50 THE ANALYST. ORGANIC ANALYSIS. Studies on Formaldehyde in Aqueous Solution. F. Auerbaeh. (Arbeit. Kaiser2. Gesundheitsamte, 1905, vol. 22, pp. 1-46.)-The formaldehyde solutions employed in the experiments described were prepared by subliming tri-oxymethylene in an atmosphere of nitrogen, and collecting the vapours in water. For the deter- mination of formaldehyde the sulphite method of Lumihre and Seyewetz (ANALYST, 1904, xxix., 288) was found to be very useful, but for small quantities the iodine method was preferred. The results of determinations of the molecular weights of solutions containing varying amounts of formaldehyde lead the author to conclude that formaldehyde exists in aqueous solutions for the greater part in a hydrated or polymerized form. As the temperature is increased the simple molecule predominates.On distilling formaldehyde solutions of any strength whatsoever the distillate is always poorer and the residue richer in formaldehyde than the original solution. This fact should be taken into consideration in the detection and determination of formalde- hyde. I t is further shown that the boiling-point at normal pressure of formaldehyde solutions falls from 100" to 99" C. as the concentration increases. w. P. s. Distinction between Phenol and Cresols. C. Arnold and G. Werner. (Apoth. Zeit., xx., 925 ; through Phamz. Jouurm., 1905, vol. 75, p. 837.)-The following reactions are suitable for the identification of phenol and cresols : Phenol.-To 10 C.C. of the solution to be tested are added 10 C.C. each of potassium hydroxide solution and alcohol and 1 drop of aniline.After shaking the mixture, 5 drops of hydrogen peroxide and 10 drops of sodium hypochlorite are addedTHE ANALYST. 51 and the whole again shaken, Phenol gives a transient dirty red coloration, changing to yellow ; o-, m-, and tricresol, a violet, changing at once to green; p. cresol, a violet, which at once disappears. o-CresoL-With ferric chloride o. cresol gives a blue colour, rapidly changing to green ; phenol, m. cresol, and tricresol, a violet ; p. cresol, a blue colour. m-CresoL-On warming with a little phthalic acid and 5 drops of sulphuric acid m-cresol gives a cherry-red, phenol and tricresol a dark red, 0. cresol a cherry-red, p-cresol an orange colour. When diluted with water and rendered alkaline with sodium hydroxide phenol gives a magenta, o-cresol and tricresol a violet-red, m-cresol a bluish violet, and cresol a yellowish colour.CresoL-A dilute ammoniacal solution of cresol when boiled and treated with bromine water gives no coloration. Phenol and o-cresol treated in this way give a blue colour ; m-cresol and tricresol a bluish-green. If a trace of potassium nitrate be added to a solution of a little of the substance in sulphuric acid, a dark-red colour is obtained in the case of cresol, all the others giving an emerald-green coloration. When diluted with water and treated with an excess of ammonia, cresol gives a yellow colour, all the others a green. W. p. s. On the Inflammability of Celluloid Articles (according to experiments made by Fr.Gervais, director of the Laboratory of the Ministry of Finance, St. Peters- burg.) J. Bronn. (Zeeits. angezu. Chenz., 1905, xviii., 1976.)-It is shown that when heated to 100” C. (for instance, by contact with a steam-heater) articles made of celluloid decompose exothermically. Although the quantity of heat developed is not sufficient to set the celluloid itself on fire, it will cause the paper used for packing to smoulder, a slight draught of air being then sufficient to cause the whole mass to ignite. The temperature of ignition of celluloid appears to be fairly high, ranging from 355’ to 457’ C. in the case of four articles examined, that of pyroxylin being only 130’ C. White celluloid (imitation ivory) is more difficult to ignite than the other kinds, Celluloid articles may now only be sent to Russia when packed in wooden or metal cases.A. G. L. The Detection of “Blown” Fatty Oils in Mixtures with Mineral Oil. J. Marcusson. (Chem. Rev. Fett- ZL. Hnrx. Ind., 1905, xii., 290-293.)--“ Blown ” rape and cotton-seed oils are stated to be the two preparations of this kind most used in lubricating mixtures. Their presence cannot be detected by determination of the iodine value or molecular equivalent of the fatty acids. ‘‘ Blown ” cotton-seed oil does not give Halphen’s or Milliau’s reactions, though it gives the brown coloration with nitric acid. The latter reaction, however, is also given by “ blown ” rape oil. The author relies upon the following tests to distinguish between the two (( blown ” oils: (1) The odour of the mixture and of the fatty acids, recalling that of the unblown oils.