(4) Estimation of Rosin in Fatty Oils.—To perform this with anything like accuracy, requires a tedious separation, in which the oil is saponified with alcoholic caustic) soda, and the soap is dissolved in water. Dilute sulphuric acid is dropped in until the liquid becomes permanently turbid, and then dilute sodium hydrate is added drop by drop till it just clears again. The whole is then mixed with sand, and evaporated to perfect dryness, to ensure which it is necessary to moisten the apparently dry residue with absolute alcohol, and again dry. The residue is packed into a stoppered percolator, and extracted with a mixture of 5 parts by volume of absolute ether and 1 of absolute alcohol. The solvent is distilled off, and the residue is dissolved in water, and warmed with a slight excess of sulphuric acid, when the rosin separates in viscous drops, which are collected and weighed. These drops still contain a little oleio acid, and, if perfect accuracy be desired, must be dissolved in alcohol, and the solution polarized, which, however, is a process requiring special training and appliances, and is therefore beyond the scope of the present article. Indeed the estimation of a very small percentage of rosin in boiled linseed-oil; for example, is a problem requiring the highest skill and practice, and if under 2 per cent., it is, iu the present state of science, practically impossible.
(5) Estimation of Tallow-oil (free Oleic Acid) in an Ordinary 011.—This point is an important one because oils containing free oleic acid are unsuitable for lubricants. 50 grm. of the oil and 100 c.c. of alcohol are placed in a flask with a few drops of tincture of turmeric, or an alcoholic solution of phenol-phthallein, and well shaken.
A normal volumetric solution of sodium hydrate (40 grm. per litre, each c.c. of which repre sents 0'282 oleic acid) is dropped in until a red colour is produced, and the whole is again shaken. This is repeated until the red is permanent ; the number of c.c. of sodium hydrate used are read off, and multiplied by 0.282, and then by 2, which gives the percentage of free oleic acid.
Before this process is undertaken, a little of the oil should be shaken up with alcohol of 60 o.p., and the alcoholic solution, when clear, mixed with a few drops of alcoholic solution of acetate of lead. If no precipitate be produced, no free fatty acid is present.
(6) Mutual Detection of the various Fatty Oils.—Having discussed all the cases of mixtures with other oils, we now come to the actual identification of the various fatty oils themselves.
The first step is to train the nose to distinguish between certain main groups. To do this, take some oil in a small flat porcelain basin, warm it up to about 142° (300° F.), and observe the smell. Then, as soon as sufficiently cool, rub some into the palm of the band, and again smell. A little practice will thus permit the easy detection and distinction between (1) marine animal oil, (2) terrestrial animal oil, (3) vegetable oil. The odours of these three classes are entirely sui generis,
and it is safe to pronounce on the main question by this test. The marine oils have all the repulsive fishy odour in various degree, the sperm requiring most practice ; the other animal oils have all the peculiar sourish smell of cooking animal fat, soon learned by experience ; the vegetable oils, on the other hand, have a more or less sweetish odour, and practice will even enable most of them to be named.
Case 1. The oil is evidently a marine animal oil. Take the "actual density" at 38° (100° F.) (see p. 1465), and compare with the following table :— To confirm this, take 10 drops of the oil in a porcelain capsule (No. 18, p. 1467), add (from a dropping-tube) 5 drops of barium polysulphide (reagent No. 6), and stir rapidly with a small rod, when sperm-oil will become golden-yellow, and remain so, while the others will be very pale after a few strokes of the rod and setting aside for 5 minutes. Now take a similar quantity of oil, add 5 drops of zinc chloride (reagent No. 7), and stir, when whale and cod-liver will not change, or will only become pale-violet, while seal and cod-fish will be yellow or orange, the former exhibiting brown spots, the latter not. Next, to another similar portion of oil, add 5 drops of sulphuric acid (reagent No. 8), when cod-liver will alone give a violet, the others going brown at once. Then, to another similar portion of oil, add 5 drops stannic chloride (reagent No. 9), when whale-oil will only turn orange-yellow, seal and fish becoming red-brown, and cod-liver violet and then red. Lastly, to another similar portion of oil, add 5 drops mercuric nitrate (reagent No. 11), and after stirring, add a drop or two of sulphuric acid, when seal-oil will effervesce, and give off red fumes.
Case 2. The oil is apparently of terrestrial animal origin. Take, as before, the "actual density," when a pure oil of this class will never vary more than from 0.9050 to 0.9082. Tallow-oil (free oleic acid) is put out of the question in the preliminary examination by alcoholic acetate of lead, and so there can be only :— All the members of this division bleach to a very pale-yellow with barium polysulphide, while lard-oil becomes perfectly white, and gives off a slight smell of sulphuretted hydrogen. They give scarcely any colour with zinc chloride, and become dark reddish-brown with sulphuric acid. A persistent yellow with the polysulphide, a green or brown with zinc chloride, and a greenish tint or too light a brown with sulphuric acid, would indicate impurities of vegetable oil. The only ones which, however, could be mixed without raising the density would be rape, nut, and olive, while sperm would lower the density. Characters of spdeial varieties of neats'-foot-oil will be found in the general tables (pp. 1472-5).