When the thermometer is sealed it must next be graduated; the materials used for scales are various, and of these, perhaps, ivory is the best : metal scales injure the delicacy of the instrument by their high conducting power. Perhaps the best of all is that engraved on the tube, which is sometimes done in chemical thermometers. We have already dis &issed the fixed points of thermometers, but we must here say a few words upon the precautions in using that of boiling water, requiring an equation, to which we have alluded in a former part of this article.
From various causes, the precise temperature of boiling water may be subject to considerable varia tion, for instance, from the nature of the vessel employed, Gay Lussac having foundtt that while water in a metallic vessel vaporizes at 212°, it will sustain a heat of 214° or even 216' in one of glass. It was with great judgment, therefore, that the Royal Society of London, about the middle of the last century, appointed a committee, at the head of which was Mr. Cavendish, to investigate the fixed points of thermometers, and their elaborate report published in the Philosophical Transactionstt seems calculated to put to rest all further doubts on the subject. They recommend a tall vessel of tin plate, with a tight cover, having two apertures, one through which the thermometer is introduced, and fitted by a pierced cork ; the other to which a chimney is affixed, two or three inches long, for the escape of steam, and over which a thin piece of metal is laid to act as a valve to prevent its too rapid escape. The thermometer is recommended to be suspended in steam a little above the surface of the water, it is therefore necessary that the steam should be kept of uniform tension. Fahrenheit first observed about 1724 the effect of atmospheric pres sure on the boiling points of liquids, which ren dered it necessary to graduate thermometers at a fixed state of the barometer, which ought to be 29.50, when the temperature of the water is taken and 29.80 when that of the steam. Deluc after wards investigated the laws of this variation, and his formula for the equation of it, reduced to Eng lish measures, is the following : Where /: denotes the temperature of the boiling point above the freezing point, in degrees of Fah renheit, and z, the height of the barometer in tenths of an inch. As the logarithms here used are those of Briggs, in which all the figures are con sidered as integers, and containing seven places be sides the indices, the formula in the common loga rithms is thus expressed : 99000 , h --= --- • log. z— 92.804.
899 The following Table shows the equation of the boiling point thus derived, and also the later num bers of Sir George Shuekburgh, which differ but little from those of Deluc.
The numbers refer to the temperature of the wa ter, that of steam is about 0°.4 lower.
Upon the principle suggested by this table of measuring the pressure of atmosphere by the ther mometer, Cavallo (though Fahrenheit had before hinted at it) proposed his thermometrical barome ter, in the execution of which, several practical dif ficulties occurred till they were overcome by the Rev. Mr. Wollaston, who has more lately con structed a very elegant instrument on this principle. It consists of a bulb a, Plate DXXIV. Fig. 12, one inch in diameter, the tube of which d c contains a few degrees near the boiling point, of such a length, that in the most delicate instruments, each of them admits of a division into 1000 parts by means of a scale e f with a vernier ; one of these parts cor responds to a difference of level of six incites. When in use, the thermometer is placed as shown in the figure in the cistern A containing water, which is made to boil by means of the spirit lamp B. When the observation is finished, the mercury contracts into the dilated part of the tube b, and being then inverted into the cistern and fixed by the screwing plate g g into which the stem of the ther mometer is secured, a cap being screwed over the bulb which is then exposed, the instrument becomes extremely portable, the boiler being only 5.5 inches deep, and 1.2 in diameter. In order to shorten the tube without diminishing the range or the sensi bility, a cap c is added by which some of the mer cury may be expelled from the tube, or united to the column if the heights are great, and the de pression likely to be considerable. A steam-pipe is employed similarly to the instrument of the Royal Society's Committee above described. The prin cipal defect of Mr. Wollaston's thermometer is its liability to breakage in carriage, on account of the great weight of the large mercurial bulb. We are disposed to think that some fluid, such as sul phuric acid, which has a moderate specific gravity, and boils at a very high temperature, might, by being substituted for mercury, obviate this defect; any small inequalities of expansion (if such exist) would be quite insensible in so small a range.