FOOD AND DRUGS ANALYSIS 347 ABSTRACTS OF PAPERS PUBLISHED IN OTHER JOURNALS. ANALYSIS OF FOOD AND DRUGS. Examination of Copaiba Oil. E. Deussen and B. Eger. (Chem. Zeit., 1912, 36, 561-562.)-0n treating 6-caryophyllene with nitrogen peroxide, a com- pound, CI,H,,N,O,, which may be termed P-nitro-caryophyllene (m.-pt. 159.5" to 160° C.), is obtained. This reaction may be used in the examination of copaiba oil in the following manner: Three grms.of the sample are dissolved in 27 C.C. of anhydrous ether, and the solution treated with nitrogen peroxide until the precipitated P-nitro-caryophyllene begins to agglomerate. The precipitate is then separated, washed with ether, dried on porous porcelain, and weighed (cj. Annalen, 1912, 388, 138). The following yields of P-nitro-caryophyllene were thus obtained : Caryophyllene (from clove oils) 50 to 52; Para copaiba oil (a), 9-5 to 10; (b, c, d,) 15 to 16; Maracaibo copaiba oils, 3 to 6; Maranham copaiba oil, 8 to 9; Para copaiba oil (d) adulterated with 10 per cent.of gurjun oil, 13.3 to 14.3; with 20 per cent, of gurjun oil, 11.7 to 12.7 ; with 30per cent., 10.7 to 11-7 ; and with 50 per cent., 7.7 to 8.3 per cent.For the detection of 10 per cent. of African copaiba oil in Para oil, the optical rotation may be determined. The melting-point of the dihydrochloride also rises with the increase in the proportion of African copaiba oil. Thus, on passing hydrochioric acid gas into pure Para oil, a mixture of P-caryophyllene dihydro- chloride (m.-pt. 69' to 70" C.) and of Z-cadinene hydrochloride (m.-pt.117' to 118" C.) is obtained, melting at about 75" C., whilst in the presence of African copaiba oil, which is rich in cadinene, the m.-pt. of the mixed hydrochlorides is raised to about 85" C. by 5 per cent.; about 111' C. by 10 per cent.; and about 114' to 115" C. by 20 per cent. I n testing Para copaiba oil for gurjun oil by Turner's reaction 3 drops of the sample are dissolved in 3 C.C.of glacial acetic acid, and the solution treated with 2 drops of freshly prepared 10 per cent. sodium nitrite solution, and then cautiously with concentrated sulphuric acid. In the presence of gurjun oil the acetic acid solution becomes dark violet within thirty minutes. Since certain pure copaiba balsams give a faint positive reaction in this test, the author modifies the method by using 1 drop of the sample and 1 per cent.sodium nitrite solution, and only taking note of colorations produced within five minutes. When a positive result is obtained by this method, 170 grms. of the sample are separated by distillation into three fractions of 50 grms. each up to 145" C. Each of these fractions is dissolved in 300 grms.of acetone (containing water), and the solution oxidised by adding, with constant shaking, 160 grms. of powdered potassium permanganate in portions of about 3 grms. every fifteen minutes. After oxidation is complete, the acetone solution is decanted, and the manganese residue shaken, first with 200 grms. of acetone, and then with 200 grms. of ether. The solvent is evaporated, and the residual oil dried and fractionated under a pressure of 2 to 3 mm., so as to obtain three fractions, boiling respectively below 130" C., 140° C., and 165" C.The second and, especially, the third fractions are then subjected to the semi-carbasone test as follows :348 ABSTRACTS OF CHEMICAL PAPERS Two grms. of the fraction are added, drop by drop, to a concentrated aqueous solution of 1.1 grm.of semi-carbazide hydrochloride, to which has been added 1 grm. of potassium acetate dissolved in 3 grms. of alcohol and 1 drop of acetic acid, and the final mixture is cleared by the addition of sufficient alcohol. The semi-carbazone is collected after two days, washed with water, dried on porcelain, and weighed. Its m.