ANALYTICAL CHEAIISTRT. 83 An a 1 y t i c a1 C h e m i s t r y. Estimation of Phosphorus in Phosphor-tin. By W. HEMPEL (Her., 22, 2478).-Phosphor-tin is best analpsed by Wohler’s chlo- rine method as employed in the analysis of Fahl-ore. The stannic chloride and phosphoric chloride which are formed are collected in concentrated nitric acid (about 10 c.c.) ; the vessel is washed out with dilute nitric acid (1 : 2),. and the phosphoric acid is precipi- tated with ammonium molybdate and estimated directly. F. S. K. Recognition of Phosphoric Acid of Mineral Origin. By J. STOKLASA (Chern. Centr., 1889, ii, 57, from Listy. Chem., 13, 153-154) .-The author contends that the percentage of fluorine in bones as given by different authors is decidedly too high. Raw bones do not Rhow any fluorine by the Loren2 reaction, whilst incinerated bones give only a slight indication of the presence of this element Lvhen tested in the same way. On the other hand, the author found the fossil bones of Elephas primigenius to contain 3.54 and 4.36 per cent.of ferric oxide and 2.08 and 52-98 per cent. of fluorine respectively in tlwo different specimens. Bones found in the older alluvids on the island of Festigos contained 2.42 per cent. of‘ ferric oxide, 1.58 of fluorine, 80.04 of tricalcium phosphate. Superphosphates prepared from this phosphate and also from bones were testetl by the Lorenz reaction, with the result that fluorine was only found in the former, whilst of the several samples of the latter class of super- phosphates only that from hone-ash gave any indication of fluorine, and then but slightly.On the other hand, remarks the author, Lorenz seems to have overlooked the fact that a series of mineral phosphates exist which contain but very little fluorine. J. W. L. Rapid Method of Estimating Arsenic. By E. POLENSKE (Chem. Centr., 1889, ii, 58-59, from Pharm. Zeit., 34, 299-300). -The method consists in evolving the arsenic as hydrogen arsenide in a Marsh apparatus, deposition of the arsenic in a tube having three bulbs blown on it, and weighing first that part of the tube containing the ‘‘ mirror,” and secondly the tube after dissolving OE the arsenic. The evolution flask is recommended to be of a capacity of 250 c.c., and to contain 80-100 grams of zinc. The apparatus includes an acid funnel 30 cm.long, having a U-formed bend, and the evolution flask is also connected with a washing flask containing lead nit.rate solution, to which is attached a tube containing calcium chloride, and a t the farther end, potash. The decomposing tube is attached to this. All air is expelled from the apparatus, first, by adding 5 C.C. of concentrated sulphuric acid and 20 C . C . of wat(’r, and the re- agents are at the same time tested for arsenic by heating the tube. The solution should not contain more than 4-5 milligrams of arsenic, and is allowed to drop into the acid funnel from a burette a t the rate of 0.5-1.0 C.C. per minute. The gas escaping from the end of the 9 284 ABSTRACTS OF CHEMICAL PAPERS. tube is passed through a solution of silver nitrate, and the evolut.ion should not be too rapid for the bubbles to be counted.During the evolution of the hydrogen arsenide, two hunsen lamps are placed under the second and third bulbs : in order to test whether all the arsenic is evolved from the flask, the first is heated, and i f a " mirror " forms it must be driven onwards into the second bulb, and after a time the same test may be applied again. When it is thus proved that arsenic no longer escapes from the flask, the tube is disconnected and reversed,and any arsenic which may have been deposited iu the third bulb is driven into the narrow tube between the second and third bulbs. All the arsenic having been thus collected in this part of the tube, it is cut off and weighed, and after dissolving off the arsenic with nitric acid, it i s weighed again. the difference giving the weight- of arsenic.Metallic salts aria organic substances interfere with the acciiracy of the method. Behaviour of Silicates when Fused with Phosphates. By K. HAUSHOFER (Chem. Centr., '1889, ii, 53, from Sitzungsber. der mat h.-naturzu. Abt. ba?yr. Akrrd. Wiss., 1889, 8-1 l).-Many silicates, when fused to a bead with an alkaline phosphate in the blowpipe, swell up, evolve gases, and finally insoluble silica, either in the form of a skeleton or as flakes, is deposited. The evolution of gases is referred by the author to the presence of chlorine, sulphates, or water, and h e draws attention to this part of the reaction as an aid in the recognition of the silicate under examination. For instance, hauyn and sodalite, owing to the evolution of the chlorine and $ulphuric anhydride, may be distinguished from nepheline, which is but slowly attacked.In like manlier, humite is distinguished from olivine by the evolution of hydrofluoric acid, and tourmaline and axinite from beryl. The micas of the scapolite group, epidote and vesiivian, lose their water of eonstitu t'ion, whereas the felspars, amphibole, and granites are but slowly attacked. Similar distinctions may be drawn between crystallised kaumerite and the dense rhodo- chrome, and between pyrophyllite and agalmstolite. Technical Analysis of Commercial Sodium Sulphide. By B. SHETLIK (Chem. Centr., 1889, ii, 211, from Listy. Chem., 12, 205- 206).-10 grams of the sulphide is dissolved in water, the solution diluted to i$ litre, and 50 C.C.titrated with normal sulphuric acid, phenolphthalein being used as an indicator. If the titration is made in the cold, the quantity of acid required must be doubled, whereas if it is carried out a t a boiling heat and the acid added until the red colour does not reappear on further boiling the solution, the acid used is equivalent to the sulphide. Insoluble sulphides which are de- composed by dilute acid may be titrated in this way. Qualitative Analysis of the Ammonium Sulphide Re- cipitate. By P. MATER (Ber., 22, 2627--26:10).-The presence of chromium in t,his precipitate renders impracticable the separation by solntion in hydrochloric acid and precipitation of the iron and aluminium by boiling with sodium acetate, since the chromium is sometimes wholly and sometimes partially prccipitated, whilst if no J.W. L. J. W. L. J. w. L.ASALYTICAL CHEMISTRY. 85 iron is present, it all remains in the filtrate. The author has found that the presence of excess of iron (at least 5 atoms to every atom of chromium) ensures the total precipitation of the chromium. He, therefore, recommends t h a t where chromium is suspected and iron is not present in large quantities, ferric chloride should be- added in excess before boiling with sodium acetate. L. T.. T. Analysis of Aluminium Sulphate. By F. BETLSTSIN arid T. GROSSET (Chem. Centr., 1889, ii, 60; from Melanges, Yhys. Chim. Bull., St. Pe'tersbourg, 13, 42--47).--The authors recommend the fol- lowing method. 1-2 grams of substance is dissolved in 5 C.C. of water, 5 C.C.of a cold saturated solution of ammonium sulphate added, and the mixture well stirred during a quarter of an hour. 50 C.C. of 95 per cent. alcohol is added, and the precipitated ammonia alum filtered off and washed with 50 C.C. of alcohol. The filtrate contains all the free acid, which is deteimined by concentratiiig and titrating with decinormal alkali. The whole of the aluminium sulphate is precipitated as ammonia alum. Estimation of Chromium and Copper in Iron and Steel. By C. REIXHARDT (Clrenz. Centr., lt89, ii, bU-61, frcim Stahl w. Eistn 9, 404--405).-For the determination of chromium, 10 grams of borings or filings are dissolved with 100 C.C. hjdrochloric acid in a covered beaker of 500 C.C. capacity, first without heat, then at a boil- ing heat, oxidised with potassium clilorate, concentrated to one half the volume, filtered into a &litre Erlenmeycr flask, and the insoluble residue washed several times with dilute hydrochloric acid on the filter, and finally with water.The solution is now reduced a t a boiling heat hy the a,ddition of 10-20 C.C. of sodium hypophosphite solutiou (200 grams in 400 C.C. of water), and afternards the chromium is pre- cipitated by the addition of zinc oxide in excess. The precipitate is dissolved in hydrochloric acid, a little more hypophosphite added, and the precipitation repeated. The chromium is separated from the zinc by precipitation with ammonia, which precipitation must be re- peated. The chromic oxide, after ignition, is iused, together with the insoluble portion of the material, with 8 grams of a mixture of 4 parts of sodium chloride, 1 part of sodium carbonate, and 1 part of potassium chlorate.From the dissdvad flux, the manpnese is pre- cipitated with alcohol, the silicic acid with lijdrochloric acid and a little sulphurous acid, and the chromium finally precipated as oxide with ammonia. The copper is determined iii 10 grams of the material : the hydrochloric acid solution is reduced with sodium hypophosphite, and the copper precipitated with hydrogen sulphide. Volumetric Estimation of Chromium in Iron and Steel. By E. WAHLHERG (Chem. Centr., 1889, ii, 194, from Bcrg. a. Huttenni. Zeit., 48, 18G-181).-0~5 gram of the metal is dissolved in boiliud. nitric acid, sp. gr. 120, evaporated to dryness, ground up, transfern*d to a platinum crucible, mixed with a mixture of 2 grams of magnesm, 1 gram of potassium chlorate, and 1 gram of fiodium carbonate, aud the whole heated, at first gently, then in the blast flame for one hour.J. W. L. J. W. L.86 A BSTKACTS OF CHEMICAL PAPERS. The mass is dissolved out with 50-100 C.C. of water, any s m a l l quantity of manganic acid reduced by a drop or two of alcohol, acidified with sulphuric acid, arid the chromic acid estimated by adding a known quantity of ferrous sulpbate and determining the amount unosidised by titrating with potassium perrnanganate. J. W. L. Microscopical Test for Tantalum and Niobium. By K. HAUSHOFER (Chem. Centr., 1889, ii, 62-63, from Sitzungsber. der math.-nnturzu. Abt. bayr. Akad. Wiss., 1889, 3-8).-The substance to be tested is fused with a very small bead of sodium carbonate in the hottest part of the bunsen flame €or 30-40 seconds.It is then treated on the object glass with one drop of water, and the form of the crystals left as the water evaporates is noted; i f tantalic acid is present in excess, these are hexagonal plates, wliereas an excess of niobic acid causes the formation of hexagonal prisms. If the residue is treated with hydrochloric acid, the colombite acids ci-gstallise out. Addition of sodium hydroxide, slightly warm, causes the formation of hexagonal plates, consisting partly of stars and prisms. A simple test for the columbite acids consists in boiling 20 millgrams of the mineral with 0.8 C.C. of concentrated sulphuric acid, the solution being poured off from the insoluble part, diluted to 2-3 c.c., and a little zinc-dust added ; the solut,ion becomes sapphire- blue in a few minutes.J. W. L. By E. WALLER (A?zaZ!lst, 14, 108-112).