102 TEE ANALYST. ORGANIC ANALYSIS. Estimation of Benzene Vapour in Illuminating Gas. 0. Pfeiffer. ( J . Gasbeleuclzt, 1899, xlii., 697 ; through Clzem. Zeit. Rep., 1899, 333.)-The author suggests a modification of Harbeck and Lunge’s nitration process ( A 4 ~ ~ ~ ~ ~ 1898, xxiii., 101) which makes it far more simple to carry out. The gas is measured and nitrated in a large separating funnel holding 3 or 4 litres; it is filled through the stopper-hole by displacement, and 10 C.C. of the mixed nitrating acids are introduced through the ordinary exit tube.* The acids are made to spread over the internal surface, and are allowed to act for an hour, when the whole of the benzene will be absorbed. The vessel is then rinsed out with 100 C.C. of water, which is brought into a small separating funnel, and neutralised with crystallised sodium hydroxide (40 grammes).It is extracted two or three times for five minutes with quantities of 50 C.C. of ether, each of which is afterwards washed two or three times with 10 C.C. of water to remove some brown substance. The solvent is finally evaporated, the residue taken up in a little anhydrous ether, filtered through freshly calcined sodium carbonate into an evaporating basin, and dried over sulphuric acid--it should neither be warmed nor placed in vacuo. To convert the weight (9) of dinitrobenzene into the percentage by volume of benzene vapour, the author gives the following simplified formula; C is the capacity of the measuring vessel, t the temperature, and b the height of the barometer : 36080 273+t c x g x b .Having a, complete analysis of the gas, and knowing its specific gravity, it is Calling the specific gravity of also possible to calculate the proportion of benzene. * The stopcocks should be L b lubricated ” with strong sulphuric acid.THE ANALYST. 103 the gas S, the specific gravity of the 6 t heavy hydrocarbons" (C,H,) s, and the total volume of the latter ZJ, we have : s = 100s - (CO x 0.9621 + CH, x 0.5530 + H x 0,0692 + N x 0-9701 + CO, x 1.5197--- -- C*,Hll, ' > whence the percentage by volume of benzene vapour becomes : ( s - 0.9674)~ _____ 1.7367 As an example of this method, the author quotes a sample of gas in which the proportion of benzene mas calculated to be 1.57 per cent., and that of ethylene 1.83 per cent. ; by direct analysis the benzene was returned at 1-52 per cent., and the ethylene 1.88 per cent.F. H. L. The Volumetric Estimation of Quinones of the Benzene Series. A. Valeur. BUZZ. SOC. Chim., 1900, xxiii., 58-61.)-This method is based on the reduction of the quinones by hydriodic acid.* When the latter is replaced by an equivalent quantity of hydrochloric acid and potassium iodide, the reaction takes place in accordance with the equation : In the determination a quantity of the pure dry quinone, sufficient to liberate from 0.20 to 0-50 gramme of iodine is dissolved in a little 95 per cent. alcohol. To this solution is added a rapidly-prepared mixture of 20 C.C. of a 10 per cent. solution of potassium iodide with 20 C.C. of concentrated hydrochloric acid previously diluted with an equal volume of 95 per cent.alcohol and cooled. The liberated iodine is titrated with standard thiosulphate, and the corresponding amount of quinone calculated. C,'H,O, + 2HC1+ 2KI = C,H,O, + 2KC1 +I,. The following results show the accuracy of the method : Tolucjuinone, Thymoquinone, C,H,( CH,) (C,H,)O,. C,H,(CH3)0,. I. 11. I. 11. --- -- /-- - ,.--, Weight of substance, gramme 0.2057 0-2707 0.2130 0.1663 Iodine per cent. . . . ... 