-pt. ranges from 215" to 228" C., rising to 234" C.after recrystallisation from boiling alcohol. The high specific rotation ( + 317') in concentrated aqueous chloral hydrate solution is characteristic of the semi-carbazone from gurjun oil. C. A. M. Estimation of Hydrocarbons in Beeswax and Carnauba Wax. A. Leys. (J. Pharm. Chim., 1912,5,577-591.)-A method of estimating hydrocarbons in waxes is based upon the fact that the higher alcohols in beeswax, etc., are soluble in a mixture in equal parts of amyl alcohol and concentrated hydrochloric acid, whilst hydrocarbons may be separated as a solid cake from the solvent.Ten grms. of the wax are saponified with 25 C.C. of alcoholic potassium hydroxide (45 grms. per litre of absolute alcohol) and 50 C.C. of crystallisable benzene in a flask, with a tubulus and tap at the side, so that it may also serve as a separating funnel. After heating the mixture for twenty minutes, 50 C.C.of hot water are added, the boiling continued under a reflux condenser for another ten minutes, and the flame then withdrawn. The liquid soon separates into two layers, and the lower soap solution is then drawn off without cooling. The residual benzene is heated again for ten minutes with 50 C.C.of water, which is subsequently drawn off as before, and the benzene is then evaporated, leaving a residue of alcohols and hydrocarbons. This is treated little by little with 100 C.C. of hot &my1 alcohol, and the whole afterwards mixed with 100 C.C. of concentrated hydrochloric acid, and gently heated, with continual stirring, until the mixture has become completely fluid.The flask is now slowly cooled, and the cake of hydrocarbons at the top is separated from the semi-crystalline mass below, redissolved in the 50 C.C. of the mixture of amyl alcohol and hydrochloric acid, separated as before, pressed between filter-paper, heated on the water-bath in a tared basin, and weighed. The wax alcohols in the solvent are separated by heating the liquid with an excess of water, cooling the mixture, drawing off the lower layer of diluted hydrochloric acid, evaporating the amyl alcohol, taking up the residue with benzene, and evaporating the solution in a tared basin. As a further test the liquid fatty acids of the wax may be estimated by treating the lead salts with benzene.Two samples of pure beeswax thus examined gave the following results : Alcohols, 39.21 and 39.60 (melting at 77" C.) ; hydrocarbons, 10.44 and 13-03 ; and liquid fatty acids, 8-40 and 8.52 per cent. respectively.I n the presence of ceresin or other foreign hydrocarbons, the amount of beeswax is calculated from the ester value, and the corresponding quantity of hydrocarbons deducted from the total hydrocarbons separated.Carnauba wax thus examined yielded 49.22 per cent. of alcohols (m.-pt. 81° C.), but no hydrocarbons. C. A. M. Refractive Index of Beeswax. L. Feldstein. (Chem. Engineer, 1912, 15, 211-212.)-To obtain data as to the refractive index of pure beeswax at temperaturesFOOD AND DRUGS ANALYSIS 349 above the m.-pt., twenty-seven samples of known purity were examined.These gave results falling within the limits of 1.4398 and 1.4451 at 7 5 O C., whilst samples known to contain small quantities of foreign substances gave values outside these limits. Since all pure waxes give a sharp reading at 7 5 O C., it is recommended that that temperature should be adopted for the estimation, and that the calculation of results to the corresponding values at 40" C.should be abandoned. C. A. M. Detection of Benzoic Acid in Foods. 0. Biernath. (Ver5fentZ. a. d . Gebiete d. Militarsanitlitswesens, 1912, Heft 25 ; through Chem. Zentralbl., 1912, I,, 1929.)