156 ABSTRACTS OF CHEMICAL PAPERS ORGANIC ANALYSIS. Influence of Hydroxy Acids and Lactones on Estimations of the Chemical Constants of Fatty Acids. C. A. Browne. ( J . Id. and Eng. ChLem., 1915, 7, 30-34.)-The author discusses the influence of laotone-forming hydroxy acids (more particularly the y-acids) on estimations of the acid, saponification, and acetyl values of fatty acids. For instance, if one-half of a sample of hydroxystearic acid should undergo lactonisation, the acid value would be depressed from 186.85 to 96.31, and the saponification value increased from 186-85 to 192.62, whilst the ester value would become 96.31 instead of zero.When a difference between the acidORGANIC ANALYSIS 157 Zanzibar Copal. Madagascar Copal. and saponification values shows that lectones are present in a fatty acid, the mean molecular weight (m) of the fatty acids before lactone formation may be Demerara Copal.calculated by the formula : m= %lo8+ 18'016 (' - '), where a is the acid value and S s the saponification value of the mixed fatty acids. Since lactones in a mixture of fatty acids are derived from pre-existing hydroxy acids, the acetyl value of Benedikt-Ulzer (ester value of the acetylated acids) is a truer measure of the original hydroxyl content of fatty acids than the acetyl number of Lewkowitsch.(potassium hydroxide required to saponify the combined acetyl in the acetylated acids). Thus, y-hydroxystearic acid in the glyceride molecule can be acetylated, and would show an acetyl number by Lewkowitsch's method ; whereas the separated fatty acids would show no acetyl value on account of the formation of stearo-lactone. The original acetyl value, calculated to the acetylated hydroxy acids before lactone formation, may be found from the formula, : + 0,00~07~c - %), where r is the acetyl ester value (Benedikt-Ulzer) and ( the acetyl value (Lewkowitsch).w. P. s. Action of Chlorhydrocarbons on Hard Copals.C. Coflignier. (Bull. SOC. Chim., 1914, 15, 780-781.)-The following amounts of insoluble matter were separated on boiling the copals with the solvents, and allowing the solutions to stand : __ . -. . . . .- . . . . - . - . -. . . . Dichlorethylene . . . ... ... Trichlorethylene . . . ... ... Te trachloret hylene ... ... Tetrachlorethane . . . I.. ... Pentachlorethane .. . ... ... I I Per Cent. 78-70 83.20 79-20 66.50 78.40 Per Cent. 70.90 70-80 88.20 37.80 63.00 Per Cent. 70.50 79.20 64.20 47.70 53.10 Madagascar copal' may be distinguished from the other two copals by the fact that it alone gives a practically complete solution when boiled for not too prolonged a period with tetrachlorethane (cf. ANALYST, 1914, 39, 13). C. A. M. Polymerised Drying Oils.R. S. Morrell. (J. SOC. Chem. Ind., 1915, 34, 105-109.)-Linseed oil from various sources was thickened in bulk or in smaller quantities in the laboratory, and to avoid oxidation this was done in an atmosphere of carbon dioxide or hydrogen. In the case of hydrogen it was found that, contrary to the usual statements, a very slight addition occurred, and eventually only carbon dioxide was used.The oils were heated for twenty-eight to sixty hours at 260" C., and the thickened oil was completely soluble in light petroleum. In the case of tnng oil the temperature and time of heating were 240" C. and twenty minutes respectively, and the heating was stopped as soon as any of the substance insoluble in light petroleum was formed, or, in other words, as soon as gelatinisation begun.158 ABSTRACTS OF CHEMICAL PAPERS 47.5 (240'0.) { 27.8 (2100 C.) - - - 0.567 198.5 765-151 The fall in the iodine value is essentially a function of the temperature, for linseed oil which has been kept at 260' C.for twenty-eight and a half hours can be heated for another twenty hours at this temperature without change; but on raising the temperature to 293' C.for another two and a half hours the iodine value fell to 89-97, and a substance is found insoluble in light petroleum and in carbon tetra- chloride. It is evident that one phase of a change in linseed oil is complete between 260' and 280' C., and above this another appears. Linseed and tung oils, thickened between 260' C. and 280' C., contain two modifications, one insoluble in acetone and the other soluble, but both soluble in light petroleum; but olive oil does not give an insoluble foEm, and in the case of poppy-seed oil the insoluble modification is not formed below 290' C.TABLE I. 46.6 (240' 0.) 72 % (210' C.) 0.9542 1.5114 (11.5' C.) 0.3143 (11.5' C.) 876.2 2.47 192.6 93-140-157 sp. gr. (15'C.) ... ... rnin .. ... ... [TI* ... ...... Molecular weight (in benzene) Acidity ... ... ... Sa onification valire ... IocPine value . . . . . . . Glyceryl per cent. ... Crystalline hexabromide ... 42-60 % 0-9763 1.4964 0.2892 Raw. 41-52 % 0-9527 1.4846 0.3035 0.933 1.4831 0.3053 (15' 0.) 805 0.4 197-203 185 9-62 (19O 0.) 33.5 % Linseed Oil. Thickened in Bulk 260.280" C. 0.9527-0*969 - - 5.15 188-193 103-106 - None Thickened in- Hydrogen (260-280" C.)... -- 0.9654 - - - 2.4 195-196 93-128 - None Linseed Oil. Solubility in acetone . . . . . . Sp. g. (15'C.) . . . . . . . . . CnlD . . . . . . . . . . . . [.ID . . . . . . . . . . Molecular weight (in benzene) ... Acidity . . . . . . . . . . . . Sa onification value ....... I o L e value . . . . . . . . . 100 % 0.933 1.4831 0*3053 805 , 0 '4 TABLE 11.C02 (2& 280" C.). -. -. 0.969 1.4915 0.2998 (15' 0.) 1.7 10.4 None (19' 0.) 1686-1704 200-203 118-134 Thickened Linseed Oil. Insoluble Soluble in Ac'etone. Tung Oil. ._ - - - - . - - Ram. . 