62 ABSTRACTS OF CHF,MICAL PAPERS. A n a1 y t i c a 1 C h e m i s t r y. Apparatus for the Volumetric Estimation of Nitrogen. By H. SCHWARZ (Ber., 13, 7’71--773).-This is a convenient moclification of Zulkowsky’s apparatus for collecting the nitrogen evolved in organic analyses of nitrogenous compounds by Dumas’s method. w. c. w. Action of Organic Acids on Minerals. By H. C. BOLTON (Bey., 13, 726--734).-l’The author points out that a large number of minerals are decomposed, either by a hot concentrated solution of citric acid, or by the combined action of citric acid and sodium nitrate, potassium iodide or ammonium fluoride, and he suggests that this acid would, on account of its safe portability, prove a valuable reagent during mineralogical excursions. Tables are given of this reaction with a large number of minerals.w. c. w. Estimation of Retrograde Phosphoric Acid. By A. GRUPE and others (Bied. Centr., 1880, 582--584).-The solvent action of ammonium citrate is greater a t 35” C. than at the ordinary tempera- ture, and in using this as a reagent for the estimation of retrograde phosphate, care must be taken to first remove by washing soluble salts of magnesia, which have the effect of rendering bicalcium phosphate insoluble in ammonium citrate ; the presence of calcium carbonate also has the same effect. J. K. C. Retrograde Phosphoric Acid. By M. A. MILLOT (Biecl. Centr., 1880, 577--580).-The author finds that when mineral superphos- phates contain a considerable quantity of ferric oxide, a part of the phosphoric acid combines with the iron, forming a compound insoluble in ammonium citrate, even when a sufficient quantity of sui- phuric acid has been used in the decomposition of the phosphate.Calcium monophosphate also after a time becomes converted into &phosphate, but not into triphosphate. The presence of alumina also gives rise to the formation of phosphates soluble in ammonium citrate. When calcium carbonate is mixed with superphosphate in sufficient quantity, tricalcium phosphate is formed, insoluble in ammonium citrate ; if, however, a neutral solution of potassium citrate is employed, instead of ammoniacal potassium citrate, the retrograding of the phosphoric acid appears to be much greater than it really is. J. K. C. Alkanet Red, a Test for Magnesium Salts. By F. v. LEPEL (Ber., 13, 763-766).-Alkanet red, the colouring matter obtained from the roots of the false alkanet, Anchusa tinctoria, dissolved in a mixture of alcohol (2 parts), and ether (1 part), and diluted with water, exhibits an ahsorption spectrum composed of three bands.The violet-coloured solution produced by the addition of ammonium car- bonate to this mixture, is changed to blue by a minute quantity of aANALYTICAL CHEMISTRY. 63 magnesium salt, and a characteristic fourth absorption band is formed. By this test, 0.05 mgrm. in 1 C.C. of liquid can be detected. w. c. w. Vegetable Colouring Matters as Tests for Magnesium Salts. By I?. v. LEEPEL (Ber., 13, 766-’768).-Magnesium salts alter the absorption spectra of the colouring matters derived from (I) the roots of Aizchusa tiizcforicc and common beet (Beta uzdgaaris) ; (2) the flowers of the dahlia, dragon’s mouth (Antirrhiimm majm), horse chestnut, hyacinth, violet, rhododendron, the purple aster, and PhzuZa fnrinosa ; Chromium Sesquioxide.By T. WILM (BUZZ. XOC. Chirr~. [2], 34. 86) .-By reducing potassium dichromate or chrome alum with alcohol and hydrochloric acid, and then precipitating the sesquioxide by ammonia, a weight of the oxide is obtained greater than that required by theory. This is attributed to a partial oxidation and formation of chromic acid. (3) the juice of elderberries. w. c. w. V. H. V. The Officinal Test of Quinine and Water of Crystallisation of Quinine Sulphate. By 0. HESSE (Bey., 13, 1517--1520).-The method of testing quinine for cinchonidine adopted in Germany is that described in this Journal, 34, 434.The auth0.r points out that when cinchonidine sulphate is not added as an adulterant, but is mixed by crystallisation, it then takes a form which allows some to escape detection by the above method. The author has discovered an optical method of testing quinine, by which small amounts of cinchonidine n a y be detected. As cinchonidine sulphate crystallises with 6 mols. H20, or 13.7 per cent., and pure quinine sulphate with 8 mols. HzO, or 16-17 per cent., so the presence of cinchonidine may be detected by its influence on the percentage of water of crystallisation of a specimen. The purer the quinine sulphate, if perfectly dry and unweathered, the nearer is the percentage of water of crystallisation to 16-17.Korner states that he has obtained pure, unweathered crystals of quinine sul- phate, with only 14.4 per cent. of water. The author’s experience is contradictory to this statement, and he would regard such a low per- centage of water as presumptive evidence of the presence of cinchoni- dine sulphate. P. P. B. Testing Commercial Sulphate of Quinine for Foreign Alkaloids. By G. KERNER (Arch. Pharm. [ 3 ] , 16, 186--205).