22 THE ANALYST. ORGANIC ANALYSIS. General Reaction for the Identification of Multiple Bonds in Un- saturated Compounds of the Aromatic and Aliphatic Series. E. Molinari. (Ber. deut. Chem. Ges., 1907, 40, 4154-4161.)-1~ the aromatic series it is necessary to distinguish between compounds which are definitely unsaturated and contain double bonds in the nucleus, as formulated by Kekulk, and compounds which combine with halogens only by substitution, having a structure more in accordance with Baeyer’s centric formula. I n the aliphatic series it is necessary to distinguishTHE ANALYST. 23 between unsaturated compounds with double bonds of the ethylene type and those with triple bonds of the acetylene type. The test proposed by the author depends on the fact that unsaturated compounds containing double bonds, whether aliphatic or aromatic, combine rapidly with ozone, forming ozonides.Aliphatic compounds con- taining triple bonds do not absorb ozone, but they possess a high iodine absorption value, Aromatic compounds to which Baeyer’s centric structure must be assigned do not absorb ozone. I n applying the test qualitatively, a few decigrams of the substance are dissolved in a few C.C. of a liquid which does not absorb ozone. A stream of ozonised air is passed through the solution, and its absorption or non-absorption is ascertained by potassium iodide and starch paper. I n the case of aliphatic compounds with double bonds, the absorption of ozone may be determined quantitatively by increase of weight. The absorption takes place in the ratio of 1 molecule of ozone (0,) to each double bond.If the iodine absorption value is distinctly greater than its equivalent ozone absorption value, it may be inferred that the substance is a mixture of compounds with double bonds and triple bonds. Amongst the aromatic compounds which absorb ozone are the polyhydric phenols, benzoquinone, phenanthrene, anthracene, naphthalene, quinoline, etc. Benzene and its homologues, phenol, pyrocatechol, anthraquinone, etc., do not absorb more than traces of ozone. J. F. B. Alkaline Saponification of Alkyl Nitrates in Presence of Hydrogen Peroxide. T. Carlson. (Bcr. deut. Chcnz. Ges., 1907, 40, 4191-4194.)-ln a previous paper, Klason and Carlson had shown that alkyl nitrates, on saponification by alcoholic potash, behave partially as if they were nitrites of alkyl peroxides.These peroxides, under the influence of the alcoholic alkali, tend to break down into aldehydes and water, and the aldehydes then undergo resinification and other profound changes. The simultaneous presence of a mercaptan prevents this change, the peroxide being reduced and the normal alcohol regenerated. The production of a peroxide and a nitrite as primary products of alkaline saponification of an alkyl nitrate is the converse of Baeyer and Villiger’s observation that ethyl nitrate is formed by the action of nitrous acid on ethyl peroxide. With the nitrates of monovalent alcohols, the production of nitrite during saponification increases with the molecular weight of the alkyl complex, whether mercaptan be present or not, but the nitrite obtained in presence of inercuptsn is always higher than in its absence.The nitrates of polyvalent alcohols, such as glycerol or cellulose, yield very large pro- portions of nitrite on saponification by alcoholic potash. With these nitrates the alkyl complex is almost entirely destroyed, owing to the action of the alkali on the peroxide first formed. The peroxide, however, may be decomposed in the nascent state by carrying out the saponification by means of alcoholic potash with the addition of hydrogen peroxide. By the interaction of the two peroxides, oxygen is liberated, and the alcohol, glycerol or cellulose is regenerated without destruction. The quantity of oxygen evolved is equivalent to the quantity of nitrite produced.J. F. B. Estimation of Halogens in Organic Substances. James Moir. (Proc. Chm. SOC., 1907, 23, 233.)