0-9405 1'5172 (14' 0 . ) 0.3140 (14' 0 . ) 797 3 *3 192-194 166-168 10.47 None Tung Oil. 100 % 0.9405 1-5174 (12.5' C.) 0'3204 (12.5' 0 . ) 797 3.3 192 168 Thickened. -- 0.956 1.5134 (15' C.) 0'3117 (15' C.) 1431 2.3 8.3 None 190-201 143-171 Thickened Tung Oil.Insoluble Soluble 7 z G FORGANIC ANALYSIS 159 In view of the results obtained from the study of poppy-seed oil, it is concluded that linseed oil (thickened and soluble in acetone) contains the components of poppy- seed oil, which polymerises only slightly at 2508 C., but these components have undergone isomeric change. The chemical properties of thickened linseed oil (soluble in acetone) are similar to, and in some respects identical with, those of heated poppy-seed oil under the same conditions.When linseed oil (soluble in acetone) is heated to 290" C., the part which becomes insoluble in acetone has the same quad- ruple molecule as in the case of the poppy-seed oil. The author concludes that linseed and poppy-seed oils contain mixed glycerides of variable amounts of unsaturated acids in addition to small quantities of saturated glycerides, the acids being inter- changeable and their amounts dependent on the sources, and probably the ripeness, of the seed.Thickening by polymerisation without change in chemical composition occurs when there are at least two pairs of doubly linked carbon atoms in the mole- cules of the acids of the glycerides.Before polymerisation occurs, there is a shifting of the linkages in the molecule, and in the case of tung oil there is some slight evidence of ring formation. The thickening is due to polymerisation of the mixed glycerides, and the first stage is the formation of a product insoluble in acetone, which may be a double molecule of linked glycerides in the case of linseed oil, or a quadruple molecule in poppy-seed oil.The change is dependent on the temperature, ind an equilibrium is established when 50 per cent. of the modification has been formed. At higher temperatures the final stage is the formation of a polymer insoluble in light petroleum, which seems to be determined in linseed oil by the linolenic acid glyceride.H. F. E. H. Estimation of Unsaponiflable Matter Applicable to Ether Extracts, Fats, Oils, and Waxes. J. B. Rather. (J. Ind. and Eng. Chem., 1915, 7, 34-35,)-The following method has been devised for the estimation of unsaponifiable substances in the ether extracts of fodders, etc. It differs from the usual methods in that the fatty acids are separated from the unsaponifiable substances by precipitation from ethereal solution.About 0.4 grm. of the substance is boiled for one hour with 20 C.C. of f alcoholic sodium hydroxide solution; the solution is then evaporated nearly to dryness, 3.5 C.C. of glacial acetic acid are added, and the mixture is warmed with 50 C.C. of ether. Twentyfive C.C. of water are now added, the mixture is warmed for a minute, transferred to a separating funnel, and the flask is rinsed with five successive portions of 20 C.C.of ether. After shaking, the aqueous layer is drawn off, the ethereal solution is shaken with 10 C.C. of warm sodium hydroxide solution (1 : 2), 25 C.C. of warm water are then added, and the aqueous layer is drawn off; the washing is repeated five times with 30 C.C.portions of cold water. The ethereal solution is then transferred to a weighed flask, the ether evaporated, and the residue dried at looo C. to constant weight. When large quantities of fatty acids axe present together with small amounts of unsaponifiable substances, fatty acids may not be completely removed from the ethereal solution by the above treatment; in this case the ethereal solution, before evaporation, is shaken with 20 C.C.of hydrochloric acid, the acid solution is drawn off, the ethereal portion evaporated, and the residue weighed. This residue is dissolved in alcohol and titrated with & sodium hydroxide160 ABSTRACTS OF CHEMICAL PAPERS solution; the number of C.C. of alkali solution required is multiplied by 0.028 to obtain the weight of fatty acid which is deducted from the weight of the residue obtained on evaporating the ethereal solution.With samples containing more than 25 per cent. of unsaponifiable substances the amount of retained fathy acid may be disregarded; it amounts to about 2 mgrms., and is balanced by a corresponding amount of unsaponifiable substance in the soap solution. w.P. s. Proposed Uniformity in Methods of Fat Analysis. W. Fahrion. (J. Amer. Chem. ASSOC., 1915, 10, 7-18; through J. SOC. Chcm. Ind., 1915, 34, 185.)-A report of the work of a Commission appointed by the International Association of Leather Trades Chemists to study analytical methods for oils and fats used in the leather industry. Each member received samples of nine oils.The concordance in results was not 80 good as was expected. The acid values agreed well except in the case of cod oils, which alter during keeping. Mixtures of alcohol and petroleum ether, and alcohol and ether, were tried as Bolvents, using either aqueous or alcoholic alkali for titration. The results were the same as with the usual method. The author recom- mends titrating 5 grms.of oil in alcohol with or alkali. Saponification values are greater the longer the boiling, and the lower the concentration of the alkali. The author recommends using at least 3 grms. of oil and alkali containing less than 10 per cent. water, and claims that saponification is complete with five minutes boiling after a clear solution is obtained. Iodine values by the Hiibl, Wijs, and Hanb methods were reported.Results by the Wijs method were in excellent agreement, whilst the Hub1 results showed differences up to 26. The author recommends the abandonment of the Hub1 method. The values by the Hanus method lie between the results given by the other two methods. In the determination of unsaponifiable matter the author washes the petroleum spirit with 50 per cent.alcohol, which does not dissociate the soap. Unsaponifiable matter from fish oils should not be dried above 100" C. The results obtained in the analysis of dkgras show good concordance. Study of Various Tests on Glue, particularly the Tensile Strength. A. H. Gill. (J. Id. and Eng. Chem., 1915, 7,102-106.)