174 ABSTRACTS OF CHEMICAL PAPERS ORGANIC ANALYSIS. Analysis of Commercial Benzols. P. E. Spielmann and E. G. Wheeler. ( J . SOC. Chem. Ind., 1916, 35, 396-399.)--The following method is described for estimating the percentages of benzene, toluene, paraffin, and carbon bisulphide, in a properly washed commercial beneol : Estimation of ToZuene.-One hundred C.C. of the sample are distilled from an Engler flask at the rate of 7 C.C.per minute, proper precautions being taken with FIG. 1. regard to draughts and the efficient condensation of the distillate; the volume in C.C. passing over up to 90" C. indicates the quantity of toluene present when reference is made to a previously constructed curve. Paraffin and carbon bisulphide, in the quantities occurring normally in commercial benzols, do not interfere.The method is reliable up to a toluene content of 20 per cent. ; above this quantity the methods described by Colman (ANALYST, 1915, 40, 166, 170) and Northall-Laurie (ibid., 1915, 40, 384) are the more accurate. Estimation of Carbon BisuZphide.-The benzol is treated with alcoholic potassium hydroxide solution, washed to remove the xanthate, alcohol, etc., and dried ; its sp.gr.ORGANIC ANALYSIS 175 is then determined. The difference between the sp. gr. of the original sample and the sp. gr: thus obtained is proportional to the quantity of carbon bisulphide present, and can be found by reference to a curve (see Fig. 1). Estimation of Parafin.-After removal of the carbon bisulphide, the mixture consists essentially of benzene, toluene, and paraffin.The quantity of toluene has Specific gravity at 15'5" C. bcfore removing Carbon Bisulphide. Volume in c.@. distilling below 90* C. FIG. 2. already been found, and the sp. gr. of a mixture of benzene and this quantity of toluene can be found from a previoudy constructed curve, so that the difference between the sp. gr. of the benzene-toluene mixture and that of the sample after removal of the carbon bisulphide is proportional to the quantity of paraffin present, the actual value of which is found from the curve given,176 ABSTRACTS OF CHEMICAL PAPERS Estzmttion of Benzene.-This constituent is taken by difference.Some benzols contain thiophene, but this is rarely present to the extent of more than 0.5 per cent., a quantity which would cause an error of 0.3 per cent.in the amount of paraffin found; as this is within the limits of accuracy of the method, it can be neglected. The average limit of accuracy of the method is-toluene, 1 per cent. ; paraffin, 0.5 per cent. ; carbon bisulphide, 0.1 per cent. I n Fig. 1 are collected the curves showing the effects on the sp. gr. of benzene, toluene, carbon bisulphide, and paraffin, together with the toluene distillation curve.For thg sake of convenience in working, the curves shown in Fig. 2 have been con- structed. Here it is not even necessary to determine the various differences in sp. gr. ; only the observed figures need be dealt with. The analytical results are interpreted as follows : The number of C.C. distilling below 90" C.represents the percentage of toluene. This number of C.C. has been plotted against the sp. gr. found after removal of the carbon bisulphide in such a manner that the percentage of paraffin is read off directly. The sp. gr. of the sample has been plotted against the sp. gr. after removal of the carbon bisulphide, whereby the percentage quantity of the latter can be read off.w. P. s. Optical Dispersion of Chinese Wood Oil as an Index of Purity. E. E. Ware. (J. I n d . and Eng. Chem., 1916, 8, 126-128.)-The method of examina- tion used by Brier (ANALYST, 1916, 13) may be simpiified by making a direct reading of the dispersion of the oil, which for this purpose is placed in a hollow prism in a spectrometer, with a brass electrode arc lamp as the light.An estimation is made of the relative dispersion for two lines given by this light in the red and the blue part of the spectrum. The angular difference for Chinese wood oil contained in a, prism with an angle of 20" 1' was found to be 26' 54", while for soya bean oil at the same temperature it was 12' 54", or, expressed in vernier units of 6" each, the rertdings were 269 and 129 respectively.The dispersion of Chinese wood oil adulterated with varying proportions of soya bean oil may be expressed as a straight line, when percentage adulteration and angular dispersion are plotted respectively aB co-ordinates. I t is suggested that by fitting the spectrometer with a constant temperature device adulteration of Chinese wood oil could be rapidly detected by the variations in the angular readings.The correction for temperature is about one second per 1" C., which corresponds to about 0.1 per cent. of adulteration. C. A., M. Heat of Bromination of Fats and Oils. J. W. Marden. (J. I n d . and Eng. Chem., 1916, 8, 121-126.)