium oxalate is added, and then a sufficiency of dilute sulphuric acid (1 : 1) so as to have about 5 C.C. of the acid in excess. After heating at 60--80°, the excess of oxalic acid is estimated by standard permanganate.C. F. B.46 ABSTRACTS OF CHEMICAL PAPERS. The authors find that the presence of chlorides does not interfere if before the final titration about 1 gram of crystallised mangnnous chloride is added. Bromides are also harmless if the same precaution is taken and the temperature not allowed to rise above 2 5 O . I n the case of iodides, the process breaks down, so another method was successfully tried. The alkaline solution is mixed with a known volume of potassium permanganate, previously standardised with a standard solution of arsenious acid, and, after a while, a slight excess of dilute sulphuric acid is added, with, if necessary, some more potass- ium iodide; after adding a slight excess of potassium hydrogen car- bonate, the liberated iodine is titrated with the standard solution of arsenious acid, the end reaction being the disappearance of the yellow colour ; starch need not be used as an indicator.Supposing the two solutions balance each other-the difference in the two titrations equals the number of C.C. of permanganate consumed by the tellurous acid. L. DE I(. Detection of Nitric Acid in Cadaveric Matter. By DIOS- CORIDE VITALI (Zeit. anal. Chem., 1899, 38, 539-541 ; from Oesterr. Chemikerxeit., 1, 330).--For the detection of free nitric acid in animal matter in which nitrates may already be present, either as normal constituents or by introduction in food, it is useless to distil with water, for the mineral acids combine with albumin, forming acid- albumins, which in many cases are not decomposed at 190'.The substance is therefore digested on the water-bath with freshly preci- pitated barium carbonate, the solution evaporated to dryness, and the residue boiled repeatedly with absolute alcohol for the removal of calcium and magnesium nitrates. The alcoholic solution will also contain the soluble acid-albumin. It is evaporated to dryness, the residue dissolved in water, and exactly neutralised with barium hydroxide solution, which decomposes the acid-albumin. The solution is again evaporated and the nitrates separated by absolute alcohol. The residual barium nitrate is then dissolved in water, decolorised with lead acetate, freed from lead by hydrogen sulphide, and the concentrated solution finally crystallised in the desiccator. The residue from the first treatment with alcohol contains the barium nitrate derived from the presence OF free acid as well as insoluble acid-albumin, and possibly some calcium nitrate.It is treated with a dilute solution of sodium carbonate until an alkaline reaction is just obtained. The acid-albumin and the barium nitrate are by this means converted into sodium nitrate, the calcium nitrate remaining undecomposed. The filtered solution is evaporated to dryness and the residue boiled with absolute alcohol. The sodium nitrate dissolves, but it is stated that the calcium nitrate remains undissolved. The sodium nitrate is purified and identified in the same way as the barium nitrate above, [W. Fresenius expresses grave doubts as to the correctness of some of the above reactions, and points out that the author contradicts himself, as well as established fact, in stating that calcium nitrate is undissolved by boiling alcohol.] Another method consists in treating the animal matter with freshly precipitated strychnine, when strychnine nitrate is formed.ThisANALYTICAL CHEMISTRY, 47 can be purified by solution in boiling alcohol, the colour removed with lead acetate, and the nitrate finally obtained as crystals. A very sensitive reaction for nitric acid is obtained by adding salicin and 8-10 drops of concentrated sulphuric acid to the residue of an evaporated solution. A blood-red coloration is produced, which becomes violet on dilution. By THORNSMITH (J. A.mei*. Chem. Soc., 1899,21, 769--772).-Two grams of the sample are boiled with 100 c.c.of water and 2 grams of sodium hydroxide; when cold, the liquid is made up t o 250 c.c., well shaken, and filtered through a dry filter.50 C.C. are then concentrated t o 25 c.c., and after cooling to 80°, 25 C.C. of hydrochloric acid and 3 grams of potassium iodide are added. After remaining for 10 minutes, any arsenic acid will be completely reduced to the arsenious state and the liberated iodine is removed by diluting with water and carefully decolorising with solution of sodium thiosulphate. After first neutralising with, and then adding a n excess of sodium hydrogen carbonate, the arsenious acid is titrated in the usual maner with N/lO solution of iodine. M. J. S. Estimation of Arsenic in Paris Green. L. DE K. Estimation of Boric Acid in Tourmaline. By GEORGE W. SARGENT (J.Amer. Chem. Xoc., 1899,21,858-887).-The paper contains a summary of all the chief methods proposed for the estimation of boric acid, and the author’s opinion as to their respective merits. When dealing with tourmaline and similar minerals containing alum- inium, the chief difficulty arises from the incomplete separation of boric acid and umina, the only way of effecting complete separation is by the method proposed by Gooch, namely, volatilisation of the acid by means of methyl alcohol, coupled with the titration of the volatilised acid by By LOUIS C. JONES (Amel.. J. Xci., 1899, 8, 127-132. Compare Abstr., 1899,ii, 332).