I do not propose spirits of wine, because that liquor decomposes in a very short time, and cannot be used for experiments of a strong heat.[G] [G] Many travellers have told and written to me, that Reaumur's thermometers of spirit of wine, became quite useless to them, because this liquid lost its colour, and became charged with a sort of mud in a very short time. When on the scale of these thermometers filled with oil or mercury, the first divisions 1, 2, 3, 4, &c. are marked to indicate the double, treble, quadruple, &c. augmentations of heat, we must search after the aliquot parts of each division; for example, of the point 1-1/4, 2-1/4, 3-1/4, &c. or 1-1/2, 2-1/2, 3-1/2, &c. and 1-3/4, 2-3/4, 3-3/4, and which will be obtained in an easy manner, by covering the 1/4, 1/2, or 3/4, of the superfices of one of those small mirrors; for then the image which it reflects, will contain only the 1/4, 1/2, or 3/4, of the heat which the whole image will contain, and, consequently, the division of the aliquot parts will be as exact as those of the whole numbers. If once we succeed in this real thermometer, which I call real, because it actually marks the proportion of the heat, every other thermometer whose scale is arbitrary and different, will become not only superfluous, but even inimical, in many cases, to the precision of natural truths sought after by these means. 5. By means of three mirrors we may easily collect in their entire purity, the volatile parts of gold, silver, and other metals and minerals; for, by exposing to the large focus of those mirrors a large piece of metal, as a dish, or silver plate, we shall see smoke issue from it in great abundance, and for a considerable time, till the metal is in fusion; and by giving only a smaller heat than what fusion requires, we shall evaporate the metal so as to diminish the weight considerably. I am certain of this circumstance, which also elucidates the intimate composition of metals. I was desirous of collecting this plentiful vapour, which the pure fire of the sun causes to issue from metal, but I had not the necessary instruments, and I can only recommend to chemists and naturalists to follow this important experiment, the results of which would be as much less equivocal as the metallic vapour is pure; whereas, in all like operations made with common fire, the metallic vapour is necessarily mixed with other vapours proceeding from combustible matters, which serve for food to this fire.