FOOD AND DRUGS ANALYSIS 345 ABSTRACTS OF PAPERS PUBLISHED IN OTHER JOURNALS. FOOD AND DRUGS ANALYSIS. Estimation of Caffeine and Antipgrin in Admixture. W. P. Emery and S. Palkin. ( J . Ind. and Eng. Chem., 1915, 7, 519-521.)-The method proposed depends on the conversion of the antipyrin into iodo-antipyrin (cj. ANALYST, 1914, 39, 480) ; the latter substance, together with the caffeine, is then extracted with chloroform, the extract evaporated, and the residue weighed. Iodine is now estimated in this residue, and the quantity found is a measure of the antipyrin; the caEeine is obtained by difference.A quantity of about 0-25 grm, of the caffeine-antipyrin mixture is rinsed into a separating funnel with two 5 C.C. portions of alcohol-free chloroform (it is essential that the chloroform be free from alcohol, otherwise iodoform is formed which interferes with the estimation), 10 C.C.of water and 1 grm. of sodium bicarbonate are added, and 15 C.C. of iodine solution are introduced in small quantities at a time, while the mixture is shaken continuously; the presence of decided excess of iodine in the mixture should be apparent at the end of the operation. The excess of iodine is destroyed by the addition of a crystal of sodium thio- sulphate, the mixture then extracted with several successive quantities of chloroform ; the chloroform extracts are evaporated in a weighed flask, the residue is dried for thirty minutes at 105' C., and weighed.This residue, consisting of iodo-antipyrin a d caffeine, is dissolved in 5 C.C.of glacial acetic acid, 10 C.C. of concentrated sulphurous acid are added, the solution is diluted to about 200 c.c., and the iodide precipitated by the addition of silver nitrate. After adding a few drops of nitric acid, the mixture is boiled, and the silver iodide collected, and weighed. The weight of the silver iodide multiplied by 0.8012 gives the quantity of antipyrin. The quantity of caffeine is found by subtracting the product obtained by multiplying the weight of silver iodide by 1.3374 from the weight of the caffeine plus the iodo-antipyrin.When the caffeine and antipyrin are mixed with the usual excipients of tablet and pill com- binations, the material should be extracted with chloroform, the chloroform solution evaporated, and the residue obtained used for the estimation, as described.The method yields from 99.2 to 100-5 per cent. of antipyrin actual present in a mixture. w. P. s. Studies on Chicory. V. Grafe. (Biochem. Zeitsch., 1915, 68, 1-22; through Monthly Bull. of Agric, Intelligence and Plant Diseases, 1915, 6, 584-585.)-The chief food value of chicory is due to inulin, and its use as a drug to a bitter principle which it contains.The rootscontain more inulin and bitter principle when grown in sandy or loamy soils than when grown in humus or peaty soils ; at the same time, this differ- ence may also be influenced by general conditions of growth, climate, and weather. Attempts to isolate the bitter principle from the unroasted roots did not yield a chemically pure product owing to the fact that the substmce decomposes very readily. It was found, however, that the substance is a glucoside, the constituent346 ABSTRACTS OF CHEMICAL PAPERS sugar being laevulose, and the non-sugar component a derivative of catechol (pyro- catechin), probably the aldehyde.According to several authorities, the roasting of chicory results in an increase in the amount of reducing sugar present at tbe expense of the inulin; this is not quite correct, as products similar to dextrin are formed instead, and also products of the decomposition of inulin similar to (‘ assamar,” The empyreumatic oil which is formed during roasting is analogous to the caffeol resulting from the roasting of coffee, but of essentially different composition ; the writer proposes the name of chicoreol.Its chief constituent is acetic acid (63.5 per cent.); it also contains valerianic acid (5.4 per cent,), and acrolein (2.5 per cent.), the remainder being furfuraldehyde and furfuralcohol, Roasted chicory yields from 0.08 to 0.1 per cent. of this oil. w. P. s. Tests for Hashish. W. Beam. (TVelZcome Trop. Res. Lab., 1915; Bull. No. 3, Chem. Section.)-In a former publication (Fourth Rep.Wellcome Trop. Res. Lab., 1911, p. 25) the author described a test for hashish which he then believed was oharacteristic for samples of Cannabis indka, whatever their origin. I t consisted of extracting the materia€ with petroleum ether and evaporating the extract so obtained to dryness. To the residue is added a few drops of approximately & alcoholic potash or soda, when, in the presence of hashish, a rich purple or reddish-purple colour gradually develops, which on dilution with water takes on a bluish shade.Petroleum ether is, on the whole, the best solvent to use, but all the ordinary resin solvents are satisfactory; whilst if the material, as is often the case, is sold dissolved in fat or oil, alcohol is the best