- Attention is directed to the fact that in cases where, either from exces- sive hardness or from the presence of magnesium salts, it is neces- sary to dilute a water before applying the soap test, the results may vary widely according to the degree of dilution employed, especinllj- i f no deduction is made for the soap required to give a lather with pure water. The hardness of a mixture of calcium and magnesium solutions appears to be less than that of either of the individual solu- tions apart. M. J. S. Estimation of Chlorine in Water. By A. HAZEN (Amer. Ghem. J., 11, 409-414).-An investigation of the ordinary method of esti- mating chlorine in water by titration with a silver solution, using potassium chromate as an indicator.It is found that an excess of silver is always required to make the colour reaction apparent ; this excess is smaller the greater the amount of chromate used, provided that this does not colour the liquid so much as to obscure the end- point; i t is also smaller when the rolume of the liquid titrated is small. The amount of silver chloride precipitated also influences the result, and, other things being equal, the excess of silver solution used is nearly proportional to the amount of silver precipitated. To correct for this, the use of a silver solution 1 per cent. stronger than its normal value is recommended. It is still better to staudardise the silver solution against a solution of sodium chloride; with such a solution, and making a correction for the volume of liquid titrated, accurate results were obtained.If the amount of chlorine is small, the water must be concentrated, a very little sodium carbonate being added to prevent loss of liydrochloric acid on boiling. Hardness of Water. C. F. B.ANALYTICAL CHEMISTRY. 87 Dynamical Theory of Albuminoi’d Ammonia. By R. B. WARDER (Amer. Cheni. J., 11, 365-:378).-The integral calculus is applied t’o obtain formulae representing the distillation of an aqueous solution of ammonia, and the conversion of albuminoid matter into ammonia by alkaline permanganate. It is assumed that the “ coefficient of vola- tility”-that is, the ratio of the conceritration in any small portion of the distillate to that of the liquid in the retort-is constant.I n the case of the formation of albuininold ammonia, the law of mass action is applied, and the particular formula is investigated which represents the reaction between one molecule each of three different substances (permanganate, ,potash, and a nitrogenous substance). Curves are given representing the formulaz obtained. It is found that the rate of formation of albuminoid ammonia varies with the amounts of per- manganate and of potash present, and also with the rate of distilla- tion, and with the concentration of the original solution. The calcu- lated ratios of the am0unt.s of ammonia in successive portions of the distillate do not agree with those obtained by experiment ; this dis- crepancyis attributed to the fact that there is not one simple reaction taking place, but several ; ,and hence the curve actually obtained is the resultant of a number of curves.It is also found that the amount of ammonia left in the retort when the distillation is stopped, as calculated from the formula, is much less than that actually left. This is attributed to the formation of intermediate compounds which only yield ammonia with great diffi- culty. The author finally concludes that Wanklyn’s ammonia process gives valuable but purely cornparatire results, and is useless for the absolute estimation of organic nitrogen. Estimation of Ferrocyanide in Gas-lime. By 0. KXORLAUCH (Chem. Centr., 1889, ii, 211-212, from J.Gasbpleucht. u. Wasser- z-ersorg, 32, 450-459).-10 grams of the well mixed and finely ground-up gas-lime is digested, with frequent agitation, for 15-16 hours with 50 C.C. of 10 per cent. potassium hydroxide in a, flask graduated on the neck at 250 C.C. and a t 255 C.C. The volume is thsn made up to 255 c.c., the whole well mixed, and filtered; 100 C.C. of the filtrate is added to a hot solution of ferric chloride (containing 60 grams of ferric chloride and 200 C.C. of hydrochloric acid in the litre), the precipitate collected and washed with hot water, the funnel being covered meanwhile. The, filter-paper with the precipitate is agaiii transferred to the beaker in which the precipitation took place, the precipitate treated with 20 C.C. of 10 per cent. potassium hydroxide, and the whole then transferred to a 250 C.C. flask and made up to that volume.50 to 100 C.C. of the solution is filtered from the insoluble ferric hydrate and paper, 3-5 C.C. dilute sulphuric acid added, and the solution titrated with standard solution of cupric sulphate, which has been standardised with a solution of potassium ferrocyanide (4 grams in 1 litre). If hydrogen sulphide is present, it must be removed before the titration by adding 1-2 grams of lead carbonate. In applying t h i s volumetric method for determining hydroferrocyanic acid with cupric sulphate, the indicator used is a drop of ferric chloride on a piece of filter-paper to which is applied a drop of the solution C. F. B.88 AHSTRACTS OF OHEAIICAL PAPERS. under experiment ; so long as an excess of potassium ferrocyanide is present, the formation of prussian-blue will at once take place.Towards the end of the titration it is necessary to filter very small quantities of the solution into a dilute solution of ferric chloride, when the last traces of soluble ferrocyanide can be observed. Absorption of Bromine by Fatty Acids. By G. HALPHEN (-7. Pharm. [ 5 ] , 20, 247--249).-The process may be applied either to fats or to the fatty acids obtained from them; the results differ in the t w o cases, but are comparable amongst themselves. A saturated aqueous solution of bromine atid one of sodium hvdroxide coloured with eosin are required. 20 C.C. of soda-lye at 36" B. is added to 980 C.C. of water and 2 grams of eosin. 20 C.C. of carbon bisulphide and 10 C.C.ot bromine solution of known strength are placed in a flask provided with a cork. The soda fiolution is run in gradually from a burette. After each addition, the flask is closed and shaken four or five times, and the addition repeated until the brown colour passes to a salmon tint. The bromine solution is titrated by means of the sodium sollx- tion before each estimation, as its strength varies constantly. 'LO C.C. of carbon bisulphide is placed in a 250 C.C. flask, 1 gram of fatty acid is added, and an excess of bromine to the amount of about 0.5 gram. The flask is shaken up and allowed to remain for five hours ; at the end of this time the absorption is complete. The excess of bromine is titrated by means of the soda solution; the brown mass formed gradually passes to a white, soapy solution which becomes rosy on the addition of a few drops of the soda solution in excess.The vegetable oils absorb much more bromine than does lard, so that their presence in lard can thus be easily detected.-J. T. Note.-The standard solutions could not be originated by the method given. J. 'l'. Estimation of Citric Acid in Lemon Juice. By R. WILLIAMS (Annalyst, 14, 25-29).-The object of this paper is to recommend the use of sodium hydroxide with phenolphthalein as indicator for de- termining the acidity of lemon juice, instead of sodium carbonate with litmus-paper. Normal sodium citrate blues litmus-paper, but has no effect on yhenolphthaleih ; accordingly titrations of' pure citric acid made with sodium hydroxide and the latter indicator give numbers agreeing closely with theory, whilst those with the carbonate and litmus are low.Kevertheless, for some unexplained reason, the carbonate gives higher results than the hydroxide when applied to lemon juice, and estimations by precipitation as calcium salt agree better with the latter than with the former, being in fact generally lower than either. M. J. S. J. W. L. Impurities in Commercial Salicylic Acid. By B. FISCHER (J. Pharnz. [5], 20, 258-261 ; from Pharm. Zeit., 1889, 329, after Pharm. Zeit. RUSR., 2889, 28, 378) .-Salicylic acid contains cresotic acid when manufactured from impure phenol containing cresol. The presence of potash in the sodium hydrate employed occasions the formation of parahydroxjbenzoic acid ; this acid is also produced ifASALYTICAL CHENIS'I'RY.89 the temperature is too low at the time when the current of carbonic anhydride is passed, whilst too high a temperature at this stage results in the production of hydroxyisophthalic acid, due to the action of the gas on tlie sodium sslicylate already formed. Lastly, particularly in presence of iron salts, brown or yellow compounds are formed by oxidation, which are insoluble in water, and give a yellow colour to the salicylic acid. In a well-conducted process, parahydroxybenzoic and hydroxyisophthalic acids are usually not formed in quantities exceed- ing 0.4 per cent., and the first is easily removed by washing, as it is readily soluble in water. The second acid is less soluble in water, and may amount to 1 per cent.in certain cases. Cresotic acid is tho most important impurity, as apart from its obscure physiological action, its presence is very objectionable. The amount of cresotic acid present may be estimated by titrating with decinornial baryta solution, using phenolphthalein as indicator. Owing to the difference in their molecular weights, less solution is required to saturate cresotic acid than is required by salicjlic acid, but great care is needed to obtain satiafactory results, and certain accidental impuri- ties should be previously sought for, namely, water, colouring matters, and sodium chloride. With this view, dissolve in ether ; if the solu- tion is not clear, filter, evaporate, and dry first at 6U", then in a vacuum over sulphuric acid. In the absence of these impurities, it is necessary to dry the sample.The bsryta solution is standardised by the use of pure salicylic acid obtained by converting the commercial acid into the calcium salt, recrystallising, and then deconiposing the salt by means ofhydrochloric acid. For the detection of cresotic acid, 15 C.C. of waCer and 1 t o 2 grams of calcium carbonate are boiled in a 200 C.C. flask ; 3 grams of the salicylic acid is added, and the flask is agitated over a flame until the volume is reduced to about 5 C.C. By this time some crystals have formed. After cooling, the mother- liquor is transferred to a test tube and evaporated to 1 C.C. On rubbing this with a glass rod, crystallisation sets in. 1 C.C. of water is added, and the liquid filtered through a small plug of cotton.The filtrate is made up to 1 C.C. and hydrochloric acid is added ; if the original acid contained 3 t o 5 per cent. of ci*esotic acid, there separates out a mixture of acids which fuses in boiling water and collects at the bottom of the test tube in the form of thick, oily drops. The test does not succeed with less than 1 per cent. Hydroryisophthalic acid may be separated from salicylic acid by distillation in a current of steam. The first acid remains in the still as a light-grey powder or as small lumps. By dissolving i t in sufficient hydrochloric acid and filtering through charcoal, i t can be obtained in the form of slender, white needles, which fuse with decomposition about 300-305". The author has found in one sample of commercial salicylic acid 0.3 per cent.hydroxyisophthalic acid, and in another 5.3 per cent. of cresotic acid. By F. JEAN (2 Pharm. [ 5 ] , 20, 337--341).