-Under certaindefinite conditions, a mixture of boric acid and mannitol possesses an acidic power sulticiently strong t o liberate iodine from a mixture of potassium iodide and iodate.The iodine is readily estimated by means of sodium thiosulphate, and represents the amount of boric acid. To obtain correct results, the following mode of procedure should be used: the borate is dissolved in the smallest possible quantity of hydrochloric acid and then diluted with water so that 50 C.C. of liquid shall contain about 0.1 gram of boric acid. The greater part of the free acid is then neutralised with sodium hydroxide and 5 C.C. of a 40 per cent. solution of potassium iodide and 10 C.C. of a 5 per cent. solution of potassium iodate are added and the liberated iodine at once removed by a few drops of a stror;g solution of sodium thiosul- phate. After restoring a very faint yellow colour by means of stand- ard iodine, 15 gramsof mannitol are added, and when dissolved, a stand- ard solution of sodium thiosulphate is added until the colour is bleached, when an extra 10 C.C.are added. A little more mannitol is added and after standing in a cool place for an hour, the excess of thiosulphate is carefully titrated by means of standard iodine solution. Thornson’s glycerol method. L. DE K. Iodometric Method for the Estimation of Boric Acid. L. DE K.4s ABSTKACTS OF CHEMICAL PAPERS. Stutzer and Hartleb’s Process for the Estimation of Com- bined Carbon Dioxide (Calcium Carbonate) in Soils. By H. SCHUTTE (Zeit. angew. Chem., 1899, 854--858).--The author criticises this process (Abstr., 1899, ii, 521) and states that, although good enough for marls, it is not sufficiently accurate for the estimation of small quantities of calcium carbonate in soils.The chief objection to the process is that ammonium chloride also acts on calcium in combination with silicic acid, and so causes the results to be in excess of the truth. Ferrous carbonate is also stated t o interfere with the accuracy of the method, and cannot always be Gas Washing Apparatus : Rapid and Accurate Process for the Estimation of Carbon Dioxide. By ROBERT SCHALLER (Zeit. angew. Chem., 1899,878-880).-The apparatus is essentially a U-tube 17 cm. long, one side of which is very narrow and serves to admit the current of the gas. The other side is filled with glass beads 2 mm. in diameter, which are moistened with the absorbing liquid. I n the new apparatus for estimating carbon dioxide, the substance is decomposed in a kind of test-tube with hydrochloric acid, and the carbon dioxide is swept out by means of a current of air which has been freed from this gas by passing through two of the U-tubes containing a solution of potash.The gas then passes through a similar tube containing sulphuric acid before it reaches the absorber. The latter consists of one or two similar weighed U-tubes con- taining 10 C.C. of 50 per cent. solution of potassium hydroxide. These tubes are, in addition, fitted with a small test-tube 8 cm. long con- taining 1 C.C. of strong sulphuric acid, and provided with a doubly perforated rubber cork. Through one of the holes passes a tube with a bulb in its centre and drawn to a narrow point a t the bottom, the object being t o prevent any escape of moisture from the apparatus.The other hole is fitted with a small beut tube t o allow the air to By HERMANN NOLL (Zeit. angew. Chem., 1899,859 -86O).-The process recommended by Tacke is found to answer for the estimation of calcium carbonate in marls, as it is not affected by the presence of small quantities of ferrous and aluminium compounds or of calcium silicate. The details of this process are briefly as follows : 0-2 gram of the powdered sample is mixed with 200 C.C. of warm water, 25 or 50 C.C. of N/5 normal sulphuric acid are added, and the carbon dioxide expelled by boiling. The excess of acid is then titrated with standard baryta, using phenolphthalein as indicator. New Volumetric Method for the Estimation of Magnesium.By I~ICHARD K. MEADE (J. Amer. CILem. Soc., 1899, 21, 746-752).- The process is fin indirect one. The ammoniacal solution free froin calcium is precipitated with a solution of sodium arsenate, and the excess of the reagent removed by washing with dilute aqueous ammonia (1 : 3). The arsenical precipitate is then dissolved in 75-100 C.C. of dilute completely decomposed by boiling. L. DE R. escape. L. DE I(. Estimation of Calcium Carbonate in Marls. L. DE K.ANALYTICAL CHEMISTRY, 49 hydrochloric acid (1 : l), potassium iodide added, and the liberated iodine titrated with standard solution OF sodium thiosulphate. L. DE K. Double Ammonium Phosphates of Beryllium, Zinc, and Cadmium in Analysis. By MARTHA AUSTIN (Amer. J. Sci., 1899, 8, 206--216).--When salts of these three metals are precipitated by boiling with excess of sodium ammonium hydrogen phosphate, double ammonium phosphates are obtained, which, on ignition, yield the corresponding pyrophosphatee.It has, however, been found impossible to accurately estimate beryllium in t h i s manner. Zinc may be estimated if care be taken to precipitate the metal from a neutral solution containing 10 per cent. of ammonium chloride. The process is also well adapted for cadmium ; the solution should contain 10 per cent. of ammonium chloride, be practically neutral, and be allowed t o remain several hours before By P. A. MACKAY (J. Amer. Chem. Soc., 1899, 21, 940-941).-The metal is treated with dilute hydrochloric acid in sufficient quantity t o dissolve the bulk of the zinc ; the undissolved portion then contains all the lead and cadmium present.After dissolving the residue in nitric acid, the lead is recovered in the usual way as sulphate, and in the filtrate the cadmium is pre- cipitated by hydrogen sulphide; the cadmium sulphide is then col- lected, washed, and redissolved in boiling hydrochloric acid and titrated with potassium ferrocyanide. It has been noticed, however, t h a t small quantities of cadmium can only be accurately estimated in the presence of zinc, and also that 1 gram-mol. of cadmium requires 3.676 gram-mols. of potassium ferrocyanide, instead of 3.767, as required by theory. A solution containing 10 grams of zinc per litre is prepared, and 50 C.C. of this titrated with a solution containing 36.76 grams of potassium ferrocyanide per litre, using uranium acetate as indi- cator. Another 50 C.C.is then added to the liquid containing the cadmium, and the titration is repeated; each C.C. excess of ferrom cyanide solution used in this titration represents 0.01 gram of cad- Electrolytic Estimation of Zinc in the Presence of Manganese. By EMIL J. RIEDERER (J. Amer. Chm. SOC., 1899, 21, 789-792).