THERMOMETRY (from Gk. (Wpm?, theme, heat + -metria, measurement, from Alrpow, metron, measure). The science of as signing proper numerical values to the tempera ture of bodies. As explained in HEAT, this in volves several steps, the arbitrary selection of: (1) some substance as the thermometric body, (2) some property of it which changes with the temperature, (3) two standard thermal states, e.g. melting ice and boiling water at standard pressures, (4) the scale or the number of degrees between these standard temperatures, and (5) a number to give one of these standard tempera tures.
By definition the 'temperature on the standard hydrogen thermometer,' Centigrade scale, is as follows: 'Hydrogen is the gas used, the initial pressure being 100 cm. of mercury. Change in pressure at constant volume is the effect measured.
The two standard thermal states are (1) melt ing ice and (2) vapor rising from boiling water at normal atmospheric pressure.
One hundred degrees are chosen between these two temperatures. The temperature of melting ice is called 0°. Then, if is the pressure of the gas at 0°, that at 100°. and p that at the temperature (t) to which a number is to be given, 1=100 Other gases than hydrogen might be used; and the same definition would apply to temperature On the new scale as on the hydrogen scale. Slight differences, however, would always he ob served between the numerical values for the same temperatures.
If the change in volume is the heat-effect ob served as the temperature is changed, the ob vious definition of the numerical value of the temperature is t — 100 where v are the measured values of the volume of the gas or liquid at 0°, 100°, t'.
If the change in the electrical resistance of a given conductor is the heat-effect measured as the temperature is changed, the definition of t =100 p p where R. are the electrical resistances at 0°, 100°, t°.
Similarly almost any heat-effect of any body may be taken as the one to measure in order to secure a number for temperature; but the num bers so found all differ in general; and if any thermometer other than the hydrogen one is actually used, the observed numbers must all he corrected by a calibration table connecting the scale of the thermometer used with 'the hy drogen one.
For measuring ordinary laboratory tempera tures either mercury-in-glass or air-thermome ters are used; for measuring extremely low tem peratures, e.g. —200° C., a hydrogen thermome ter or a platinum-resistance thermometer is gen erally used; for high temperatures, e.g. 300° C., •a platinum-resistanee thermometer or a thermo couple is used. (See THERMOMETER and THERMO ELECTRICITY.) For a full discussion of these and other methods reference should be made to the Reports of the International Congress of Physics, Paris, 1900, vol. i., articles by Chappuis and Barns.
The scale of temperature as defined above is known as the 'Centigrade' or 'Celsius' (q.v.) scale. There are several other scales inn use. The `Fahrenheit' scale has 32° for the melting point of ice and 212° for the boiling point of water; the `Ilisaumue (q.v.) *Tale has 0° for the former temperature and 80° for the latter ; the 'absolute gas scale' adds to the temperature on any scale the reciprocal of the coefficient of cubical ex pansion of hydrogen as determined on that scale—on the Centigrade scale this reciprocal is 273.04, and so t° C. is (t X absolute. It is shown 111 THERMODYNAMICS (q.v.) that there is a method of defining temperature which is quite independent of the thermometric substance, and in which the only arbitrary things are the choice of a number for the difference between the temperatures of any two thermal states. This ( scale is called `Thomson's absolute scale,' be cause it was proposed by William Thomson (now Lord Kelvin). If it is agreed to have the temperature of freezing and boiling water 100° apart, it is found that the numbers on the Thomson absolute scale agree most remarkably with those given on the absolute gas scale as defined above.