-The author described in 1862 the ‘‘ ammonia-method ’’ for testing the purity of commercial sulphate of quinine: the method has been adopted very widely. A fuller account of his process is here given, together with criticisms on a method proposed by Hesse. Hessc’s process resembles the author’s in not detecting less than 1 per cent. of cinchonidine sulphate, but from published results it appears to yield only an approximate estimation, whereas the ammonia- method as now described yields numbers of almost absolute accuracy.After stating his reasons for considering Hesse’s process unsatisfactory, the author describes the ammonia-method with recent improvements. The “ammonia-method” depends on the facts that a cold satu-64 ABSTRACTS OF CHEMICAL PAPERS. rated sclution of quinine sulphate contains an invariable quantity of the alkalo?d, and that the quantity of ammonia required to precipitate this as hydrate is constant, and that the excess of ammonia required to re-dissolve this precipitate is also constant. If the quantity of ammonia necessary to form the precipitate, and then re-dissolve it in a known volume of pure quinine sulphate solution a t normal or known temperatures has been determined, the excess of ammonia required for the same quantity of a saturated solution of commercial quinine sulphate gives the means of calculating the quantity of associated alkaloids, of which cinchonidine alone is usually preseut.Since quinine solution readily alters by becoming mouldy, and ammonia solution also alters in strength by keeping,. it is best to prepare a solution of pure quinine snlphate for each estimation, and titrate with it the ammonia to be used. Pure Quiniiae Xul@hate.-It is often necessary to recrystallise from three to six times, with addition of several drops of sulphuric acid in excess, in order t o get rid of the last traces of cinchonidine.The purity can be tested by treating portions with different proportions of cold water and titrating the solutions with ammonia: solutions of commercial samples require different amounts of ammonia when the proportion of solvent to solid has been varied ; but when pure, varia- tions of solid to solvent between 1 : 10 and 1 : '700 produce no altera- tion in the quantity of ammonia required, and the excess of solid remaining undissolved in making the solutions yields also solutions of precisely the same ammonia titre. Preparation, of Solutions and Process of Titratiort.-The pure quinine sulphate prepared as above is rubbed into a homogeneous paste with water in a mortar and rinsed into a stoppered vessel, in which it is frequently agitated during from twelve to eighteen hours : the pro- portion of quinine to water used being about 1 : 100.To prepare the solution of the sample to be tested, 5 grams are similarly treated with 50 C.C. of water. The vessels containing the quinine solutions and the vessel containing the ammonia solution (of 0.92 sp. gr.) are placed in cold water, and as soon as their contents have reached the same temperature, the quinine solutions are filtered through dry filter-papers. The temperature need not be normal, provided both quinine solutions are prepared a t the same temperature. The undissolved pure quinine is dried and kept, for future use. 10 C.C. of* each of the quinine solutions are then measured off into t,est-tubes, and each is titrated with the ammonia solution. 5 C.C. of ammonia solution are run in, and the test-tube is closed with the finger, and its contents are mixed by several times inverting without shaking it.The quinine is thus precipitated and almost entirely redis- solved, the liquid remaining but slightly turbid ; by gradually dropping in ammonia, mixing, and waiting several seconds after each addition, the moment when the liquid becomes perfectly clear is easily noted. The excess of ammonia required gives the quantity of cinchonidine sulphate present; on an average 0.288 c.c., or roughly 0.3 C.C. of ammonia solution of 0.92 sp. gr. were found to correspond to 1 mgrm. of crystallised cinchonidine sulphate. The error in the process cannot exceed 0.05 per cent.ANALYTICAL CHEMISTRY. 