- Zinc may be conveniently separated from manganese by electrolysis. The solution, which should not contain chlorides or nitrates, but only sulphates, is mixed with 5 grams of ammonium lactate, 0.75 gram of lactic acid, 2 grams of ammonium sulphate, and then made up t o 230 C.C. ; the amount of zinc should not exceed 0.17 gram. The cathode, which should be 1& cm. distant from the anode, should consist of a platinum dish on which silver has been deposited, and during the action of the current it is advisable to use a mechanical stirrer so as t o obtain an even and uniform deposit of zinc ; the current may vary from 0.20-0*26 ampkres and the temperature from 15-26'.The time required for the complete deposition of the zinc does not VOL. LXXVIII. ii. 4 filtering. L. DE K. Analysis of Zinc for Cadmium and Lead. mium. L. DE K.50 ABSTRACTS OF CHEMICAL PAPERS. exceed 54 hours; the deposit is free from manganese. It is rapidly washed first with water, then successively with alcohol and ether, dried Direct Estimation of Aluminium in the Presence of Iron, Manganese, Calcium, and Magnesium, By WILLIAM H. HEM and EDWARD D. CAMPBELL (J. Amer. Chem. SOC., 1899, 21, 776-780).-A convenient bulk of the liquid, preferably containing the metals as chlorides, is heated to boiling and dilute ammonia is added so long as the precipitate readily redissolves; to reduce the ferric salts, a saturated solution of ammonium hydrogen sulphite is added drop by drop until the liquid is colourless. The aluminium is now precipitated by adding a few C.C. of phenylhydrazine and is washed with hot water containing about 10 per cent. of phenylhydrazine hydrogen sulphite free from excess of sulphur dioxide, On ignition, it leaves pure alumina. Chromium may also be separated from iron, calcium, mag- nesium, cobalt, or nickel in this manner. If, however, the mixture also contains phosphoric acid, the alumina will be contaminated with phosphoric oxide, which must be estimated By RODOLFO NAMIAS (Chem.Centr., 1899, i, 1224; from Annzcar. SOC. chirn. Milano, 1899, 54--56).-The author uses Volhard’s process when estimating man- ganese in samples of iron or steel. Use is made of carefully selected crystals of potassium permanganate which are reduced by boiling with hydrochloric acid ; the manganous chloride so obtained is precipitated with sodium carbonate, the precipitate is converted into manganoso- manganic oxide, and this is titrated with ferrous sulphate. It is recommended to first dissolve the steel, or iron, in hydrochloric acid before oxidising with nitric acid t o prevent the formation of By PORTER W. SHIMER (J. Anzer. Chern. Xoc., 1899, 21, 723-’724).-A new form of the reductor used in the estimation of iron and phosphomolybdate solutions.It is essentially a plain glass tube g-inch in diameter and 20 inches long drawn out and cut off at the lower end. It is united with a 4-inch funnel by means of wired rubber tubing fitted with a screw clamp. The lower end passes through a soft two-hole stopper of a thick-walled pint gas bottle connected with a filter pump through an intermediate safety bottle and valve. The passage of the solution through the reductor may be effected either by use of the pump or by Bunsen’s device of condensing steam. The tube is filled by placing a few small pieces of broken glass in the drawn out portion, then a little clean sand and finally about 3 ounces of amalgamated zinc. The speed of filtration isregulated by the upper clamp; a reduction takes about five minutes.i n the water oven for 5 minutes, and weighed. L. DE K. and allowed for. L. DE K. Volumetric Estimation of Manganese. organic compounds, which reduce permanganate. L. DE K. A Simplifled Reductor. L. DE K. Separation of Iron from Chromium, Zirconium, and Beryl- lium by the Action of Hydrogen Chloride on the Oxides. By FRANKE S. HAVENS and ARTHUR F,WAY (Am. J. Sci., 1899,8,217--215). -Gooch and Havens (Abstr., 1897, ii, 232) have shown that ironANALYTICAL CHEMISTRY. 51 oxide may be readily separated from alumina by heating the mixed oxides in a current of hydrogen chloride a t a temperature of 450-500' ; if to the mixture of sodium chloride and sulphuric acid a small quantity of manganese dioxide is added so as to also generate a little chlorine, the decomposition may be effected a t 200-300'.The authors now state that this method may be used successfully for the separation of oxide of iron from the oxides of chromium, zirconium, By GEORGE W,,SARGENT (J. Amer. Chem. Xoc., 1899, 21, 854-857).--The metal is dissolved in hydrochloric acid, oxidised with nitric acid, and the solution evaporated to dryness; the residue is taken up with a little dilute hydrochloric acid, and filtered into a separating funnel. The bulk of the ferric chloride is now removed by agitating the liquid with ether (Chase's method) and the solution, after removal of the dissolved ether, pre- cipitated by bromine water and ammonia ; the precipitate is freed from nickel by a second solution and precipitation.The combined filtrates containing the nickel are acidified with hydrochloric acid, freed from any copper by hydrogen sulphide, the filtrate boiled to expel hydrogen sulphide, and, when cold, a slight excess of ammonia added, and the liquid titrated with standard potassium cyanide, with silver iodide emulsion as indicator (Campbell and Andrew's method, Abstr., 1895, ii, 421). L. DE K. [Estimation of Chromic Acid in Acetylene Ririfiers.] By FRITZ ULLMANN and IRMA GOLDBERG (Chem. Centr., 1899, ii, 19-20 ; from J. Gashel., 42, 374--377).-See this vol., i, 1. By THOMAS BROWN, jun. (J. Arrrer. Chem. Soc., 1899, 21, 780--789).-Assu~ of soluble oves such, as Stibnite and Senarnzontite.