-1n a summary, the author says his experiments described in this paper seem to justify the following conclusions : (1) That the tensile strength, jelly test, and viscosity of glue, bear no relation to each other ; (2) that a, particular method described for measuring tensile strength may be expected to give individual results not differing by more than 10 per cent.from the mean of a, series ; and (3) that the method of testing the strength of glue by measuring the strength which it imparts to bibulous paper is dependable, and gives fairly con- cordant results.Conclusion 1 is fully justified, the viscosity in particular being shown to be in no sense a measure of the useful properties of glue, the best and worst of a, series of glue solutions having viscosities which stood to each other in the close ratio of 100 : 96, whilst other more practical tests pointed to their values standing in the ratio of 100 : 60, or thereabouts.The jelly test, on the other hand, appears to exaggerate the difference in value of two glues, if careful measurements of the tensileORGANIC ANALYSIS I61 strength of glued joints be taken as the true measure of their value. For example, the jelly test suggested that one sample of glue was twice as strong as another, whereas the tensile strength of joints glued with each stood in the comparatively close ratio of 15 : 13.The jelly tests were few in number, and appear to have been carried out before the tensile strength test (applied as standard) had been so far perfected that duplicate results never differed from each other by more than 20 per cent.Conclusion 3 is important in view of the fact that measurements of the tensile strength of joints may be quite misleading unless conducted with elaborate pre- cautions. The author obtained erratic results when attempting to measure the tensile strength of bibulous paper impregnated with glue, and dried. He measures the breaking strength with the Mullen paper-tester, the paper being first dipped in 25 per cent.glue solution, dried in the air for thirty-six hours, and the amount of glue taken up per square inch being estimated by weighing pieces 2 inches square before and after treatment. The results are expressed in pounds per square inch per 100 mg. glue on 4 square inches surface. They are much more concordant than measure- ments of tensile strength of joints, except when the latter are made with elaborate precautions, and in general follow the tensile strength ; but unfortunately there are notable exceptions, Conclusion 2 is supported by the author's experimental results.As worded in the original paper, it might be taken to mean that the valuation of glue by measurement of the tensile strength of glued joints was easy, whereas the main value of the paper lies in the demonstration it affords that such work, to have any value at all, must be conducted with extraordinary care.The author uses a special machine for sawing his test-block in two, the saw, frequently filed, being carefully trued on a mandrel, the block, securely held in a rest, being sawn at a definite rate. Nearly two days are required to make the glue solution.Meanwhile the blocks are dried at 60" C. for two hours, sized and set aside to dry at room temperature for two days. They are again heated to 60" C., glued, trued up in a frame to secure perfect alignment, and placed at once in a drying frame of special construction, such that a uniform pressure of 30 pounds per square inch is secured. I n this frame they are supported on the points of adjustable screws, fitted to centre marks on the lower half blocks, the pressure of a 3-pound weight at the end of a 10-inch lever being transmitted through a steel point 1 inch from the hinge of the lever, this steel point pressing centrally on a steel pin, which in its turn is centred on the upper half block.After twenty-four hours in this frame, the blocks are allowed to dry for 100 hours at; room temperature and then tasted.With all these precautions, duplicate determinations may differ 10 per cent. from the mean ; by the neglect of any one of them duplicate determinations may stand in the ratio 7 : 25. I t appears to be established beyond doubt that a pressure of about 30 pounds per square inch gives the strongest joint, notably stronger than a joint obtained with pressures of 10 or 100 pounds.If the pressure be maintained for only three instead of twenty-four hours, the joint is on the average only 3 per cent. weaker. Artificial drying of the joint at 6 5 O C. tends to weaken the joint, and, for test purposes, is less satisfactory than drying at room temperature for 100 hours. The drying apparatus is figured in detail.G. C . J.162 ABSTRACTS OF CHEMICAL PAPEES Composition of Paint