(7) Estimation of the Glycerine formed by Saponification.—This is not generally necessary, except for scientific purposes, and it is customary to deduct the percentage of total fatty acids found from 100, and consider the difference as glyceryl. Up till the beginning of this year (1881), no ready process for the estimation of glycerine had been proposed ; but since then, Dr. Muter has published the preliminary notice of a process which is likely to give good results. The author takes advantage of the power of glycerine in arresting the precipitation of cupric hydrate from cupric sulphate by potassium hydrate. He takes a definite quantity of the solution of glycerine in one of the oleine tubes already described (Fig. 1045), and to it, he adds an excess of potassium hydrate, and drops in a solutioo of cupric sulphate with constant shaking, until a permanent precipitate is produced. The whole is then made up to a definite bulk, and, when settled, a portion of the blue liquid is run off through the stopcock ; the amount of dissolved copper is estimated by neutralizing with nitric acid, then adding excess of ammonium hydrate, and running in A volumetric solution of potassium cyanide till decolorized. By doing this on a solution of glycerine of known strength, the value of the cyanide in glycerine is ascertained. Those interested will find details in the Analyst for March 1881.
(8) Testing a Solid Fat for Paraffin-wax and Mineral Oils.—The mixture of fats with solid mineral hydrocarbons has become of late years quite an acknowledged custom. Therefore, any fat should always be submitted to the " actual density" process described for butter, on p. 1465. Ordinary fats have an "actual density" at 38° (100° F.) rarely below 0'9032, while solid paraffin-wax treated in the same way shows nothing above 0.8810. If, therefore, an apparent fat shows less than the latter figure, it is probably all mineral ; but if somewhere between the two, it is most likely a mixture, and must be treated in the same manner as hereinafter described for the detection and estimation of heavy mineral oil in ordinary fatty oils. The test for paraffin-wax in oils is, as here after shown, so simple, that it is advisable always to apply it, uuless the fat shows a density so high as to positively preclude its presence.
B. InEwmtmovriorr OF 01LS IN Mtxmin.—The next anbject claiming attention is the identifi cation and testing of oils, especially when mixed, a point of the greatest difficulty, and one which eminently requires experience. It does not, like the subject just finished, rest on a definite chemical basis ; and although many processes have been from time to time advocated, none has really stood the test of repetition by other hands. The peculiarity of oils is that one analyst may have methods which may and do give fair results in his own hands, but which, repeated by others without his special experience, become not only inaccurate, but positively misleading. It would be quite possible to make an apparently valuable résumé of the subject, by giving all the processes above referred to, and still to leave the reader really no nearer his desired object than at the commencement. The
aim of the present article, however, is to avoid this beaten track, and to omit everything but the few definite points which, intelligently followed up and backed by practice, may lead to the fairest deduction possible in the present state of science. The first essential in setting about the study of oils is the possession of a set of really genuine standard samples ; this is very difficult to procure, as the oil-trade is so permeated by the principle of admixture, that the refiners have too often good reason to shun any attempts to render its detection more easy. To test the real state of matters in this respect, the writer once applied to seven leading houses to assist him with standards; only one came forward in response. Nothing, however, can be done without standards of, at least, the following kinds :— Many other oils exist, but the foregoing represent such as are commonly met with in commerce. If, however, the reactions of others should be desired in any special case, it will be easy to follow out the methods hereinafter given, and fix one's own reactions, if not found in the present article, which must necessarily be limited.
The following reagents and special articles are required :—(1) Pure carbon bisulphide ; (2) petroleum-spirit, rectified under 88° (190° F.); (3) alcoholic caustic soda, made by dissolving 115 grm. of sodium hydrate in 1 litre of redistilled methylated spirit ; (4) sodium bicarbonate, pure ; (5) silver sand, well washed, dried, and ignited ; (6) barium polysulphide, made by dissolving barium oxide in boiling distilled water, cooling, pouring off the mother-liquor from the crystals, boiling it with excess of sulphur, and filtering ; (7) syrupy zinc chloride, made by saturating hydro chloric acid with pure zinc oxide, and evaporating till the liquid assumes the consistence of golden syrup ; (8) sulphuric acid, sp. gr. P843, which has been shaken up with a little mercury occasion ally during some hours ; (9) stannic chloride, the fuming perchloride of tin of commerce ; (10) syrupy phosphoric acid, ordinary phosphoric acid evaporated to a sp. gr. of 1.72; (11) mercuric nitrate, made by dissolving mercury to saturation in cold nitric acid, and then boiling for ten minutes with as much more nitric acid ; (12) absolute sulphuric, acid, the strongest acid, which has been recently heated for some time to 316° (600° F.), then boiled, and quickly secured in an air tight vessel ; (13) amylic alcohol, sp. gr. 0.818; (14) a sp. gr.-bottle, fitted with a thermometer stopper from 60° to 120° F. ; (15) a long delicate thermometer, graduated in single degrees from 0° to 149° (32°400° F.); (16) some long test-glasses of 1 in. diameter, capable of holding 150 c.c., and made to stand heat ; (17) some dropping-tubes, delivering slowly drops of water weighing gr.; (18) some white porcelain capsules, semicircular in form, and 2 in. diameter, without spouts ; (19) some small glass rods, 3 in. long.