--The author's method comprises the determination of the density, melting point of the fatty acids, the elevation of temperature under the induence of sulphuric acid, and the refractive power. To determine the density, Wesphal's balauce is employud. To determine tlie J. T. Oil Testing.sn AHS'L'RAGTS OF CHEMICAL PAPERS. nielting point, a special apparatus is employed consisting of a thin U-shaped tube, in the two limbs of which are platinum wires nearly touching the bottom of the tube. A lager of solid acid comes between t'he ends of the wire, but this is displaced by mercury which has been charged in one side of the tube, when the temperature of a surrounding beaker of water has reached the melting point of ttle enclosed acid.The mercury causes electrical contact between the wires, and the transmitted current rings a bell when the temperature is read off, as given by a thermometer immersed in the bath. 'I'o determine the rise in temperature when mixed with sulphuric acid, a small beaker 4 cm. diameter and 6 cm. high is marked to contain 15 c.c., and in this is placed an acid tube provided with a stopper having a small tube through which air can be blown into the interior, and a small glass tube reaching from the bottom of the acid tube, and just passing through its side towards the upper end, so that on blowing into the acid tube its cont'ents are expelled and mixed with oil i n the beaker.15 C.C. of the oil to be tested is placed in the beaker and heated to 40", the acid tube is charged with 5 C.C. of sulphuric acid a t 65" B. and placed within the beaker; the whole is allowed to cool down to SO", and is then placed in a felt-lined box, when the acid is transferred to the oil by blowing and well mixed with it, the temperature is carefully observed, and the maximum reached is noted. I n general, when this temperature and the density of the oil are satisfactory, t,he sample may be regarded as pure. Oils which have been oxidised or otherwise changed require treatment with alcohol, or, better still, saponification, before determining the rise in temperature. Sometimes when the rise of temperatare is nearly the same for two oils, that of their two fatty acids may differ much more.One or two results may be given of oils and their acids:-Olive oil 41*5", acid 45"; linseed oil G l O , acid 109"; colza (Pas-de-Calais) 37", acid 44" ; ditto (India) 37", acid 46". To determine the refractive power of the oil, a, special oleorefractometer is employed which is not described. The index of refraction only vmies within narrow limits for the same species if care be taken to remove excess of acid by treatment with alcohol. The purity of a sample may be safely affirmed when the index of' refraction, the rise in temperature, and the density agree with a standard oil of known purity. J. T. By W. BISHOP (J. Pharm. [ 5 ] , 20, 244-247).- If this oil is shaken for a short time with pure hydrochloric acid cf 21-22" B. in the proportion of 8 of oil to 12 of acid, no special effect is produced, but if the oil is exposed to air and solar light for some days, and the same test is applied, the mixture becomes green and, after rt time, the colour is found to be con6ned to $he acid layer.I f the action of air and light be much prolonged, the green colour is intensified, and after a still longer period, a bluiali- violet, flocculent precipitate is produced. The green acid solution gives an absorption-spectrum almost exactly coinciding with that of chlorophyll. The application of this reaction will serve to indicate, when the results are positive, that a sample of sesame oil has been Oil of Sesame.ANALYTICAL CHIC31 ISTHY. 91 exposed to light and air for some time, and is not probably of recent production.Such an oil added to olive oil in the proportion of 5 to 10 per cent. can be easily detected by this method, whilst 10 to 20 per cent,. of oil of sesame may be detected in the same way after soiiie days’ exposure. 3 . T. By E. H. AMAGAT and F. JEAN (Compt. rend., 109, 616--617).--L)eterrnination of the refractive index in a refractometer of special construction is a delicate and trustwortthy means of detecting adulterations in oils and fats. The variations in the refractive indices of samples of the same oil from different sources are very slight, and distinctive differences are observed between vegetable oils, animal oils, and mineral oils. As little as 10 per cent. of oleomargarin can be detected in bnttei-. Optical Examination of Oils and Fats.U. H. B. Analysis of Fats and Oils. By J. MUTER and L. DE KOKINGH ( A d y s t , 14, 61--65).-The authors’ object has been to re-determine the iodine absorption (Hubl’s) of the liquid fatty acids from various oils and fats under conditions which should be as nniform as possible, and should exclude any alteration of the acids either by exposure to air or by drying a t a high temperature. A weighed portion of the fat is saponified with alcoholic potash, and the solution accurately neutralised with acetic acid. It is then poured into an excess of a boiling solution of lead acetate. The pre- cipitate is washed, then transferred to a stoppered bottle and treated with ether. The ether solution is filtered from lead stenrate, &c., into a Muter’s “olei’n tube,” in which it is decomposed by dilute hydrochloric acid.The volume of the ethereal solution of the fatty acids having been read, an aliquot part is run into a flask and most of the ether distilled off. The ether vapour protect’s the fatty acids from the air. Alcohol is then added, and the solution is titrated with soda ; this gives the total amount of the liquid fatty acids, calculatinv them as oleic acid. Another portion of the ethereal solution contain- ing 0.5 gram of the fatty acid is then evaporated in a bottle through which a stream of carbonic anhydride is being passed. When the last traces of ether are removed, 50 C.C. of Hubl’s reagent is instantly added, and the bottle, having been stoppered, is placed i n the dark for 12 honrs, side by side with a blank, after which the excess of iodine is titrated by thiosulphate.The authors anticipate that the “ iodine absorbing power ” thus ascertained will permit the amount, of any admixture of fats to be calculated with more precision than has hit herto been possible. M. J. S. Extraction of Fat from Milk Solids. By H. D. RICHMOKI) (AnaZyst? 14, 121--130).-0f the 15 or more msthods which havcb been proposed for the extraction of the fat from the dry residue of milk, those of Adams (paper coil), Soxhlet (plaster of Paris), and Storch (pdmice) give the highest and most concordant, but yet not identical results. The author has reinvestigated these three methods, using kieselguhr in place of pumice. I n Adams’ method, some analysts extract the paper coils with ether for a short time before92 ABSTRACTS OF CHEMICAL PAPERS.using them ; others apply a correction based on blank experiments wihh the,same batch of paper. The author finds that the complete extraction of the paper with ether requires a, very prolonged treat- ment; the total extract in 7& hours being more than three tirlles as much as that obtained in the first 1i hour. The matter extracted consists chiefly of the calcium salt of a resinous acid. The most complete and rapid extraction is obtained by the use of alcohol con- taining 10 per cent. of acetic acid. After 3 or 4 hours' treatment with this reagent in a Soxhlet's apparatus, nothing soluble in &her remains. With the plaster and kieselguhr methods, the chief requisite is to grind the dried residue to a very fine powder, and to extract it with ether for at least 3 hours.Working in this waj-, the three methods agree closely. From the results of numerous determi- nations by the three methods, the author has developed a new formula for deducing the percentage of f a t from that of total solids and the specific gravity : T = 1.17 P - 0.263 - (apparently a mis- print for -+ 0.263 -), where T is the percentage of total solids, F that of fat, D is the specific gravity of the milk, and G = 1000 (D - I). This formula gives results which do not differ materially from those of Hehner and Richmond's older formula (AnaZyst, 13, 32). The most satisfactory method of estimating the total solids appears to be the evaporation of not more than 2 grams of milk in a flat-bottomed basin and drying for 1 or 14 hour.Volumetric Method for the Estimation of Fat in Milk, &c. By C. L. PARSONS (Analyst, 14,181-187).-This method is proposed as one which can be carried out a t the dairy by unskilled persons. 100 C.C. of the milk is placed in a, bottle 11 inches high and 14 in diameter. 10 C.C. of soda solution (made by dissolving 1 part of commercial caustic soda in 2 parts of water) is added, then 5 C.C. of alcoholic soap solution (1 ounce of Castile soap to the gallon). 50 C.C. of gasoline (free from residue) is next added; the bottle is corked and shaken hard five or six times during half-an-hour. The petroleum solution of the fat is then allowed to rise to the surface. Should it fail t o do so, 5 C.C. or more of the alcoholic soap solution may be added and gently mixed in.When the upper layer is perfectly clear, 25 C.C. of it is withdrawn and evaporated in a small flask, which has its neck cut off obliqiiely. Two drops of strong acetic acid is added to the fat, which is then dried at 120" for 14 hours and drained from the flask into a measuring tube graduated in twentieths of a cubic centimeter. A table given in the paper converts the readings of the volume of the fat into percentages. G D G D M. J. S. The necessary precautions are fully described. M. J. S. Condensed Milk and the Estimation of Casein and Lact- albumen. By H. FABER (Analyst, 14, 141--147).