PYROMETER 111 range of available temperatures ; hence, to ob- tain accurate measures with Daniell's pyrome- ter, it should be graduated by noting its indica- tions at successive high temperatures, the ther- mornetric values of which have been determined with an air thermometer. We should also be sure that successive heatings and coolings of the metallic bar do not change its coefficient of expansion. 2. Wedgwood's pyrometer, using the contraction of baked clay as a measure, con- sists of a metallic groove or gauge, the sides of which gradually converge ; pieces of very pure clay are made into small cylinders, having one side flattened, and, being heated to redness, made just to fit the larger extremity of the groove. It is a property of clay permanently to contract and become harder by exposure to a high temperature, in consequence of its losing a portion of the water with which it is com- bined ; and it was supposed that the amount of the contraction was exactly proportioned to the intensity of the heat to which it is exposed. The amount of contraction in the clay cylinders, after being exposed to the tem- perature which it was desired to measure, was easily determined by allowing the cylin- ders to slide from the top of the groove down- ward, till they arrived at a point beyond which they would not pass. Mr. Wedgwood divided the whole length of this gauge into 240 parts or degrees, each of which he calcu- lated to be equal to 130 of Fahrenheit's scale ; and the zero of his scale, indicating a red heat, corresponded, according to his experiments, to 1,077. The difficulty of obtaining clay of uniform composition is of itself an almost in- superable objection, to this method of estima- ting high temperatures ; and it has been since ascertained that the observation upon which it is founded is not correct, for clay will con- tract as much by the long continuance of a comparatively low heat as by a short continu- ance of a high one. Hence the degrees of heat which Wedgwood's pyrometer has been applied to measure have been enormously ex- aggerated. Thus, Daniell's pyrometer shows that the melting point of cast iron is 2,786, and the highest temperature of a good wind furnace about 8,300 ; points which were esti- mated by Wedgwood at 20,577 and 32,277 respectively. In other words, Wedgwood's pyrometer gave figures nearly ten times higher than those obtained by Daniell's. 3. The ex- pansion of air is used in pyrometers contrived by Pouillet, Regnault, and Jolly. If thermom- eters carefully made of any number of solids and liquids are all simultaneously exposed to the same successive elevations of temperature, it will be found that no two of them agree in their readings throughout the range of tem- peratures ; but if we at the same time expose a set of thermometers made of the permanent gases (i. e., of gases which have never been liquefied by pressure and cold), as oxygen, ni- trogen, air, hydrogen, and carbonic oxide, we will find that they all agree in their readings. 6G2 VOL. xiv. 8 For this reason, if for no other, these perma- nent gases should have the preference as bodies to form the expanding material of thermome- ters ; but the theory of the thermodynamics of gases shows that from the expansion of these gases alone can we arrive at the knowl- edge of true temperatures. In short, it ap- pears that while the coefficients of expansion of solids and liquids increase with the temper- ature, the permanent gases have the same co- efficient of expansion, which also remains con- stant throughout the range of available tem- peratures. Indeed, men of science have agreed that the determination of temperature rests upon the assumption that the permanent, or, as they are now called, perfect gases, when subjected to a constant pressure, expand pro- portionately to the rise of temperature. This expansion, in the case of dry air, amounts for each degree centigrade to 0-003665, or ^-fg of the volume at C. ; or what is the same, the pressure of a mass of air kept at a constant volume increases 0'003665 of its pressure at C. for each rise of 1 C. in temperature. The simplest air thermometer, and the one best adapted to practical purposes, is that of Jolly. Its action depends on the law just given. A hollow globe of hard porcelain (platinum cannot be used by reason of its permeability to gases at high temperatures), A, fig. 3, communicates through the capillary tube E with the fixed vertical glass tube B. The tube B commu- nicates with the open glass tube C through the rubber tube D. The tubes B and and the connecting rubber tube contain mercury. The tube C moves upward and downward in a ver- tical direction, and carries with it the rubber tube D, and thus the surface of the mercury in B can always be brought to. coincide with a mark, R, on the capillary tube E ; so that the air in A and E is al- ways observed under a constant volume after it has been heated to any tem- perature. The height of the mercury in above R is read off on scale.s formed by cut- ting lines in the silvering of slips of glass mir- rors placed behind the tubes C and R. When the centre of the pupil of the eye is seen in the plane passing through the surface of the mer- cury, we know that the line of sight is perpen- dicular to the length of the columns of mercu- Fra. 8. Jolly's Air Ther- mometer.