-The process previously described by the author24 THE ANALYST. (ANALYST, 1906, 31, 420) is now improved by adopting the Volhard method of back titration with standard thiocyanate solution. The contents of the nickel crucible after heating are dissolved and treated with a crystal of bisulphite to reduce manganate. A measured quantity of standard silver solution is added, and the mixture then acidified with nitric acid, heated to 70" C., filtered through a Buchner funnel, using hardened paper, and the residue on the filter extracted with hot dilute nitric acid. The filtrate is cooled t o 40' C., and then titrated with standard thio- cyanate and iron alum.Each C.C. of decinormal silver solution, by difference, is equal to 0.0080 gram of bromine, A. R. T. Estimation of Hydrazine. A. W. Browne and F. F. Shetterlg. (Bey. dezk Chent. Ges., 1907, 40, 3953-3962.)-The usual method of estimating hydrazine is that proposed by Hofmann and Kiispert (ANALYST, 1898, 23, 95), which consists in oxidising the hydrazine in acid solution by means of ammonium metavanadate, ac- cording to the equation : N2H4.H,S0, + 2 0 = N, + 2H20 + H,SO,. The quantity of metavanadate reduced is then determined by titration with permanganate. An alternative method proposed by the same chemists consists in performing the oxida- tion in a current of carbon dioxide, and collecting and measuring the nitrogen evolved in a nitrometer. The authors have found, however, that the reaction does not proceed entirely according to the above equation, but that larger or smaller quantities of hydraeoic acid and ammonia, depending on the conditions, are produced, thus : The error thus introduced in the method of estimation described (Zoc. cit.) amounts to about 4.5 per cent.in the nitrometer method, but only to about 1 per cent. in the permanganate method. The end-point with the permanganate is not very sharp, and consequently a small excess of permanganate has to be added before the end of the titration is certain; this slight excess makes the error of the per- manganate method less than it would otherwise have been. 2N2H, + 2 0 = N,H + NH, + 2H20. J. F. I%. The Reaction of Phloroglucinol-Hydrochloric Acid with Essential Oils.K. Kobert. (Zezt. a n d Chcm., 1907, 46, 711-714.)-1t has been shown previously by Rosenthaler (ANALYST, 1905, 30, 247) that certain essential oils give colorations when treated with vanillin and hydrochloric acid, and the author now describes tho reactions which take place when these oils are mixed with phloroglucinol and hydro- chloric acid. The reagent is prepared by dissolving 1 gram of phloroglucinol in 10 C.C. of alcohol; 1 C.C. of this solution is then mixed with 9 C.C. of concentrated hydrochloric acid. On shaking a few drops of the essential oil with 0.5 C.C. of the reagent, a bright red coloration is obtained in the case of mustard oil, clove oil, pimento oil, dill oil, orange-blossom oil, tarragon oil, basil oil, bay oil, lavender oil, Peru balsam oil, geranium oil, parsley oil, sassafras oil, and the oil from jaborandi leaves.A brownish-red coloration is given by cassia oil, bergamot oil, aniseed oil, eucalyptus oil, mint oil, rosemary oil, and lemon oil. I t appears that the red colora- tion is given only by those essential oils, or their constituents, which contain an ally1 group in their molecule. w. P. s.THE ANALYST. 25 Note on Fucose and the Estimation of Methylpentosans in Natural Products. W. Mayer and B. Tollens. (Journ. Landw., 1907, 55, 261 ; Chem Z e d Rep., 1907, 31, 531.)-Tragaca,nth is not suitable for the preparation of fucose, because the product is contaminated with large quantities of arabinose, but from 19 kilograms of seaweed the authors obtained 70 grams of pure fucose. The osazone melts at 177.5' C.; thus it has the same melting-point as rhodeosazone. Fucose is decomposed by hydrochloric acid more slowly than rhamnose, and for the estimation of the methylfurfural it requires a longer distillation. For this reason the yield of methylfurfural is lower than with rhamnose, since the aldehyde is gradually converted into other products. With 0.05 to 0.1 gram of fucose it is necessary to distil over more than the usual 360 to 400 C.C. before the distillate ceases to give the methylfurfural reaction. A quantity of phloroglucinol equal to that of the fucose taken is added to the distillate, and the liquid is put aside for one and a half to two days until the precipitate has settled. The phloroglucide is collected in a Gooch crucible, washed with 150 C.C.of water, dried in the water-oven for four hours, and weighed in a stoppered bottle. The quantity of fucose is calculated from the formula already given (ANALYST, 1907, 32, 300), or from tables drawn up by the authors. When arabinose is present the phloroglucide consists of a mixture of the furfural and methylfurfural derivatives. These may be separated, after drying and weighing the precipitate, by repeatedly extracting the latter in the crucible with hot alcohol. The meth ylfurfural compound is dissolved, and the insoluble furfural phloroglucide is then dried and weighed. J. F. B. Melting-Point of d-Phenyl-Glucosazone. Frank Tutin. (Proc. Chenz. Soc., 1907, 23, 250.)