65 It must be understood that this process is only directly applicable to samples which answer t o the qualitative ammonia test, and which contain not more than 1.5 per cent.of cinchonidine sulphate ; if more than 2 per cent. is present, the final reaction cannot be obtained, since either insoluble flocks appear or the solution gelatinises when near the clarifying point : if from a preliminary qualitative test the nature or intensity of the turbidity renders it probable that 2 per cent. or more is present, the solution to be titrated may be prepared as directed above, and then be diluted with known quantities of the pure quinine solution, or larger proportions of water to quinine may be employed in making the solution : in the latter case, it is better to warm during the process of solution. It is nofed that chemically pure quinine hydrate crystallises out from the titrnted solution on standing. Afz approximate estimation of the cinchonidine sulphate can be made by introducing 5 C.C.of the water extract (1 : 10, prepared at 15" C.) into a 10 C.C. cylinder graduated to tenths of a c.c., and adding 3 C.C. of ammonia of 0.92 sp. gr. On mixing by inversion, the liquid Kill usually remain very turbid ; ammonia is then gradually added with constant mixing until the liquid becomes perfectly clear, and the total volume of ammonia added is read off. Assuming that 5 C.C. of ammonia indicate 1 per cent. of cinchonidine sulphate, and 3 C.C. indicate none, the percentage can be ascertained. This method gives with great accuracy relative values for quinine samples examined under similar conditions. This method is suaciently accurate for practical purposes, the former more exact method being resorted to for disputed cases and for scientific purposes only.Quinidine is seldom present, except as an aclulterant purposely added ; it dissolves somewhat more readily in excess of ammonia than cinchonidine does. The percentage of water, present as water of cry stallisation or other- wise, is of considepable importance, both for commercial reasons and to ensure the correctness of quantities of the alkalo'id used as doses. The presence of a small quantityof sulphuric acid tends to promote crumbling of the crystals with loss of their water of crystallisation, whilst a small amount of uncombined water tends to prevent this change. Uncombined water is estimated by the difference bet ween the quantities of water found on drying a portion of the original sample and another portion which has been pressed between soft blotting-paper. The author, after a long experience in estimating wa$er in quinine, considers that in the crystalline condition its formula is 2CzoHz4N,O,.H,SO4 -+ 7H,O: when dried at 115" C.it loses 14.45 per cent. of water: in practice the loss varies between 14.38 and 14.8. A good sample will usually not lose more than 13.8 to 14.4 per cent. by drying, but no sample should lose more than 14.6 per cent. without exciting suspicion. If left for some time in a dry and moderately warm situation, the crystallised sulphate loses nearly 5H20 (= 10.32 per cent.), leaving 2CzoH,4N,0z.H2S04 + 2Hz0, which contains 4.60 per cent.of water. In this form, the salt is less presentable in appearance, but is per- manent at temperatures below 100" C.: the sulphate would be well VOL. XL. f66 ABSTRACTS OF CHEMICAL PAPERS. suited for pharmaceutical purposes in this condition, since it is not liable to loss or absorption of moisture, and contains a maximum amount of 5 per cent. of water : it would also be impossible to moisten it without altering its appearance. Another means of avoiding the inconvenience arising from the variation in composition which is noticed in the ordinary commercial sulphate, would be to replace it by the hydrochloride, which is a far less variable salt, and is also more easily assimilated. As long as the crystallised sulphate is used, it is necessary to estimate the percentage of water in every sample in determining its value.The percentage of water is estimated either by finding the quantity of anhydrous alkaloid, and then calculating from the formula the quantity of water as recommended by Dwars (Arch. Phamz., 11, 149), or better, by directly estimating the loss of water, when from 1 to 2 grams are dried by heating very gradually to 115” C. 1’. c. Butter Analysis. By L. MEDICUS and others (Bied. Cerztr., 1880, 615-617). -The authors have examined Reichert’s method of esti- mating adulteration in butter, and confirm the results already obtained by him. Donny recommends as an easy method of detecting adultera- tion to heat a sample of butter in a test-tube at 160” ; if the butter is pure it foams very much, and becomes of a uniform brown colour ; if impurities are present there is very littlle swelling, but the liquid bumps violently, and casein separates out on the sides of the tube in brown masses, whilst the rest of the butter retaius its original colour.J. I(. C.62 ABSTRACTS OF CHF,MICAL PAPERS.A n a1 y t i c a 1 C h e m i s t r y.Apparatus for the Volumetric Estimation of Nitrogen. ByH. SCHWARZ (Ber., 13, 7’71--773).