208.5 207.9 155.4 154-7 (Theory 208.2) (Theory 154.8) Iodine liberated I 9 0.4290 0.5629 0.3311 0.2574 I t is necessary to mix the hydrochloric acid and the potassium iodide, and not to add them separately to the quinone, since the acid would immediately react on the quinone, and the iodide would cause partial oxidation.This method appears to be applicable to the majority of true quinones. I t can be used to determine their solubility in different solvents, which is not easily determined by other methods, on account of the volatility of the quinones and the difficulty of drying them without loss. It can also be employed €or the determination of quinones in somewhat unstable combinations, such as the phenoquinones and quinhydrones. In the case of * Reduction with HI yields a mixture of hydroquinone and quinhydrone ; from what is stated in the paper it appears that with KI and HC1 the reduction is carried o step further, and gives only hydro- quinone.104 THE ANALYST. ordinary quinhydrone the following percentages of iodine were obtained : 116.6,116*3 and 116.1, whilst the calculated percentage for C,H,O;CGHGO, is 116.5.The author therefore concludes that this compound, contrary to the view of Wichelhaus, results from the union of equal molecules of quinone and hydroquinone. C. A. M. The Properties of the Oils of Lemon, Bergamot, and Orange. A. Soldaini and E. Berth. (Boll. chim. farm., 1899, xxxviii., 537; through Chew. Zeit. Rep., 1899, 323.)-Some constants of these three oils in the pure state are given in the annexed table : Lenion. Bergamot . Orange. Specific gravity at 15' ... ... . . . 0.854-0.860 0.882-0.886 0*847-0+353 Rotatory power at 20" (100 millimetres tube) 56-66' 8-20" 96-98" Boiling-point at ordinary pressure . . . ... 171-172" - 173-174" Lemon Oil.-In lemon oil the proportion of citral should not be below 6.5 per cent.; and when 20 grammes are fractioned at a pressure of 20 to 30 miilimetres, the first 10 C.C. of the distillate should have as high a rotatory power as the original oil. The presence of orange oil is shown (a) by a yellow colour when a drop of the sample is mixed with 15 or 20 drops of brominated chloreform, ( b ) by a yellow flocculent precipitate instead of a white crystalline one when sodium bisulphite solution is added. To determine the percentage of citral, a 5 C.C. pipette graduated in fortieths is required, and also a pear-shaped flask, the neck of which has the same diameter as the pipette, and which has a lateral tube bent upwards at a right angle to carry a funnel. Five C.C. of the sample are run from the pipette into the flask, 25 C.C. of a solution of potassium bisulphite containing an excess of sulphur dioxide are introduced, the lower end of the pipette is connected to the neck, the whole is shaken and warmed for twenty minutes on the water-bath, then cooled and warmed again for five minutes.When finally cold, the volume of oil still remaining liquid is read off in the pipette by adding water through the side funnel ; and the difference between it and the 5 C.C. taken represents the volume of the aldehyde. Bergamot Oil.-The proportion of linalyl acetate usually varies between 21 and 22 per cent. When 15 C.C. are fractioned at the above pressure, the first 5 C.C. of the distillate should have a rotatory power 2$ times as great as the oil itself; and the next 9.5 C.C. should be almost inactive. Evaporated on the water-bath, the residue should be between 5 and 6 per cent.The oil should be soluble in ij vol. of 90 per cent. alcohol, and the clear solution shouldnot be rendered turbid on dilution. SchiiY's reagent for aldehydes should fail, or at most give a faint tint in half an hour ; an immediate colour, or a strong red in half an hour, indicates lemon oil. To estimate the linalyl acetate, 1.5 grammes are saponified with an excess of seminormal alcoholic potash, diluted with a little 80 per cent. spirit, and titrated with semi- normal sulphuric acid and phenolphthalein ; the volume of alkali multiplied by 0.09775 (mol: wt. of the ester, 195.5) gives the acetate. OrcLnge Oil.-When 20 C.C. are fractioned at a pressure of 10 or 20 rnillimetres, the distillate should have a rotatory power from I" to 3" higher than the original. Schiffs reagent should give no colour.F. H. L.THE ANALYST. 105 Determination of the Solidifying-point of Fatty Acids. I. Freundlich. (Chem. Zeit , 1899, xxiii., 1014.)