FOOD AND DRUGS ANALYSIS 503 ABSTRACTS OF PAPERS PUBLISHED IN OTHER JOURNALS. FOOD AND DRUGS ANALYSIS. Deteetion and Estimation of Benzoic Aeid in Foods. K. Baumann and J. Grossfeld. (Zeitsch. Untersuch. Nahr. ‘Genussnz., 191 5, 29, 387-409 ; through Chem. Zmztr., 1915, ii., 202-203.)- Several methods for the detection of benzoic acid were investigated, and Mohler’s method was found to be the most trustworthy (cj.Halphen, ANALYST, 1908, 33, 420; Von der Heide and Jacob, ibid., 1910, 35, 166). The following procedure is recommended for the estimation, of benzoic acid in various animal foods. In milk : 100 C.C. of the sample are mixed with a few C.C. of phenolphthalein solution and concentrated potassium carbonate solution is added until a permanent pink coloration is obtained ; the mixture is then diluted to 200 c.c., 20 C.C.of 20 per cent. calcium chloride are added, the mixture is heated for fifteen minutes on a waterbath, cooled, diluted to 250 c.c., and filtered. Old milks tend to yield a turbid filtrate, but this dces not interfere with the process. A quantity, say 175 c.c., of the filtrate is transferred to a 200 C.C. flask, a few drops of copper sulphate solution are added (to remove any sulphide present), followed by 10 C.C.of sodium phosphotungstate solution (200 grms. of sodium tungstate and 120 grms. of sodium phosphate per litre), and the mixture is diluted with sulphuric acid (1 : 3) to the mark. After standing for eighteen hours, the mixture is filtered, 100 C.C. of the filtrate are percolated for six hours with carbon tetrachloride, the latter is then504 ABSTRACTS OF CHEMICAL PAPERS drawn off, mixed with an equal volume of neutral alcohol containing phenolphtha- lein, and the solution titrated with TG alkali solution.Under these conditions, milk free from benzoic acid requires 0.1 C.C. of the alkali solution for neutralisation ; allowing for this correction, the quantity of benzoic acid present is found from the equation : x= 0.03486 (n - 0.1) grm., where n= C.C.of Fa alkali solution used. In fats, etc. : 50 grms. of the fatty food, such as butter or margarine, are melted in a beaker, dissolved in benzene, the solution is washed into a separating funnel with ether (any insoluble portion may be left in the beaker), and shaken with 20 C.C. of warm potassium carbonate solution after the addition of phenolphthalein ; if the red coloration disappears, more of the potassium carbonate solution must be added.The aqueous layer is now drawn off into the beaker, and the fatty layer is washed several times with warm water. The united extracts are then treated with 20 C.C. of calcium chloride solution, evaporated until all odour of benzene has disappeared, the solution next transferred to a 200 C.C.flask, and the process continued as described under milk. The amount of benzoic acid present is calculated from the equation : x= 27.88 (n - p ) mgrms., where p is the number of C.C. of & alkali solution used by 100 grms. of different fats free from benzoic acid. For lard the correction is 0.05 c.c., for butter 0.50 c.c., and for palm oil 0.30 C.C.w. P. s. Note.-An almost similar method has been previously described by Revis (ANALYST, 1912, 37, 346) for the detection of benzoic acid in milk. He also points out that in the case of fats ‘‘ violent shaking I’ with the alkali is necessary for the removal of the benzoic acid.-ED. Caffeine in Java Tea. J. J. B. Deuss, (Chem. WeekbZad, 1915, 12, 938-943.) -The amount of caffeine in twenty-nine samples of Java tea, was estimated by van Romburgh and Nanninga’s method of extraction with chloroform, the extracted alkaloid being weighed and the result controlled by the refractometer reading of the aqueous caffeine solution. The proportion ranged from 2.68 to 5.58 per cent., and the conclusion is drawn that 3 per cent.of caffeine may be taken as the minimum pro- portion of caffeine in good Java tea.In the case of certain samples of good quality with low caffeine content (e.g., 2.85 per cent. in Java orange pekoe) the leaves came from Chinese plants or hybrids. C. A. M. Detection of Citric Aeid in Wine. E. Baier and P. W. Neumann. (Zeitsch. Untersuch. Nahr. Geizussm., 1915, 29, 410-411 ; through C‘hcm.Zentralbl., 1915, ii., 205-206).-The test depends on the formation of the mercury salt of acetone- dicarboxylic acid. Twenty-five C.C. of the wine are neutralised, acidified with acetic acid, then shaken for ten minutes with the addition of 3 grms. of blood-charcoal (free from calcium carbonate), and filtered. To 10 C.C. of the filtrate is added 1 C.C. of DenigBs’ reagent (5 grms.of mercuric oxide dissolved in 20 C.C. of sulphuric acid and 100 C.C. of water), the mixture is boiled, cooled, filtered if any precipitate has formed, and 1 per cent. potassium permanganate is added drop by drop until manganese dioxide begins to precipitate; the precipitate is then dissolved by the addition of hydrogen peroxide. The solution exliibits a white turbidity, or a precipitate forms,FOOD AND DRUGS ANALYSIS 505 i Cat0 oil, I if the wine contained citric acid.The precipitate may be identified as the acetone- dicarboxylic acid salt by collecting it on a filter, washing it with cold water, dissolving it in 3 C.C. of 10 per cent. sodium chloride solution, and adding a few drops of very dilute ferric chloride solution. A raspberry-red coloration, due to the formation of the corresponding iron salt, is obtained.w. P. s. Pili Oil. Long Nuts. I Short Nuts. C ~ l ~ ~ m ~ a n g - Oil. Increase in the Total Solids of Milk after Separation of the Cream. (Ann. Chim. anal., 1915, 20, 209-212.)