The thermometer having- been time carried' by eucceseirseteps to. perfection. it was found by dell' rate experiments, that, between the points of boiling and freezing, the expansion• of mercury, to the fifty.fburth part of its- bulk, or that it acquires for each degree of heat; an increase• of volume amounting• to the 54126 part on the. centesimal or the stud part ea the- scale of ahren belt.. A amnion, therefore, on the height of the mercurial column in the barometer, becomes neces sary according to the changes of temperature which it undergoes. In this climate, the extreme varia don arising from that cause will seldom exceed two-tenths of an inch. But if the barometer be suspended in a room, kept at an agreeable tempe rature, the error occasioned by the expansion of the mercury may, in ordinary cases, be disregarded, since it will scarcely amount to the twentieth part of an inch.
Since the barometer marks the condition of the distant atmosphere, and intimates those internal al terations which are generally cdnnected with the change of the weather, it is particularly valuable at sea, by forewarning the mariner of the approach of *storm. But an instrument of the ordinary con etruction would. not answer this purpose, the agita tion of vessel on a tempestuous ocean. being such as will not only throw the ponderous mercurial co lumn into violent oscillation, but communicate those sudden shocks which must infallibly break the tube. Variousatteraptshaveaccordinglybeeninade to obtain a Marine Barometer, exempt from risk, and yet suf ficiently sensible to the of the atmosphere. The conical or pendant barometerbeing, from the nar rowness of its bore, rather sluggish, was first recent mended for that purpose, though never adopted into practice. About the beginning of tile eighteenth century, Dr Hook and Amentonsseverally proposed to employ for a barometer on board ship, the mano meter or air-thermometer. To obviate the derange ment arising from the influence of heat, there was to be placed betide it a spirit-of-wine-thermometer, with s ball so large as to give expansions equal to those of the portion of air confined within the. bulb. The difibrence between the two adjacent columns of liquid would therefore measure the variation of ester. nal pressure. But procure such a nice adapt*. don would prove so extremely difficult in practice, that most probably this instrument was seldom or ever actually constructed. Besides, the liquid co lumn of the manometer,. though light and narrow, would yet be much shaken by the rolling and pitch: ing at sea. Notwithstanding these weighty °Woo
dons, however, this compound manometer was tried in England, mercury being employed as the Raid both of expamion and pressure, and various adjustments applied by means of a complex machi nery.
Au ingenious and very substantial kind of marine barceoner was above twenty younsince recommended by Blondeau; one of the professors of the naval amide my at Brest. Vesfig.11. Plato XXXII.) It consisted of an Iron fad*, bets below into a syphon, and filled carefully with mercury, which carried a float. For this purpose, a musket-barreli.about three feet long, was chosen, having• a very smootha d nd even bore, sean iron. breech 'closely welded to it, instead of being en. dered• with bras*, which might become °arrant, by the action of the mercury. The lower end of the tube had a• collar of leather, to which was mewed a piece of iron, perforated through its whole length, and ' bent•into an arch, having screwed with a collar of leather at its other extremity, a vertical cylinder Jf iron, four inches high, and of the same bore exactly as the tube. The contracted aperture at the end of the tube, not being exactly in the mid was not always opposite that of the arch ; and, therefore, by turning it occasionally aside, the commu nication could be contracted at pleasure, or even ob structed entirely. The cylindrical appendix was taper ed at the top to a narrow orifice, through which an iron wire, attached to a small ivory float, had been intro duced. To prepare this instrument for action, the mercury was first boiled in the tube ; then the arch, filled with hot mercury, was screwed to the end, the cock opened, and the surplus mercury allowed to flow over ; next the vertical piece, with its float, was screwed on, and a little mercury added, to give it due play. The origin of the scale was to be deter mined from the comparison with another good baro meter of the ordinary construction; but, owing to the equality of the bores of the opposite tubes, the divisions were only half the usual size, or the inches were exhibited by half-inches.
This species of barometer is certainly free from all sort of risk, while the facility which, by means of turning the arch, it affords in checking the ascent and descent of the mercury, prevents in a great measure the oscillations of that fluid. If the instrument were properly suspended, therefore, its indications would be tolerably steady and regular. The chief objection to it consists in the diminutive range of its scale.