-This is a convenient moclificationof Zulkowsky’s apparatus for collecting the nitrogen evolved inorganic analyses of nitrogenous compounds by Dumas’s method. w. c. w.Action of Organic Acids on Minerals. By H. C. BOLTON (Bey.,13, 726--734).-l’The author points out that a large number ofminerals are decomposed, either by a hot concentrated solution ofcitric acid, or by the combined action of citric acid and sodium nitrate,potassium iodide or ammonium fluoride, and he suggests that this acidwould, on account of its safe portability, prove a valuable reagentduring mineralogical excursions.Tables are given of this reactionwith a large number of minerals. w. c. w.Estimation of Retrograde Phosphoric Acid. By A. GRUPEand others (Bied. Centr., 1880, 582--584).-The solvent action ofammonium citrate is greater a t 35” C. than at the ordinary tempera-ture, and in using this as a reagent for the estimation of retrogradephosphate, care must be taken to first remove by washing soluble saltsof magnesia, which have the effect of rendering bicalcium phosphateinsoluble in ammonium citrate ; the presence of calcium carbonate alsohas the same effect.J. K. C.Retrograde Phosphoric Acid. By M. A. MILLOT (Biecl. Centr.,1880, 577--580).-The author finds that when mineral superphos-phates contain a considerable quantity of ferric oxide, a part ofthe phosphoric acid combines with the iron, forming a compoundinsoluble in ammonium citrate, even when a sufficient quantity of sui-phuric acid has been used in the decomposition of the phosphate.Calcium monophosphate also after a time becomes converted into&phosphate, but not into triphosphate. The presence of alumina alsogives rise to the formation of phosphates soluble in ammonium citrate.When calcium carbonate is mixed with superphosphate in sufficientquantity, tricalcium phosphate is formed, insoluble in ammoniumcitrate ; if, however, a neutral solution of potassium citrate is employed,instead of ammoniacal potassium citrate, the retrograding of thephosphoric acid appears to be much greater than it really is.J.K. C.Alkanet Red, a Test for Magnesium Salts. By F. v. LEPEL(Ber., 13, 763-766).-Alkanet red, the colouring matter obtainedfrom the roots of the false alkanet, Anchusa tinctoria, dissolved in amixture of alcohol (2 parts), and ether (1 part), and diluted withwater, exhibits an ahsorption spectrum composed of three bands. Theviolet-coloured solution produced by the addition of ammonium car-bonate to this mixture, is changed to blue by a minute quantity of ANALYTICAL CHEMISTRY. 63magnesium salt, and a characteristic fourth absorption band is formed.By this test, 0.05 mgrm.in 1 C.C. of liquid can be detected. w. c. w.Vegetable Colouring Matters as Tests for Magnesium Salts.By I?. v. LEEPEL (Ber., 13, 766-’768).-Magnesium salts alter theabsorption spectra of the colouring matters derived from (I) the rootsof Aizchusa tiizcforicc and common beet (Beta uzdgaaris) ; (2) the flowersof the dahlia, dragon’s mouth (Antirrhiimm majm), horse chestnut,hyacinth, violet, rhododendron, the purple aster, and PhzuZa fnrinosa ;Chromium Sesquioxide. By T. WILM (BUZZ. XOC. Chirr~. [2], 34.86) .-By reducing potassium dichromate or chrome alum with alcoholand hydrochloric acid, and then precipitating the sesquioxide byammonia, a weight of the oxide is obtained greater than that requiredby theory.This is attributed to a partial oxidation and formationof chromic acid.(3) the juice of elderberries. w. c. w.V. H. V.The Officinal Test of Quinine and Water of Crystallisationof Quinine Sulphate. By 0. HESSE (Bey., 13, 1517--1520).-Themethod of testing quinine for cinchonidine adopted in Germany isthat described in this Journal, 34, 434. The auth0.r points out thatwhen cinchonidine sulphate is not added as an adulterant, but is mixedby crystallisation, it then takes a form which allows some to escapedetection by the above method. The author has discovered an opticalmethod of testing quinine, by which small amounts of cinchonidinen a y be detected. As cinchonidine sulphate crystallises with 6 mols.H20, or 13.7 per cent., and pure quinine sulphate with 8 mols.HzO, or16-17 per cent., so the presence of cinchonidine may be detected by itsinfluence on the percentage of water of crystallisation of a specimen.The purer the quinine sulphate, if perfectly dry and unweathered, thenearer is the percentage of water of crystallisation to 16-17. Kornerstates that he has obtained pure, unweathered crystals of quinine sul-phate, with only 14.4 per cent. of water. The author’s experience iscontradictory to this statement, and he would regard such a low per-centage of water as presumptive evidence of the presence of cinchoni-dine sulphate. P. P. B.Testing Commercial Sulphate of Quinine for ForeignAlkaloids. By G. KERNER (Arch. Pharm. [ 3 ] , 16, 186--205).