He therefore thought of altering the relative po sition of the mercury and water, in the following manner: ABC, Fig. 6. is a bent tube hermetically sealed at C, and open at A. At DE and FG the tube is swelled out into two equal cylindrical vessels, which are about 29 inches asunder. The diameter of the bore of the tube is about 1 line ; that of the cy linders 15 lines, and their depth 10. • The limb BC is then filled with mercury, and the barometer be ing placed in a proper position, as much mercury is re tained as occupies EBF. Oil of tartar, (a solution of tartrate of potass,) or any other liquor which does not freeze readily, or act upon the mercury, is pour ed into AE till it rises to a suitable height above the surface of the mercury at E.
Since the two cylinders are equal, and since their diameters are to the diameter of the tube in any ratio whatsoever,'it is evident, that, by this construction, the smallest difference in the atmospherical pressure may be estimated. The scale, however, owing to the nature of hydrostatical pressure, is not capable of being extended beyond certain limits. The liquor will only rise above .D, till its weight, together with that of the air, becomes a.counterpoise to.the mercu ry in the other limb. Its evaporation is, in some measure, prevented by a thin film of oil on its surface.
This barometer has several defects. The column of mercury is supported above its level at the lower surface, not only by the weight of the air,. but also by that of the liquor above it, which increases with the height : when the weight of the air diminishes,. the pressure of the liquor increases, and conversely ; for the motion of the mercury causing the liquor to increase or diminish in altitude, the apparent effect of, will be increased or. diminished in like man ner; the pressure of fluids being in the compound, ratio of their bases and altitudes. The friction, vary ing with the height of,•he liquor, must be another• source of inaccuracy. The liquor itself must also he considerably affected in bulk by heat and cold ; and to thesb. inconveniences it may be added, that, notwith standing the film of oil on its surface, it will gradual ly. evaporate, and render the scale erroneous..
The barometer of Hooke may be considered 23 an improvement on that of Huygens. This barometer, represented Fig, 7, is composed of two tubes ABC. The parts AD and EF are equally wide, and the bore of CE is made as much narrower as it is propo sed to enlarge the scale.. IBG is filled with mercu
ry, the part Al forming a vacuum. HG is occu pied by some fluid lighter than mercury, as a solution of tartrate of potass, and CH by some fluid still lighter, as petroleum, which has little tendency to mix with the other. The cistern C is of the same diameter with AD.
From this construction we may readily perceive the following advantages : The height of the two liquors above the mercury is always the same, whatever be the weight of the air ; and the straight tube CE be ing always filled by the liquors, the friction to which it gives rise must be constantly the same.
The range of the scale, which will be determined by the motions of the line of separation H of the two liquors, will have the same extent as that of the baro meter of Descartes and Huygens. Though this ba rometer is among the best of those with an enlarged scale, it is not free from imperfections : the weight or pressure of the fluids on the surface of the mercu ry will differ on account of their difference of speci fic gravity ; and they will gradually mix together, so that the line of separation will at last become incapa ble of being distinguished.
M. de la Hire, in the memoirs of the Academy o. Sciences for 1708, speaks of the above method em ployed by Hooke for correcting the defects of the double barometer ; and he says that he mentioned it to Huygens in 1690. M. Amontons also affirms, in a work which he printed in 1695, that the same me thod had occurred to him eight or ten years before ; and that he disclosed it to M. Hubin, who had exe cuted a barometer of the same kind,•without having previously communicated with any person on the sub ject. When Hubin went afterwards to England, Hooke proposed the same thing to him ; and indeed the honour of the invention is due to this philoso pher, if we adjudge it by priority of claim, for it appears that he had suggested the double barometer which bears his name, in 1668. Phil. Trans. No. 185. . We are indebted also to Hooke for the wheel-ba rometer, which he invented the same year. This form of the barometer, (See Fig. 11.) on account of its exhibiting the rise and fall of the mercury in a very conspicuous manner, is become extremely com mon. The tube is generally concealed in the frame ; but, for the sake. of representing the whole in one figure, we have made it to appear in front ; it is about 40 inches long, but 6 inches of the lower end is bent upwards, so as to become parallel to the rest of the tube.