solvent to employ. I t has since been observed that the ordinary British Pharmacopoeia extract of Cannabis indica reacts but feebly, if at all, with this test, and samples from Ceylon and some from the Sudan and Southern Uganda also failed to give the coloration.I t is thus evident that the influence of soil, climate, and method of cultivation, etc., have more influence on the chemical com- position of the plant than has been suspected.As the cultivation or importation into Egypt of hashish in any form is prohibited, a test for the varieties that do not react in the above described manner becomes desirable, Such a test has yet to be supplied, but the following presumptive test has found to answer with all hashish and cannabis preparations so far examined from India, Sudan, Egypt, Greece, and Uganda.The petroleum ether extract is made as usual, and is evaporated in a short test-tube. To the residue are added a few C.C. of a reagent made by saturating absolute alcohol with hydrochloric acid gas. I n the presence of cannabis extract the liquid assumes a, bright cherry red colour, which disappears on dilution with alcohol or water. Trials made with a number of plant extracts and over 200 alkaloids, glucosides, etc., failed to give a similar reaction.Certain volatile oils-e.g., origanum and santal-give a similar reaction, but the colour is far less intense for similar amounts of material. H. F. E. H. Estimation of Small Quantities of Heroine. R. Miller. (Amer. J. Pharm., 1915, 87,248-250.)-Small quantities of heroine in admixture with cocaine and lactose may be estimated as follows, the method depending on the coloration obtained when heroine is mixed with sulphuric acid and formaldehyde : The substance is extractedFOOD AND DRUGS ANALYSIS 347 in the usual way with an immiscible solvent, the residue of heroine and cocaine is weighed, and then dissolved in such a quantity of 1 per cent.sulphuric acid that each C.C.of the resulting solution shall contain from i& to One C.C. of this solution is placed in a tube and treated with 3 C.C. of a mixture consisting of 600 C.C. of commercial sulphuric acid, 300 C.C. of water, and 25 C.C. of 40 per cent. formaldehyde solution. After twenty minutes, the coloration produced is compared with those obtained with definite amounts of heroine under similar conditions.The coloration varies from straw-yellow for i+a grain of heroine to deep cherry red for + grain ; a difference of GV grain will produce a very perceptible change in colour when Morphine and certain other substances interfere with the reaction. w. P. s. of a, grain of heroine. grain is used. Notes on the Hydrocyanic Acid Content of Sorghum.J. J. Williamson and R. M. West. (J. Agric. Research, 1915, 4, 179-185.)-For the determination of hydrooyanic acid the colorimetric method of Francis and Connell (ANALYST, 1913, 38, 569) was used, with one modification. I t was found that when the macerated tissue was distilled with sulphuric acid according to Francis and Connell's method the distillate became yellow, and when subsequently treated with ferric chloride a greenish or brownish precipitate was formed which masked the colour of the thio- cyanate.Enzyme hydrolysis was therefore resorted to. Slade (J. Amer. Chenz. SOC., 1903,25,55-59) digested the ground tissue for twelve hours at room temperature, making use of the enzyme which is always found in a plant in conjunction with a cyanogenetic glucoside.I t was found, however, that at 40" to 45' C. complete hydrolysis was obtained in two hours or less, as portions of the same sample gave the following results : HCN in 10 grms. of Time of Digestion. Ground Material. ... 2 hours *.. ... 0.00040 grm. 4 ,, ... ... ... 0.00040 ,, 6 J 9 ... ... ... 0*00025 ,, The authors digested in all cases for two hours, and the hydrocyanic acid was distilled and estimated in the usual way.The following conclusions are drawn : When sorghum is grown on poor, infertile soil, added nitrogen may slightly increase the amount of hydrocyanic acid in the plant. With a fertile soil and abundant nitrogen this effect may not be produced. During the first three or four weeks of the plant's life the hydrocyanic acid is concentrated in the stalks.Then it rapidly decreases and disappears there, but apparently persists in the leaves in decreasing quantities until maturity. Climate and variety may be more important factors than soil-nitrogen in determining the amount of the acid in the plant. Complete hydrolysis of the glucoside is obtained by digesting the macerated tissue for two hours at 40" to 45" C. H.F. E. H. Estimation of the Degree of Homogenisation of Milk. Von Sobbe. Milchw. Zentyalbl., 1914, 43, 503-506 ; through Monthly Bull. of Agric. Intelligence and Plant Diseases, 1915,6,622-623.)