-The author proposes to employ the estimation of the relative proportions of ca,sePn and lactalbumen as a means of distinguishing between fresh milk and that which hae been condensed arid afterwards diluted with water.ANALT TlCAL CHERll STKY.93 Fresh milk contains from 0.35 to 0.45, or perhaps more, of lactalbumen. By boiling the milk about two-thirds of this is coagulated, or SO modified that it is precipitated together with the case'in. The heating to ahout 75"' to which condensed milk must be subjected in order to sterilise it, has a similar effect. The two albuminoids can be sepa- rated by Sebelien's method. The casein is first precipitaked by magnesium sulphate, 2 vols. of the saturated solution of that salt being first added, and then as much of the powdered crystals as the mixture is able to dissolve ; the precipitate is washed with a saturated solution of magnesium sulphate ; and the lactalbnmen i s precipitated from the filtrate by either tannic acid or phosphotuiigstic acid.In these precipitates, the nitrogen is estimated by Kjeldahl's method. Test analyses show that the separation is very exact. Estimation of Soluble and Insoluble Fatty Acids in Butter. By W. JOHNSTONE (Analyst, 14, 113-114) and H. D. RICHMOND (ibid., 153--155).--Instead of estimating the volatile fatty acids by the Reichert process, the author prefers the following method. The butter is saponified with a known quantity of alcoholic potash and the excess found by titration. The alcohol having been removed by boiling, an excess of acid is added, and the insoluble fatty acids are filtered off and washed. After ail*-drying they are dissolved by ether and weighed after evaporation. They are now again saponified by standard potash, and the amount they neutrafise is ascertained.The difference between these two titrations gives the amount of fatty acid soluble in water, which thus estimated is considerably higher than is shown by Reichert's process. The author hints that the results of the latter may be vitiated by the production of propionic, acetic, and formic acids by the action of potash on the glycerol. Richmond, commenting on the above process, shows that the results given cannot possibly be correct, the total fatty acids, together wit11 the glycerol residue corresponding t o the potash neutralised, adding up to more than the weight of the butter taken, and that this is due to the titration of the insoluble acids being performed in aqueous solution.The results by Reichert's process, when corrected for the recognised average deficiency of &, add up almost exactly to 100 per cent. He points out that at the temperature of the water-bath potash has no action on glycerol. M. J. 8. M. J. S. Examination of Lard for Adulteration. By T. S . GLADDING (Analyst, 14, 32--.34).-The following tests should all be applied to a, suspected sample:-(1) specific gravity at 100'; (2) Hiibl's iodine test ; (3) BecLi-Millian test (Abstr., 1889,319) ; (4) Dalican's " Titre " test; (5) Belden's microscopic test for beef fat (Analyst, 13, SO). Dalican's " titre " is the temperature of crystallisation of the fatty acids. These are to be prepared from the sample by saponification, washed well with hot water, and filtered through dry paper into a test- tube.The crystallising point is then taken with a thermometer graduated t o tenths of a degree. The titre of lard may range from 36.4" to 41.4" ; iodine absorption from 57 to 68.4 per cent., a high titre being associated with a low kdine absorption, and vice vers6.9 4 ABSTRACTS OF CHEMICAL PAPERS. The titre of beef fat is about 41.6 to 44; iodine absorption, 43.8 to 40; that of cotton-seed oil, 33.3, iodine absorption, 108. The one adultmerant will therefore to some extent mask the other; they are, however, respectively revealed by Bechi’s and Belden’s tests, The high specific gravity of cotton-seed oil affords the onlv means of esti- mating the amount of i t present. (See also Abstr., 1889, 319, 659.) M. J. S. Action of Acids on Benzoic Sulphinide and Analysis of By I.REMSEN and W. &I. BURTON (Anzer. Chewi. J., u Saccharin.” l1,403408).-When benzoic sulphinide, C6H4<co >NH [ = 1 : 21, so2 is bniled with dilute acids, hydrogen ammonium orthosulphobenzoate, COOH.C,H4*S03NH, [= I : 21 is formed, toqether with a little ( wthosulphaminebenzoic acid, C OOH.C6H,.S02NH2. The best strength of acid is that obtained by diluting strong hydrochloric acid of sp. gr. 3 -17 with 8 to 10 times its volume of water. Commercial “ saccharin ” is found to be a mixture of parasulph- aminebenzoic acid, benzoic sulphinide, and hydrogen potasqinm ortho- sulphobenzoate, the amount of sulphinide present being somewhat less than 50 per cent. To analyse it, 2 grams are boiled for one hour with 100 C.C.of dilute hydrochloric acid (1-8) in a flask of 250 C.C. capacity, provided v-ith a reflux condenser. The clear solution is then evaporated to nbont 15 c.c., wheii the parasulphaminebenzoic a,cid separates out ; it is dried a t 80” and weighed. The filtrate, containing hydrogen am- monium orthosulphobenzoate (from the decomposition of the sul- phinide) and the hydrogen potassium salt of the same acid, is evaporated ; the residue is weighed, and the amount of potassium in it is estimated by heating a portion with sulphuric acid, and weighing the potassium sulphate formed. Two samples of saccharin were analysed, each five times ; the mean percentage composition of each is given below. Benzoic COOH*C,H, *S02NHZ. sulphinide. C00hT*C,H4.SO3K. I ...... 50.00 42.86 7.12 II ......44.49 48.3:3 7.99 C. F. B. Estimation of Morphine in Opium. By F. A. FL~CKIGER (Arch. Pharm. [ 3 ] , 27, 721-732, ’7t 9-i72).-The author discusses various points which arise in the estimation of morphine, and arrives a t the following fairly good, although not qnite perfect, method. 8 grams of opium powder is placed in a folded filter of 12 cm. diameter with a little tapping, and is dried a t 100”. After half an hour 10 C.C. of ether mixed with 10 C.C. of chloroform is poured over it, the covered funnel being frequently struck, and finally 10 C.C. more of chloroform is poured on. After all possible liquid has run through, tlhe filter with its contents is opened out and dried a t a gentle heat. Next the powder is vigorously and repeatedly shaken in a flask with 80 C.C.of water and filtered after two hours. 42.5 grams of the filtrate is well and often shaken in a weighed flask with 7-5 C.C.AShLY TICAL CHEMISTRY. 95 of alcohol (0.83 sp. gr.), 15 C.C. of ether, and 1 C.C. of ammonia (0.96). After six hours, the contents of the flask are poured on to a double- folded filter of 10 cm. diameter, and the morphine is washed on to the filter with about 10 C.C. of water. This is dried, returned to the dried flask, and dried at 100" until its weight becomes constant. This pro- cess with a particular sample gave 12.90, 13.12, 13.35 per cent. of morphine, which was not pure white, but which dissolved completely in lime-water with very little colour.. As an appendix the author criticises in some detail an article by E.R. Squibb in the "Ephemeris '' for July, on morphine estimation, and, although be sees many defects in the process given, he remarks that the comprehensive paper desei-veu the fullest consideration. J. T. Analysis of Pepper and the occurrence of Piperidine in the same. By W. JOHNSTONE (Analyst, 14, 41-49).-Moisture and ash. -A weighed portion is dried at 100' and then incinerated in a muffle. The ash is treated successively with water and hydrochloric acid, and the amount of insoluble matter noted. Oil.-20 grams is distilled with water ; the distillate is shaken with ether, the ethereal solution is evaporated at a very low temperature, and the residue is dried over sulphuric acid. Piperidine.-20 grams is dist'illed as for the oil determination, and the distillate is titrated with N/lO sulphuric acid (compare Abstr., 1889, 298).That the piperidine is not derived from the hydrolysis of piperine is shown by the fact that pure piperine yields no piperidine when distilled with water, also that in distilling pepper with water, piperidine soon ceases to come over, although the amount obtained is very small in comparison with the piperine present. Piperine.-10 grams is digested at 100" in a closed bottle with 3 grams of potash dissolved in 25 C.C. of water and 25 C.C. of alcohol. The bottle (4 oz.) should have tlhe neck ground flat and be closed by a plate of caoutchouc pressed tightly upon it by a screw-frame. After 4-6 hours' digestion, the bottle is cooled, the contents are washed into a large flask and distilled as long as the distillate is alkaline.The theoretical yield of pipedine is obtained. Crude Fibre.-A small quantity is boiled for half an hour in a flask with inverted condenser with 200 C.C. of dilute sulphuric acid (12.5 grams per litre). The residue is twice boiled with water, then with 200 C.C. of potash (12.5 grams per litre), and again twice with water. It is collected on a tared filter, dried and weighed, and any ash it contains deducted. Nitrogen.-Determined by soda-lime, as usual. Alcoholic Extract.-10 grams is extracted with 95 per cent, alcohol in a Soxhlet's apparatus for 24 hours. The alcohol is distilled off and the extractive matters dried at 100". St,arch.-The exhausted residue from the preceding is, without drying, washed into a flask with 200 C.C. of water and 20 C.C.of hydrochloric acid (1.121) and heated in boiling water for three hours. After cooling, the liquid is filtered, neutrdised with soda, made up to 500 c.c., and titrated with Fehling's solution. In 13 genuine samples from various localities, the moisture ranged9 ri ABSTRACTS OF CHEMICAL PAPERS. from 12 to 15 per cent., ash 1.07 to 4.46 (long pepper 7 57), oil 0.53 to 1.87, piperidine 0.21 to 0.77, piperine 5.21 to 13.03, fibre 4.2 to 15.05, starch 29.6 to 53.5, ash insoluble in acid 0.06 to 0.62 (long pepper 1-47), Any larger amount of insoluble ash would probably be the best indication of a fraudulent addition. M. J. S. Detection of Cocai'ne Hydrochloride. By M. GOELDNER (Arch. Phai-m. [31, 27, $99; from Pharm. Zeit., 34, 471).-The author believes that the following is a charactelistic test for cocaine. 6 or 7 drops of pure, strong sulphuric acid are added to some crystals of resorcinol in a porcelain basin, and the latter is moved to and fro a little, then a litt,le cocaine hydrochloride is added to t'he yellow liquid.A som+ what strong reaction follows, and a splendid, blue coloration is imme- diately obtained; a drop of sodium hydroxide changes this to a light- rose colour. The reaction goes more quickly with powdered resorcinol in place of crystals. Verysmall quantities of the reamgent give no colour reaction. Other alkaloids give nothing approaching to this reaction. J. T. Estimation of Indigotin for Commercial Purposes. By F. A. OWEN (Chem. Centr., 1889, ii, 217-213; from J. dmer. Chem. SOC., 10, 178).-J gram of the substance is weighed on a watch-glass, dried a t lOO", finely powdered, rubbed with water to a thin paste, and washed into a 250 C.C.flask. 