-The ordinary Dalican process for determining the exact solidifying-point of fatty acids is not quite accurate, for the temperature to which the thermometer finally rises is partly dependent on that to which it was made to fall during the stirring of the fat. The following modified way of carrying it out leads to absolutely concordant results, or at the worst to differences of 0.05 to 0.1" C. At the lowest temperature above the expected solidifying-point at which the sample round the thermometer bulb remains perfectly liquid, the thermometer is moved quickly two or three times backwards and forwards through the fat, and the mercury is observed ; if it falls sharply, the operation is repeated until the column remains constant for thirty or forty seconds; then the sample is stirred fifteen or twenty-five times, and if the mercury falls during the agitation and risee immediately afterwards to a maximum s t which it stands unchanged for three or five minutes, that maximum is the true solidifying-point.The great thing to avoid is too prolonged stirring. F. H. L. The Becchi and Halphen Colour Reactions for Cotton Oil. P. N. Raikow and N. Tscherweniwanow. (Chem. Zed., 1899, xxiii., 1025.)-At the present time there exist some ten different modifications of the Becchi test for cotton oil ; and statements as to its utility and the best proportions for its several ingredients vary enormously.Benedikt and other authorities question whether cotton oils do not occur which cause no reduction of the silver nitrate at all; this, however, seems problematical, unless the samples have been specially treated in order to prevent their giving the Becchi reaction. I t is not possible to render cotton oil indifferent to the Becchi test by blowing air through it in the cold; nor can the same object be attained by repeated extractions with alcohol, the latter fact being in contradiction to Benedikt's assertion that the true cause of the reaction is an aldehyde-like body which is readily soluble in spirit. Even the fatty acids of cotton oil after alcoholic saponification and liberation from their barium salts, or after aqueous saponification and liberation from the sodium salts, give the test, while the small proportion of unsaponifiable matter in the original oil does not give it.Treated with ordinary steam even for long periods of time, cotton oil still retains its usual properties ; but superheated steam, or a simple heating of the oil to between 210" and 220°, quickly destroys its reducing power, which also vanishes more slowly at 150", the residual material being almost unaltered, except that its colour is slightly darkened, and that it possesses a faint burnt odour. Precisely the same remarks apply to the Halphen test : one hour at 150" hardly affects the reaction, five hours reduce its intensity to one-half, ten hours to one-third ; in time, probably, it would fail altogether.[Cf. Holde and Pelgry, ANALYST, 1899, xxiv., 214.1 After elaborate experiments, the authors find that the details of the Becchi test are most important : the proportion of silver nitrate to the oil, the proportion of free acid to the silver, ought to be kept uniform ; and they decide that the method adopted by the Italian Commission (cj. ANALYST, 1895, xx., 222, last paragraph) should be taken as the standard. The latest modification of the Becchi test described by106 THE ANALYST. Tortelli and Ruggeri (ANALYST, 1898, xxiii., 179) was not examined; in comparison with the more certain and very delicate Halphen reaction, it is too complicated to be generally useful. I t appears to be universally admitted that the Halphen reaction (ANALYST, 1897, xxii., 326 ; 1898, xxiii., 131) is better than Becchi’s ; the only points of divergence are its degree of delicacy and the best method of carrying it out.The actual temperature employed is only a matter of convenience; a cherry-red colour is pro- duced in the cold by bright sunlight in three hours. The part played by the amylic alcohol is obscure, but it does not serve simply to retain the carbon disulphide in the hot liquid; it should, therefore, not be omitted. A faint Halphen red is manifested