-The calorimeter used for measuring the heat of bromina- tion consisted of a large Dewar vacuum-tube containing a special mixing chamber of copper, coated inside with paraffin wax, and fitted with a glass cover which rested on a flange about halfway up, so as to divide the chamber into two compartments.In the lower of these was placed the oil, while a solution of bromine in three times its weight of carbon tetrachloride was placed in the upper chamber. The liquids were brought into contact by breaking the cover-glass by means of a paddle-stirrer within the ohamber, and mixture was effected by rotating the mixing chamber.AORGANIC ANALYSIS 177 smaller calorimeter, consisting of a Dewar tube (3 x 20 cm.) containing a thin glass tube (0.8 x 23 cm.), with copper wire stirrers in both tubes, was used for measuring the specific heat of brominated oils, and both calorimeters were standardised by determining therein the heat of dilution of standard sulphuric acid. The calories per grm.of oil were calculated by adding the heat capacity of the apparatus to that of the mixture of bromine solution and oil, multiplying by the total rise of tempera- ture, and dividing the result by the grms. of oil used. The following table gives results thus obtained, together with the relationship of the values to the iodine valuee : Oil or Fat.Sandalwood . . . ... Raw linseed ... Boiled linseed ... Chinese wood ... Maize ... ... ... Sesame ... ... Rapeseed ,.. ... Cot tonseed . . . ... Sperm.. . ... Castor (purified) . . . ,, (commercial) Arachis ... ... Olive, I. ... ... 7, 11. ... ... ,) 111. ... ... Neat's foot . , . ... Lard ... ... ..Mineral ... .", Heat of Bromination. Cals. per Grm. 274.0 206.0 204.6 150.0 146.2 126.4 120.8 117.0 109.0 104.1 102.2 102.2 100-7 96.6 95.4 83.0 79.3 nil Iodine Value found. 270.0 173.5 169.0 156.0 123.1 108-2 105.8 101.7 93.4 88.8 87.5 83 *2 84.0 80 -0 80-3 68.7 68.6 - Iodine Value calculated. Factor, 0'846. 232.0 174.2 173.0 127.0 1238 107.0 102.2 99.0 92.2 88.1 86.5 86.5 85.2 81.7 80.7 70.2 67.2 - Deviation per Cent. - 14.1 0.98 2.36 0.57 - 1.12 - 3.4 - 2.6 - 1.28 - 18.6 - 0.79 - 1.14 3.96 1.43 2.12 0.5 2.2 - 2-04 - The method cannot be used for testing oil of turpentine, owing to the solubility of the paraffin wax in that liquid.The cases where there was marked deviation between the calculated and determined iodine values was attributed to there being a greater amount of substitution of the bromine (see Hehner and Mitchell, ANALYST, 1895, 20, 148).C. A. M. Oils of the Conifers. V. The Leaf and Twig and Bark Oils of Incense Cedar. A. W. Schorger. ( J . I n d . and Erzg. Chem., 1916, 8,22-24.)-1ncense cedar (Libocedrus decurrens) is largelyrestricted in its range to the State of California. The leaves and twigs, if stored for a few weeks, show a slight increase in the yield of oil as compared with the fresh material.The majority of leaf and twig oils were pale greenish-yellow in colour, a few being dark greenish-brown ; their properties were: Specific gravity at 15" C., 0.8655 to 04766. Refractive index, at 15"C., 1.4754 to 1.4778. Optical rotation, [ a ] ~ 2oOc., -3.2 to + 38.7. Acid number, 0.48178 ABSTRACTS OF CHEMICAL PAPERS to 1.3.Ester number after acetylation, 28.6 to 46.2. Acetate per cent., 6.47 to 9.74. Free alcohols per cent., 2-26 to 5-14. Total alcohols per cent., 8.28 to 12.8. The bark oil had the following properties : Colour, faint greenish-yellow. Specific gravity at 15" C., 0.8621. Refractive index at 15" C., 1.4716. [a]D20°c. + 1.1. Acid number, 0.6.Ester number, 3.22. Ester number after acetylation, 9.53. The following table shows the approximate percentage composition : Ester number, 18.5 to 27.82. 1 Leaf and Twig Oil. I Bark Oil. ... ... ... Furfural ... la-Pinene ... ... ... ... d- Sylvestrene d-Limonene ... ... ... ... ... ... ... ... 1 Dipentene 1 Bornyl acetate ... ... ... ... Free borneol ... ... ... ... '' Librocedrene " ...... ... -.. '' Green oil " ... ... ... ... trace 12-16 54-58 - 8 4 6-7 a trace 75-85 __ 5-6 1 2 3 H. F. E. H. Comparative Actions of Cane Sugar and Invert Sugar on Alkaline Copper Solutions. L. Maquenne. (Comptes rend., 1916, 162, 145-149.)