3 grams of zinc-dust and 6 grams of sodium hydroxide are added, the solution dilut,ed to a little above the mark, shaken up now and then, and after the reaction is complete (during which the solution must remain green, red or brownish streaks indicate that the reduction has been carried too far ; a froth indicates the presence of too much zinc), 50 C.C. of the clear liquid is exposed to the air for half an hour, acidified with hydrochloric acid, and filtered through a well-wnshed filter, dried a t loo", and weighed. J. W. L.ANALYTICAL CHEAIISTRT. 83An a 1 y t i c a1 C h e m i s t r y.Estimation of Phosphorus in Phosphor-tin.By W. HEMPEL(Her., 22, 2478).-Phosphor-tin is best analpsed by Wohler’s chlo-rine method as employed in the analysis of Fahl-ore. The stannicchloride and phosphoric chloride which are formed are collectedin concentrated nitric acid (about 10 c.c.) ; the vessel is washed outwith dilute nitric acid (1 : 2),. and the phosphoric acid is precipi-tated with ammonium molybdate and estimated directly.F. S. K.Recognition of Phosphoric Acid of Mineral Origin. ByJ. STOKLASA (Chern. Centr., 1889, ii, 57, from Listy. Chem., 13,153-154) .-The author contends that the percentage of fluorine inbones as given by different authors is decidedly too high. Raw bonesdo not Rhow any fluorine by the Loren2 reaction, whilst incineratedbones give only a slight indication of the presence of this elementLvhen tested in the same way.On the other hand, the authorfound the fossil bones of Elephas primigenius to contain 3.54 and4.36 per cent. of ferric oxide and 2.08 and 52-98 per cent. of fluorinerespectively in tlwo different specimens. Bones found in the olderalluvids on the island of Festigos contained 2.42 per cent. of‘ ferricoxide, 1.58 of fluorine, 80.04 of tricalcium phosphate. Superphosphatesprepared from this phosphate and also from bones were testetl bythe Lorenz reaction, with the result that fluorine was only found inthe former, whilst of the several samples of the latter class of super-phosphates only that from hone-ash gave any indication of fluorine,and then but slightly.On the other hand, remarks the author,Lorenz seems to have overlooked the fact that a series of mineralphosphates exist which contain but very little fluorine.J. W. L.Rapid Method of Estimating Arsenic. By E. POLENSKE(Chem. Centr., 1889, ii, 58-59, from Pharm. Zeit., 34, 299-300).-The method consists in evolving the arsenic as hydrogen arsenidein a Marsh apparatus, deposition of the arsenic in a tube having threebulbs blown on it, and weighing first that part of the tube containingthe ‘‘ mirror,” and secondly the tube after dissolving OE the arsenic.The evolution flask is recommended to be of a capacity of 250 c.c.,and to contain 80-100 grams of zinc. The apparatus includes anacid funnel 30 cm. long, having a U-formed bend, and the evolutionflask is also connected with a washing flask containing lead nit.ratesolution, to which is attached a tube containing calcium chloride, anda t the farther end, potash.The decomposing tube is attachedto this. All air is expelled from the apparatus, first, by adding 5 C.C.of concentrated sulphuric acid and 20 C . C . of wat(’r, and the re-agents are at the same time tested for arsenic by heating the tube.The solution should not contain more than 4-5 milligrams of arsenic,and is allowed to drop into the acid funnel from a burette a t the rateof 0.5-1.0 C.C. per minute. The gas escaping from the end of the9 84 ABSTRACTS OF CHEMICAL PAPERS.tube is passed through a solution of silver nitrate, and the evolut.ionshould not be too rapid for the bubbles to be counted.During theevolution of the hydrogen arsenide, two hunsen lamps are placedunder the second and third bulbs : in order to test whether all thearsenic is evolved from the flask, the first is heated, and i f a " mirror "forms it must be driven onwards into the second bulb, and after a timethe same test may be applied again. When it is thus proved thatarsenic no longer escapes from the flask, the tube is disconnected andreversed,and any arsenic which may have been deposited iu the thirdbulb is driven into the narrow tube between the second and thirdbulbs. All the arsenic having been thus collected in this part of thetube, it is cut off and weighed, and after dissolving off the arsenicwith nitric acid, it i s weighed again.the difference giving the weight-of arsenic. Metallic salts aria organic substances interfere with theacciiracy of the method.Behaviour of Silicates when Fused with Phosphates. ByK. HAUSHOFER (Chem. Centr., '1889, ii, 53, from Sitzungsber. dermat h.-naturzu. Abt. ba?yr. Akrrd. Wiss., 1889, 8-1 l).-Many silicates,when fused to a bead with an alkaline phosphate in the blowpipe,swell up, evolve gases, and finally insoluble silica, either in theform of a skeleton or as flakes, is deposited. The evolution of gasesis referred by the author to the presence of chlorine, sulphates, orwater, and h e draws attention to this part of the reaction as an aid inthe recognition of the silicate under examination. For instance,hauyn and sodalite, owing to the evolution of the chlorine and$ulphuric anhydride, may be distinguished from nepheline, which isbut slowly attacked.In like manlier, humite is distinguished fromolivine by the evolution of hydrofluoric acid, and tourmaline andaxinite from beryl. The micas of the scapolite group, epidote andvesiivian, lose their water of eonstitu t'ion, whereas the felspars,amphibole, and granites are but slowly attacked. Similar distinctionsmay be drawn between crystallised kaumerite and the dense rhodo-chrome, and between pyrophyllite and agalmstolite.Technical Analysis of Commercial Sodium Sulphide. ByB. SHETLIK (Chem. Centr., 1889, ii, 211, from Listy. Chem., 12, 205-206).-10 grams of the sulphide is dissolved in water, the solutiondiluted to i$ litre, and 50 C.C.titrated with normal sulphuric acid,phenolphthalein being used as an indicator. If the titration is madein the cold, the quantity of acid required must be doubled, whereas ifit is carried out a t a boiling heat and the acid added until the redcolour does not reappear on further boiling the solution, the acid usedis equivalent to the sulphide. Insoluble sulphides which are de-composed by dilute acid may be titrated in this way.Qualitative Analysis of the Ammonium Sulphide Re-cipitate. By P. MATER (Ber., 22, 2627--26:10).-The presence ofchromium in t,his precipitate renders impracticable the separation bysolntion in hydrochloric acid and precipitation of the iron andaluminium by boiling with sodium acetate, since the chromium issometimes wholly and sometimes partially prccipitated, whilst if noJ.W. L.J. W. L.J. w. LASALYTICAL CHEMISTRY. 85iron is present, it all remains in the filtrate. The author has foundthat the presence of excess of iron (at least 5 atoms to every atom ofchromium) ensures the total precipitation of the chromium. He,therefore, recommends t h a t where chromium is suspected and iron isnot present in large quantities, ferric chloride should be- added inexcess before boiling with sodium acetate. L. T.. T.Analysis of Aluminium Sulphate. By F. BETLSTSIN arid T.GROSSET (Chem. Centr., 1889, ii, 60; from Melanges, Yhys. Chim.Bull., St. Pe'tersbourg, 13, 42--47).--The authors recommend the fol-lowing method. 1-2 grams of substance is dissolved in 5 C.C.of water,5 C.C. of a cold saturated solution of ammonium sulphate added, andthe mixture well stirred during a quarter of an hour. 50 C.C. of95 per cent. alcohol is added, and the precipitated ammonia alumfiltered off and washed with 50 C.C. of alcohol. The filtratecontains all the free acid, which is deteimined by concentratiiig andtitrating with decinormal alkali. The whole of the aluminiumsulphate is precipitated as ammonia alum.Estimation of Chromium and Copper in Iron and Steel.By C. REIXHARDT (Clrenz. Centr., lt89, ii, bU-61, frcim Stahl w. Eistn9, 404--405).-For the determination of chromium, 10 grams ofborings or filings are dissolved with 100 C.C. hjdrochloric acid in acovered beaker of 500 C.C. capacity, first without heat, then at a boil-ing heat, oxidised with potassium clilorate, concentrated to one halfthe volume, filtered into a &litre Erlenmeycr flask, and the insolubleresidue washed several times with dilute hydrochloric acid on thefilter, and finally with water.The solution is now reduced a t a boilingheat hy the a,ddition of 10-20 C.C. of sodium hypophosphite solutiou(200 grams in 400 C.C. of water), and afternards the chromium is pre-cipitated by the addition of zinc oxide in excess. The precipitate isdissolved in hydrochloric acid, a little more hypophosphite added, andthe precipitation repeated. The chromium is separated from thezinc by precipitation with ammonia, which precipitation must be re-peated. The chromic oxide, after ignition, is iused, together withthe insoluble portion of the material, with 8 grams of a mixture of 4parts of sodium chloride, 1 part of sodium carbonate, and 1 part ofpotassium chlorate.From the dissdvad flux, the manpnese is pre-cipitated with alcohol, the silicic acid with lijdrochloric acid and alittle sulphurous acid, and the chromium finally precipated as oxidewith ammonia. The copper is determined iii 10 grams of the material :the hydrochloric acid solution is reduced with sodium hypophosphite,and the copper precipitated with hydrogen sulphide.Volumetric Estimation of Chromium in Iron and Steel. ByE. WAHLHERG (Chem. Centr., 1889, ii, 194, from Bcrg. a. Huttenni.Zeit., 48, 18G-181).-0~5 gram of the metal is dissolved in boiliud.nitric acid, sp.gr. 120, evaporated to dryness, ground up, transfern*dto a platinum crucible, mixed with a mixture of 2 grams of magnesm,1 gram of potassium chlorate, and 1 gram of fiodium carbonate, audthe whole heated, at first gently, then in the blast flame for one hour.J. W. L.J. W. L86 A BSTKACTS OF CHEMICAL PAPERS.The mass is dissolved out with 50-100 C.C. of water, any s m a l lquantity of manganic acid reduced by a drop or two of alcohol,acidified with sulphuric acid, arid the chromic acid estimated byadding a known quantity of ferrous sulpbate and determining theamount unosidised by titrating with potassium perrnanganate.J. W. L.Microscopical Test for Tantalum and Niobium. By K.HAUSHOFER (Chem. Centr., 1889, ii, 62-63, from Sitzungsber. dermath.-nnturzu.Abt. bayr. Akad. Wiss., 1889, 3-8).-The substanceto be tested is fused with a very small bead of sodium carbonatein the hottest part of the bunsen flame €or 30-40 seconds. It isthen treated on the object glass with one drop of water, and theform of the crystals left as the water evaporates is noted; i ftantalic acid is present in excess, these are hexagonal plates, wliereasan excess of niobic acid causes the formation of hexagonal prisms. Ifthe residue is treated with hydrochloric acid, the colombite acidsci-gstallise out. Addition of sodium hydroxide, slightly warm, causesthe formation of hexagonal plates, consisting partly of stars andprisms. A simple test for the columbite acids consists in boiling20 millgrams of the mineral with 0.8 C.C.of concentrated sulphuricacid, the solution being poured off from the insoluble part, diluted to2-3 c.c., and a little zinc-dust added ; the solut,ion becomes sapphire-blue in a few minutes. J. W. L.By E. WALLER (A?zaZ!lst, 14, 108-112).