-The reaction of A. Jonescu (J. Pharm. Chim., 1909, 29, 523) with 1 drop of 1 per cent. ferric chloride solution and 3 or more drops of 1 per cent. hydrogen peroxide solution is very sensitive ; it renders possible the detection of 0.001 grm.benzoic acid in foods, by distillation, in fifteen minutes. Mineral acids, volatile fatty acids, and other volatile acids hinder the reaction. If the first distillate from material containing benzoic acid does not yield a reaction within fifteen minutes, four subsequent distillates from the residue will each give the reaction within fifteen minutes.Salicylic acid, accompanying the benzoic acid, may be destroyed by means of alkaline permanganate in the liquid obtained by distilling the sample with 20 C.C. of water and 0-5 C.C. of sulphuric acid. The benzoic acid is then distilled off, and its presence determined as before. 0. E. M. Analytical Constants of Some Vegetable Fats. H. Sprinkmeyer and A.Diedrichs. (Zeitsch. Untersuch. Nahr. Genussm., 1912, 23, 581-596.)-The following results were obtained with samples of different vegetable fats which are now being used for edible and other purposes : Mowrah butter- Crude, 23 samples : Lowest ... ... Highest.. . . . . Refined, 9 samples : Lowest ... ... Highest.. . Sh ea butter- 13 samples : Lowest ... ... Highest.. . .. . Adjab fat- Stillingia tailow- Crude, 1 sample : Refined, 3 samples : Lowest ... ... Highest.. . . . . Borneo tallow- Tulucuna fat- Dika fat- Malukang butter- Befined, 1 saniple Refined, 1 saniple 1 saniple ... ... 1 sample ... ... "ID 40" C. 50.7 53.2 52.8 54 -2 58.2 60.7 52.4 46'7 47'1 46 *7 51.7 3 6 5 44.2 Iodine Value. 61.2 64.2 60.2 63.8 54'4 60-0 59.0 39 5 40.5 29-7 64.9 2'9 19.4 3sponifi- cation Value.190 '4 192.7 190'0 192.0 177.0 188-0 182-8 200.9 202-4 196.6 194'8 242-0 253'0 Acid Value. 15.0 58.1 1.3 3 -5 9.0 66.6 29.0 0.6 0.9 1 -0 -_ 4-0 11.4 Melting- Point 0 c. 24.5 28'4 26-4 28'4 43 *4 45'2 42.5 41.3 42 5 28 -7 37 '5 33.0 - Solidify- ing Point O c. 16.6 22.2 18 -7 19.6 21.0 25 '2 39-3 26 *7 27 -5 22.6 32.3 - - Reichert- Meissl Value. 1 -70 - - - 1-40 2.50 - 0 -20 0 -80 - 2.30 0.22 15-50 ?olenske Value.0 '40 - -- - 0 '45 - 0.50 0.50 0.60 - 0'50 5-50 0.60 LJnnsaponi- fiable Matter. Per Cent. 1'82 -- - - 3.60 10.01 7 '40 0 -30 0'49 0.50 - - -350 ABSTXACTS OF CHEMICAL PAPERS Mowrah butter is obtained from the seeds of Bassia longifolia, adjab fat from the seeds of Mimusops djave, shea butter from the seed kernels of Bassia Parkii, tulucuna fat from the seeds of Carapa procera, dika fat from the seeds of Irvingia gabonensis, and malukang butter from the seeds of Polygctla bulyracea. With the exception of tulucuna fat, and possibly stillingia tallow, all the above-mentioned fats are suitable for edible purposes.Tulucuna fat possesses a bitter taste. W. P. S. Behaviour of Milk when treated with Ammonium Sulphate, and a New Method for the Estimation of Laetose.E. Salkowski. (Zeitsch. physiol. Chem., 1912, 87, 89-95.)-When milk is saturated, or nearly so, with ammonium sul- phate, the proteins are precipitated completely together with the fat, and a clear solution is obtained in which the lactose may be estimated polarimetrically. Fifty C.C. of the milk are shaken with 17.5 grms.of ammonium sulphate, and, when the salt has dissolved, the mixture is diluted to 100 C.C. with saturated ammonium sulphate solution. The whole is shaken and passed through a dry filter, the filtrate being examined polarimetrically. w. P. s. Detection of Saponin in Beverages and Foods by Haemolysis. C. Sor- mani. (Zeitsch. Untersuch. Nahr. Genussm., 1912, 23, 661-566.)-A rhsumh is given of the papers published by A.Rusconi (Boll. SOC. Med. Chir., Pavia, 1910, 1911,1912) on the detection of saponin by hemolysis-that is, the property possessed by certain substances of dissolving the red colouring matter from blood-corpuscles when the latter are suspended in a suitable medium. The blood-corpuscles are obtained by submitting defibrinated blood to centrifugal action with 1 per cent.sodium chloride solution, separating the corpuscles, and suspending them in physiological salt solution. The liquid or substance to be tested must not contain alcohol or carbon dioxide, and must be neutral in reaction ; before being added to the corpuscle liquid it is rendered isotonic or slightly hypertonic by the addition of 2 per cent. of sodium citrate.The test may be used for the detection of Agrostemma githago in flour, the injurious properties of these seeds being due to the presence of saponin. I n this case the flour is made into a dough with 2 per cent. citrate solution, and the dough is then kneaded with a further quantity of the citrate solution, the extract obtained being filtered, and B portion of the filtrate used for the test.The mixture of corpuscle liquid and the solution to be tested is placed in a thermostat for three hours, at the end of which time the red colouring matter will have been dissolved, and the super- natant liquid will appear coloured, should haemolysis due to the presence of saponin have taken place. Red wine contains a substance which inhibits hemolysis by saponin, but which may be removed by precipitation with albumin. The test will detect the presence of 1 part of saponin in 10,000 parts of beer, or in 40,000 parts of aerated water.w. P. s. Detection of Saponin. J. Ruhle. (Zeitsch. Untersuch. Nahr. Genussm., 1912 23, 566-577.)-The author has investigated various tests for the detection of saponin, and finds that Vamvakas’ reaction, and colour reactions with sulphuric acid, with Frohde’s reagent, and with naphthol, thymol, and ox-gall, are all unreliable, chieflyFOOD AND DRUGS ANALYSIS 351 owing to the presence of interfering substances in the articles under examination (cf.ANALYST, 1912, 22). Saponin may, however, be readily detected, even in the presence of glycyrrhizin, by means of its haemolytic properties.The action of the saponin is dependent on the amount of cholesterol contained in the corpuscles, those containing the largest quantity being most resistant to the haemolytic action of the saponin. If cholesterol be added to the corpuscle liquid, a point is reached finally at which saponin has no action, and in this way saponin may be differentiated from other haemolytic substances. In carrying out the test, 1 C.C.of a dorpuscle liquid (obtained by mixing 1 C.C. of defibrinated blood with 99 C.C. of a 0.9 per cent. sodium chloride solution, submitting the mixture to centrifugal action, pouring off the super- natant liquid, and diluting the residue to 100 C.C. with the sodium chloride solution) is mixed in a test-tube with from 1 to 3 C.C. of the solution to be tested.If, after the lapse of some length of time, the corpuscles have dissolved, a further quantity of the liquid under examination is mixed with cholesterol and the test repeated; the absence of hemolytic action in this test indicates definitely that saponin is present. About 4 parts of cholesterol will inhibit the action of 20 parts of saponin. The cholesterol is most conveniently added to the saponin solution in the form of an ethereal solution, the mixture being then heated to 36" C.for some hours in order to expel the ether. When only very small quantities of saponin are being dealt with, the test may be carried out on a microscope elide, the haemolytic action of the saponin being observed under the microscope. w. P. s. Polarimetric Estimation of Starch.C. J. Lintner. (Zeitsch. angew. Chem., 1912, 25, 1177.)-Wenglein's method of hydrolysing the starch with sulphuric acid of sp. gr. 1.7 is liable to give irregular results in hot weather owing to the hydrolysis pro- ceeding too far. The substitution of hydrochloric acid for sulphuric acid is not practi- cable, since, even without heating, the stage of gelatinisation followed by colloidal solution may be rapidly passed.Under the conditions found most suitable for obtaining constant results, barley starch gave [.ID = 198" either with sulphuric acid of sp. gr. 1-7 kept cool, or with acid of sp. gr. 1.4 at 22" C., the latter being found the more reliable. The method finally based on these experiments wag as follows : 28 grms. of the finely- ground barley are thoroughly mixed with 5 C.C.of 96 per cent. (by vol.) alcohol in a 100 c.c.