10. The variation of viscosity in varying temperatures is less in mineral oils than in animal or vegetable oils.
75. Methods of Testing Oils are necessary in order to determine which may be adapted to a specific purpose. Their peculiar characteristics must be studied in order to know which will best fulfill the conditions arising in actual practice. Experiments are necessary in which the oil is subjected to conditions approximating, as nearly as possible, to the conditions proposed in its actual use.
Saunier states[17] that "success depends largely on the skill of the manipulator; and if he is not endowed with the power of judging, mainly by the taste, whether oil satisfies certain prescribed conditions, he can never be certain of the result." As the author's abilities in this regard are not up to the required standard, and as some oils are sometimes in such a state of decomposition that even the odor is unpleasant, he has used other, and perhaps more satisfactory, methods of determining the relative values of the various oils.
The following experiments show the relative values of oils that have been, or may be, used in horology:
J. J. Redwood has made experiments on the action of oils upon metals, especially for the purpose of determining which oils were best adapted for use on the various metals and for ascertaining which oils were most suitable for mixing as lubricants. He has tabulated the results of his researches in two tables, which show that:[18]
Mineral oil has no effect upon copper and zinc, and attacks lead most.
Olive oil attacks copper most, tin least.
Sperm oil attacks zinc most, copper least.
The experiments show, on the other hand, that:
Brass is attacked most by olive oil.