PYROMETER (literally " fire-meaaurer " from Tip, "fire," and p‘rpos," a measure "). No fluid has hitherto been found applicable to the construction of thermometers capable of indicating higher tempe rature; than that of boiling mercury (about 660° Fahr.). The term pyrometer was first employed by Muschenbreek to desigoate an instru ment invented by him for measuring the effects produced in the dimensions of solid bodies by the application of heat; but the signifi cation of the term has since been extended so as to include those instruments the object of which is to measure all gradations of temperature above those which can be indicated by the mercurial thermometer.
Mturchenbroek's pyrometer consisted of a metallic bar, about six inches In length, one extremity of which was fined, while the other was left free to advance as the metal elongated from the effect of the flames of several spirit-lamps placed beneath, which at each experiment were charged with the same quantity of highly rectified spirit of wine. The advance of the moveable extremity gave motion to a pinion and wheel, the latter of which drove an index over a graduated circle, each degree of which corresponded to a linear expansion of 12,500th of an inch. The Instrument, as it was originally constructed, is described in the second part of 31teschenbroek's translation of thee Saggi di Naturali Eeperienze fatte cull' Academia del Chnento,' Leyden, 1731 ; and as improved by Desaguliers (who substituted fine cords and friction rollers for the wheel and pinion), in his ' Experimental Philosophy,' vol. I.
To Muscheribroek's pyrometer succeeded those of Ellicott (described in the Phil. Trans.' for 1736 and 1751), Graham (' Phil. Trans.,' 1754), Smeaton, Ferguson (' Lectures", &e., which, like those that have prince leen constructed, with few exceptions, down to the present time. evince but little originality in the principle upon which they rest. A bar of metal is in most cases subjected to the direct action of flame, or immersed in a fluid of convenient temperature. The minute resulting
expansion is multiplied, and thereby rendered appreciable by the inter vention of a succession of levers or a system of wheels and pulleys.
Supposing this intervening machinery to perform with theoretical accuracy, and that the same quantity of heat is successively communi cated to different substances, the indications of such an instrument would give the relative expansions of those substances under the same circumstances. But where wheels, pinions, levers, &c., are employed, there must be considerable liability to error, arising from flexure, obliquity of action, and other causes, the magnitude of which it would be difficult to estimate, and which, even if it be supposed small in the first instance, will be magnified almost in the same proportion as the delicacy of the instrument is increased. Moreover the substance itself, if its nature be such as to be softened by heat, is very liable at high temperatures to undergo compression in giving motion to the machinery.
Even, therefore, as measures of expansion they cannot be considered as deserving of much confidence. A similar remark is applicable, though in a less degree, to the contrivance employed by Lavoisier and Laplace, in which the expansion of the metal deflected a telescope from the position that it had at the commencement of the experiment, and the absolute expansion was deduced from the extent of this deflexion, which was read off upon a graduated scale placed at a considerable distance in front of the telescope. A description of the apparatus employed is given in Biot's Physique Exp6rimentale,' tome i., and a table of the expansions of the several substances experimented on between the temperatures of 32° and 212' Fahr. Troughton, in 1794, constructed an instrument which bore some resemblance to the pre ceding, the principal difference consisting in the employment of a spirit-level, the deviations of which from the horizontal determined the expansion of the metal.