--l gram of the finely powdered sample is boiled in a covered beaker with 25 C.C.of strong hydrochloric acid until about 15 c.c are left. 2 grams of tartaric acid are then added, and when dissolved, about 4-6 drops of strong nitric acid are intro- duced, and the boiling continued for a minute. The mixture is then diluted with water, the insoluble residue well washed on a filter, and the filtrate diluted to about 250 C.C. The liquid is now saturated with hydrogen sulphide and gently heated, and after remaining for an hour, the antimony sulphide is collected and washed with cold water, If there is any probability of lead being present', the precipitate must be dissolved in ammonium sulphide and the filtrate reprecipitated with dilute hydrochloric acid. The precipitate is collected on a counterpoised (double) filter, washed, dried at 1 lo', and weighed ; the bulk of the crude antimony sulphide is transferred to a weighed platinum boat, and then placed inside a piece of combustion tube and heated in a current of carbon dioxide, so as to convert i t into black antimony sulphide, which is then weighed.A correction is finally made for the small amount of sulphide not removable from the filter. Assay of insoluble ores such us Cewantite.--l gram of the powdered ore is mixed in a covered porcelain crucible with 10 grams of a mixture of equal parts of sulphur and dry sodium carbonate, covered with a little more of this flux, and fused for 10 minutes in a muffle. The mass and beryllium. L. DE K. Estimation of Nickel in Nickel-steel. Estimation of Antimony in Ores.52 ABSTRACTS OF CHEMICAL PAPERS. is extracted with water, and in the filtrate the antimony is precipitated as sulphide by means of hydrochloric acid ; the crude sulphide is then treated as previously directed.Assay of pccrtiably soluble ores : 0xysdphides.-These are assayed by the combination of the two processes described. The fire assay (fusion with potassium cyanide), although easy of execution, seldom agrees with the wet assays, and is particularly un- suitable for ores containing sulphur. Apparatus for the Preparation of Plant-Ash for Analysis, By G. M. TUCKER (Ber.. 1899, 32, 2583-2585).-This is A simplified form of Shuttleworth's apparatus (Dim. Gottingen, 1899, and J . Landwirtlmhajt, 1899, 173) designed to give a control over any loss of potash, chlorine, &c., by vola,tilisation. The platinum vessel, in which the plant is burnt, is of a conical form and is provided with a closely-fitting cover having two apertures.I n the central aperture is soldered a tube passing t o the bottom of the vessel and down this tube, by means of which air enters, passes the handle of a stirrer. By means of the other aperture, the gases from the combustion chamber are led through a small wash-bottle, beyond which is an aspirator which draws air through the whole apparatus. Any ash mechanically carried by the air stream is stopped by a loose movable cap fitted to the exit tube. The conical shape of the vessel allows of more uniform heating, which is carried out first in a conical sand-bath, and, later, over a bare flame, the vessel being then fitted into a hole in a n asbestos card.T. H. P. Ammoniacal Copper Solutions and Hydroxylamine as a Test for Acetylene. By LUDWIG ILOSVAY YON NAGY ILOSVA (Bey., 1899, 32, 2697-2699).-Ammoniacal copper solutions decolorised by reduc- ing with hydroxylamine can be used for detecting acetylene, but the value of the reagent and the coIour of the precipitate depend on the proportions used. The quantities recommended are, for 50 C.C. of solution, (I) cupric chloride ( + 3H,O), 0.75 gram ; ammonium chloride, 1.5 grams ; aqueous ammonia (20-21 per cent.), 3 c.c., hydroxylamine hydrochloride, 3 grams. (11) cupric nitrate ( + 5H,O), 1 gram ; aqueous ammonia, 4 c.c., hydroxylamine hydrochloride, 3 grams. (111) copper sulphate (+ 5H,O), 1 gram ; aqueous ammonia, 4 c.c., hydroxylamine hydrochloride, 3 grams.The reagent gives a splendid red precipitate with acetylene, but begins to deteriorate after three days. By EDMOND JANDRIER (Chem. Centr., 1899, i, 1296; from Ann. chim. anal. appl,, 4, 156).-Resorcinol recommended by Mulliken and Scudder (Abstr., 1899, ii, 388) gives, with acraldehyde, almost the same colour as with formaldehyde, and is consequently not a safe test for the detection of methyl alcohol ; the test is also interefered with by the presence of furfuraldehyde. It is proposed to use gallic acid instead, as recommended by Barbet and Jandrier (Abstr,, 1898, ii, 265). By WILHELM PRESENIUS and LEO GRUNFIUT (Zeit. and. Chem., 1899, 38, 472-512). -Wines obtained by the fermentation of solutions of glucose in L. DE K. T. M. L. A Colour Test for the Detection of Methyl Alcohol.L. DE K. Recognition of Marc Wines (Tresterweine).ANALYTICAL CHEMISTRY. 53 contact with grape marc ( T ~ e s t e ~ ) have hitherto not been distinguish- able with certainty by chemical analysis from natural wines. The authors believe they have discovered a trustworthy criterion of mch mines in the fact that they contain little or no tartaric acid combined with alkaline earths, whilst natural white wines always contain calcium tartrate to an extent which permits a limit of about 0.1 gram of tartaric acid per 100 C.C. to be established. The recognition of this fact in their composition depends on the estimation of the ratio of the total tartaric acid to the total alkalinity of the ash and the alkalinity of the portion of the ash soluble in water. With reference to the observations of Kulisch, Kohlmann, and Hoppner (Abstr., 1899, ii, 341 ; also 1895, ii, 465) on this subject, they point out that, in the estimation of the soluble constituents of the ash, a very limited amount of washing (30-80 C.C.of hot water for the ash of 100 C.C. of wine) must be employed, so as to leave the calcium carbonate of the ash as far as possible undissolved, Formuls are given for calculating the analytical data. An exception to the above rule occurs in the case of marc mines to which tartaric acid has baen added during