-Attention is directed to the fact that in cases where, either from exces-sive hardness or from the presence of magnesium salts, it is neces-sary to dilute a water before applying the soap test, the results mayvary widely according to the degree of dilution employed, especinllj-i f no deduction is made for the soap required to give a lather withpure water. The hardness of a mixture of calcium and magnesiumsolutions appears to be less than that of either of the individual solu-tions apart. M.J. S.Estimation of Chlorine in Water. By A. HAZEN (Amer. Ghem.J., 11, 409-414).-An investigation of the ordinary method of esti-mating chlorine in water by titration with a silver solution, usingpotassium chromate as an indicator. It is found that an excess ofsilver is always required to make the colour reaction apparent ; thisexcess is smaller the greater the amount of chromate used, providedthat this does not colour the liquid so much as to obscure the end-point; i t is also smaller when the rolume of the liquid titrated issmall. The amount of silver chloride precipitated also influences theresult, and, other things being equal, the excess of silver solutionused is nearly proportional to the amount of silver precipitated.Tocorrect for this, the use of a silver solution 1 per cent. stronger thanits normal value is recommended. It is still better to staudardise thesilver solution against a solution of sodium chloride; with such asolution, and making a correction for the volume of liquid titrated,accurate results were obtained. If the amount of chlorine is small,the water must be concentrated, a very little sodium carbonate beingadded to prevent loss of liydrochloric acid on boiling.Hardness of Water.C. F. BANALYTICAL CHEMISTRY. 87Dynamical Theory of Albuminoi’d Ammonia. By R. B. WARDER(Amer. Cheni. J., 11, 365-:378).-The integral calculus is applied t’oobtain formulae representing the distillation of an aqueous solution ofammonia, and the conversion of albuminoid matter into ammonia byalkaline permanganate.It is assumed that the “ coefficient of vola-tility”-that is, the ratio of the conceritration in any small portion ofthe distillate to that of the liquid in the retort-is constant. I n thecase of the formation of albuininold ammonia, the law of mass actionis applied, and the particular formula is investigated which representsthe reaction between one molecule each of three different substances(permanganate, ,potash, and a nitrogenous substance). Curves aregiven representing the formulaz obtained. It is found that the rateof formation of albuminoid ammonia varies with the amounts of per-manganate and of potash present, and also with the rate of distilla-tion, and with the concentration of the original solution.The calcu-lated ratios of the am0unt.s of ammonia in successive portions of thedistillate do not agree with those obtained by experiment ; this dis-crepancyis attributed to the fact that there is not one simple reactiontaking place, but several ; ,and hence the curve actually obtained isthe resultant of a number of curves.It is also found that the amount of ammonia left in the retort whenthe distillation is stopped, as calculated from the formula, is muchless than that actually left. This is attributed to the formation ofintermediate compounds which only yield ammonia with great diffi-culty. The author finally concludes that Wanklyn’s ammonia processgives valuable but purely cornparatire results, and is useless for theabsolute estimation of organic nitrogen.Estimation of Ferrocyanide in Gas-lime.By 0. KXORLAUCH(Chem. Centr., 1889, ii, 211-212, from J. Gasbpleucht. u. Wasser-z-ersorg, 32, 450-459).-10 grams of the well mixed and finelyground-up gas-lime is digested, with frequent agitation, for 15-16hours with 50 C.C. of 10 per cent. potassium hydroxide in a, flaskgraduated on the neck at 250 C.C. and a t 255 C.C. The volume is thsnmade up to 255 c.c., the whole well mixed, and filtered; 100 C.C. ofthe filtrate is added to a hot solution of ferric chloride (containing60 grams of ferric chloride and 200 C.C. of hydrochloric acid in thelitre), the precipitate collected and washed with hot water, the funnelbeing covered meanwhile. The, filter-paper with the precipitate isagaiii transferred to the beaker in which the precipitation took place,the precipitate treated with 20 C.C. of 10 per cent.potassium hydroxide,and the whole then transferred to a 250 C.C. flask and made up to thatvolume. 50 to 100 C.C. of the solution is filtered from the insolubleferric hydrate and paper, 3-5 C.C. dilute sulphuric acid added, andthe solution titrated with standard solution of cupric sulphate, whichhas been standardised with a solution of potassium ferrocyanide(4 grams in 1 litre). If hydrogen sulphide is present, it must beremoved before the titration by adding 1-2 grams of lead carbonate.In applying t h i s volumetric method for determining hydroferrocyanicacid with cupric sulphate, the indicator used is a drop of ferric chlorideon a piece of filter-paper to which is applied a drop of the solutionC.F. B88 AHSTRACTS OF OHEAIICAL PAPERS.under experiment ; so long as an excess of potassium ferrocyanide ispresent, the formation of prussian-blue will at once take place.Towards the end of the titration it is necessary to filter very smallquantities of the solution into a dilute solution of ferric chloride,when the last traces of soluble ferrocyanide can be observed.Absorption of Bromine by Fatty Acids. By G. HALPHEN(-7. Pharm. [ 5 ] , 20, 247--249).-The process may be applied either tofats or to the fatty acids obtained from them; the results differ inthe t w o cases, but are comparable amongst themselves. A saturatedaqueous solution of bromine atid one of sodium hvdroxide coloured witheosin are required. 20 C.C.of soda-lye at 36" B. is added to 980 C.C.of water and 2 grams of eosin. 20 C.C. of carbon bisulphide and 10 C.C.ot bromine solution of known strength are placed in a flask providedwith a cork. The soda fiolution is run in gradually from a burette.After each addition, the flask is closed and shaken four or five times,and the addition repeated until the brown colour passes to a salmontint. The bromine solution is titrated by means of the sodium sollx-tion before each estimation, as its strength varies constantly. 'LO C.C.of carbon bisulphide is placed in a 250 C.C. flask, 1 gram of fatty acidis added, and an excess of bromine to the amount of about 0.5 gram.The flask is shaken up and allowed to remain for five hours ; at theend of this time the absorption is complete.The excess of bromineis titrated by means of the soda solution; the brown mass formedgradually passes to a white, soapy solution which becomes rosy onthe addition of a few drops of the soda solution in excess.The vegetable oils absorb much more bromine than does lard, sothat their presence in lard can thus be easily detected.-J. T.Note.-The standard solutions could not be originated by themethod given. J. 'l'.Estimation of Citric Acid in Lemon Juice. By R. WILLIAMS(Annalyst, 14, 25-29).-The object of this paper is to recommend theuse of sodium hydroxide with phenolphthalein as indicator for de-termining the acidity of lemon juice, instead of sodium carbonate withlitmus-paper.Normal sodium citrate blues litmus-paper, but has noeffect on yhenolphthaleih ; accordingly titrations of' pure citric acidmade with sodium hydroxide and the latter indicator give numbersagreeing closely with theory, whilst those with the carbonate and litmusare low. Kevertheless, for some unexplained reason, the carbonategives higher results than the hydroxide when applied to lemon juice,and estimations by precipitation as calcium salt agree better withthe latter than with the former, being in fact generally lower thaneither. M. J. S.J. W. L.Impurities in Commercial Salicylic Acid. By B. FISCHER(J. Pharnz. [5], 20, 258-261 ; from Pharm. Zeit., 1889, 329, afterPharm. Zeit. RUSR., 2889, 28, 378) .-Salicylic acid contains cresoticacid when manufactured from impure phenol containing cresol. Thepresence of potash in the sodium hydrate employed occasions theformation of parahydroxjbenzoic acid ; this acid is also produced iASALYTICAL CHENIS'I'RY.89the temperature is too low at the time when the current of carbonicanhydride is passed, whilst too high a temperature at this stage resultsin the production of hydroxyisophthalic acid, due to the action ofthe gas on tlie sodium sslicylate already formed. Lastly, particularlyin presence of iron salts, brown or yellow compounds are formed byoxidation, which are insoluble in water, and give a yellow colour to thesalicylic acid. In a well-conducted process, parahydroxybenzoic andhydroxyisophthalic acids are usually not formed in quantities exceed-ing 0.4 per cent., and the first is easily removed by washing, as it isreadily soluble in water.The second acid is less soluble in water,and may amount to 1 per cent. in certain cases. Cresotic acid is thomost important impurity, as apart from its obscure physiologicalaction, its presence is very objectionable. The amount of cresoticacid present may be estimated by titrating with decinornial barytasolution, using phenolphthalein as indicator. Owing to the differencein their molecular weights, less solution is required to saturatecresotic acid than is required by salicjlic acid, but great care isneeded to obtain satiafactory results, and certain accidental impuri-ties should be previously sought for, namely, water, colouring matters,and sodium chloride.With this view, dissolve in ether ; if the solu-tion is not clear, filter, evaporate, and dry first at 6U", then in avacuum over sulphuric acid. In the absence of these impurities, it isnecessary to dry the sample. The bsryta solution is standardised bythe use of pure salicylic acid obtained by converting the commercialacid into the calcium salt, recrystallising, and then deconiposing thesalt by means ofhydrochloric acid. For the detection of cresotic acid,15 C.C. of waCer and 1 t o 2 grams of calcium carbonate are boiled ina 200 C.C. flask ; 3 grams of the salicylic acid is added, and the flaskis agitated over a flame until the volume is reduced to about 5 C.C.By this time some crystals have formed.After cooling, the mother-liquor is transferred to a test tube and evaporated to 1 C.C. Onrubbing this with a glass rod, crystallisation sets in. 1 C.C. of wateris added, and the liquid filtered through a small plug of cotton.The filtrate is made up to 1 C.C. and hydrochloric acid is added ; if theoriginal acid contained 3 t o 5 per cent. of ci*esotic acid, there separatesout a mixture of acids which fuses in boiling water and collects atthe bottom of the test tube in the form of thick, oily drops. The testdoes not succeed with less than 1 per cent. Hydroryisophthalicacid may be separated from salicylic acid by distillation in a currentof steam. The first acid remains in the still as a light-grey powderor as small lumps. By dissolving i t in sufficient hydrochloric acidand filtering through charcoal, i t can be obtained in the form of slender,white needles, which fuse with decomposition about 300-305".Theauthor has found in one sample of commercial salicylic acid 0.3 per cent.hydroxyisophthalic acid, and in another 5.3 per cent. of cresotic acid.By F. JEAN (2 Pharm. [ 5 ] , 20, 337--341).