-Two hundred end fifty C.C. of the homogenised348 ABSTRACTS OF CHEMICAL PAPERS milk are treated with 2 drops of formaldehyde solution, then placed in a, graduated cylinder and allowed to stand for seventy-two hours at the ordinary temperature.At the same time 250 C.C. of the un-homogenised milk are treated similarly. After seventy-two hours the two samples are each divided into three layers-an uppermost layer of 50 c.c., an intermediate layer of 150 c.c., and a bottom layer of 50 C.C. The fat-content of each of these layers is determined separately. The degree of homo- genisation is the fat-content of the lowest 50 C.C.expressed as a percentage of the fat-content of the original milk. w. P. s. Origin of Methyl Alcohol in Brandies. T. von Fellenberg. (Mitt. Schweiz. Ges.-Amt, 1914, 5, 172; through J. SOC. Chem. Id., 1915, 34, 677.)-Brandies prepared from grape-juice fermented in presence of the marc contain more methyl alcohol than those from juice fermented after removal of the marc.I t is concluded that the pectin and protopectin contained in the marc produce methyl alcohol, under the influence of pectase, during fermentation. I n an investigation of a number of cellulosic drugs (Membranindrogen) methoxy-groups were found only in the cellulosins (ligno- and pecto-membranins) containing polysaccharides and in gum tragacanth ; those drugs which contained no polysacchsrides were free from methoxy-groups.Cause of Acidity of Fresh Milk of Cows, and a Method for the Estima- tion of Acidity. L. L. Van Slyke and A. W. Bosworth. (New York Agric. Exp. Stat., Tech. Bull. No. 37, December, 1914, 4-6; through J. SOC. Chem. I~zd., 1915, 34, 678.)-In the estimation of the acidity of whole milk by titration with alkali in presence of phenolphthalein, the calcium should first be precipitated by treating 100 C.C.of the milk with 2 C.C. of a saturated solution of neutral potassium oxalate; otherwise high results are obtained, because dicalcium phosphate present in the milk and formed during titration is hydrolysed to form calcium hydroxide and phos- phoric acid, and the calcium hydroxide reacts with more dicalcium phosphate to form tricalcium phosphate.If whole milk be titrated direct, about twice as much alkali is required as when the serum obtained by filtering the milk through a porous porcelain filter is titrated. This discrepancy is due, not, as is ordinarily assumed, to the acidity of milk casein, but to retention of dicalcium phosphate by the filter; casein is present in fresh milk as a calcium caseinate neutral to phenolphthalein.Methods of Estimation of Lecithin in Milk. N. A. Brodrick-Pittard. (Biochem. Zeitsch., 1914, 67, 382-390 ; through J. Chem. SOC., 1915, 108, ii., 293-294.) -Burow’s method (J. Chem. SOC., 1901, 80, ii., 30) of determining lecithin in milk cocsists in dropping the sample into a mixture of equal parts of alcohol and ether acidified with acetic acid, evaporating the filtrate at a low temperature, extracting with dry ether, and determining the phosphorus in the ethereal extract.The author finds that the process is liable to give inaccurate results if the evaporation residue is not dried before extraction ; drying is conveniently effected with anhydrous sodium sulphate.FOOD AND DRUGS ANALYSIS 349 Estimation of Citral in Concentrated Oil of Lemon.E. Biicker. J. prakt. Chem., 1914, [ii.], 90, 393-404 ; through J. Chem. SOC., 1915, 108, ii., 294.)- Concentrated oil of lemon is adulterated principally by addition of citral derived from lemon-grass oil. The citral-content of these oils may be determined with a considerable degree of accuracy by the sodium sulphite method devised by Burgess (ANALYST, 1901,26,260) ; methods in which sodium bisulphite is used give erroneous results.The proportion of citral present is not sufficient to indicate addition of citral, but, taken in conjunction with the hydrocarbon-content of the oil (compare Bocker, ANALYST, 1914, 37, 185), it will show if the proportion added is large. The highest percentage of citral found in oil of lemon free from terpenes and sesquiter- penes is 66, the content varying inversely with the hydrocarbons present.Con- sequently, if an oil of lemon contains either no hydrocarbons and more than 66 per cent. of citral, or 50 per cent. of hydrocarbons and more than 33 per cent, of citral, addition of the latter must have occurred. Since, however, the genuine hydrocarbon-free oil usually contains less than 66 per cent.of citral, the above method gives no more than the minimal value of the adulteration, and may, unless the latter amounts to about 20 per cent., not detect it at all. Palm-Kernel Oil. A. Heiduschka and A. Burger. (Zeitsch. ofentl. Chem., 1914, 20,361-369 ; through J. SOC. Chem. Ind., 1915, 34,668.)