Vapours. C. A. Klein. (J. Id. and Eszg. Chem., 1915, 7, 99-102.) A criticism of the experimental methods employed by Gardner (ANALYST, 1914, 39, 266) in investigating this subject. This criticism is particularly directed to the evidence adduced by Gardner in support of his alleged discovery that paint vapours contain carbon monoxide, a statement that has been repeated, solely on his authority, in the public Press of Europe and America, as well as in technical journals (ANALYST, Zoc.cit. and Oil and Golour Trades Journat, 1914, 45, 1000). For the detection of carbon monoxide, Gardner employed the iodine pentoxide method and the formation of potassium formate by tho interaction of carbon monoxide with heated solid potassium hydroxide, with subsequent confirmation of the formate by the mercuric chloride reaction.But first he sought to free the vapours from any other substances which might render the results of these tests ambiguous (or as he put it, 4 L to destroy all organic vapours ''1 by passing them through fuming sulphuric acid.Now, he himself showed that these vapours included formic acid, and it is well known that formic acid at least is destroyed by fuming sulphuric acid with liberation of carbon monoxide. Gardner's deduction that paint vapours contain carbon monoxide is therefore invalid, and the present author states that no other evidence has ever been adduced in support of such a statement. G.C. J. Determination of Gasoline Vapour in Air. G. A. Burrell and I. W. Robertson. (J. Ind. and Eng. Chem., 1915, 7, 112-113.)-The first and more accurate method depends on liquefaction and fractional distillation of the air. The apparatus illustrated is exhausted, then filled at atmospheric pressure with the air to be analysed, and, after closing the stop-cock, is immersed in liquid air contained in a Dewar flask.After ten minutes, the air is removed with a vacuum pump, the cock is again closed, and the apparatus is removed from the liquid air-bath and allowed to attain the room temperature. The pressure of the residual gasoline is then read off on the manometer shown, the ratio of this pressure to that of the atmosphere giving the volume of gasoline vapour in one volume of the original air.The alternative method described is ordinary combustion anaIysis with calculation of the volume percentage of gasoline vapour from the contraction, or from the carbon dioxide pro- duced, on the assumption that gasoline vapour is substantially pentane. The ratio of the contraction to the carbon dioxide pro- duced, a8 well as the agreement of the results with those obtained by the method first described, show that this assumption is usually justified.In cases where the mean molecular weight of the vapours differed notably from that of pentane, the ratio of the contraction to the carbon dioxide would indicate nearly enough for most purposes how the results should be calculated. G. C. J.ORGANIC ANALYSIS 163 Separation of the Illuminants in Mixed Coal and Water Gas.G. A. Burrell and J. W. Robertson. (J. Id. and Eng. Chem., 1915, 7, 17-21.)-The method of fractional distillation in vacuo at low temperatures previously described (ANALYST, 1914,39, 414) has been used in estimating the illuminants in Pittsburgh artificial gas (3 parts of coal gas with 1 part of carburetted water gas). The con- stituents of such gases may thus be separated into the following groups : Distillates.Liquid Air Temperatures, - 185' C. B. -pt. OC. Methane ... ... -165 Nitrogen ... ... -195 Oxygen ... ... -183 Hydrogen.. . ... -253 Carbon monoxide - 190 Below -140' C. Ethylene . . . . . a -103 Ethane ... ... - 51 Below -120' C. Propane ... ... - 45 Prop ylene ... - 51 Below- 78O C. %-Butane ...... 1 Is0 bu tane ... - 10 Isobutylene ... - 4 - Residues. Ethane Propane %-Butane Isobutane Propane N-butane Isobut ane n-Butane Is0 butane Benzene B.-yt. "C. ... -93 ... -45 ... 1 ... -10 ... -45 ... 1 ... -10 ... 1 ... -10 ... 80 B;-pt. Isobutylene . . . - 4 C. Ethylene ... - 103 Propylene . . . - 51 Benzene ... 80 Propylene ... - 51 Isobutylene ... - 4 Benzene ... 80 Isobutylene .., - 4 Benzene ...80 Complete analysis of the artificial gas gave the following results : Carbon dioxide, 2.63 ; oxygen, 0.81 ; carbon monoxide, 13-25 ; hydrogen, 37-33 ; methane, 31.31 ; ethane, 2.10 ; propane, 0.43 ; ethylene, 6.05 ; propylene, 0.60 ; butylene, 0.11 ; benzene, 1-33 ; and nitrogen, 4.23 per cent. The method is recommended as a simple' means of estimating benzene in artificial gas.C. A. M, Comparison of Various Modifications of the Kjeldahl Method with the Dumas Method of Determining Nitrogen in Coal, with Notes on Errors in the Dumas Method due to Nitrogen evolved from the Copper Oxide. A. C. Fieldner and C. A. Taylor. (J. Id. and Eng. Chem., 1915, 7, 106-112.)