--Theauthor's method comprises the determination of the density, meltingpoint of the fatty acids, the elevation of temperature under theinduence of sulphuric acid, and the refractive power. To determinethe density, Wesphal's balauce is employud. To determine tlieJ. T.Oil Testingsn AHS'L'RAGTS OF CHEMICAL PAPERS.nielting point, a special apparatus is employed consisting of a thinU-shaped tube, in the two limbs of which are platinum wires nearlytouching the bottom of the tube.A lager of solid acid comesbetween t'he ends of the wire, but this is displaced by mercury whichhas been charged in one side of the tube, when the temperature of asurrounding beaker of water has reached the melting point of ttleenclosed acid. The mercury causes electrical contact between thewires, and the transmitted current rings a bell when the temperatureis read off, as given by a thermometer immersed in the bath. 'I'odetermine the rise in temperature when mixed with sulphuric acid, asmall beaker 4 cm. diameter and 6 cm. high is marked to contain15 c.c., and in this is placed an acid tube provided with a stopperhaving a small tube through which air can be blown into the interior,and a small glass tube reaching from the bottom of the acid tube,and just passing through its side towards the upper end, so that onblowing into the acid tube its cont'ents are expelled and mixed withoil i n the beaker.15 C.C. of the oil to be tested is placed in thebeaker and heated to 40", the acid tube is charged with 5 C.C. ofsulphuric acid a t 65" B. and placed within the beaker; the whole isallowed to cool down to SO", and is then placed in a felt-lined box,when the acid is transferred to the oil by blowing and well mixedwith it, the temperature is carefully observed, and the maximumreached is noted. I n general, when this temperature and the densityof the oil are satisfactory, t,he sample may be regarded as pure.Oils which have been oxidised or otherwise changed require treatmentwith alcohol, or, better still, saponification, before determining therise in temperature.Sometimes when the rise of temperatare isnearly the same for two oils, that of their two fatty acids may differmuch more. One or two results may be given of oils and theiracids:-Olive oil 41*5", acid 45"; linseed oil G l O , acid 109"; colza(Pas-de-Calais) 37", acid 44" ; ditto (India) 37", acid 46". Todetermine the refractive power of the oil, a, special oleorefractometer isemployed which is not described. The index of refraction only vmieswithin narrow limits for the same species if care be taken to removeexcess of acid by treatment with alcohol. The purity of a samplemay be safely affirmed when the index of' refraction, the rise intemperature, and the density agree with a standard oil of knownpurity.J. T.By W. BISHOP (J. Pharm. [ 5 ] , 20, 244-247).-If this oil is shaken for a short time with pure hydrochloric acid cf21-22" B. in the proportion of 8 of oil to 12 of acid, no specialeffect is produced, but if the oil is exposed to air and solar light forsome days, and the same test is applied, the mixture becomesgreen and, after rt time, the colour is found to be con6ned to $heacid layer. I f the action of air and light be much prolonged, thegreen colour is intensified, and after a still longer period, a bluiali-violet, flocculent precipitate is produced. The green acid solutiongives an absorption-spectrum almost exactly coinciding with that ofchlorophyll. The application of this reaction will serve to indicate,when the results are positive, that a sample of sesame oil has beenOil of SesameANALYTICAL CHIC31 ISTHY.91exposed to light and air for some time, and is not probably of recentproduction. Such an oil added to olive oil in the proportion of 5 to10 per cent. can be easily detected by this method, whilst 10 to 20per cent,. of oil of sesame may be detected in the same way aftersoiiie days’ exposure. 3 . T.By E. H. AMAGATand F. JEAN (Compt. rend., 109, 616--617).--L)eterrnination of therefractive index in a refractometer of special construction is a delicateand trustwortthy means of detecting adulterations in oils and fats.The variations in the refractive indices of samples of the same oil fromdifferent sources are very slight, and distinctive differences areobserved between vegetable oils, animal oils, and mineral oils.Aslittle as 10 per cent. of oleomargarin can be detected in bnttei-.Optical Examination of Oils and Fats.U. H. B.Analysis of Fats and Oils. By J. MUTER and L. DE KOKINGH( A d y s t , 14, 61--65).-The authors’ object has been to re-determinethe iodine absorption (Hubl’s) of the liquid fatty acids from variousoils and fats under conditions which should be as nniform as possible,and should exclude any alteration of the acids either by exposure toair or by drying a t a high temperature.A weighed portion of the fat is saponified with alcoholic potash,and the solution accurately neutralised with acetic acid.It is thenpoured into an excess of a boiling solution of lead acetate. The pre-cipitate is washed, then transferred to a stoppered bottle and treatedwith ether. The ether solution is filtered from lead stenrate, &c.,into a Muter’s “olei’n tube,” in which it is decomposed by dilutehydrochloric acid. The volume of the ethereal solution of the fattyacids having been read, an aliquot part is run into a flask and most ofthe ether distilled off. The ether vapour protect’s the fatty acidsfrom the air. Alcohol is then added, and the solution is titrated withsoda ; this gives the total amount of the liquid fatty acids, calculatinvthem as oleic acid. Another portion of the ethereal solution contain-ing 0.5 gram of the fatty acid is then evaporated in a bottle throughwhich a stream of carbonic anhydride is being passed.When thelast traces of ether are removed, 50 C.C. of Hubl’s reagent is instantlyadded, and the bottle, having been stoppered, is placed i n the darkfor 12 honrs, side by side with a blank, after which the excess ofiodine is titrated by thiosulphate. The authors anticipate that the“ iodine absorbing power ” thus ascertained will permit the amount,of any admixture of fats to be calculated with more precision than hashit herto been possible. M. J. S.Extraction of Fat from Milk Solids. By H. D. RICHMOKI)(AnaZyst? 14, 121--130).-0f the 15 or more msthods which havcbbeen proposed for the extraction of the fat from the dry residue ofmilk, those of Adams (paper coil), Soxhlet (plaster of Paris), andStorch (pdmice) give the highest and most concordant, but yet notidentical results.The author has reinvestigated these three methods,using kieselguhr in place of pumice. I n Adams’ method, someanalysts extract the paper coils with ether for a short time befor92 ABSTRACTS OF CHEMICAL PAPERS.using them ; others apply a correction based on blank experimentswihh the,same batch of paper. The author finds that the completeextraction of the paper with ether requires a, very prolonged treat-ment; the total extract in 7& hours being more than three tirlles asmuch as that obtained in the first 1i hour. The matter extractedconsists chiefly of the calcium salt of a resinous acid. The mostcomplete and rapid extraction is obtained by the use of alcohol con-taining 10 per cent.of acetic acid. After 3 or 4 hours' treatmentwith this reagent in a Soxhlet's apparatus, nothing soluble in &herremains. With the plaster and kieselguhr methods, the chiefrequisite is to grind the dried residue to a very fine powder, and toextract it with ether for at least 3 hours. Working in this waj-, thethree methods agree closely. From the results of numerous determi-nations by the three methods, the author has developed a newformula for deducing the percentage of f a t from that of total solidsand the specific gravity : T = 1.17 P - 0.263 - (apparently a mis-print for -+ 0.263 -), where T is the percentage of total solids, F thatof fat, D is the specific gravity of the milk, and G = 1000 (D - I).This formula gives results which do not differ materially from thoseof Hehner and Richmond's older formula (AnaZyst, 13, 32).Themost satisfactory method of estimating the total solids appears to bethe evaporation of not more than 2 grams of milk in a flat-bottomedbasin and drying for 1 or 14 hour.Volumetric Method for the Estimation of Fat in Milk, &c.By C. L. PARSONS (Analyst, 14,181-187).-This method is proposed asone which can be carried out a t the dairy by unskilled persons. 100 C.C.of the milk is placed in a, bottle 11 inches high and 14 in diameter.10 C.C. of soda solution (made by dissolving 1 part of commercialcaustic soda in 2 parts of water) is added, then 5 C.C.of alcoholicsoap solution (1 ounce of Castile soap to the gallon). 50 C.C. ofgasoline (free from residue) is next added; the bottle is corked andshaken hard five or six times during half-an-hour. The petroleumsolution of the fat is then allowed to rise to the surface. Should itfail t o do so, 5 C.C. or more of the alcoholic soap solution may be addedand gently mixed in. When the upper layer is perfectly clear, 25 C.C.of it is withdrawn and evaporated in a small flask, which has its neckcut off obliqiiely. Two drops of strong acetic acid is added to the fat,which is then dried at 120" for 14 hours and drained from the flaskinto a measuring tube graduated in twentieths of a cubic centimeter.A table given in the paper converts the readings of the volume of thefat into percentages.GDGDM.J. S.The necessary precautions are fully described.M. J. S.Condensed Milk and the Estimation of Casein and Lact-albumen. By H. FABER (Analyst, 14, 141--147).