354 ABSTRACTS OF CHEMICAL PAPERS ORGANIC ANALYSIS. Estimation of Methyl Alcohol in Presence of Ethyl Alcohol. W. A. R. Wilks. (Sudan Government, Wellcome Tropical Research Laboratories, Chemical Section, Bull. No. 1, 1914, pp. &)--An improvement in the method of Thorpe and Holmes (J. Chem. Soc., 1904,85,1) is described. The author finds that the correction, necessitated by the fact that the formation of carbon dioxide from pure ethyl alcohol cannot be wholly inhibited, is not EL constant, as the experiments of Thorpe and Holmes led them to believe, but depends on the exact experimental conditions.Thorpe and Holmes gave this correction as 0.01 grm. carbon dioxide per grm. of ethyl alcohol present, neglect of the correction leading to the apparent discovery of 0.7 per cent.of methyl alcohol in pure ethyl alcohol. M7ithout departing from Thorpo and Holmes’ experimental conditions, so far as these authors thought it necessary to define them, pure ethyl alcohol may be made to give rise to 2.4 times as much carbon dioxide as is stated above, with the result that the application of the method of Thorpe and Holmes, including their supposed constant ” correction, may lead to the apparent discovery of 1 per cent.of methyl alcohol where none is, in fact, present. Such results as these are obtained when the materials are rapidly mixed at the ordinary temperature, so that the heat of reaction gives rise to a considerable rise of t emperst ure.ORGANIC ANALYSIS 355 The author has tried to reduce as far as practicable the correction due to the fact that some ethyl alcohol is always oxidised to carbon dioxide when treated with dichromate and sulphuric acid, Low temperatures, low acid concentration, and the substitution of a weaker acid for sulphuric acid, all tend to reduce the magnitude of the correction.The order in which the materials are mixed is also not without influence. The method finally recommended is a compromise, some alternative methods, which kept the correction somewhat lowor, requiring too much time for the complete oxidation of methyl alcohol, or introducing other sources of error, or being inconvenient in their application.The procedure recommended is as follows : Thirty grms. of potassium dichromate are dissolved in 150 C.C. of water in the distillingflask, and the solution cooled in a mixture of ice and salt.To the cooled mixture, containing much crystallised dichromate, 50 C.C. of the suitably diluted spirit to be examined (cf. Thorpe and Holmes, Zoc. cit.) are added, and finally, in small quantities at a time, a mixture of 20 C.C. sulphuric acid and 20 C.C. water, previously cooled in ice and salt. The flask, connected of course to the train of drying and absorption apparatus, is allowed to stand until next day, the freezing mixture not being renewed.Finally the mixture is heated to the boiling-point, and the carbon dioxide swept out of the apparatus and into the soda-lime tubes by means of a current of air (cf. Thorpe aud Holmes, Zoc. cit.). Under the above conditions the subtractive correction for each grm.of ethyl alcohol present is 0*0040 grm. carbon dioxide. The test numbers are excellent. G. C . J. Detection of Methyl Alcohol in Ethyl Alcohol. V. Pazienti, (Annuli Chim. Applic., 1915, 3, 279-281.)-Five C.C. of the alcohol are diluted to 50 C.C. with water, and oxidised by means of 3 grms. of sodium persulphate and 10 C.C. of 20 per cent. sulphuric acid. The flask is connected with a condenser, the liquid dis- tilled, the distillate collected in fractions of about 2 c .~ . , and the fractions tested for formaldehyde with Schryver’s reagent. To 10 C.C. of distillate containing formalde- hyde are added 2 C.C. of a 1 per cent. recently filtered solution of phenylhydrazine hydrochloride, 1 C.C. of freshly prepared 5 per cent. potassium ferricyanide solution, and 5 C.C.of strong hydrochloric acid. h roae coloration is obtained in the presence of 1 : 1,000 of methyl alcohol. By treating the unfractionated distillate with the reagent 4 per cent. of methyl alcohol can be detected. When the colour reaction is very faint owing to the presence of excess of formaldehyde, the test should be repeated after dilution of the distillate.C . A. M. Method for the Titration of Small Amounts of Halides. F. C. MeLean and D. D. Van Slyke. ( J . Amer. Chem. Soc., 1915, 37, 1128-1134.)-For the titration of small quantities of halides Volhard’s method is not entirely satisfactory, requiring one drop of Tc solution to determine the end-point, The authors describe a modified method for titrating back the excess of silver, in which the far more delicate reaction between iodine and starch is employed as indicator, using & or i& potassium iodide standardised against the silver solution.The halide is precipitated in presence of a known amount of free nitric acid (preferably about 1 grm. of HN0,- 1 C.C. of the acid of 1-42 sp. gr.). In the case of chlorides and bromides, it356 ABSTRACTS. OF CHEMICAL l’APEl18 is necessary to remove the precipitated silver halide by filtration to avoid its subse- quent reaction with the iodine.Coagulation of the colloidal precipitate is most effec- tively brought about by shaking the liquid with one or two drops of caprylic alcohol, and a perfectly clear filtrate obtained. To the filtrate is added a rnixed reagent con- taining sodium’ nitrite to liberate iodine, starch, and a sodium salt of a weak acid equivalent to