36,4 THE ANALYST. ORGANIC ANALYSIS. Titration of Amino-Acids in Presence of Formaldehyde ; Decolorisation of the Solutions. S. P. L. Ssrensen and H. Jessen-Hansen. (Conzptes Rend. Tmv. Lab. de Carlsberg, 1908, 7, 58-71.)-The titration of rtmino-acids in presence of formaldehyde by the method previously described (,~NALYST, 1908, 20) becomes praotically impossible when the solution to be tested is strongly coloured, as is generally the case with the products of the hydrolysis of proteins by acids. The authors have found, however, that very dark-coloured solutions may be sufliciently decolorised by the precipitation therein of silver chloride, by the addition of silver nitrate, and slightly more than the equivalent quantity of barium chloride. The coloured bodies are carried down by the precipitate without appreciable precipita- tion of nitrogen, and the titration may be carried out on the filtered solution.The procedure is as follows : If the liquid to be titrated be of suitable acidity (about $*), 25 C.C. are placed in a marked 50-C.C. flask; if the acidity requires correction, only 20 C.C. of the liquid are taken and treated with 5 C.C. of $ hydrochloric acid or sodium hydroxide, as the case may be. In the case of a solid substance, 1 to 3 grams are dissolved in the flask in 25 C.C. of Fa hydrochloric acid. There are then added to the liquid in the flask about 4 C.C. of a 2N solution of barium chloride, after which about 20 C.C. of a solution of silver nitrate are added, little by little,THE ANALYST 365 with frequent agitation. Finally, the contents of the flask are diluted to 50 C.C.with water free from carbonic acid; 4 drops of water may be allowed for the valume of the precipitate. The liquid is poured, together with as much of the precipitate as possible, on to a dry filter, and the filtrate is returned until it runs through clear. A suitable quantity (15 to 30 c.c.) is then taken for the titration in presence of formaldehyde, as previously directed, phenolphthalein being used as the indicator, and the blank being tinted with orange and brown dyestuffs, to match the shade of the solution under titration. If it be desired to ascertain the quantity of nitrogen present in the precipitate (generally negligible), the latter is washed on the filter with a dilute solution of barium chloride (about ;), and subjected to the Kjeldahl process.When large quantities of acid liquid have to be decolorised, silver nitrate is scarcely suitable, on account of the large quantity of nitric acid (or njtrates) left in the liquid., In such cases silver sulphate may be used, accompanied by the addition of an equivalent quantity of baryta, water, to precipitate the sulphuric acid, if necessary. Silver sulphate, however, must be employed in the form of a solution, since the solid salt causes the precipitation of nitrogenous matters in considerable quantities. J. F. B. A New Method of Estimating Casein. Determination of its Equivalent Weight. G. T. Matthaiopoulos. (Zed. anal. Chem., 1908, 47, 492-501.)-The method is based upon the fact that, if a definite quantity of very dilute acid be added to milk, and the liquid then titrated with standard sodium hydroxide solution, with phenolphthalein as indicator, the whole of the substances acid to the indicator, including the casein, are neutralised.But if the precipitated casein be filtered off before the titration, it smaller quantity of alkali will be required, the difference corresponding to the amount of casein. Various values have been given as the equivalent weight of casein, ranging from 5000 to 6000 [Salkowski (Zeits. fiir Biolog., 1899, 37, 401), Hammarsten (Konig. Gesell. der Wissenschaft zu Upsala, 1877)] to 1135 [Laqueur and Sackur (Hofmeister’s Beitrage, 1903, 3, 193)]. In the author’s experiments, the amount of sodium hydroxide solution, corresponding to the casein in a definite quantity of milk, was found as described below, and the quantity of the casein in the same milk subsequently estimated by Hoppe-Seyler’s gravimetric method.From the mean results of many estimations, it was found that 1 C.C. of ;k sodium hydroxide solution corresponded to 0.11315 gram of casein, and the equivalent weight of casein in its sa1t.s with bases was therefore taken as 1131-5. In the author’s opinion, the higher values of 5000 to 6000 are incorrect, and were possibly obtained with hydrolysed or