-The authorproposes to employ the estimation of the relative proportions of ca,sePnand lactalbumen as a means of distinguishing between fresh milk andthat which hae been condensed arid afterwards diluted with waterANALT TlCAL CHERll STKY. 93Fresh milk contains from 0.35 to 0.45, or perhaps more, of lactalbumen.By boiling the milk about two-thirds of this is coagulated, or SOmodified that it is precipitated together with the case'in. The heatingto ahout 75"' to which condensed milk must be subjected in order tosterilise it, has a similar effect.The two albuminoids can be sepa-rated by Sebelien's method. The casein is first precipitaked bymagnesium sulphate, 2 vols. of the saturated solution of that saltbeing first added, and then as much of the powdered crystals as themixture is able to dissolve ; the precipitate is washed with asaturated solution of magnesium sulphate ; and the lactalbnmen i sprecipitated from the filtrate by either tannic acid or phosphotuiigsticacid. In these precipitates, the nitrogen is estimated by Kjeldahl'smethod. Test analyses show that the separation is very exact.Estimation of Soluble and Insoluble Fatty Acids in Butter.By W. JOHNSTONE (Analyst, 14, 113-114) and H. D. RICHMOND(ibid., 153--155).--Instead of estimating the volatile fatty acids bythe Reichert process, the author prefers the following method.Thebutter is saponified with a known quantity of alcoholic potash andthe excess found by titration. The alcohol having been removed byboiling, an excess of acid is added, and the insoluble fatty acids arefiltered off and washed. After ail*-drying they are dissolved by etherand weighed after evaporation. They are now again saponified bystandard potash, and the amount they neutrafise is ascertained. Thedifference between these two titrations gives the amount of fattyacid soluble in water, which thus estimated is considerably higherthan is shown by Reichert's process. The author hints that theresults of the latter may be vitiated by the production of propionic,acetic, and formic acids by the action of potash on the glycerol.Richmond, commenting on the above process, shows that the resultsgiven cannot possibly be correct, the total fatty acids, together wit11the glycerol residue corresponding t o the potash neutralised, addingup to more than the weight of the butter taken, and that this is dueto the titration of the insoluble acids being performed in aqueoussolution.The results by Reichert's process, when corrected for therecognised average deficiency of &, add up almost exactly to 100 percent. He points out that at the temperature of the water-bathpotash has no action on glycerol.M. J. 8.M. J. S.Examination of Lard for Adulteration. By T. S . GLADDING(Analyst, 14, 32--.34).-The following tests should all be applied toa, suspected sample:-(1) specific gravity at 100'; (2) Hiibl's iodinetest ; (3) BecLi-Millian test (Abstr., 1889,319) ; (4) Dalican's " Titre "test; (5) Belden's microscopic test for beef fat (Analyst, 13, SO).Dalican's " titre " is the temperature of crystallisation of the fattyacids.These are to be prepared from the sample by saponification,washed well with hot water, and filtered through dry paper into a test-tube. The crystallising point is then taken with a thermometergraduated t o tenths of a degree. The titre of lard may range from36.4" to 41.4" ; iodine absorption from 57 to 68.4 per cent., a hightitre being associated with a low kdine absorption, and vice vers69 4 ABSTRACTS OF CHEMICAL PAPERS.The titre of beef fat is about 41.6 to 44; iodine absorption, 43.8 to40; that of cotton-seed oil, 33.3, iodine absorption, 108.The oneadultmerant will therefore to some extent mask the other; they are,however, respectively revealed by Bechi’s and Belden’s tests, Thehigh specific gravity of cotton-seed oil affords the onlv means of esti-mating the amount of i t present. (See also Abstr., 1889, 319, 659.)M. J. S.Action of Acids on Benzoic Sulphinide and Analysis ofBy I. REMSEN and W. &I. BURTON (Anzer. Chewi. J., u Saccharin.”l1,403408).-When benzoic sulphinide, C6H4<co >NH [ = 1 : 21, so2is bniled with dilute acids, hydrogen ammonium orthosulphobenzoate,COOH.C,H4*S03NH, [= I : 21 is formed, toqether with a little( wthosulphaminebenzoic acid, C OOH.C6H,.S02NH2.The best strengthof acid is that obtained by diluting strong hydrochloric acid of sp. gr.3 -17 with 8 to 10 times its volume of water.Commercial “ saccharin ” is found to be a mixture of parasulph-aminebenzoic acid, benzoic sulphinide, and hydrogen potasqinm ortho-sulphobenzoate, the amount of sulphinide present being somewhat lessthan 50 per cent.To analyse it, 2 grams are boiled for one hour with 100 C.C. ofdilute hydrochloric acid (1-8) in a flask of 250 C.C. capacity, providedv-ith a reflux condenser. The clear solution is then evaporated tonbont 15 c.c., wheii the parasulphaminebenzoic a,cid separates out ; itis dried a t 80” and weighed. The filtrate, containing hydrogen am-monium orthosulphobenzoate (from the decomposition of the sul-phinide) and the hydrogen potassium salt of the same acid, isevaporated ; the residue is weighed, and the amount of potassium init is estimated by heating a portion with sulphuric acid, and weighingthe potassium sulphate formed.Two samples of saccharin wereanalysed, each five times ; the mean percentage composition of eachis given below.BenzoicCOOH*C,H, *S02NHZ. sulphinide. C00hT*C,H4.SO3K.I ...... 50.00 42.86 7.12II ...... 44.49 48.3:3 7.99C. F. B.Estimation of Morphine in Opium. By F. A. FL~CKIGER(Arch. Pharm. [ 3 ] , 27, 721-732, ’7t 9-i72).-The author discussesvarious points which arise in the estimation of morphine, and arrivesa t the following fairly good, although not qnite perfect, method.8 grams of opium powder is placed in a folded filter of 12 cm.diameter with a little tapping, and is dried a t 100”.After half anhour 10 C.C. of ether mixed with 10 C.C. of chloroform is poured overit, the covered funnel being frequently struck, and finally 10 C.C.more of chloroform is poured on. After all possible liquid has runthrough, tlhe filter with its contents is opened out and dried a t agentle heat. Next the powder is vigorously and repeatedly shakenin a flask with 80 C.C. of water and filtered after two hours. 42.5 gramsof the filtrate is well and often shaken in a weighed flask with 7-5 C.CAShLY TICAL CHEMISTRY. 95of alcohol (0.83 sp. gr.), 15 C.C. of ether, and 1 C.C. of ammonia (0.96).After six hours, the contents of the flask are poured on to a double-folded filter of 10 cm.diameter, and the morphine is washed on to thefilter with about 10 C.C. of water. This is dried, returned to the driedflask, and dried at 100" until its weight becomes constant. This pro-cess with a particular sample gave 12.90, 13.12, 13.35 per cent. ofmorphine, which was not pure white, but which dissolved completelyin lime-water with very little colour.. As an appendix the authorcriticises in some detail an article by E. R. Squibb in the "Ephemeris ''for July, on morphine estimation, and, although be sees many defectsin the process given, he remarks that the comprehensive paper desei-veuthe fullest consideration. J. T.Analysis of Pepper and the occurrence of Piperidine in thesame. By W.JOHNSTONE (Analyst, 14, 41-49).-Moisture and ash.-A weighed portion is dried at 100' and then incinerated in a muffle.The ash is treated successively with water and hydrochloric acid, andthe amount of insoluble matter noted.Oil.-20 grams is distilled with water ; the distillate is shaken withether, the ethereal solution is evaporated at a very low temperature,and the residue is dried over sulphuric acid.Piperidine.-20 grams is dist'illed as for the oil determination, andthe distillate is titrated with N/lO sulphuric acid (compare Abstr.,1889, 298). That the piperidine is not derived from the hydrolysisof piperine is shown by the fact that pure piperine yields nopiperidine when distilled with water, also that in distilling pepperwith water, piperidine soon ceases to come over, although the amountobtained is very small in comparison with the piperine present.Piperine.-10 grams is digested at 100" in a closed bottle with3 grams of potash dissolved in 25 C.C. of water and 25 C.C. of alcohol.The bottle (4 oz.) should have tlhe neck ground flat and be closed bya plate of caoutchouc pressed tightly upon it by a screw-frame. After4-6 hours' digestion, the bottle is cooled, the contents are washedinto a large flask and distilled as long as the distillate is alkaline.The theoretical yield of pipedine is obtained.Crude Fibre.-A small quantity is boiled for half an hour in a flaskwith inverted condenser with 200 C.C. of dilute sulphuric acid (12.5grams per litre). The residue is twice boiled with water, then with200 C.C. of potash (12.5 grams per litre), and again twice with water.It is collected on a tared filter, dried and weighed, and any ash itcontains deducted.Nitrogen.-Determined by soda-lime, as usual.Alcoholic Extract.-10 grams is extracted with 95 per cent, alcoholin a Soxhlet's apparatus for 24 hours. The alcohol is distilled off andthe extractive matters dried at 100".St,arch.-The exhausted residue from the preceding is, withoutdrying, washed into a flask with 200 C.C. of water and 20 C.C. ofhydrochloric acid (1.121) and heated in boiling water for three hours.After cooling, the liquid is filtered, neutrdised with soda, made up to500 c.c., and titrated with Fehling's solution.In 13 genuine samples from various localities, the moisture range9 ri ABSTRACTS OF CHEMICAL PAPERS.from 12 to 15 per cent., ash 1.07 to 4.46 (long pepper 7 57), oil 0.53to 1.87, piperidine 0.21 to 0.77, piperine 5.21 to 13.03, fibre 4.2 to15.05, starch 29.6 to 53.5, ash insoluble in acid 0.06 to 0.62 (longpepper 1-47), Any larger amount of insoluble ash would probably bethe best indication of a fraudulent addition. M. J. S.Detection of Cocai'ne Hydrochloride. By M. GOELDNER (Arch.Phai-m. [31, 27, $99; from Pharm. Zeit., 34, 471).-The authorbelieves that the following is a charactelistic test for cocaine. 6 or 7drops of pure, strong sulphuric acid are added to some crystals ofresorcinol in a porcelain basin, and the latter is moved to and fro a little,then a litt,le cocaine hydrochloride is added to t'he yellow liquid. A som+what strong reaction follows, and a splendid, blue coloration is imme-diately obtained; a drop of sodium hydroxide changes this to a light-rose colour. The reaction goes more quickly with powdered resorcinolin place of crystals. Verysmall quantities of the reamgent give nocolour reaction. Other alkaloids give nothing approaching to thisreaction. J. T.Estimation of Indigotin for Commercial Purposes. By F. A.OWEN (Chem. Centr., 1889, ii, 217-213; from J. dmer. Chem. SOC.,10, 178).-J gram of the substance is weighed on a watch-glass,dried a t lOO", finely powdered, rubbed with water to a thin paste, andwashed into a 250 C.C. flask. 3 grams of zinc-dust and 6 grams ofsodium hydroxide are added, the solution dilut,ed to a little above themark, shaken up now and then, and after the reaction is complete(during which the solution must remain green, red or brownishstreaks indicate that the reduction has been carried too far ; a frothindicates the presence of too much zinc), 50 C.C. of the clear liquid isexposed to the air for half an hour, acidified with hydrochloric acid,and filtered through a well-wnshed filter, dried a t loo", and weighed.J. W. L