- The most important observation is that low results were obtained in every case where the Kjeldahl digestion was stopped as soon as the liquid was colourless.Heating should be.continued for about two and a half hours after the liquid becomes colour- less, but an hour map be saved, without loss of nitrogen, by allowing to cool for ten minutes when colourless, adding permanganate in small crystals, and reheating for one to one and a half hours. Potassium sulphate and a drop of mercury should be164 ABSTRACTS OF CHEMICAL PAPERS used.Omitting the potassium sulphate not only prolongs the period required to obtain a colourless solution, but also the subsequent period which is necessary to convert all nitrogenous substances into ammonium sulphate. About 15 grm. potassium sulphate per 30 C.C. acid appears to be best ; smaller additions slow the process, whereas the use of much larger proportions is attended with risk of loss of nitrogen.Omission of mercury lengthens the process some one and a half to two hours, whilst the substitution of copper sulphite also loses about forty-five minutes as compared with the use of mercury. Phosphoric anhydride was found less effective than potassium sulphate. None of the coals examined contained nitric nitrogen, the results obtained by Jodlbauer’s method being identical with those obtained by the method above described.Identical results were obtained by a modification of Dumas’ method, which is described in detail. I t involves heating the copper oxide for several hours in vacuo, and cooling it in carbon dioxide before using it for a nitrogen determination. With- out these precautions, Dumas’ method seriously overestimates the nitrogen in coal.Whereas the difficulties, such as they are, of applying Kjeldahl’s process to coal depend on the nature of the latter, those connected with the use of Dumas’ method have their root in the method itself. But whilst the difficulty of freeing copper oxide from nitrogen affects the accuracy of any results obtained by Dumas’ method, the incidence of this source of error is especially heavy in the case of coal, since the percentage of nitrogen in coal is usually small, and Dumas’ method precludes the use of large quantities of substance. G .C. J. New Reaction of /3-Naphthol. J. Katayama and B. Ikeda. (Yakugakuxasshi, October, 1914; through J. Pharm. Chim., 1915, 11, 73-74.)-0n mixing 1 C.C.of a dilute (0.01 to 0.001 per cent. solution of &naphthol with a few drops of strong sulphuric acid, and adding 0-05 C.C. of a 0.01 per cent. solution of sodium nitrite, a violet coloration is produced. The reaction, which appears to be due to the forma- tion of a quinone derivative of &naphthol, is faintly perceptible in solutions con- taining only 0*0002 grm.of P-naphthol per C.C. C. A. M. Volumetric Fehling Method Using a New Indicator. A. M. Breckler. (J. lnd. artd Eng. Chem., 1915,7, 37-38.)-The method is dependent on a constant volume at the end of the titration, a constant time of boiling, and the use of sodium monosulphide as indicator. Ten C.C. of mixed Fehling solution are placed in a large test-tube and the sugar solution is run in, starting with 8.5 c.c.; after the first addition, the mixture is boiled for one minute, counting from the time a bubble of steam first traverses the liquid.The sugar solution is now added 2 C.C. at a time, boiling fifteen seconds after each addition. When the mixture becomes faintly blue, a drop of it is added to a drop of 4 per cent. sodium sulphide solution placed on a tile; the tile is given a slight rotary movement and the colour of the spot noted.The suspended cuprous oxide settles at once as black copper sulphide, leaving a, yellow supernatant liquid. The sugar solution is now added in smaIler quantities, boiling each time for fifteen seconds, until the test with the indicator gives a colour- less supernatant liquid after the copper sulphide has settled.The estimation is thenORGANIC ANALYSIS 165 repeated, adding to the 10 C.C. of Fehling solution sufficient water to make the final volume 30 C.C. and abont 97 per cent. of the sugar solution required in the first experiment. This mixture is boiled for ninety seconds and the titration completed, which can usually be done with two additions of the sugar solution.The presence of proteins and metals which give coloured sulphides interfere with the indicator ; the former may be removed from the sugar solution by means of alumina cream. w. P. s. Estimation of Sulphur in Motor Spirit. W. A. Bradbury and F. Owen. (Chem. News, 1915, 111, 39-41.)-A modification of the apparatus previously described (ANALYST, 1914, 39, 30) enables a volumetric estimation of sulphur in motor spirit to be made within about two hours.The burner passes upwards through the base of an absorption vessel, while an inverted round-bottomed flask with a shortened neck is used as the combustion chamber. Round the neck of the flask is fitted a disc of tinned copper (2g inches in diameter) with rings soldered on each side, the lower one being made to rest upon a constriction near the bottom of the absorption vessel.The circumference of the disc is perforated with holes, and the space between the ring and neck of the flask is fitted with a cork. The two carburettore are respectively charged with 10 C.C. of the motor spirit and with pure benzene, whilst about 50 C.C. of & sodium carbonate solution and 20 C.C.of neutral (10 volumes) hydrogen peroxide are placed in the absorption vessel. Air for the combustion is supplied by means of a water-blower, and by turning the taps may be passed through either carburettor, the connections being secured by mercury joints. The pure benzene is lighted at the burner, and the combustion chamber (flask) lowered over it, until the disc is covered by the liquid in the absorption vessel, and the gaseous products of combustion are bubbling gently through the holes in the disc.The taps are then turned so that the motor spirit is farced from the other carburettor into the burner, where it should burn with a small blue flame, while the platinum coil above the gauze cap of the burner is maintained at its maximum state of incandescence.The water in the beaker is now gradually heated from 27O C. to 43O to 