when the free sulphur is left out of the regular mixture; an excess of sulphur is use- less, and in the testing of oils containing only a little cotton oil, it tends to decrease the delicacy of the reaction. The present authors accordingly reject Soltsien’s proposals ; there is no objection to the use of plain water instead of the brine bath, but in other respects Halphen’s prescription should be retained in its integrity.The behaviour of olive, been studied under bath nor in sunlight oil can be readily sensitiveness. walnut, linseed, poppy, and arachis oils towards the test has various conditions ; neither by prolonged heating on the water- did they respond. I n pale-coloured oils 0.5 per cent. of cotton detected, and this proportion may be taken as its limit of F. H. L. The Determination of the Bromine Absorption of Oils. P. C. McIlhiney. (Journ. Amer. Chem. SOC., 1899, xxi., 1084-1089.)-1n a former communication (ANALYST, xix., 141) the author described a process for the determination of this con- stant in which a distinction was made between the bromine addition and bromine substitution values.He has now simplified his process by making use of the iodometric method of Schweitzer and Lungwitz (Journ. SOC. Chem. Ind, 1895, 130), and shortening the time in accordance with the fact that the addition of bromine to fats is practically instantaneous (ANALYST, xx., 146). I n the modified process a weighed quantity of the oil is dissolved in 10 C.C. of carbon tetrachloride in a stoppered bottle and 20 C.C. of one-third normal bromine in carbon tetrachloride added. Simultaneously a blank determination is made, and subsequently titrated with standard thiosulphate to determine the strength of the bromine solution. After the lapse of one or two minutes, 20 to 30 C.C.of a 10 per cent. solution of potassium iodide are introduced, the bottle shaken to insure the absorption of the bromine and hydrobromic acid, and the iodine titrated with deci- normal t hiosulphat e. After the titration, 5 C.C. of a neutral 2 per cent. solution of potassium iodate are introduced, and the amount of iodine liberated, which is equivaleut to the hydro- bromic acid formed, is titrated, and gives the bromine substitution figure. In order to prevent a loss of bromine or hydrobromic acid on removing the stopper to introduce the potassium iodide, a piece of wide indiarubber tubing is slipped over the neck of the bottle, thus forming a well round the stopper. TheTHE ANALYST. 107 potassium iodide is poured into this well and the stopper slightly opened, preferably after the bottle has been cooled in ice to create a partial vacuum in the interior.The following figures were thus obtained with various representative oils. The difference between the figures given in the last column and 1.000 are intended to represent the degree of substitution which occurred in the determination of the Hub1 value : I HUN l Value. - _ - - Raw linseed oil, several years old ,? ,, ,, average of seven samples ... ... ... Boiled linseed oil, average of eight samples ... .., ... Third run rosin oil ... ... ‘‘ Java” boiled rosin oil ... ... Menhaden oil, average of three samoles ... .._ . _ . 157.3 183.8 - 63.9 73.3 174.9 Maize oit average of three samples- I - Cotton-seed oil ”.. ... - Turpentine ... ... ... ... - Ceylon cocoanut oil ... - Tallow rendered in laboratory ... - Hard paraffin ... ... Black rosin ... ... ... -- ... ... ... - Bromine I I I 99.1 99.2 115.7 1 112.0 46.2 ~ 101.9 110.2 I 110.6 - 75.8 - I 65.8 - 1 266.1 - 5-36 - 1 24.0 - , 3.55 I 135.4 92.0 106.6 103.0 7.7 8.3 95.6 72.9 3.6 2.7 3.2 42.3 46.8 7.5 1-5 62.2 1.8 166.1 50.0 4.7 0.33 21.481 1.26 1 -06 65.0 Calcu- lated Bromine Value divided Bromine Addition Valne. 1.077 1.083 bY ___-- - 5.231 5-685 1.154 - - - - - - - The advantages claimed for this method are that the bromine solution is readily pre- pared and does not change on keeping, that it is exceedingly rapid, and that it dis- tinguishes between the halogen absorbed by addition and by substitution. C. A. M.