-Theauthor described in 1862 the ‘‘ ammonia-method ’’ for testing thepurity of commercial sulphate of quinine: the method has beenadopted very widely.A fuller account of his process is here given,together with criticisms on a method proposed by Hesse.Hessc’s process resembles the author’s in not detecting less than1 per cent. of cinchonidine sulphate, but from published results itappears to yield only an approximate estimation, whereas the ammonia-method as now described yields numbers of almost absolute accuracy.After stating his reasons for considering Hesse’s process unsatisfactory,the author describes the ammonia-method with recent improvements.The “ammonia-method” depends on the facts that a cold satu64 ABSTRACTS OF CHEMICAL PAPERS.rated sclution of quinine sulphate contains an invariable quantity ofthe alkalo?d, and that the quantity of ammonia required to precipitatethis as hydrate is constant, and that the excess of ammonia requiredto re-dissolve this precipitate is also constant.If the quantity ofammonia necessary to form the precipitate, and then re-dissolve it ina known volume of pure quinine sulphate solution a t normal or knowntemperatures has been determined, the excess of ammonia requiredfor the same quantity of a saturated solution of commercial quininesulphate gives the means of calculating the quantity of associatedalkaloids, of which cinchonidine alone is usually preseut. Sincequinine solution readily alters by becoming mouldy, and ammoniasolution also alters in strength by keeping,.it is best to prepare asolution of pure quinine snlphate for each estimation, and titrate withit the ammonia to be used.Pure Quiniiae Xul@hate.-It is often necessary to recrystallise fromthree to six times, with addition of several drops of sulphuric acidin excess, in order t o get rid of the last traces of cinchonidine. Thepurity can be tested by treating portions with different proportions ofcold water and titrating the solutions with ammonia: solutions ofcommercial samples require different amounts of ammonia when theproportion of solvent to solid has been varied ; but when pure, varia-tions of solid to solvent between 1 : 10 and 1 : '700 produce no altera-tion in the quantity of ammonia required, and the excess of solidremaining undissolved in making the solutions yields also solutions ofprecisely the same ammonia titre.Preparation, of Solutions and Process of Titratiort.-The pure quininesulphate prepared as above is rubbed into a homogeneous paste withwater in a mortar and rinsed into a stoppered vessel, in which it isfrequently agitated during from twelve to eighteen hours : the pro-portion of quinine to water used being about 1 : 100.To prepare thesolution of the sample to be tested, 5 grams are similarly treated with50 C.C. of water. The vessels containing the quinine solutions andthe vessel containing the ammonia solution (of 0.92 sp. gr.) are placedin cold water, and as soon as their contents have reached the sametemperature, the quinine solutions are filtered through dry filter-papers.The temperature need not be normal, provided both quinine solutionsare prepared a t the same temperature.The undissolved pure quinineis dried and kept, for future use.10 C.C. of* each of the quinine solutions are then measured off intot,est-tubes, and each is titrated with the ammonia solution. 5 C.C. ofammonia solution are run in, and the test-tube is closed with thefinger, and its contents are mixed by several times inverting withoutshaking it. The quinine is thus precipitated and almost entirely redis-solved, the liquid remaining but slightly turbid ; by gradually droppingin ammonia, mixing, and waiting several seconds after each addition,the moment when the liquid becomes perfectly clear is easily noted.The excess of ammonia required gives the quantity of cinchonidinesulphate present; on an average 0.288 c.c., or roughly 0.3 C.C.ofammonia solution of 0.92 sp. gr. were found to correspond to 1 mgrm.of crystallised cinchonidine sulphate. The error in the process cannotexceed 0.05 per centANALYTICAL CHEMISTRY. 