To make thermometers self-recording, various schemes have been proposed, of which we shall notice only one or two. Those most commonly used indicate only maximum and minimum temperature during each 24 hours; or during the interval which has elapsed since they were last set. The usual arrangement consists of two thermometers, a mercurial and a spirit one, fixed horizontally to the same frame, with their bulbs at opposite ends of. the frame. Above the mercury is a small piece of steel or ivory, and in the spirit a small and light float of glass or enamel. Capillary forces prevent the steel from entering the mercury, and the enamel from leaving the spirit. As the mercury expands, it pushes the steel before it, and when it again contracts, it leaves it behind; the end nearest the mercury thus remaining at the highest or maximum indication which that thermometer has given. In the the liquid, as it expands, freely passes the enamel, and leaves it undisturbed; but it can never contract so as to leave it dry. It therefore pulls the enamel back when it contracts, and thus the extremity furthest from the bulb marks the lowest point which the spirit has reached, or the mini mum temperature. To set this instrument, incline it so that the steel falls back to the surface of the mercury.—the enamel at the same time comes to the surface of the spirit.
The best mode of registration is undoubtedly the photographic. For this purpose, a mercurial thermometer is placed vertically before a narrow slit, in such a way that no light can pass through the slit save above the level of the mercury in the tube. A gas flame is kept burning at some distance in front of the slit, the bulb of the thermometer being protected from its radiation ; and behind the slit a sheet of prepared photographic paper is exposed to the narrow line of light which passes above the mercury. This paper is fixed on a cylinder with a vertical axis, which is made to revolve uniformly by clockwork. Lines are drawn by the clockwork on the paper, giving the position of the slit at each hour of the 24, or the gas-flame is mechanically reduced or eclipsed at inter vals of an hour; so that the record, when photographically developed, gives the tempera ture for every minute of the day and night; the portion of the paper which has been exposed to the light is blackened.
Among ordinary meteorological instruments the wet-bulb thermometer is deserving of notice. It is simply an ordinary thermometer, with the bulb covered with paper or cotton-wool, kept constantly moist by the capillary action of a few fibers connecting it with a small vessel of water. If the air be saturated with moisture (see DEW, EVAPO RATION), there will be no evaporation, and the wet-bulb thermometer will give the same indication as the dry-bulb. But the drier and the warmer the air is the faster does the water evaporate, and (the latent heat of evaporization being mainly taken from the moist bulb) the lower does the mercury sink in the moist-bulb instrument. The difference between the readings of the two instruments, compared with the actual temperature, as shown by the dry-bulb, thus leads to a determination of the hygrometric state of the air.
So far, we have spoken of the instruments now in common use. But the air-ther mometer was probably the oldest form; and possesses a scientific superiority over those just described. Theoretical and experimental investigations, connected with the modern dynamical theory of heat (see FORCE, HEAT), show that equal increments of heat pro duce almost exactly equal changes of bulk in a nearly perfect gas, such as air, if the pressure to which it is exposed be constant. Hence, temperature, as measured by an air thermometer, gives a true indication of the quantity of energy present in the form of heat. As the comparison of an air-thermometer with a mercurial one shows that, for
temperatures not greater than 300° C., or 572° Fahr., the indications of the two agree very closely, the ordinary mercurial thermometer practically possesses within these limits the same advantage.
As the pressure of a gas depends on the amount of heat it contains, the absolute zero of temperature, or the temperature of a body wholly deprived of heat, may be determined by finding the temperature at which a perfect gas would cease to exert pressure. For ordinary temperatures, it is found (see HEAT) that air increases in bulk by .3665, and hydrogen by .3668 of its bulk, when heated under constant pressure from 0° to 100° C. Again, by Boyle's law, if the air be compressed again, at constant temperature 100° C., to the bulk it had at 0° C., its pressure is increased by .3665 of its former amount. Thus, being the pressure at temperature 0° C., p that at t° C., we have, when the volume is kept constant, Pt= P. If we assume this to hold for all temperatures, vanishes when 1 +.003665t = 0; or t° = — 274° C. very nearly.
• That is to say, at 274°C., under the freezing-point of water, a perfect gas ceases to exert pressure on its containing vessel—i.e., is deprived of that thermal energy on which pressure depends.
The air-thermometers in common use are affected by the pressure, as well as the tem perature of the atmosphere. To avoid this inconvenience, Leslie and Rumford in the pres ent century revived the differential thermometer of Sturmius. In this instrument, in one of its common forms, a bulb is blown at each end of the tube (which is bent into a U-form), and the liquid in the stem is used merely as an index, both balls being full of air. The length of the column of fluid is usually adjusted so that it can just fill one of the vertical arms and the horizontal portion of the tube; and the quantities of air in the two balls are so adjusted that the column will take this position when the two balls are at the same temperature. If the one ball be heated more than the other the liquid index will take a new position, and this is read off by a scale applied to either of the vertical arms. The graduation of this instrument may be effected by calculation, but it is usually done experimentally. Leslie made good use of it in his investigations on heat; and, with various adjuncts, such as coloring the glass of one ball while that of the other was left white; silvering or gilding one of the balls; covering one of them with moist silk or linen, etc., this instrument became in his hands a photometer, an cethrioscope, a hygrom eter, etc.
To thermometers which depend for their action on the expansion of solids, the name PYROMETER (q.v.) is frequently given; but that of Breguet, as delicate as a good ordi nary mercurial thermometer, is not alluded to in that article. The principle of this very beautiful instrument may easily be explained thus. In bending a slip of wood, the fibers on the convex side are necessarily more extended than those toward the concave side. Conversely, if the fibers on one side of a slip of wood were to expand more than those on the other, the slip would bend. Breguet solders together two thin strips of gold and platinum, or platinum and silver; for portability and concentration bends the compound strip into a helix, fixes its upper end, and attaches a horizontal index to the lower end. The least change of temperature in the surrounding air changes the length of one side of the compound slip more than the other, and the helix twists or untwists through an angle very nearly proportional to the change of temperature.
For measuring radiant heat, the most delicate instrument is the thermo-multiplier. See TILERMO-ELECTRICITY