-The experiments were carried out with 10 C.C. each of the copper and alkaline tartrate solutions, mixed and made up with sugar to a total volume of 36.2 c.c.; the flasks were placed cold in water at the given temperatures, the level of the water being 1 to 2 em.above that of the liquid in the flasks, and kept immersed for stated times. In the case of invert sugar, for a period of ten minutes, the increase in cupric reduc- tion with increasing temperatures is small and regular from 70" C. upwards, and the influence of variations of 2" or 3" C.is practically negligible. At 65" C. there is a, larger difference, and it is preferable either to work at 70" C. or to warm the contents 3f the flask first to 40" C., or else to keep it in the bath for fifteen minutes instead of ten. In the case of cane sugar the influence of increasing temperatures is relatively very great, and, in order to avoid as far as possible the disturbing effect of cane sugar isa the estimation of reducing sugars, it is desirable to work at comparatively low temperatures, preferably at 70" C., at which point the reducing power of cane sugar is 8,000 to 10,000 times less than that of invert sugar, and its influence only one-tenth of what it is at 100' C.Under the conditions noted, the presence of 10 grms.of cane sugar will produce an error of about 1 mgrm. in the estimation of invert sugar. An error of this magnitude may escape detection in methods based on the filtration of the cuprous oxide, but is easily detected by the thiosulphate method. The time of heating has but little influence in the case of invert sugar, except that when working at 65" C. (measured outside) ten minutes is barely sufficient to attain regu- larity of results.In the case of cane sugar, on the other hand, bhe reduetior,ORGANIC ANALYSIS 179 increases largely with the time of heating; and at moderate temperatures with quantities of sugar not exceeding 15 grms., the increase in reducing power is nearly proportional to the time. Hence any sample of sugar, tested at 65" C.under the conditions noted, which shows in twent,y minutes a value sensibly less than double that found for ten minutes' heating, contains invert sugar, and this fact may be utilised for the estimation of very small quantities of invert sugar; for instance, 0.5 mgrm. of invert sugar in 10 grms. of cane sugar is clearly shown. At tempera- tures above 65" C, this proportionality is not so marked.At 100" C. the reduction decreases with increase in the quantity of sugar, whereas at 65' and 70" C. it increases; consequently at some temperature near 100" C. there must be a point at which the reduction is independent of the quantity of sucrose present. The extreme sensitiveness of the reducing power of cane sugar to small changes of temperature may account for variations and errors in results obtained when working at the boiling-point, owing t o the fluctuations of barometric pressure. J.F. B. Estimation of Reducing Sugars in Presence of an Excess of Cane Sugar. L. Maquenne. (Comptes rend., 1916, 162, 207-213.)-Owing to the appreciable and extremely sensitive cupric reducing power of cane sugar and its influence on the reducing power of the invert sugar itself, certain very elaborate precautions are necessary for the accurate estimation of invert sugar, etc., in presence of an excess of cane sugar.As a general rule, in order to restrict the influence of the cane sugar, the reduction is carried out at a relatively low temperature ; the lower the tempera- ture, the longer the time. The absolute minimum is fixed by the author at ten minutes at 65" C.(measured in the water-bath), but even under these conditions the influence of the cane sugar must be determined and accounted for. When working a t ebullition (three minutes boiling) the variations of temperature due to barometric pressure may introduce an error up to 0.01 per cent. ; this may be compensated by working always against a control experiment made at the same time with pure cane sugar or a sugar of known composition.The ebullition method, in spite of its defects, possesses the advantage that, when an approximation of 1 in 10,000 is sufhient, the same tables map be used as those showing pure invert sugar in absence of cane sugar, since when 10 grms. of sugar are taken at a total volume of 36.2 C.C. the reduction due to the cane sugar is nearly independent of the amount of invert sugar present.I n working against a control experiment with pure cane sugar, it is advisable in the analysis of products pour in invert sugar to employ constant quan- tities--e.g., 20 grms. of cane sugar for each determination at 65" C., or 10 grms. at ebullition-constructing tables for these quantities and making up with pure sugar to the standard quantities whenever it is not convenient to take so much of the original sample.When working at 65' C., or any other temperature which is independent