-flask with a wide neck, and 50 C.C. of sulphuric acid of sp. gr. 1.4 (= 50 per cent. of sulphuric acid) introduced with continual shaking. The flask is then left for an hour, with frequent shaking, at 22" C., after which 10 C.C. of a 2 per cent. phospho- tungstic acid solution are added, and the contents made up to 100 C.C.with sulphuric acid of sp. gr. 1.3 ( = 40 per cent. acid). After thorough shaking, the mixture is filtered, and examined polarimetrically, using sodium light. The rotation multiplied by 10-1 gives the amount of starch in the air-dried barley. C. A. M. New Method for the Estimation of Glycerol in Wines. C. Beis. (Bull. SOC. Chim., 1912, 11, [iv.], 618.)-Existing methods for the estimation of glycerol in wines usually involve removal of the sugar by means of barium or calcium hydroxide, variations in the quantities of which lead to inconsistent results : too little allows sugar to pass into the glycerol, while too much causes the retention of glycerol, To352 ABSTRACTS OF CHEMICAL PAPERS meet this difficulty the following method was devised: 100 C.C.of ordinary wine, 50 C.C. of sweet wine, or 25 C.C. of wine over 1.035 sp. gr., are neutralised in a porcelain basin with barium hydroxide and concentrated at 70' C. to a syrup, mixed with sand, and extracted by warming with acetone below its boiling-point. The first portion of acetone is sufficient in quantity to prevent its dilution by the water in the residue below 95 per cent.of acetone ; succeeding portions are smaller. (40 to 50 c.c.) Each portion is cooled and filtered ; the process is carried on until the volume of the filtrate reaches 200 C.C. Two aliquot parts of the filtrate are evaporated without ebullition. TO the other is added five times its volume of water, and powdered barium hydroxide; a few milligrams in excess of the weight of sugar, if this is not more than 0.050 grm. : a weight equal to that of the sugar if this is between 0.050 and 0.300 grm.; and four-fifths of the weight of tbe sugar if between 0.300 and 0.500 grm. Sand is added, and the mixture extracted wit.h 40 C.C.of acetone, followed by several portions of 25 C.C. Each extract is filtered, and the united filtrate evaporated in a suitable vessel and the residual glycerol weighed.The sugar is estimated by Fehling's solution in one of them. 0. E. M. Analyses of Wines from Eastern France, G. Filaudeau. (Ann. FaZsi@., 1912, 5, 296-303.)-The wines examined comprised samples from the departments of Meurthe-et-Moselle (10 samples), Meuse (9), Vosges (8), Haute-Sa6ne and Doubs (lo), and Jura (10). The results obtained were normal as compared with those found in other years for these wines. w. P. s. Analyses of Wines from the Department of Gironde. E. Dubourg. (Ann. Falsific., 1912, 5, 281-296.)-Results of analyses of 206 samples of red wines and 71 samples of white wines from different districts in the department of Gironde (France) are recorded, the wines all being of the 1911 vintage. The average percentage results for all the red wines were: Alcohol, 9.8; total solids, 23-58; ash, 2 4 2 ; reducing sugars, 3-75 ; volatile acid, 0-85 ; non-volatile acid, 2.70 ; potassium hydrogen tartrate (from the total potassium), 4.60; (from the tartaric acid), 1-95. In the case of the white wines (excluding the sauternes) the average results were : alcohol, 10.1 ; total solids, 28-47 ; ash, 1.74 ; reducing sugars, 7-74 ; volatile acid, 0.55 ; non-volatile acid, 3-19 ; potassium hydrogen tartrate (from the total potassium), 2.69 ; (from the tartaric acid), 2-20, As compared with other years, the red wines contained less non- volatile acid and a correspondingly larger amount of potassium. These irregularities were less accentuated in the case of the white wines. w. P. s.