acid salts of casein. In making estimations by this method, 20 C.C. of milk are diluted with 80 C.C. of water in a 200-C.C. flask, and & sulphuric acid added drop by drop from a burette with continual shaking, until the casein separates in large flakes.From 23.1 to 23.5 C.C. of the acid are usually aufficient to enable a clear filtrate to be obtained, though any excess must be avoided. One hundred C.C. of the clear filtrate are then titrated with TG sodium hydroxide solution, with 1 C.C. of a 1 per cent. alcoholic solution of phenol phthalein as indicator, and the number of C.C. required calculated for the total quantity of water, milk, and added acid. A second portion of 20 C.C. of the milk is treated in the same366 THE ANALYST. way as the first portion, except that it is titrated without previously filtering off the casein, and the difference between the number of C.C. of alkali consumed in the two titrations, multiplied by the factor 0.11315, gives the amount of casein in the 20 C.C.of milk. Experimental estimations are described to show that the method gives very accurate results; and it is suggested that, since the quantity of casein in milk varies within smaller limits (1.91 to 4.65 per cent.) than that of the fat, it might be advisable to base conclusions as to the purity of a milk on the proportion of the former rather than upon the amount of fat. Moreover, in the case of milk which has become sour, it is, as a rule, neceesary to filter off the casein before the free acid can be titrated, although, if the original acidity of the milk be known, that which has been formed may be estimated by the method of Soxhlet and Henkel, since the casein is precipitated by the equivalent amount of acid which has been produced. C. A. &I. Quantitative Estimation of Active Hydrogen in Organic Compounds. T.Zerewitinoff. (Ber. deut. Chenz. Ges., 1908, 41, 2233-2243.)-1n a previous paper (ANALYST, 1907, 32, 299) the author described 8r method for the quantitative estiniation of hydroxyl groups in organic compounds, based on the measurement of the volume of methane given off when the substance is treated with magnesium- methyl iodide in presence of a suitable solvent (amylic ether or pyridine). This method has been employed by the author for the solution of a number of constitutional questions in connection with complex substances, chiefly colouring matters of the flavone group. A few notes are added dealing with details of manipulation necessary to ensure good results. The method tends to give numbers which are higher than the theoretical numbers, chiefly owing to the dificulty, previously mentioned, caused by the secondary action of the pyridine on the magnesium-methyl iodide.This secondary action tends to be increased by the heat, which is developed in the reaction with the hydroxyl compound, The improved procedure is as follows : A s soon ae the solution of magnesium-methyl iodide is brought into contact with the pyridine solution of the hydroxylated substance, the contents of the reaction flask must be shaken vigorously so long as the mercury in the measuring-tube falls rapidly, but when it begins to fall more slowly, the reaction flask must be at once cooled in a vessel of water in order to restrict the secondary reaction with the pyridine. The stock of pyridine employed must be kept in closely-stoppered vessels over lumps of barium oxide, but the greatest care must be taken to prevent particles of the barium hydroxide from entering the reaction flask.For this reason the pyridine is best stored in tall, stoppered cylinders. The method described for the estimation of hydroxyl groups may be employed for tbat of other active groups which yield methane with magnesium-methyl iodide--for instance, the SH group of mercaptans, for which either amylic ether or pyridine may be used as the solvent. The imines and acid imides evolve only one molecule of methane for each imino group, but the amines and acid amides may evolve two molecules, according to the tempera- ture. In the cold, only one hydrogen atom of the amino group reacts, and the restriction to this stage may be made quantitative by preventing the riss of tempera- Imino and amino groups also react quantitatively.THE ANALYST.367 ture which tends to take place, by keeping the flask immersed in a vessel of cold water and shaking well during the reaction. I n order to obtain a