manufacture. Red wines also contain much smaller proportions of calcium tartrate than white mines. The above criterion must therefore in all cases be considered in connection with other characteristics of marc wines, namely, a high proportion of ash to total solids, high percentage of tannin, and occasionally also of volatile acids.It can likewise only be employed in cases where the ash of the wine exhibits a normal total alkalinity and the proportion of sulphate is not abnormal, since it is obvious that excessive sulphuring would disturb the relations of the various tartrates in the wine. The alkalinity factor of natural wines, that is, the alkalinity of 0.1 gram of the ash expressed in cubic centirnetres of normal alkali, seldom exceeds 0.8 or 1.0. Wines exhibiting a much lower alkalinity factor should be examined for excessive sulphuration. h much higher factor would suggest a search for added organic acids. The ratio of glycerol to alcohol in marc wines seems generally t o exceed the limit of 7 : 100 laid down for natural wines.With regard to the use of the large proportion of tannin as a criterion of marc mines, it is evident that no limit can be laid down, since the proportion of tannin in natural wines varies widely. Barth has argued that the ratio of total extract (after deduction of sugar, ash, and non-volatile acids) to tannin may be regarded as constant, and employs the factor 5 for multiplying the tannin. The authors show, however, that in a great number of natural wines, even the factor 4 would be too high and point out the need for further in- vestigations of this sub ject. The authors also give a complete analysis of the ash of a marc wine prepared by themselves, and some analyses of raisin-wines, one of which they also manufactured, M.J. S. Volumetric Estimation of Alcohols, especially of Fusel-Oil in “Brandies.” By FRANZ ADAM (Chem. Centr., 1899, i, 1226-1227; from Uesterr. Chem. Xeit., 2, 241-243).-A solution of acetyl chloride in chloroform, after it has been well shaken with water and so54 ABSTRACTS OF CHEMICAL PAPERS. decomposed into acetic and hydrochloric acids, is titrated with normal alkali. Another portion of the chloroform solution is allowed to act on the alcohol, whereby it is partly decomposed into hydrogen chloride and alkyl acetate ; after decomposing the remainder with water, the titration is repeated, and the difference in alkali represents the amount of the alcohol. To estimate amyl alcohol in brandy, the sample is diluted to 20 per cent.strength by volume, and the fuse1 oil then extracted by repeated agitation with pure chloroform, This is re- peatedly shaken with water, then dried over burnt gypsum and treated with standard solution of acetyl chloride, &c. The process is also suitable for the estimation of the alcohol number of ethereal oils. L. DE K. Action of Bromine on Phenol and Cresols with Reference to the Analysis of Mixtures of these Compounds. By Huao DITz and FRANZ CEDIVODA (Zeit. angew. C'hem., 1899, 873-817 and 897-903).-An alkaline solution of a phenol is mixed with a known excess of bromine dissolved in caustic potash, acidified with dilute HCl (1 : l), shaken vigorously for 1 minute, and titrated with standard thiosulphate solution in the presence of potassium iodide ; under these conditions, 1 mol.of o-cresol or p-cresol takes up 2 atoms of bromine, whilst m-cresol and phenol each combine with 3 atoms. I n the case of p-cresol, it is essential that the duration of the experi- ment should not exceed the prescribed time limit, otherwise the amount of bromine required exceeds 2 atoms. When the alkaline solution of a phenol and excess of bromine is acidified with concen- trated sulphuric acid, allowed to remain for 10 minutes, subsequently shaken for 5 minutes, and then filtered through glass wool or sand, it is found, by treating the filtrate with potassium iodide and titrat- ing with thiosulphate, that the amount of bromine taken up corresponds with 3 atoms in the case of o-cresol and p-cresol, and with 4 for m-cresol and phenol.These results are applied to the analysis of mixtures of two or more of these constituents. The mixture of the phenols is separated from other impurities by extraction with ether ; the ethereal extract is dried over calcium chloride or dry sodium sulphate, distilled up to 180' to remove ether and last traces of moisture, the fraction distilling over above this temperature is treated with bromine by the two methods just described, and the amount of phenol, m-cresol and mixed ortho- and para-cresol determined by means of the following equations : (i) x + y + x = a ; (ii) 3Br/94.06.x + 2Br/108*08.y + 3Br/108'08.x = 6; and (iii) 4Br/94*06.x + 3Br/108.08.y + 4Br/l08-08.x = c, where x and x represent the amounts of phenol and m-cresol, y the amount of the mixture of the ortho- and para-isomerides, a the total quantity of phenolic compounds, b the amount of bromine absorbed when the solution is acidified with dilute hydrochloric acid, and G the amount taken up when concentrated sulphuric acid is employed, and the bromo-derivatives filtered off before titration.G. T. M. Simpliftcation of the Phenylhydrazine Test [for Sugar in Urine.] By ALBERT KOWARSKY (Chem. Centr., 1899, i, 1294; from Bed. klin. Wochschr., 36, 412-414)-5 drops of pure phenyl-ANALYTICAL CHEMISTRY, 55 hydrazine are mixed with 10 drops of glacial acetic acid and 1 C.C. of brine, and the mixture boiled for 2 minutes with 3 C.C. of the supposed diabetic urine and then allowed to cool slowly. If the characteristic osazone crystals form after a few minutes, the urine is practically free from sugar.If appreciable traces of albumin are present, these must be first removed by coagulation at the boiling Stability of Gun-Cotton and Smokeless Powder. By C. HOITSENA (Zeit. urzgew. Chem., 1899, 705-710).