the quantity of free nitric acid known to be present.The reagent is composed of trisodium citrate (Na,C,H,O, + 54H,O), 446 grms. ; sodium nitrite, 19.0 grms. ; soluble starch, 2-5 grms., with water to 1 litre. Of this, 4 C.C. are added to the clear filtrate, immediately before titration, for each 1 grm.of free nitric acid present, the total volume of the liquid being preferably not more than 50 C.C. Under them conditions, on titration of the excem of silver nitrate with zG potassium iodide the end-point is indicated with perfect sharpness, one drop producing a distinct coloration owing to the liberation of iodine and its reaction with the starch. The titrations may be made either with the entire filtrate and washings of the precipitate, or with an aliquot portion of the filtrate passed through a dry filter, discarding the first portion.Results of analyses quoted show a maximum error of 3 per ccnt. in the case of small amounts (0.7 mgrm. of C1 and 2-6 mgrms. of Br) down to 0.1 per cent. in the case of substantial amounts (70 mgrms. of C1 and 33 mgrms.of Br). J. F. B. Optical Activity of Mineral Oils. C. Engler and W. Steinkopf. (Ber., 1914, 47, 3358-3362 ; through J. Chcm. Soc., 1915, 108, i., 205-206.)-Sporadic optical activity in mineral oils can be attributed to contact with optically active animal or vegetable matter, but universal activity is a strong argument in favour of an organic origin. In examining oils for optical activity they should be separated itno as large a number of fractions as possible, and these frequently require to be redistilled ; otherwise optically active constituents disseminated through the mass of the oil may escape detection, particularly if both dextro- and laevo-rotatory con- stituents are present.Mineral oils readily lose their activity, wholly or in part, when exposed to a high temperature ; distillation should be effected, therefore, under the smallest possible pressure and from small vessels, superheating being carefully avoided.The authors have found some portions with distinct optical activity in every mineral oil which has been investigated. Use of Nitric Acid as a Solvent for Compounded and Vulcanised Rubbers. H. W. Jones.(Rubber Industry, London, 1914, 189-190 ; through J. SOC. Chenz. Tnd., 1915, 34, 671.)--The author suggests the use of nitric acid of sp. gr. 1.42 as a means of separating carbon black from rubber and of obtaining an indication of the presence of (‘ hydrocarbons.” Estimation of Bitumen in Rubber Mixings. B. D. Porritt and E. Ander- son. (Rubber Indzislry, London, 1914, 181-188 ; through J.SOC. Chcm. Ind., 1915, 34, 672.)-Neither pyridine nor carbon disulphide gives sufikiently trustworthy results when used as solvent for the estimation of bitumen in rubber mixings.ORGANIC ANALYSIS 357 Attempts were made to effect a separation of the bitumen based upon the observa- tion that it is apparently unaffected by heating with nitric acid (sp. gr. 1.355 at 19-1O C.) for fifteen minutes at 100" C.A portion of the bitumen, however, is rendered insoluble in solvents by the action of the acid, and tho method fails, partly, at least, for this reason. Method of Estimating Small Amounts of Carbon Dioxide in Rubber Goods in Presence of Sulphides. (Iiubber Industry, London, 1914, 191-192 : through J. SOC. Chem. Ind., 1915, 34, 672.)-The separated insoluble portion of the rubber sample is heated in a flask with 10 C.C.of hydrochloric acid and 10 C.C. of water. A current of air, free from carbon dioxide, is aspirated through the flask, and the carbon dioxide is collected in two flasks, the first contain- ing 30 C.C. and the second 15 C.C. of the following reagent: barium chloride, 30 grms. in 180 C.C. of water ; ammonia (sp.gr. OoS80) 36 C.C. ; water '( sufficient to measure 300 c.c." The precipitated barium carbonate is rapidly collected, dissolved in hydroohloric acid, and converted info barium sulphate. H. W. Jones. I n this reagent sulphides are held in solution. A Simple Method f o r the Estimation of Mineral Natter in Vuleanised Rubbers. H. W. Jones. (-Rubber Idustry, London, 1914, 199-201 ; through J.SOC. Chem. Ind., 1915,34, 672.)-Two grms. of the sample are heated with 40-50 C.C. of nitrobenzene in a 200-300 C.C. flat-bottom flask, connected to a reflux air-con- denser. When solution of the rubber is complete, the flask is allowed to 0001, the contents diluted with acetone, stirred with a glass rod, and allowed to stand. The mineral matter is deposited rapidly and is separated by decantation, transferred to a weighed filter-paper, and washed well with acetone.I n some cases it is advanta- geous to wash further with alcohol and chloroEorm, then to moisten the filter with water so as to obtain a moist and uncaked residue for further examination. Little or no carbon dioxide is eliminated from calcium or magnesium carbonate by boiling for one hour in nitrobenzene. Estimation of the Mineral Matter in Rubber Mixings. B. D. Porritt and R. Wheatley. (Rubber Idastry, London, 1914, 193-199; through J. SOC. Chem. Ind., 1915,34, 672.)-Ozone was passed over dry, powdered, vulcanised rubber in a flask, which was shaken repeatedly in order to expose a fresh surface; the sample was then extracted with acetone, and the ash estimated in the extracted residue. I t was not found possible to render the whole of the organic matter soluble, and in many instances the mineral matter was chemically altered.