-The increase in the total solids G. Tellera. of milk after separation of the cream may be calculated by means of the formula- r‘ 1 0 0 - 9 - loo’ where r represents the dry total solids-not-fat of the original milk, g the fat removed with the cream, and r’ the dry fat-free solids of the soparated milk. As a rule, the actual result agrees with the calculated amount within 0.25 per cent.in the case of samples of milk taken from the farm, but with commercial samples the increase varies in an irregular manner.If a commercial separated milk gives a dry residue decidedly in excess of the maximum found by Formenti (10.17 per cent.), it must be regarded as of abnormal composition. C. A. N. Sp. gr. a t 15O-C. ... ... ... ,, 30’ C. ... ... ... Butyro-refractometer reading at Iodine value (Hanvus) ... ... Reichert-Neissl value . . . ... Acid value ... ... ... ... Saponification value . .. ... ... ... 30’ C. ... ... I 0.9203 0.9188 60-61 80.78 7 *34 7-06 192.02 63-64 76.14 2.10 0.30 212.01 1 0-9067 54-54.2 61-25 3.3 2.70 192.6 - 0.9067 54-54.2 59.61 2.2 3.13 186.8 The oils of Aleurites moluccana (“lumbang bat0 oil ”) and A. trisperma (“ lumbang brtnuoalag oil ”) have excellent drying properties. They may be distinguished from506 ABSTRACTS OF CHEMICAL PAPERS each other by the fact that the former is soluble in four parts and the letter in its own volume of absolute alcohol. C.A. M. Detection of Phytosterol in Animal Fats by Bomsr's Method, using Digitonin to precipitate the Phytosterol and Cholesterol. B. Kuhn, F. Bengen, and J. Werwerinke. (Zeitsch. Umters. Nahr. Genussm., 1915,29, 331-329 ; through Chem. Zentr., 1915, ii., 50-51).-The separation of the phytosterol and cholesterol from the fat by precipitation with digitonin (cf.Windhaus, ANALYST, 1910,35, 256 ; Klostermann, ibid., 1914, 39, 32) may be carried out rapidly by adopt- ing the following procedure : The mixed fatty acids obtained from 50 grms. of the fat are melted and treated with digitonin solution ; after standing for one hour at 70" C., the mixture is diluted with 20 C.C.of chloroform, the precipitate collected on a filter, washed with chloroform, then with ether, and dried for ten minutes at 100" C. Ths precipitate is then separated from the filter, heated with acetic anhydride, the hot solution mixed with four times its volume of 50 per cent. alcohol, cooled, the pre- cipitate collected, washed with 50 p3r cent. alcohol, and then recrystallised from alcohol in the usual way. In the case of animal fats, almost pure cholesteryl acetate is obtained after the first crystallisation ; should the fat contain vegetable oils, the presence of these is generally indicated at this stage.w. P. s. Determination of Volatile Esters in Citrus Oils and Extracts. A. R. Albright and C. 0. Young. (J. Anter. Chem.SOC., 1915, 37, 2382-2387).--In essential oils containing aldehydes the condensation of the aldehydes with semi- carbazide is recommended as a general procedure preliminary to the determination of the saponification value by alkali. The semicarbazones are not hydrolysed during the saponification of the esters, and the resinification of the aldehydes is prevented. The procedure recommended in the case of lemon oils is as follows : The terpene fraction is first removed from 100 grms.of the oil by distillation under diminished pressure (2-5 MM.), in a three-bulb Ladenburg flask, at a rate of 18 to 20 drops per minute, and distillation is continued until tho limonene ceases to come over. The residual oil is then distilled with a current of steam until the volume of distillate amounts to 200 C.C.This operation is regulated so as to last for at least thirty to forty-five minutes, and the volume of liquid in the flask is kept as nearly as possible constant by heating with a flame. The proportion of aldehydes having been determined in a previous operation, the calculated quantity of semicarbazide hydrochloride in the form of a concentrated solution is added with an equivalent amount of crystalline sodium acetate.When an insufficient amount is used the end-point is not sharp. One hundred C.C. of 95 per cent. alcohol are added, the mixture is shaken and allowed to stand for ten to fifteen minutes. Citral semicarbazone may separate. The solution is then made neutral to phenolphthalein, 50 C.C. of $ alcoholic potassium hydroxide are added, and the mixture is boiled under a reflux condenser for two hours. I t is cooled rapidly and the excess of alkali titrated with $ hydrochloric acid.A large amount of phenolphthalein must be used. I n the case OE lemon extracts, 400 grms. of the sample are distilled from an ordinary flask until the volume is reduced to 50-75 C.C.BACTERIOLOGICAL, PHYSIOLOGICAL, ETC.507 Steam is then passed through until no more volatile oil comes over. The combined distillates are treated in the same manner as the steam-distillate from lemon oils, the amount of semicarbszide hydrochloride being calculated st the rate of 0.75 grm. per grm. of citral. The distillation with steam before saponification is necessary because lemon oils contain a non-volatile residue of resinous matter, which itself possesses a, considerable saponification value, and which cannot be credited &S linalyl or geranyl ester. Experiments made on the sapmification of pure linalyl acetate, under the conditions described above, showed that the saponification equiva- lent only corresponded to 78.4 per cent. of the actual quantiby, whereas saponification under optimum conditions should come to an equilibrium at 85.5 per cent. There is therefore a loss of about 7 per cent. in carrying through the operations. The per- centage of ester calculated from the saponification value should be multiplied by the empirical factor 1.28. J. F. B.