49O C. at the end of the estimation ; and when only about 1 C.C. of the spirit is left, two successive portions of 1 C.C. of absolute alcohol are introduced into the carburettor through a, tap above it, to remove the last traces. The air supplied is then reduced to a gentle current, which is continued until the combustion chamber is cold, after which the absorbing liquid and washings are boiled to expel oxygen and carbon dioxide, and titrated with sulphuric acid, with lacmoid as indicator.Tests made with this apparatus showed that when motor spirit is burned in a lamp the wick has a selective action, and the amounts of sulphur vary at different stages of the combustion. No such action, however, took place during the spontaneous evaporation of benzene through the wick of an unlighted lamp.In the fractional distillation of a sample of benzene containing sulphur (348 grains per gallon) 49 per cent. was present in the first fraction of 20 per cent. C. A. M. Determination of the Specific Gravity of Tars, Oils, and Pitches. J. M. (J. Id. and Eng. Chem., 1915, 7, 21-24.)-For ordinary work standardised Weiss.166 ABSTRACTS OF CHEMICAL PAPERS hydrometers (225 mm. long with bulbs about 24 mm.in diameter) give sufficiently accurate results with creosote oils, If the reading be taken above 15.5" C., it may be corrected by means of the formula- Sp. gr. at 15"/15" C. = Sp. gr. t°C/15.50C. + 0*0008(t°C. - 15.5"C.). In the case of results obtained with the pyknometer or Westphal balance the formula to be applied is- Sp.gr. 15O/15OC. = Sp. gr. to/to x sp. gr. of water at t/15*5"C. + 0*0008(t0 - 15*5'c.) When the quantity of material available is less than 100 C.C. it is advisable to use the Westphal balance, with the addition of a special plummet for fractions of less than 20 C.C. For still smaller fractions a 1 C.C.pipette may be adapted as a pyknometer, the point being closed with a piece of sealed glass tubing. The specific gravity of viscous tars and pitches may be determined by the use of a small platinum pan (20 mm. in diameter and 12 mm. in height) supported by platinum wires meeting at the top to form a hook, which is suspended from the balance by a waxed silk thread. The specific gravity of the tar is found by means of the formula- c - a x= - ..(b+c)-(a+d) where a represents the weight of the empty pan in air ; b, its weight in water ; c , its weight in air when filled with tar ; d, its weight in water when filled with tar. In the case of pitches the formation of pockets is obviated by cooling the melted material in the pan slowly and under a slight pressure.Slow cooling causes a pronounced increase iu the results given by hard pitches, but its effect is negligible in the case of soft pitohes. Only rough results are given by a hydrometer with coal- tar ; but, if used, a correction of 0*000685' for each degree C. above or below 15.5' C. is applied. C. A. M. Determination of the Percentage of Toluene in Commercial Toluol. H.G. Colman. (J. Gas Lighting, 1915, 129, 196-198.)-As the result of a large number of distillations of varying mixtures of known amounts of pure benzene, toluene, and xylene, carried out in an ordinary Wurtz distilling-flask, it was found that if the distillation is carried out under uniform conditions it is possible to ascer- tain with reasonable accuracy the percentage of toluene in these mixtures by the determination of the fractions (a) boiling below 105" C.and (b) boiling above 117' C. Further experiments with similar mixtures of pure benzene, toluene, and xylene, but also containing small quantities of the other substances often present to a limited extent in oommercial toluol, such as carbon bisulphide, paraffin hydrocarbons of similar boiling-point, and cumenes, showed that the presence of these does not materially influence the results obtained.A table constructed from the results obtained with the mixtures of pure benzene, toluene, and xylene may therefore be employed with reasonable accuracy in the determination of the percentage of toluene in commerclial toluol which has been properly washed with caustic alkali and sulphuric acid, this being distilled under the same conditions as were employedORGANIC ANALYSIS 167 with the pure mixtures and the percentages boiling below 105' C.and above 117' C. ascertained. I n the accompanying table the percentage of toluene is found in that place in the vertical column above the figure showing the percentage below 105" C., which is on 8 horizontal line with the figure showing the percentage above 117' C., the quantities of the fractions found in the test being taken as the nearest whole per cent., and the toluene result given in the nearest whole per cent.This table only holds good generally for samples containing from 50 to 75 per cent. of toluene, and for such as give at least 5 per cent. and not more than 50 per cent. either below 105* C.or above 117O C. In the great majority of cases the commercial samples fall within the limits named, but by a simple modification the exceptional samples can also be analysed in a similar manner. The distillation is conducted in a standard Engler 100 C.C. distillation-flask, as employed in the petroleum industry, having the following dimensions : Diameter, 6.5 om. ; neck, 15 x 1.6 cm.; length of side-tube, 10 cm. ; angle of side-tube, 75'. The following apparatus is also required : A round-bottomed flask of 150 to 200 C.C. capacity, a Young 12-bulb pear " fractionating column, an 18-inch Liebig con- denser, a