65It must be understood that this process is only directly applicableto samples which answer t o the qualitative ammonia test, and whichcontain not more than 1.5 per cent. of cinchonidine sulphate ; if morethan 2 per cent. is present, the final reaction cannot be obtained, sinceeither insoluble flocks appear or the solution gelatinises when nearthe clarifying point : if from a preliminary qualitative test the nature orintensity of the turbidity renders it probable that 2 per cent.or moreis present, the solution to be titrated may be prepared as directedabove, and then be diluted with known quantities of the pure quininesolution, or larger proportions of water to quinine may be employedin making the solution : in the latter case, it is better to warm duringthe process of solution.It is nofed that chemically pure quinine hydrate crystallises outfrom the titrnted solution on standing.Afz approximate estimation of the cinchonidine sulphate can be madeby introducing 5 C.C. of the water extract (1 : 10, prepared at 15" C.)into a 10 C.C. cylinder graduated to tenths of a c.c., and adding 3 C.C.of ammonia of 0.92 sp. gr. On mixing by inversion, the liquid Killusually remain very turbid ; ammonia is then gradually added withconstant mixing until the liquid becomes perfectly clear, and thetotal volume of ammonia added is read off.Assuming that 5 C.C. ofammonia indicate 1 per cent. of cinchonidine sulphate, and 3 C.C.indicate none, the percentage can be ascertained. This method giveswith great accuracy relative values for quinine samples examinedunder similar conditions. This method is suaciently accurate forpractical purposes, the former more exact method being resorted tofor disputed cases and for scientific purposes only.Quinidine is seldom present, except as an aclulterant purposelyadded ; it dissolves somewhat more readily in excess of ammonia thancinchonidine does.The percentage of water, present as water of cry stallisation or other-wise, is of considepable importance, both for commercial reasons andto ensure the correctness of quantities of the alkalo'id used as doses.The presence of a small quantityof sulphuric acid tends to promotecrumbling of the crystals with loss of their water of crystallisation,whilst a small amount of uncombined water tends to prevent thischange.Uncombined water is estimated by the difference bet weenthe quantities of water found on drying a portion of the originalsample and another portion which has been pressed between softblotting-paper. The author, after a long experience in estimatingwa$er in quinine, considers that in the crystalline condition its formulais 2CzoHz4N,O,.H,SO4 -+ 7H,O: when dried at 115" C.it loses 14.45per cent. of water: in practice the loss varies between 14.38 and14.8. A good sample will usually not lose more than 13.8 to 14.4 percent. by drying, but no sample should lose more than 14.6 per cent.without exciting suspicion.If left for some time in a dry and moderately warm situation, thecrystallised sulphate loses nearly 5H20 (= 10.32 per cent.), leaving2CzoH,4N,0z.H2S04 + 2Hz0, which contains 4.60 per cent. of water.In this form, the salt is less presentable in appearance, but is per-manent at temperatures below 100" C.: the sulphate would be wellVOL. XL. 66 ABSTRACTS OF CHEMICAL PAPERS.suited for pharmaceutical purposes in this condition, since it is notliable to loss or absorption of moisture, and contains a maximumamount of 5 per cent. of water : it would also be impossible to moistenit without altering its appearance. Another means of avoiding theinconvenience arising from the variation in composition which isnoticed in the ordinary commercial sulphate, would be to replace itby the hydrochloride, which is a far less variable salt, and is also moreeasily assimilated. As long as the crystallised sulphate is used, it isnecessary to estimate the percentage of water in every sample indetermining its value.The percentage of water is estimated either by finding the quantityof anhydrous alkaloid, and then calculating from the formula thequantity of water as recommended by Dwars (Arch. Phamz., 11, 149),or better, by directly estimating the loss of water, when from 1 to 2grams are dried by heating very gradually to 115” C. 1’. c.Butter Analysis. By L. MEDICUS and others (Bied. Cerztr., 1880,615-617). -The authors have examined Reichert’s method of esti-mating adulteration in butter, and confirm the results already obtainedby him. Donny recommends as an easy method of detecting adultera-tion to heat a sample of butter in a test-tube at 160” ; if the butter ispure it foams very much, and becomes of a uniform brown colour ; ifimpurities are present there is very littlle swelling, but the liquidbumps violently, and casein separates out on the sides of the tube inbrown masses, whilst the rest of the butter retaius its original colour.J. I(. C