(2) The consistency of the fatty acids-oily in the case of ‘‘ blown ” rape oil, semi-solid in that of ‘‘ blown ” cotton-seed oil. (3) The behaviour of the respective lead salts towards ether. Thus the amounts of fatty acids from the lead salts insoluble in ether ranged from 1.2 to 20.6 per cent. in the case of five ‘(blown ” rape oils, and 32.9 to 45.8 in the case of tws cotton-seed oils. The difference was52 THE ANALYST. more marked when the solubility of these Eatty acids in petroleum spirit was com- pared, the rape oils yielding 0 to 8-7 per cent. and the cotton-seed oils 23.3 to 32.5 per cent. The fatty acids from the insoluble salts of the (‘ blown ” cotton-seed oil’ melted at 5 4 O to 56” C., whilst those of the (( blown ” rape oil were oily or semi- solid.The following table gives results thus obtained, the amounts being calculated on the quantity of oil used in the preparation of the lead salts : Kind of Oil. Commercial (‘ blown ” rape oil ... Ditto ... ... ... ... Rape oil (( blown ” in laboratory.. . Commercial (( blown ” cotton-seed oil .. ... ... .. ... Cotton-seed oil blowu ” in labora2- tory ... ... ... ... On the Determination of Fatty Acids separated from Lead Salts insoluble in Ether. Total Fatty Acids. Per Cent. 1.2 14.5 20.6 32.9 45.8 Fatty Acids soluble in Petroleum Spirit. Per Cent. 1.2 5.7 8.7 23.3 32.5 Fatt Acids insoYuble in Petroleum Spirit. Per Cent. 0.0 8.8 11.9 9.6 13.3 C. A. M. Sulphur in Liquid Fuels (Petroleum, Oil, etc.).J. Matwin. (Zeits. angew. Chem., 1905, xviii., 1766.)-The author recom- mends a modification of Drehschmidt’s method in place of that of Goetzl (ANALYST, 1905, xxx., 376) for the determination of sulphur in liquid fuels, as much larger quantities (50 to 100 grams) may be used : The fuel is burnt in a small lamp pro- vided with a, wick, the products of combustion being led through a 5 per cent. solution of potassium hydroxide contained in three wash-bottles. The difference in weight of the lamp before and after the experiment gives the quantity of fuel burnt. No test analyses are given. A. G. L. Determination of Sulphur in Petroleum and Bituminous Minerals. F. C. Garrett and E. L. Lomax. (Journ. SOC. Chem. Ind., 1905, xxiv., 1212.)- A convenient quantity (0.7 to 1.5 gram) of the substance is intimately mixed in A small crucible with 3 or 4 grams of a mixture of 4 parts of pure lime with 1 of anhydrous sodium carbonate, and the crucible completely filled with this mixture.A larger platinum crucible is placed over the small one (mouth downward), the whole inverted, and the space between the two crucibles filled with the lime-soda mixture. The apparatus is then placed in a muffle furnace heated to bright redness, the mouth of the crucible being covered with a thick pad of asbestos board, which prevents distillation from the inner crucible before the mixture in the outer crucible is heated, due to radiation from the roof of the muffle, and the pad may be removed as soon as a flame appears. The roasting should be continued for two hours, after which the mixture is brought into water, any sulphide oxidized by bromine, andTHE ANALYST.53 the solution acidified, filtered, and precipitated by barium chloride in the usual way, the solution being allowed to stand on the water-bath for twenty-four hours before filtration, if the amount of sulphur is small. (Cf. ANALYST, 1905, xxx., 418.) W. H. S. Estimation of Tannin as Strychnine Tannate. S. R. Trotman and J. E. Hackford. (Joum. $06. Chem. Ind., 1905, xxiv., 1096.)-The present absorp- tion methods of determination are very unsatisfactory, even when the hide powder is replaced by Collin, and the authors have investigated a large number of bodies with the object of obtaining a compound of tannin which could be readily precipi- tated and weighed.Metallic salts proved useless, precipitation being incomplete, and after experimenting with many organic bases, including phenylhydrazine and various primary and secondary amines, the alkaloids were tried, and strychnine, one molecule of which combines with one molecule of tannin, finally adopted, as it does not precipitate gallic acid, and its tannate is highly insoluble in water. Sufficient material is4extracted with alcohol in a Soxhlet extractor to yield about 0.5 gram tannin, the solution evaporated to 50 c.c , transferred to a 100 C.C. flask, and made up to the mark with water, thus precipitating any resins and similar bodies. These are filtered off through a dry Gooch crucible, and the tannin estimated in the filtrate, 25 C.C. of which are placed in a 250 C.C.flask, diluted with water, the cooled strych- nine solution (prepared by dissolving 0.25 gram strychnine in 50 C.C. alcohol, and mixing with an equal volume of water) added, and the solution