set.of 100 C.C. graduated cylinders, and a thermometer registering from 50' to 150' C., and graduated in one-fifth degrees. The graduations of the 100 C.C.cylinders sold are sometimes incorrect to the extent of 1 0.0. The cylinders employed must be standardised by running into them known quantities of liquid, preferably toluene, from an accurate burette. Each day, before testing, the thermometer car- rection must be ascertained by placing it in a distillation-flask, with the top of the bulb just below the side-tube, and hoiling distilled water in the flask. The difference between the thermometer reading and looo is taken as the correction of the thermometer, This correction includes that due to variations of barometric pressure and inaccuracies of the thermometer.Should the barometric pressure vary considerably during the day, the correction should be reascertained in the same manner.The Engler flask and condenser are washed out with the toluol to be tested and allowed to drain. One hundred C.C. of the toluol to be tested are poured into the flask from the graduated cylinder, the latter being drained out only. Distillation is effected by means of a small naked flame, with a wire gauze screen to protect the flame and the bulb of the flask from draughts. The top of the thermometer bulb should be just below the side-tube of the flask, and the rate of distillation 7 C.C.per minute from the end of the condenser, the distillates being collected in 100 C.C. cylinders. When the thermometer reaches 105' C. (corrected) the heating is stopped, the condenser allowed to drain, and the receiver changed. The distillation is continued till the thermometer reaches 117' C.(corrected) ; the heating is then stopped, the condenser allowed to drain, and the residue in the flask, after complete cooling, drained into a 100 C.C. cylinder. The combined amounts of the two distillates and the residue should not amount to less than 99.5 C.C. From the amounts boiling (a) below 105' C. and ( b ) above 117' C. the percentage of toluene in the sample is found, subject to the correction for paraffin content specified below, by means of the accompanying table.TABLE FOR ASCERTAINING IRE PERCENTAGE OF PURE TOLUENE IN COMMERCIAL TOLUOL FROM THE PERCENTAGES OF THE FRACTIONS BOILING BELOW 105' C.AND ABOVE 117' C. - r 49 48 47 46 45 44 43 42 41 - Zr 38 37 36 35 34 33 32 31 29 - T ~. 28 27ORGANIC ANALYSIS 169 The table only holds for crude toluol samples containing 50-75 per cent.of toluene, and for such as give not less than 5 per cent., or more than 50 per cent., either below 105' C. or above 117' C. Most commercial samples fall within these limits, but in the exceptional cases a modification of the method must be adopted. The exceptional cases are of four classes : 1. The percentage boiling either below 105' C.or above 117' C. may be below 5. In this case 90 C.C. only of the sample is taken and mixed previous to distillation with 10 C.C. of pure benzene if in the first test the distillate below 105' C . has been found less than 5 per cent., or with 10 C.C. of xylene (boiling at 136-143" C.) if the amount, boiling above 117' C. has been found below 5 per cent. The mixture is then to be distilled exactly as prescribed, and the percentage of toluene in the mixture found by the table.The figure thus found is the number of C.C. of toluene in 90 C.C. of the original sample. 2. The percentage boiling both below 150' C. and above 117" C. may be so low that there is no corresponding entry in the table. In this class, which includes samples containing high percentages of toluene, SO C.C.of the sample is taken and mixed with 10 C.C. of pure benzene and 10 C.C. of xylene (boiling at 136-143" C.), and the mixture distilled as prescribed. The figure in the table corresponding to the amounts of distillate found below 105" C. and above 117" C. then gives the number of C.C. of toluene in 80 C.C. of the original sample. 3. The percentage boiling above 117' C. may be so high that there is no corre- sponding entry in the table.In this class, where the fraction boiling above 117' C. is very high, 80 C.C. is taken and mixed with 20 C.C. of pure benzene previous to distilla- tion. The figure found in the table from the percentages below 105' C. and above 117' C., multiplied by 100 and divided by SO, gives the percentage of toluene in the original sample, 4.The percentage boiling below 105' C. may be so high that there is no corre- sponding entry in the table. In this class, which comprises samples containing relatively high percentages of benzene and low percentages of toluene, 80 C.C. of the sample is taken and mixed with 20 C.C. of pure toluene previous to distillation, and the percentage of toluene in the mixture obtained from the percentages distilling below 105" C.and above 117" C. in the prescribed manner. From the figure thus found must be deducted the 20 C.C. of toluene added, and the difference represents the amount of toluene contained in the 80 C.C. of the original sample taken. lChe percentages of toluene determined in the prescribed manner by the dis- tillation test will include paraffin hydrocarbons of a boiling-point approximating to that of toluene, when these are present in the commercial toluol.The amount of such paraffins is determined in the following manner, and a corresponding reduction made from the percentage of toluene as found by the distillation test: 100 C.C. of the sample is placed in a round-bottomed flask of 150-200 C.C.capacity, fitted