full reaction of both hydrogen atoms of the amino group heat must be employed, and the results must be corrected by performing a blank experiment under identical conditions in the following manner : 5 C.C. of the magnesium-methyl iodide solution and 15 C.C. of pyridine are heated in the reaction flask in a water-bath at a temperature of 85" C. for exactly five minutes. During this time the flask is shaken two or three times, and then rapidly cooled to the original temperature before reading the volume.The author has observed the following peculiarity with this class of compound: Compounds which contain two amino groups, or one amino and one imino group attached to a single carbon atom, always show a deficiency of one active hydrogen atom, which escapes reaction with the magnesium-methyl iodide, Thus, urea gives three molecules of methane instead of four, and phenylurea gives two instead of three. The enolic forms of tautomeric ketones behave as mono-hydroxylated bodies, but only yield the full theoretical results when the reaction is carried out at 100" C. The enolic forms of diketones only show the presence of one hydroxyl group. J. F. B. Analysis of Inks [Estimation of Gallo-tannic Acid]. E. Kedesdy. (Papier Zeit., 1908, 33, 2297-2298, 2383.)-The German specification for normal ink states that it should contain at least 30 grams of tannic and gallic acids derived from oak galls and 4 grams of iron per litre ; it fails to state whether the presence of other tannins in excess of the above gallo-tannic acid is permissible. The estimations of the total dry substance and the iron are perfectly simple, but that of the gallo-tannic acid is less straightforward.The most convenient method is that of Finkener : The ink is acidified with hydrochloric acid, and the organic acids are extracted by shaking repeatedly with ethyl acetate, in which they are freely soluble. The extract con- taining the tannic and gallic acids is then evaporated to dryness and weighed. In order to identify these acids as the product of oak galls, certain '' constants " must be determined. One of these is the copper equivalent, but unfortunately gallic acid has a distinctly higher copper equivalent than tannic acid, and both these acids may be present.If the method be permissible, it is carried out by adding a known exce~s of copper sulphate to the aqueous solution of the tannin in presence of chalk, and titrating the excess of copper with a standard solution of gallo-tannic acid until a filtered drop of the liquid no longer reacts with potassium ferrocyanide. A better method is the iodine method, because both gallic and tannic acids combine with practically the same quantityof iodine. The aqueous solution of the acids is treated in closed vessel with an excess of standard iodine solution in presence of sodium bicarbonate.After standing for some time the excess of iodine is titrated back with thiosulphate. Tannins from other sources than galls, which are soluble in ethyl acetate, possess different iodine values from that of gallo-tannic acid. By comparing the gravimetric result with the iodine value, a very fair idea may be gained as to whether the whole of the tannin has been derived from galls, but the method does not identify mixtures of tannins. The colour of the reaction of the tannin extract with very dilute ferric chloride should be noted, since a green coloration indicates the absence of gallo-tannic acid. The last important constituent of ink is the free368 THE ANALYST. mineral acid, but no satisfactory method for its estimation is available on account of the presence of the iron salt and the organic matters. Some test-paper of standard sensitiveness might possibly be devised.The formation of sediments and crusts is generally due to a deficiency of free acid; the iron contents of the deposits are generally very small, and they probably consist chiefly of polymerised oxidation products of the tannin. The ink, when kept in an open vessel protected from dust, should gradually dry to a homogeneous mass without the preliminary formation of crusts. J. F. I3. On the Fractional Distillation of Commercial Petroleum Spirit. R. Kissling. (Chem. Zed., 1908, 32,695-697.)-The following method of examining petroleum Rpirit by fractional distillation is recommended : The apparatus of Holde and Ubbelohde (Chem. Zed.