--The phenyl-osazones obtained from the sugars present in various plants were found in many instances to melt at a temperature considerably higher (up to 219' C.) than the usually accepted melting-point of d-phenyl-glucosaxone-- namely, 205" C.Analyses of these compounds of high melting-point showed them to be osazones of hexoses, and it was at first supposed that they were identical with a-acrosazone (melting-point 217O C.). By careful recrystallisation from pyridine, however, all the osazones were obtained melting between 216" to 218" C. The method adopted was to dissolve the osazone in a small quantity of boiling pyridine, add some hot alcohol and then a little water, and allow the solution to cool. On similarly purifying a specimen of d-phenyl-glucosazone (prepared from dextrose and melting at 205' C.), this also melted-at 217O C., while another specimen of the same compound, carefully prepared with recently distilled phenyl-hydrszine, showed an initial melting- point of 216" C.Thus d-phenyl-glucosazone, when pure, melts at about 217' C., and not at 205' C., as stated by Fischer (Ber., 1884, 17, 579). A. R. T. Colour Reactions of Rosin Spirit. C. Grimaldi. (ChesIL. Zeit., 1907, 31, 1145-1146.)-Rosin spirit, and more particularly that portion of it which boils at a temperature of 170" C., gives an emerald green coloration when heated with hydro- chloric acid and zinc, whilst turpentine, camphor oil, mineral oil, etc., yield yellow or brown colorations. The sample to be tested for rosin epirit is distilled, and the portion26 THE ANALYST. passing over at 170" C. is heated with an equal portion of concentrated hydrochloric acid and a small piece of zinc for five minutes in a boiling water-bath.During the heating the mixture should be shaken frequently. Rosin spirit also gives a yellowish- green coloration with Halphen's test-phenol and bromine vrtpour (ANALYST, 1903, 28, 9)-whilst turpentine is not coloured at all, and rosin oil, camphor oil, mineral oil, etc., give violet or red colorations. w. P. s. Vulcanisation Tests with Plantation Rubbers. C. Beadle and H. P. Stevens. (Chem. News, 1907, 96, 235-236,)-A determination of the specific gravity of the vulcanised samples described in a previous communication (ANALYST, 1907, 32,337) showed that this value tended to increase on keeping, especially in the case of fully cured rubber. Over-cured samples, whether of plantation or (' hard-cure " Para rubber, showed a relatively high specific gravity after being kept for some months, and, as a rule, a decrease in tensile strength was attended by a decided increase in the specific gravity.In the case of freshly vulcanised samples of different types of rubber, some of which were under-cured and others over-cured, but small differences were observed in the specific gravities, although different methods of preparation were used. As a rule this value is not lower than 0.94, and seldom reaches 0.96, unless the rubber has been over-cured and kept for some months, in which case it may reach 0.913. Samples tested for tensile strength show a lower specific gravity than the untreated rubber if examined immediately after the test, but recover their former value when kept for some time.A specimen of the best Ceylon biscuit rubber gave the following results on analysis : Moisture, 0-5 ; albuminoids, 0.1 ; resin, 2.8 ; ash, 0.4 ; and caoutchouc (by difference), 96.2 per cent. When vulcanised under various conditions the average tensile strength of the product differed but little from the results given by "hard- cure " Para rubber (Zoc. cit.). C. A. M. Poison Sumach. A. B. Stevens and L. E. Warren. (Amer. Journ. Pllzarm., 1907, 79, 499-522.)