-A review of the methods in use for testing these explosives as t o their keeping powers. The author thinks that at present there are not sufficient grounds for substituting the Simon-Thomas test for those of Abel and of Guttmann. For comparison, a stock of matrerials of guaranteed make Chemistry of Butter- Fat. 11. Chemical Composition of Butter-Fat. By C. A. BROWNE, jun, (J. Anter. Chem. Xoc., 1899, 21, 807-827. Compare Abstr., 1899, ii, 709).-This is a long investiga- tion as to the true composition of butter-fat ; on the whole, the article does not lend itself to useful abstraction.The following analysis of the fatty acids obtained from 100 grams of butter is interesting as it confirms the result of Hehner and Mitchell (Abstr., 1897, ii, 287) as to the small proportion of stearic acid contained in butter : dihydroxystearic, 1.00 ; oleic, 32.50; stearic, 1-83 ; palmitic, 38.61 ; myristic, 9.89 ; lauric, 2.57 j capric (decoic), 0.32 ; cnprylic (octoic), 0.49 ; caproic (hexoic), 2.09 ; butyric The Reichert Number of Butter. By JAMES H. STEBBINS (J. Amer. Chem. Soc., 1899, 21, 938-940).--From the result of 317 analyses of genuine butters, the author is in favour of placing the lowest allowable limit of the Reichert number at 11.5. Analysis of Bees Wax. By ANGIOLO FUNARO (L'Orosi, 1899,22, 109--123).-The adulterants of bees wax are Japan and Carnaiiba wax, animal fats, stearic acid, resins, cerasin, and paraffin wax.The various methods for detecting and estimating these impurities are described in detail and discussed, By EARL BERNHARD SOHN (Biecl. Celztr., 1899, 28, 298-299 ; from Milchxeit., 1898, 498). -The employment of pure furfuraldehyde which has been recently distilled in a vacuum, and is colourless, is of great importance in testing for margarine by means of the Baudouin test. I n case pure furfuraldehyde is not available, the employment of furfuramide is recommended. The following process for detecting for sesame oil is given : the melted filtered butter (10 c.c.) is extracted with hydrochloric acid of sp. gr. 1.125 three or four times or until the acid remains colour- less.Five C.C. of the butter fat is then shaken with a solution of 1-08 grams of furfuramide in 100 C.C. of absolute alcohol (0.1 c.c.) and hydrochloric acid of sp. gr. 1-19 (10 c.c.) for a t least half a minute. If the fat solidifies, it is heated at 60-'70°, If the hydrochloric acid heat. L. DE K. and of known age should be kept. L, DE K. acid, 5.45; total, 94.75. L. DE I(, L. DE K. T. H. P. The Sesame Oil Reaction and Sesam6 Butter.56 ABSTRACTS OF CHEMICAL PAPERS. shows a red coloration which does not soon disappear, sesame oil is proved to be present ; with small amounts of sesame oil, the colora- tion takes hours to appear. [Testing Butter for Oil of Sesam6.1 By H. WEIGMANN (Bied. Centr., 1899, 28, 629).-See this vol., ii, 40.Detection and Estimation of Formaldehyde in the Free State and in its Compounds. By G. H. A. CLOWES and BERNIIARD TOLLENS (Ber., 1899, 32, 2841-2848).-Formaldehyde is estimated by heating with phloroglucinol and hydrochloric acid at 70-80" and weighing the phloroglucide, C,H,O,, which is produced under these conditions. In most of its compounds, the formaldehyde can be estimated in this way by using hydrochloric acid and water, but in some cases it is necessary to employ a more or less concentrated sul- phuric acid t o hydrolyse the methylene compound. A large number of analytical results are given. [Separation of Acetone from Acetoacetic and Acetonedi- carboxylio Acids]. By LUIGI ~ABBATANI (C'hem. Centr., 1899, ii, 22-23 ; from Atti Red. Accad.Torino, 34)-See this vol., ii, 32. New Reaction of Acetone and a New Method for the Detection of Aliphatic Amines. By ENRICO RIMINI (L'Orosi, 1899, 22, 40-44).-1f to a mixture of an aliphatic monamine with a 10 per cent solution of acetone a few drops of a concentrated solution of sodium nitroprusside are added, a magenta coloration is formed, which gradually becomes more and more intense and lasts for several hours ; the addition of a few drops of acetic acid changes it to an intense violet. When the hydrochloride of the amine is employed, too much alkali must not be added, as in this case the orange tint of Legal's reaction is obtained ; this, however, vanishes much more quickly than the red coloration. By using aniline or phenylhydrazine in place of the aliphatic monamine, no coloration is obtained, whilst benzylamine gives a faint violet colour which forms slowly. With secondary aliphatic monamines, Legal's reaction takes place, whilst tertiary amines act simply as alkalis.No coloration is obtained if, instead of acetone, one of the following substances is employed :-formaldedyde, acetaldehyde, paraldehyde, isobutaldehyde, or valeraldehyde, chloral, benzaldehyde, salicyl- aldehyde, anisaldehyde or cinnamaldehyde, furfuraldehyde, dextrose, acetophenone, benzophenone, ethyl acetoacetate, camphor, fenchone, or tanacetone ; with concentrated solutions of pyruvic acid, a brown coloration forms and changes to green, whilst with dilute solutions a ponceau tint is developed. The presence of 1/10 per cent. of acetone in a solution can be detected in this way,and the formation of the colour is not interfered with by the presence of alcohol or acetaldehyde.Use is made of this reaction of acetone and of that of acetaldehyde formerly described (Rendiconti Amminisir. Civile. Minister0 del Interno, 1898) by the author t o distinguish between primary and secondary aliphatic amines. The formation of an azure coloration on adding sodium nitroprusside and aldehyde to the solution of an amine, shows N. H. J. M. T. M. L.ANALYTICAL CHEMISTRY. 5 7 that the base is secondary, whilst if acetone is added in place of the aldehyde, and a magenta tint is given to the liquid, the amine is primary. These reactions are obtained in mixtures of primary, secondary, and tertiary amines. Volumetric Estimation of Quinones Derived from Benzene. By AMAND VALEUR (Compt.rend., 1899,129,552-553).