[Rep.], 1908, 32, 620) is used in conjunction with an Engler's distillation flask, and a thermometer graduated in tenths of a degree. The upper part of the thermometer-bulb should be in the same plane as the lower edge of the outlet-tube of the flask, and the rate of distillation should be regulated so that in the first ten minutes, 2 c.c.; in the next twenty minutes, 3 C.C. ; and in the last ten Ininutes, 2 C.C. distil over, the distillation being complete in forty minutes. The temperature at which the first drops fall from the condenser is taken as the lower boiling-point, and the volumes of distillate are measured every 10" C., beginning with the first completed 10" C. The temperature at which the bottom of the flask is free from liquid, or at which white vapours appear, is taken as the upper boiling-point, and the weight of the residue in the distillation flask is subsequently determined.In case of a deviation of +_ 5 mm. from the normal barometrical pressure, a correction must be applied for the influence of atmospheric pressure. This is found by adding to or subtracting from the temperature of each of the fractions the difference between 100" C. and the temperature of boiling water under the pressure in question. In testing a petroleum spirit for small quantities of constituents of high boiling-point, it is best to stop the distillation when 95 per cent. by volume has passed over, to transfer the residue to a clock-glass, and to allow it to evaporate at the ordinary temperature, which takes from five to ten hours.C. A. 1\12. . Detection and Identiflcation of the Divalent Phenols. A Delicate Reaction for Resorcinol. T. Silbermann and N. Ozorovitz. (Bull. SOC. Sci. de Bucharest, 1908, 17, 41-42.)-When the divalent phenols are heated with formaldehyde in acid solution, characteristic resinous condensation products are formed, which are insoluble in all ordinary solvents, but which undergo changes when further treated under certain conditions. The reaction with resorcinoE is extremely delicate, and is distinct with a solution containing only 1 part of the phenol per 100,000. When a solution of resorcinol is heated with formaldehyde, with the addition of a, few drops of hydrochloric or sulphuric acid, a white flocculent pre- cipitate is formed.When the product of the reaction is mixed with strong sulphuric acid or heated with strong hydrochloric acid, the mass is coloured a brilliant carmine red, which changes to orange on dilution or neutralisation. Both the carmine and the orange products are changed to a Bordeaux red colour by alkalis. If the con-THE ANALYST. 369 densation be originally effected in presence of concentrated acid, a, red product is formed directly. Pyrocatechol, when heated with formaldehyde in acid solution, yields a white resinous precipitate, which is coloured a deep brown-violet by con- centrated acids, the colour changing to fleeh-colour on dilution or neutralisation ; an excess of alkali turns it greenish-brown. In the case of qzbinol it is neoessary to use more acid to effect the condensation (10 C.C. of strong hydrochloric acid per 100 c.c.). A similar white resin, insoluble in all media, is precipitated. Concentrated acids turn it yellowish-brown, and the colour becomes paler on dilution or neutralisation ; excess of alkali turns it olive-green. The reactions with pyrocatechol and quinol are not SO delicate as in the case of resorcinol. J. F. B. A Test to Distinguish a- and B-Naphthol. Volcy-Boucher. (Repertoire 1908, 20, 289 ; Pharm. Journ., 1908,81,82.)-A solution of 0-5 gram of the naphthol in the smallest possible amount of alcohol is treated with 2 C.C. of a 10 per cent. solution of copper sulphate end 4 C.C. of a freshly-propared 10 per cent. solution of potassium cyanide, the liquid being shaken after eaoh addition. I n the case of a-naphthol a, violet-red precipitate is obtained, whilst P-naphthol gives a yellow precipitate. On treatment with just sufficient 95 per cent. alcohol, the red precipitate gives a solution which is rose-coloured by transmitted light and violet by reflected light, whilst the @naphthol precipitate forms an orange-yellow solution. The test is capable of detecting 1 part of a-naphthol in the presence of 2,000 parts of /3-naphthol, of distinguishing between the camphorated naphthols, and of detecting either naphthol in the free state in benzo-naphthol, which gives no coloration in the test. C. A. M.