of barometric pressure, a control experiment is not required, and the reduction due to 20 grms. or other standard weight of cane sugar chosen is simply deducted in con- structing the tables.Two analytical methods are available : When the amount of cuprous oxide is large, it may be separated by centrifuging or filtration, and esti- mated by any known accurate method. When the quantity is small, however, an appreciable proportion remains in solution, so that for refined products, poor in invert180 ABSTRACTS OF CHEMICAL PAPERS sugar, it is essential to proceed by estimating the excess of cupric salt after reduc- tion.For this purpose the author proceeds as follows: Ten C.C. each of copper solution, alkaline tartrate, and water, are mixed, and 10 or 20 grms. of solid sugar dissolved in the mixture; the solution is heated for the prescribed time, cooled, treated with 20 C.C. of 50 per cent. sulphuric acid, 5 C.C. of a 20 per cent. solution of potassium iodide are added, and the titration is made with standardised thiosulphate (20 grms.per litre), using starch paper as an external spot test indicator. The method is not applicable to molasses which cmtain bodies which combine with iodine. If for any reason the sugar must be dissolved in advance, great care must be taken to avoid the action of the invertase present in many samples of sugar.Especially when working at moderate temperatures, the solution should be cooled and acidified at once to avoid continuation of the reducing action. A11 the known precautions as to size of vessels, manner of heating, etc., must be closely followed, temperatures must be regulated to 0-5" C., and times to within five seconds ; the method being purely empirical and personal, every operator must work with tables constructed by himself.Sugar free from invert sugar may be prepared by three recrystallisations of refined sugar from 75 per cent. aqueous solutions at 90" C. ; alcoholic solutions should be avoided. The degree of approximation obtainable when the method is carefully fol- lowed is about one-fiftieth of the quantity of invert sugar present-ie., under favour- able conditions, two significant figures exact and the third fairly close, except with less than 1 per cent.of invert sugar, when experimental errors may increase the divergence of the third decimal, J. F. B. Presence in Industrial Sugars of Reducing Sugars other than Invert Sugar. L. Maquenne. (Comptes rend., 1916, 162, 277-282.)-The methods now available for the accurate estimation of reducing sugars in presence of a large excess of cane sugar permit closer attention to be paid to those unknown reducing sugars, whether pre-existing in the crude material or formed in the processes of manufacture, traces of which occur even in many refined products.The presence of these reducing sugars, other than invert sugar, is recognised in a general manner by the variation of their cupric reducing power according to the temperature at which the estimation is made.In this respect they resemble cane sugar rather than invert sugar, in that the cupric reducing power increases considerably at ebullition (three minutes at 103' to 104' C. under the conditions prescribed) as compared with that determined at 65" C.(ten minutes' heating in water-bath), whereas the reducing power of invert sugar shows only a, slight increase between 65" C. and the boiling-point. Consequently, in the estimation by the author's method of the cupric reducing powers of industrial sugars, by comparison with pure cane sugar, any appreciable difference between the reduction at 65" C. and that at ebullition may be attributed to the unknown reducing sugars. I t is true that their influence at 65" C. is probably not absolutely nil, and the value attributed to the invert sugar is to that extent over-estimated, but for an approximate and relative measure of the unknown reducing sugars the assumption forms the basis of a useful method. The observations here recorded have two practical consequences : (1) I n the examination of the purity of refinery products it s not sufficient merely to determine the invert sugar by cupric reduction at 65" C. ;INORGANIC ANALYSIS 181 the total non-cane sugar reducing sugars only exert their full influence at ebullition, and the ebullition test possesses the greatest practical significance. (2) In the examina- tion of sugar products for distillery value the reverse is the case. The unknown reducing sugars are unfermentable, and their estimation leads to an erroneous valuation; the cane sugar and invert sugar alone are of practical significance, and the test at 65" C. should be employed. J. F. B.