-The juice of the poison sumach (Rhus vernis), which grows in swamps in North America, has varnish-forming properties resembling those of the juice from the Japanese lac-tree (R. vernicifera), and has the same poisonous action upon the skin. The fresh juice examined by the authors was slightly acid, and had a specific gravity of 0.9976 at 20" C.I t contained from 9.9 to 18.9 per cent. of water, 76 to 84 per cent. of crude resins " soluble in alcohol, and 5.2 to 8 per cent. of a gumlike substance insoluble in alcohol, but soluble in water to the extent of 1.7 to 2.6 per cent. This gum had a strong enzymic action, turning guaiacum tincture blue, and causing the separated resins to become black in the same way as the fresh juice. The crude resins contained a poisonous oily substance to which the odour of the juice was attributed. The fruit of the tree contained a large amount of a greenish-white fat with a peculiar odour and a taste like that of tallow. It gave the following values: Specific gravity at 25'/25O C., 0.9749 ; melting-point, 38' to 39" C.; saponification value, 236.3 ; and iodine value, 13-1. The residue left after recrystallisation from petroleum spirit and from alcohol had an elementary composition agreeing with that of myristin. The The separated resins were also found to be poisonous.THE ANALYST. 27 fruit contained no starch, alkaloids, or glucosides, but the residues left after extraction of the ftlt were rich in nitrogen. No poisonous constituent was found in the ripe fruit. C. A. &I. The Titration of Tannin by Iodine : An Answer to Cormimbamfs Note. F. Jean. ( A m . Clzirn. a n d , 1907, 12, 426-427 ; compare ANALYST, 1907, 32,426.)- The author claims that his method gives exact results when properly carried out, working with a solution saturated in the cold with sodiuni bicarbonate. The iodine solution must be separately standardised against pure tannic acid and against gallic acid, as these bodies absorb different quantities of iodine. A.G. L. The Estimation of Tartaric Acid in the Presence of Malic and Succinic Acids. J. v. Ferentzy. (Chem. h i t . , 1907, 31, 1118.)-The method is based upon the insolubility of basic magnesium tartrate in a mixture, in equal parts, of alcohol and water, and the ready solubility of the corresponding salts of malic and succinic acids in the same solvent, The solution containing the three acids is concentrated to a small volume, and' sufficient alcohol added to bring the alcoholic strength to 50 per cent. I t is next treated with magnesia, mixture and 10 C.C. of ammonia solution, the alcoholic strength again brought to 50 per cent., and the whole thoroughly shaken and allowed to stand for twelve hours.The crystalline preci- pitate is then collected, washed with 50 per cent, alcohol, dried, ignited, and weighed. The amount of magnesium oxide multiplied by the factor 1.875 gives the corresponding amount of tartaric acid. C. A. If. The Examination of Turpentine Oil. W. Flath. (Farben Zeit., 1907, 13, 78; Clzem. Z e d . Rep., 1907, 31, 564.)-The specific gravity of pure turpentine oil usually lies between 0.860 and 0.870 at 15' C., but may be somewhat higher or lower. If the specific gravity of a pure oil were particularly high, the value might still remain normal after the addition of 10 per cent. of heavy petroleum, which would also not be recognissble by the odour. Even larger quantities of heavy petroleum may be mixed with refined tar oil without appreciably affecting the specific gravity. According to the author, pure turpentine oil, mixed with an equal volume of freshly distilled aniline, gives a clear solution, whereas in the presence of petroleum the liquid remains turbid. The substitution of American wood turpentine oil for Russian or Swedish turpentine oil may be detected by shaking 2 C.C. of the sample with 1 C.C. of concentrated sulphurdioxide solution. In the case of Russian or Swedish turpentine oil a bright yellow or greenish coloration is produced, whilst with pure American or French turpentine oil the liquid remains white and turbid. The addition of 5 or 10 per cent. of a Russian oil causes a perceptible greenish coloration to appear. This reaction is given by all turpentine oils obtained from wood, and is due to an apparently inseparable constituent of the oil. C. A. N.