with a suitable fractionating column such as a Young 12-bulb '( pear " column, and distilled at the rate of 1 drop per second from the end of the condenser. The fraction distilling between 107-115" C. (corrected) is collected separately, and its specific gravity at 15.5" C. ascertained by any method giving results accurate to the third place of decimals.reduction a t the rate of 9 per cent. is to be made on tho amount of toluene found by For every 0.001 that the specific gravity is found below 0.868,170 ABSTRACTS OF ~ ~ E ~ I C A L PAPERS the distillation test. Thus, if the percentage of toluene found by the distillation test is 70, and the specific gravity of the fraction 107-115" C, obtaiBed in the prescribed manner is 0.864 at 15.5" C., 4 x 9- 3 per cont.of the figure shown by the distillation test is to be deducted. The corrected ~ e r c ~ n t ~ e of toIuene, &llowing for p a r & ~ n content, is therefore found by multipl~ing 70 by 200-3, or 97, and dividing by 100 = 68 per cent, Commercial. toluene should comply with the following test: 90 C.C. of the sample, shaken with 10 C.C.of 90 per cent. sulphuric acid for five minutes, should not give more than a light brosn colour to the wid layer, G. c. J. ~ e t ~ r r n ~ n a t ~ o ~ of the Percentage of Toluene in C ~ ~ ~ e ~ ~ i a l Solvent Naphtha. H. G. Colman. (J. Gas ~ ~ g ~ t ~ ~ , 1915, 129, 314315.)-The method consists in carefully distilling off from the solvent naphtha all constituents boiling below 1 8 8 O C., the distillate containing all the toluene in the sample mixed with all other lower boiling subst&nces present in the solvent naphtha, t~gether with xylene and small quantities of higher boiling substances.The amount of toluene in the distillate, and consequently in the volume of solvent naphtha originally taken, is found by the method already described for ascertaining the percentage of toluene in com~ercial toluol.A 150-200 c,c. r o u n d - b o t t ~ ~ ~ d fiask and conde~ser are washed out with the solvent naphtha to be tested and allowed to drain. One hundred C.C. of the solvent naphtha are poured into the flask and distilled, using a Young 12-bulb c6pear" fraction- ating column, the rate of distillation being 1. drop per second from the end of the condenser. The distillate is collected in a, 100 C.C.cylinder, and the distilIatio~ stopped when the thermo~eter reaches 1 3 8 O C, (correcte~ : see preceding abstract^, and the volume of distillate noted after al~ow~ng the condenser to drain. If the amount of distillate up to 138" C. thus obtained is less than 35 c.c., a further 100 C.C. of the sample is distilled in the same manner, and the distillates are combined.If the total volume of distillate then still falls below 35 c.c., the sample may be taken as ~ractically free from toluene, If the volume of dis~illate is 35 C.C. or over, thea! 35 0.c. thereof are taken, and mixed with 50 c,c. of pure toluene and 15 C.C. of pure benzene, and the mixture distilled as described in the preceding abstract, the distillates (1) below 105' C,, (2) from 105" to 117" C., and (3) the residue above 117" C., being collected and measured, and the specific gravity at 15.5' 0.of the fraction 1 0 5 O to 117O C. also det~rmined. From the percentages boiling below 105' C, and above 117O C., the percentage of toluene in the mixture is found by means of the table reproduced in the preceding abstract, In cases where the original sample of solvent naphtha contains parafins of a, boiling-point approximating to that of toluene, the figure thus found is too high, and a correction must be made in the following ~ & n n e r : For each 0.0~1 that the specific gravity of the fraction 1 0 5 O to 117' C. is below 0.868, thr~e-quar~ers of 1 per cent. is to be deducted of the percentage of toluene found bg the table, the remainder giving the corrected percentage of toluene in the mixture. Of this amount 50 c,c. is derived from the pure toluene added to the mixture, and must be deducted, the remainder giving the number of c.0. of toluene in 35 C.C. of the dist~l~ate. This number multi-INORGANIC ANALYSTS 171 plied by the number of C.C. of distillate obtained from the original volume of solyent naphtha taken and divided by 35, gives the number of C.C. of toluene in the total distillate, and also the number of C.C. of toluene in the volume of solvent naphbha originally taken. Thus, for example, 100 C.C. of solvent naphtha on distillation in the prescribed manner gave 53.5 C.C. of distillate up to 138" C. Thirty-five C.C. of this distillate were mixed with 50 C.C. of pure toluene and 15 C.C. of pure benzene, and on distillation in a 100 C.C. distillation-flask gave the following fractions : Up to 105' C., 24 C.C. 105" to 117' C., 53 C.C. of sp. gr. 0.855 at 15.5' C. Above 117' C., 23 C.C. From the table, the percentage of toluene in this mixture is 62 per cent., but as the specific gravity of the fraction 105' to 117' C. is only 0*855--i.e., 13 points below 0.868-13 x Q= 9.8 per cent. of 62 must be deducted for paraffins. The corrected amount of toluene in the mixture is therefore 62 - 9.8 per cent. of 62 = 62 - 6.1 = 55.9 per cent., and, deducting from this the 50 C.C. of pure toluene added to the mixture, leaves 5-9 C.C. of toluene derived from the 35 C.C. of distillate. The total amount of toluene in the 53.5 C.C. of distillate is therefore 5.9 x 53*5+35=9*3 c.c., and the amount of toluene in the original solvent naphtha, taken to the nearest whole number is 9 per cent. G. C. J.