-The method described is based on the reduction of quinones by hydriodic acid, all secondary reactions being avoided by operating in the following manner. A cooled mixture of 20 C.C. of concentrated hydrocbloric acid with an equal volume of 95 per cent. alcohol is quickly added to 20 C.C. of a 10 per cent. aqueous solution of potassium iodide. The liquid is then poured into the alcoholic solution of the quinone and the iodine liberated is titrated with N/10 sodium thiosulphate solution. The process is expeditious, and is shown to give accurate results in the case of quinone, dichloroquinone, toluquinone, and thymoquinone ; i t is also applicable to unstable compounds of quinones, such as pheno- A Colour Reaction to Distinguish the Hydrochlorides of m-Phenylenediamine and p-Phenylenediamine.By L. CUNIASSE (Chem. C’entr., 1899, i, 1297 ; from Ann. chim. anal. appZ., 4,156-157). -The meta-compound, when heated in aqueous solution with a few drops of a 1 per cent. solution of acetaldehyde in proof-spirit slightly acidified with acetic acid, yields, when cold, a splendid yellow colora- tion with a strongly green fluorescence. The para-compound turns Estimation of Urea in the Tissues: and the Amount con- tained in the Liver. By RUDOLF GOTTLIEB (Chem. Centr., 1899, i, 1298 ; from Arch. Exp. Path. Pharm., 42, 238-249).-The urea is first isolated as a mercuric compound by von Schroeder’s process; finally, the alcoholic solution of the urea is precipitated by an ethered solution of oxalic acid and the acid contained in the urea oxalate found by titration.The blood of starving dogs contains 0*011-0*02 per cent. of urea ; when on meat diet, it increases to 0.033-0-056 per cent. The liver, contrary to expectation, contains only 0.008-0*02 per cent. T. H. P. quinones and quinhydrones. N. L. orange-red without any fluorescence. L. DE K. L. DE K. Detection of ‘( Saccharin ” in Wines. By DIOSCORIDE VITALI (Chem. Centr., 1899, i, 1297-1298; from Boll. Chirn. -Farm., 38, 297-300).--The only trustworthy method is based on the conversion of the sulphur of “ saccharin ” into sulphate. Small quantities of “ saccharin ” may be estimated by precipitation with mercuric nitrate. The precipitate is then weighed and the amount of mercuric oxide in it estimated, the difference representing the saccharin.” L.DE K. Employment of Chloral Hydrate in the Estimation of Alkaloids. By EDUARD SCHAER (Zed. anal. Chem., 1889, 38, 46 9-472). -The author completely confirms Len z’s opinion (A bs tr.,58 ABSTRACTS O F CHEMICAL PAPERS. 1899, ii, 391) as to the value of chloral hydrate for the ex- traction of the vegetable alkaloids from plant products, especially when large proportions of resins and ethereal oils are present. A still more powerful solvent, especially for solid fats and waxes, is a concentrated alcoholic solution of chloral alcoholate. Mauch has made the observation that thesalts of the alkaloids are dissolved by aqueous chloral hydrate as readily as the free bases.A ditliculty is, however, encountered in the extraction of strychnine from nux vomica seeds and Ignatius beans, in consequence of the presence of mucoid compounds, which (in common with certain varieties of starch and also animal gelatin) swell up in chloral hydrate, and render it viscous. Caoutchouc and gutta percha also resist completely the sol- vent action of chloral hydrate, and, when they are present t o a large extent, no advantage results from the use of chloral. Rfauch asserts that when the chloral hydrate solutions of the bases are warmed, the bases are often converted into formates. So far as the author’s experience extends, the chloral solutions of the alkaloid salts exhibit great permanence, and, may be heated on the water-bath without loss of alkaloid. The Chromic Acid Test for Cocaine.By GEORGE L. SCRAEFER (J. Amer. Chem, Xoc., 1899, 21, 936-938. Compare Abstr., 1899, ii, 715).-In reply to the criticisms of Squire (Chemist and Druggist, April 22), Cownley (Plmrm. J. Tmns., April 15), and Merck (Phawn. Zeit., No. 42), the author upholds his own chromate process. M. J. S. L. DE K. The Active Principle of Cayenne Pepper. By KARL MICKO (Chem. Centr., 1297 ; from Zeit. Unters. Nuhr.-Gennussm., 2, 411-412. Compare Abstr., 1899, i, 716).-Capsacutin described by Mijrbitz is probably the same compound as the author’s capsaicin, Ci,~iY03N, the-active principle of Capsicum unnuum, L., and Capsicum fustagautum, BZ. Reactions of cupsuicin.-When moistened with neutral solution of ferric chloride and a little alcohol, insoluble,. greenish-blue drops are noticed, but this test is not parlicularlg delicate. Strong sulphuric acid and a particle of sugar cause, after some hours, a fine violet coloration.When heated on the water-bath with strong hydrochloric acid, an alkaloidal substance is obtained which yields precipitates with platinic chloride, iodine, and potassium mercuric iodide. L. DE K. Detection of Nucleo-albumin in Urine by means of Tannin. By LEON GARNIER and L. MICHEL (J. Phurm., 1899, [vi], 10, 150-1 52).-The precipitate produced when tannin (Almh’s reagent) is added t o urine previously diluted with its own volume of a satu- rated solution of sodium chloride, consists chiefly of tannin with a small amount of sodium chloride and traces of nitrogenous substances. The phosphates present in the urine cause the precipitation of the tannin.Ott’s statement (Jahyesber. fiir Thierchemie, 1895, 567) that the formation of this precipitate is a sure test for the presence of nucleo-albumin is thus proved to be erroneous. H. R. LE S.ANALYTICAL CHEMISTRY. 59 Estimation of the Products of Digestion with Pepsin. By JEAN EFFRONT (Chem. Zeit., 1899,23, 770-771 and 783-784).