238 THE ANALYST. ON THE DETERMINATION OF MINERAL OIL I N ROSIN OIL. BY JOHN H. WALKER AND CHARLES D. ROBERTSHAW. (Read at the. Meeting, April 9, 1902.) THE correct determination of mineral oil in rosin oil is still an unsolved problem. We recently had occasion to determine small quantities of mineral oil in rosin oil, and in the course of the examination we made a number of experiments with the methods which have been proposed for the solution of this problem. McIlhiney’s method for estimating rosin oils in vegetable oils consists in ascertaining the bromine substitution figure. This is very low for vegetable oils, but somewhat considerable for rosin oils, the figures given by him for rosin oil varying from 42 to 90. Since the bromine substitution value of mineral oils must be very low, if any, it was thought that the application of this method might lead to good results.The first step was to ascertain whether the method yields constant results, and with this view a genuine rosin oil wa8 examined, with the following results : BROMINE VALUES OF ROSIN OIL. No. 1 ... 112 58.37 - 4.34 No. 2 ... 116-5 59.95 - 3.40 No. 3 ... 105.8 56.56 - 7.32 Bromine Absorption. Substitution. Addition. A second set of experiments led to absorption figures varying from 95.7 to 129.2. Another genuine rosin oil gave values varying from 77.9 to 109.1. This proves that the reaction taking place is not a very definite one, and that the results depend not only on variations of time, temperature, and other conditions, but even in comparative tests and under exactly the same conditions the results remained very unsatisfactory.In fact the numbers obtained above were got inTHE ANALYST. 239 experiments worked under exactly identical conditions, and using ice, as directed by McIlhiney. This is further’emphasized by the very erratic figures obtained when applying the Hub1 or Wijs methods to the oils in question. Similar tests made with two mineral oils led to the following figures : Bromine Absorption. Mineral oil, No. 1 ... 16.8 to 29.2 ,, ,, No. 2 ... 21.9 to 27.4 From the foregoing it would appear a hopeless Bromine Value calculated from Iodine Value. 9.6 11-7 task to determine quantitatively the rosin oil in a mixture of rosin oil with mineral oil. This is shown in several experiments carried out by mixing 90 parts of rosin oil No.1 with 10 per cent. of mineral oil No. 1. The figures obtained were as follows : Bromine Bromine Value calculated Absorption. from Iodine Value. Mixture of 90 per cent. rosin oil No. 1 and 10 per cent. mineral oil No. 1 Recently Holde proposed a method for detecting and determining rosin oil and mineral oils by their different solubility in 96 per cent. alcohol, when 10 C.C. of the sample are shaken with 90 C.C. of 96 per cent, alcohol in a stoppered measuring cylinder, the refractive index of the insoluble portion being then taken. The un- dissolved portion would indicate the presence of mineral oil. This naturally pre- supposes that pure rosin oil is completely soluble, and that may have been the case with the rosin oils examined by Holde, but the samples of pure rosin oils which we tested left a considerable residue at the bottom of the cylinder, In the case of rosin oil No.1 about thirty per cent. was left undissolved, and in the case of rosin oil No. 2 forty per cent. Whether this is due to our rosin oils having been freed from the rosin acids or not has not been investigated, but at any rate we feel justified in concluding that Holde’s method does not admit of general application. The sample of mixed rosin and mineral oils-90 and 10 respectively-gave 60 per cent. of in- soluble substance. The best method of determining the mineral oil in rosin oil was found to be Valenta’s acetic acid method. The percentage of rosin oil dissolved in the acetic acid was, in the two cases of genuine rosin oils, 16.8 and 16.6 per cent. respectively, which compares well with Valenta’s figure of 16.9 per cent.The mineral oil No. 1, used for preparing the mixture, gave 4.4 per cent. (Valenta stating 2.6 to 6.5 per cent.). A mixture prepared from equal parts of the mineral oil and the rosin oil No. 1 gave 10.84 per cent. insoluble, theory requiring 10.59 per cent. AS a preliminary test, the determination of the specific gravity will certainly give very valuable hints. The above experiments were carried out in the laboratory of Dr. J. Lewkowitsch. 70.4 to 98.6 33.8 So far, then, our experience shows that Valenta’s method is the best.240 THE ANALYST. DISCUSSION. Mr. JENKINS said that of the mineral oils, that from Borneo was the one which came nearest in its reactions to rosin oil. Its specific gravity was from 0.97 to 0.99, and it was very readily soluble in an equal volume of acetone. It did not give the Liebermann-Storch reaction, but in several other respects, and especially in solu- bility, it approximated to rosin oil. Dr. LEWHOWITSCH observed that the peculiarities of Borneo oil afforded a further proof of the danger of generalizing from the results yielded by a limited number of samples. If this short paper were merely considered as a, sort of warning, he thought the authors would have done well in drawing attention to the matter.