The influence of the predominating attraction of I the particles of mercury to themselves, above their adhesion to the sides of a glass tube, has not been considered with so much attention as it demands. Nothing is more common than to remark that the mercury in the barometer is in the act of rising,. if it • show a convex surface, but about to fall, if it should appear concave. Now, the top of the mer curial column must always remain convex, if the barometer he properly constructed, the tube per fectly clean, and the mercury purged of all impu rities. But if the inside_of the tube be any how soiled, whether covered with humidity or stained with mercurial oxyd, the metallic fluid will adhere so ob stinately to the glass, as to lose its convexity, and to subside into a flat surface, or even sink into a conca vity, like water and other liquids. Hence the danger of boiling the mercury too long in the tube, as it be comes partially oxydated, and the thin crust so form ed not only suspends the column higher, but obstructs the freedom of its motion. The same effect is produced by greasing the inside of the tube. Some respectable authors, from not attending to these facts, have hastily inferred that the convex appearance which mercury assumes in the barometer was merely accidental, and consequently removed by a more complete boiling and purification.
In the case of tubes having wide bores, the de-1 pression of the mercurial column may, without any sensible error, be disregarded. According to the curate experiments made by Lord Charles Caven dish, and published by his son, the celebrated Mr Ca vendish, the quantity of depression is only the 200th part of an inch in a tube of 6-10ths of an inch in diameter, the 28th part of an inch in a tube of 8-10ths diameter, and the 15th part of an inch in a tube of 2-10ths diameter. Wide tubes ought, therefore, to be preferred in the construction of barometers, both on account of the facility with which the mer cury moves in them, and the smallness of its depression. The only circumstance to overbalance these advan. • tages, would be the necessity and inconvenience of having a very large cistern. A quarter of an inch may be reckoned a good width of tube, and the cor responding depression is only the twentieth part of ' an inch.
In the syphon barometer, if both branches have the same diameter, the action is exerted on opposite sides, and, therefore, the effect of depression becomes en tirely lost. For accurate purposes, this original form of the instrument has been again resumed, and the inconvenience arising from the large variation of the lower level entirely obviated by an ingenious contriv ance introduced about forty years since. This con sists in the application of a leathern bag, instead of a wooden or ivory cistern, to hold the surplus suer.
Miry. Besides the barometric tube, there a placed adjacent to it another short one of the same width, communicating with the mercury contained in the bag, which being pressed by turning a screw below, is, at each observation, brought exactly to the same mark. The external atmosphere readily acts through
the substance of the leather, but the mercury, from the powerful cohesion of its own particles, cane not be squeezed through the pores of that casing without violent compression. The addition of a bag within a cylindrical box, omitting the lower tube, likewise renders the barometer easily portable; since, for safe carriage, the mercury can be screwed up tight, to fill the whole cavity of its tube, but, on turning the screw again, the column will subside and rest on a broad base.
The last object which required nice observation, was to estimate, the effect of heat in dilating the mercury, and consequently increasing the altitude of the equiponderant column. This correction could not be made with any sort of accuracy pre vious to the application of the thermometer, which, thleigh invented half a century earlier than the ba romer, was yet more than another half century in arriving at perfection. Hero, a mechanical phioso pher, who flourished at Alexandria about 130 years before Christ, has deicribed in his Spiritatia a sort of huge weather-glass, in which water was made to rise and fall by the vicissitudes of day and night, or rather the changes of heat and cold. This machine had for ages been overlooked, or merely con sidered in the light of a curious contrivance. But • Sanctorio, the inventor of the famous statical ba lance, a very learned and ingenious Italian physi cian, who was long professor of medicine in the university of Padua, and had laboured to improve his art by the application of experimental science, redu ced the hydraulic machine of Hero into a more compendious form, and thus constructed, about the close of the sixteenth century, the instrument since known by the name of the air-thermometer, which he employed with obvious advantage to examine the heat of the human body in fevers. Some years af terwards, a similar instrument was contrived, per haps without any communication, by Drebbel, a very clever and scheming Dutch artist, who visited London in the reign of James I., and introduced the know ledge of that instrument into England.
. But this air-thermometer was evidently of the same nature with what has been since called the manometer; it could measure only the dilatation or augmented elas ticity of the air confined within its bulb, whether occa sioned by heat or the diminution of external pressure. It was, therefore, considered merely as a weather glass, indicating the state of the atmosphere ; nor could its blended impressions, which might separate ly affect both the. thermometer and barometer, be then distinguished. Had it been more closely studied, it must have led, by another path, to the discovery of the latter. But those irregularities to which the air-thermometer was hence subject appear • tb have created such doubts respecting the accuracy of the instrument, as occasioned its being neglected king afterwards.