-A solution of tannin in tartaric acid may be employed for separating proteoses and peptones, as proteoses are precipitated and peptones remain in solution ; its use is suggested as a reagent in the analysis of the products of peptonisation. For a complete analysis, the follow- ing determinations should be made : 1, total nitrogen; 2, total pro- teids ; 3, syntonins ; 4, proteoses ; 5, peptones.For the total proteids, the solution is precipitated with phosphotungstic acid, the precipitate washed with N/2 hydrochloric acid and dried without removal from the filter; the nitrogen in this precipitate is estimated, and the number thus obtained multiplied by 6.25 gives the total proteids. Syntonins may be estimated by exactly neutralising the solution with sodium hydroxide (1 : lo), or even more correctly by estimating the total proteids in the original solution and then in the neutralised solution, the difference between the two determinations giving the percentage of syntonins. Proteoses may be precipitated by a tartaric acid solution of tannin after the albumin is first removed by coagula- tion, and the syntonins by neutralisation.The solution is made by dissolving tannin (50 grams) in water (500 c.c.) adding N-sodium hydroxide (50 c.c.), making up to 1 litre, and then adding 15 C.C. of a 10 per cent. tartaric acid solution. The precipitate is well washed, then dried between filter-paper, and the percentage of nitrogen deter- mined. The peptones which remain in solutlion may be precipitated by phosphotungstic acid. The percentages of peptoses obtained by salting out with zinc sulphate, and by precipitating with a tartaric solution of tannin, are not concordant when the proteid solution has undergone an appre- ciable amount of peptonisation. The results obtained also indicate that phosphotungstic acid is an uncertain reagent for estimating peptones. It appears that, after pro- longed peptonisation, the peptones are no longer precipitated by this reagent, and it is concluded that the compounds known as peptones are really substances of very different characters, some of which are precipitated by the phosphotungstic acid reagent and others not, and, further, that peptones themselves are acted on by pepsin.The process of analysis described above gives accurate results only when peptonisation has proceeded for a short time. J. J. 8. Analysis of Glue and Leather. By WILHELM FAHRION (Chem. Zeit., 1899,23, 452-453).-Glue and leather contain a small quantity of fatty matter, which, however, is partially oxidised, and incompletely soluble in light petroleum, To estimate fat in glue, 10 grams of the cut-up sample are heated with 40 C.C. of S per cent.alcoholic potash on the water-bath, adding some more alcohol if necessary, and finally evaporating to dryness. The residue is then dissolved in hot water and rendered acid with hydrochloric acid. The liquid is heated nearly to boiling for half-an.hour, then transferred to a separating funnel, and, when cold, shaken with ether and then left to settle overnight. The aqueous layer is drawn off, the ether poured off, and the hydroxy- acids which adhere to the sides of the funnel are dissolved in hot60 ABSTRACTS OF CHEMICAL PAPERS. alcohol. Both the ethereal and alcoholic solutions are then evaporated in the same tared dish. The rasped sample is first extracted with light petroleum to obtain the bulk of the fat. It is then treated with alcoholic potash as described, but on faintly acidifying with hydrochloric acid, a precipitate of tannin and its derivatives (phlobaphens) is obtained which occludes some more fat, and also any hydroxy-acids.The f a t may be extracted with ether, but as yet a process is not known for the separation of fatty hydroxy-acids from the tannin precipitate. The latter may be dried and weighed, but as it always contains mineral matters, it must be finally burnt to ash. By C. ASCHMAN and HAROLD FABER (Chem. Zeit., 1899, 23, 61).--25 grams of the finely divided air- dried sample are put into a porcelain dish and heated for 1 hour on the boiling water-bath with 100 C.C. of aqueous caustic soda (50 grams per litre); the liquid is poured off and the extraction repeated several times. The whole, including the deposit, is then made up to 510 C.C. (10 C.C. being occupied by the insoluble matter) and after the supernatant liquid has become clear, an aliquot part is syphoned off and tested as follows : 0.125 gram of humic acid (mid. huminic. pur.) is dissolved in soda solution and diluted to 500 C.C. ; 5 C.C. are diluted to 100 c.c., 10 C.C. of dilute sulphuric acid (1 : 5) added, and the liquid boiled with addition of solution of potassium permanganate (0.32 gram per litre) until the colour is no longer discharged. 10 C.C. of a solution of oxalic acid (0.63 gram per litre) are now added and the excess of oxalic acid titrated back with the permanganate solution. The usual blank experiment is then made, and having thus found the humic acid value of the permanganate, the amount of humic acid in the soil may be ascertained by titration. For the success of the ex- periment,it is necessary that the amount of the humic acid present in the aliquot part of the solution of the sample shall not materially differ from the amount used in the standardising experiment. Alcoholic potash is also useful in leather analysis. L. DE K. Estimation of Humus in Soils. L. DE K. Analyses of Ginger. By EDWY G. CLAYTON (Analyst, 1899,24, 122--125).--The author has analysed 37 samples of ginger from various sources (some of them being again analysed after steeping in water or proof spirit) and has tabulated the results. The analysis comprises the following estimations : soluble and in- soluble ash; the extracts obtained with cold water, ether, absolute alcohol (after complete extraction with ether), and light petroleum respectively ; essential oil ; alkalinity calculated as potash, and chlorine in the soluble ash ; in certain cases, the ash, alkalinity of the ash, chlorine in the cold water extract, water, and silica were also deter- mined. It appears that the commercial processes -washing, scraping, cutting, grinding, and separation of fibre-do not so alter the composi- tion of genuine ginger as to convey the idea that it has been adulterated with spent ginger, except in some cases where excessive washing had occurred. L. DE K.