made up to 250 C.C. This method of treatment is essential, as otherwise the precipitate is difficult to filter. The contents of the flask are now filtered through a weighed Gooch crucible of platinum or porcelain, having a diameter a t the bottom of 1 to 2 inches, with a thin mat of asbestos over the bottom. After filtration the tannate is partly air- dried, and the dehydration completed in a vacuum oven heated to about 60". The soluble non-tannins are determined by evaporating 25 or 50 C.C. of the filtrate from the resinous matter. The amount of tannin found by the method is invariably less than that obtained by the use of hide powder or Collin.W. H. S. The Detection of Biliary Pigments in Urine. L. Grimbert. (Jouyfi. P h a ~ n . Chim., 1905, xxii., 487-492,)-The following method combines the principles of the methods of Hammarsten and Salkowski, but it is claimed that it is more simple than either: Ten C.C. of the urine are shaken with 5 C.C. of a 10 per cent. solution of barium chloride, and the mixture subjected to centrifugal force. The precipitate, consisting of the sulphate, phosphate, and bilirubinate of barium, is mixed with 4 C.C. of 90 per cent. alcohol containing 5 per cent. of hydrochloric acid, and heated for about a minute on the water-bath, after which the tube is allowed to stand. If the supernatant liquid over the deposit is bluish green or dark green the presence of biliary pigments is indicated, whilst in their absence the liquid is colourless.If there is a slight brownish tint the hydrochloric acid may have been insufficient to oxidize the barium bilirubinate, and in that case, and that case only, 2 drops of hydrogen peroxide (10 vol. solution) are added and the tube again heated on the water-bath, after which the green colour will appear, If the brown tint continues54 THE ANALYST. after the addition of the hydrogen peroxide the presence of certain altered products of biliary pigments is indicated. These are only met with in urines that have been allowed to stand for some time. In certain pathological urines the separation of the barium bilirubinate is promoted by the addition of a few drops of a 10 per cent.solution of sodium sulphate. C. A. M. Estimation of Naphthalene in Coal Gas. C. J. D. Gair. (Jourfi. SOL Chem Ind., 1905, xxiv., 1279.)-Three methods are described, the first a slight modification of Colman and Smith's, in which a measured volume of gas is passed through three Woulff's bottles, each containing exactly 500 C.C. of $G picric acid, the contents of the bottles mixed together in a 20-ounce flask, and heated on the water- bath at 60' C. until the precipitated naphthalene and naphthalene picrate have entirely dissolved, the flask being closed with a cork through which a glass tube dips into a beaker containing a small quantity of picric acid from the Wodff's bottles. This is afterwards returned to the flask and its contents cooled, when naphthalene picrate crystallizes out, is dried in vacuo or in a warm room.and weighed, whence the naphthalene may be calculated; or an aliquot part of the clear residual picric acid may be titrated with Tn sodium hydroxide, using lacmoid as indicator, showing the amount of picric acid taken up by the napthalene. 7 % ~ method, though devised for pure coal gas, is applicable to impure gas if the ammonia be first removed by passing the gas through oxalic acid before entering the picric acid solution, The second method, due to Somerville, consists in passing a known quantity of the gas through three glass tubes, 7 inches long and 1 inch wide, fitted UP 8s Woulffs bottles, and each containing 35 C.C. of 70 per cent. alcohol. The alcoholic solutions are then mixed in a flask, oxalic acid added if necessary, until the ammonia is neutralized, as shown by litmus-paper, and filtered, the filter-paper being thoroughly washed with dilute alcohol. About 500 C.C. of concentrated picric acid are then added, the liquid agitated, and allowed to stand for half an hour, when all the naphthalene picrate will have separated, and may be filtered, slowly dried, and weighed. A third method, devised by the author, depends on the solubility of naphthalene in acetic acid. A known volume of gas, usually not exceeding 3 to 6 cubic feet, is passed through 350 C.C. of acetic acid (specific gravity 1.044), contained in two Woulff's bottles, a small Woulff's bottle with 150 C.C. of picric acid being placed after the acetic acid to catch any unabsorbed naphthalene. The contents of all three bottles are mixed in a flask, and 500 C.C. of concentrated picric acid solution added, when the naphthalene picrate separates out at once in large flocculent masses, which are easily filtered off, dried, and weighed. W. H. S.