-The following average values are given for palm-kernel oil : Saponification value, 253-4; Reichert-Meissl value, 6.6; Polenske value, 9-4 ; iodine value (Hubl), 15-02 ; and molecular weight of non- volatile fatty acids, 228.2.The non-volatile fatty acids consisted of 17.75 per cent, oleic acid, 23.27 per cent. myristic acid, and 58.98 per cent. lauric acid. The total volatile acids, soluble and insoluble (5.20 per cent.), contained capric, caprylic, and caproic acids.No stearic, palmitic, linolic, or linolenic acid was found. Fachini and Dorta's method of separating insoluble fatty acids by treatment of their potassium salts with acetone is useful as a qualitative test, whilst Heintz's method of fractional precipitation with magnesium acetate gives good results when the quantity of fatty acid is not less than 0.05 grm.Constituents of Horse-Fat. J. Klimont, E. Meisl, and K. Mayer. (Monats. Chem., 1914, 35, 1115-1127; through J. SOC. Chem. Ind., 1915, 34, 668.) -Four commercial samples of horse-fat of about the same consistence were purified, and then gave the following valueB: Iodine value, 74-9 to 78.1; acid value, 1.40 to 2-91; saponification value (three samples), 193.1 to 200.4.The sp. gr. was 0.9373 and 0.9461 at 16O C. for two samples and 0.9148 and 0.9184 at 27" C. respectively for the others; the m.-pts. were 20° to 41° C., 22' to 38' C., 29.5' C., and 33' C. respectively; the fatty acids from two samples melted in both cases at 37' to 39' C. In the liquid portion of the fat oleic, linolic, and linolenic acids were detected.The solid portion, when crystallised repeatedly from acetone to which one-twentieth of its volume of chloroform was added, yielded a glyceride of m,-pt. 60' C. and saponification value 197. The fatty acid separated from this glyceride was identified as heptadecylic acid (margaric acid), m.-pt. 5 7 O to 57fP C., neutralisation value 208. The fatty acid from goose- at, described as an350 ABSTRACTS OF CHEMICAL PAPERS eutectic mixture of stearic and palmitic acids, also actually consists of heptadecylic acid.Modification of Wichmann’s Method for the Detection of Small Amounts of Coumarin, particularly in Factitious Vanilla Extracts. J. R. Dean. (J. I n d . and Eng. Chem., 1915, 7, 519.)-The essential part of Wichmann’s method is the conversion of the coumarin (in the residue obtained by evaporating the ether-extract from a vanilla extract to dryness) into salicylic acid by fusion with potassium hydroxide, and identifying the Salicylic acid by the ferric chloride test.The method, however, fails completely when the extract contains saccharin or salicylic acid, but the following modification eliminates the interfering action of these substances : A de-alcoholised porti0.n of the sample is rendered alkaline with ammonia and extracted with 15 C.C.of ether. Vanillin, salicylic acid, and saccharin are insoluble in ether in the presence of ammonia, whilst coumarin is readily dissolved. The ethereal extract is then evaporated, the residue fused with potassium hydroxide at the lowest possible terriperature, and the mass dissolved in water.This solution is acidified with sulphuric acid, extracted with a few C.C. of chloroform, and the chloroform extract shaken with 2 C.C. of water and a few drops of ferric chloride solution. A violet coloration indicates the presence of coumarin. Attention is drawn to the fact that coumarin would interfere with Durand’s method for the estimation of saccharin (ANALYST, 1914, 39, 86).In using this method, the coumarin should be removed by rendering the substance alkaline and extracting with ether before proceeding with the estimation of the saccharin. w. P. s. Acidity and Ash of Vanilla Extract. A. L. Winton, A. R. Albright, and E. H. Berry. (J. Ind. and Eng. Chem., 1915, 7, 516-519.)-The following ranges in acidity and in ash were found in seventy-seven United States Pharmacopceia extracts prepared from different varieties (Mexican, Bourbon, Seychelles, Mada- gascar, Comores, South American, and Java), grades, and lengths of vanilla beans (all the results are calculated to 100 C.C.of the extract) : Total acidity, 30 to 52 C.C. & alkali solution ; acidity other than vanillin, 14 to 42 C.C. & alkali solution; total ash, 0.220 to 0.432 grm. ; soluble ash, 0.179 to 0-357 grm.; alkalinity of total ash, 30 to 54 C.C. & acid solution ; alkalinity of soluble ash, 22 to 40 C.C. & acid solution, The acidity of vanillin may be determined by direct titration, using phenolphthalein as indicator. Commercial vanillin seems to yield a slightly lower result than that required by theory, 1 grm. requiring 63 C.C. of & alkali solution instead of 65-8 C.C. Whilst the colour present in 10 C.C. of genuine vanilla extract does not interfere with the titration after dilution to 200 c.c., the caramel in imitation extracts may be present in such quantity as to render the end-point indistinct. In the latter case the vanillin may be approximately estimated by titration of the ether solution of vanillin and coumarin obtained by shaking out the lead filtrate, as in the Hess and Prescott method (ANALYST, 1899, 24, 162). The vanillin acidity of a vanilla extract is found by multiplying the vanillin percentage by 65.8, thus obtaining the C.C. of & alkali solution corresponding with the vanillin in 100 C.C. of the extract. w. P. s.