(262° C); (11) mineral oils alone are not suited for the heaviest machinery, on account of want of body, and higher degree of inflammability; (12) well-purified animal oils are applicable to very heavy machinery; (13) olive-oil is foremost among vegetable oils, as it can be purified without the aid of mineral acids; (14) the other vegetable oils admissible, but far inferior, stated in their order of merit, are gingelly-, ground-nut-, colza-, and cottonseed- oils; (15) no oil is admissible which has been purified by means of mineral acids.

As the result of numerous experiments, Veitch Wilson is convinced that mineral oils are, if used alone, unsatisfactory lubricants; but bearing in mind the natural and almost ineradicable tendency of mineral oils to develop acid, and of vegetable oils by the absorption of oxygen to gum and clog the bearings and to induce spontaneous combustion, bearing also in mind the fact that mineral oils can now be obtained in every respect as safe as the finest animal oils, and that the admixture of mineral oil with animal or vegetable oil neutralizes the acidity in the one case and the acidity and oxidizing tendency in the other, he is of opinion that the safest, most efficient, and most economical lubricants for all manner of bearings are to be produced from a judicious mixture of animal or vegetable with good mineral oils. With regard to cylinder lubrication, the peculiar conditions are the liberation of natural acids from vegetable and animal fats and oils by the action of steam and heat, the action of these acids on the cylinders, and the evidence that as these acids are constituents of all animal and vegetable fats and oils, they cannot be removed by any process of refining.

One of the lubricants largely in use is tallow, but there is conclusive evidence that it is the cause of considerable injury to the engine cylinders. From the mass of evidence he has been able to collect upon the subject, he is convinced that, if care was exercised in the selection of the oil, and equal care in its preparation and application, hydro-carbon oil would be found thoroughly efficient as a cylinder lubricant, absolutely harmless, and much more economical than tallow. The bulk of the cylinder oils now before the public are of American origin: they are usually sold pure, but sometimes a small percentage of animal or vegetable matter is added in order to increase their lubricating properties, and in his experience this has always been attended with most favourable results. The thickest oil that can be introduced to a cylinder is the best. Hot-air engines may be lubricated on precisely the same principles as steam cylinders, but gas engines present a new and special feature, as in their case the lubricant is not only subjected temporarily to the intense heat of the explosion, but also comes into direct contact with the flame, and is liable to be decomposed or carbonized thereby.

If, therefore, animal or vegetable fats and oils are objectionable in steam cylinders, they are much more so in the cylinders of gas. engines; and in the case of gas engines he would most emphatically protest against the use of any but pure hydro-carbon oils without any admixture. {Colliery Guardian.)

The results of W. H. Watson's ex periments upon the corrosive action of various oils on copper and iron surfaces are worthy of reproduction here. After 10 days' exposure of copper to the action of the several oils named below, the effects were evidenced by the following quantities of copper held by them:-Linseed-oil, 0' 3000 gr.; olive-oil, 0 . 2200 gr.; neatsfoot-oil, 0.1100 gr.; almond-oil, 0.1030 gr.; seal-oil, 0.0485 gr.; colza-oil, 0.0170 gr.; sperm-oil, 0. 0030 gr.; paraffin, 0.0015 gr. Iron subjected to similar treatment for 24 days was affected to the following extent:-Neatsfoot-oil, 0.0875 gr.; colza-oil, 0-0800 gr.; sperm-oil, 0.0460 gr.; lard-oil, 0.0250 gr.; olive-oil, 0.0062 gr.; linseed-oil, 0.0050 gr.; seal-oil, 0.0050 gr.; castor-oil, 0.0048 gr.; paraffin, 0. 0045 gr.; almond-oil, 0 • 0040 gr.; special lubricating-oil, 0.0018 .gr. These results show that the extent of the action of any oil on one metal is no guide to the degree in which it will affect another metal.

C. W. Volney has published the following results of an investigation of the effects of different oils upon brass. Strips of sheet brass were covered, each separately, with oil. The temperature was 81° F. (27° C). The strips of metal were weighed; the temperature was kept uniformly at 81° F.; after 16 days, the metal was removed from the oil and carefully washed with alcohol, dried and weighed. (1) Menhaden-oil: weight of metal, 0.590. The oil had become thick, gummy, and covered with a tough skin. After cleaning and drying, the metal weighed 0.587; loss, 0.003. The metal itself was covered with a green film; the colour of the oil was unchanged. (2) Crude cottonseed-oil : weight of metal when immersed, 0.574. The oil had retained its original consistency. The metal was covered with a green film; the colour of the oil was unchanged. Weight of metal after washing and cleaning, 0.572; loss, 0.002. (3) Lard-oil: weight of metal when immersed, 0.572; the oil showed no change of consistency or colour; there was only a slight tinge of green on the metal, which weighed, after washing and cleaning, 0.5715; loss, not quite 0.001. (4) Olive-oil: weight of metal before immersion, 0.794. The oil was green from dissolved oleate; the metal was thickly covered with green film.

Weight of metal after washing and cleaning, 0.790; loss, 0.004. (5) Neatsfoot-oil: weight of metal before immersion, 0.791; no change in colour or consistency of oil, but a green residue or precipitate had collected on the bottom of the glass; the metal was covered with green oleate. Weight of metal after washing and cleaning, 0.787; loss, 0.004. (6) Crude petroleum from Scio: weight of metal before immersion, 0.717. No change was observed in consistency or colour of the oil, and there was no change in the appearance or colour of the metal. Weight of metal after washing and' cleaning, 0.717; loss, none. The foregoing trials express in themselves the fact that the mineral oils form the best protectors for brass. The figures obtained by expressing the loss caused by the oils upon the metal, give also the relative value of the oils in this respect. Reduced, the following table is obtained, which may be considered as an indicator of the dissolving or corroding effect of these oils upon brass:-