The temperature is measured by means of the increased elastic force of the inclosed air, and the instrument is both more convenient and more precise than that in which the volume varies, for at all temperatures the sensibility of the instrument is the same. At high temperatures the apparatus is liable to distortion under the pressure of the inclosed air; but this may be prevented, if needful, by introducing air of a lower than atmospheric pressure at an ordinary temperature, even so low as one-fourth of an atmosphere; for, although the apparatus is less sensitive in proportion as the first supply of air is of less density and pressure, yet withal it is sufficiently sensitive. The thermometer, as employed by M. Regnault, is shown in Fig. 120. Two

glass tubes, df, cd, about half-an-inch bore, are united at the base by a stop-cock r The tube cd is open above, and df is connected to the reservoir a by a small tube ab. The cover of the boiler in which the reservoir is inclosed is shown at B, and the tubes are protected from the heat of the boiler by the partition CD. By means of a three-way connection, g, and tube h, the connecting tube ab communicates with an air pump, by means of which the apparatus may be dried, and air or other gas supplied to it. The first thing to be done is to completely dry the apparatus, and for this object, a little mercury is passed into the tube bd, and the cock is closed against it. The exhausting pump is then set to work to exhaust the tube, which is done several times, the air being slowly re-admitted after each exhaustion, after having been passed through a filter of pumice-stone in connection with the pump, saturated with concentrated sulphuric acid to absorb moisture, and thus desiccate the air. During this part of the process, the reservoir is maintained at a temperature of 130° F., or 140° F., to insure complete desiccation. Next, the reservoir is plunged into melting ice, the two vertical tubes bd, cd, are put into communication, and filled with mercury up to a suitable level ƒ, marked on the tube bd. If it is desired to establish an internal pressure less than that of the atmosphere, the air is partially exhausted by means of the pump, the degree of exhaustion being recorded by the difference of level in the two tubes. The exhausting tube h is then hermetically sealed, and the mercury adjusted to the level ƒ in the tube bd.

Fig. 120.

PYROMETERS.

Pyrometers are employed to measure temperatures above the boiling point of mercury, about 676° F. They depend upon the change of form of either solid or gaseous bodies, liquids being necessarily inadmissible. Pyrometric estimations are of three classes: First, those of which the

indications are based upon the change of dimensions of a particular body, solid or gaseous-the pyrometer; second, those based on the heat imparted to water by a heated body; third, those which are based upon the melting points of metals and metallic alloys.

Wedgwood's pyrometer, invented in 1782, was founded on the property possessed by clay of contracting at high temperatures, an effect which is due partly to the dissipation of the water in clay, and subsequently to partial vitrification. The apparatus consists of a metallic groove, 24 inches long, the sides of which converge, being half-an-inch wide above and three-tenths below. The clay is made up into little cylinders or truncated cones, which fit the commencement of the groove after having been heated to low redness; their subsequent contraction by heat is determined by allowing them to slide from the top of the groove downwards till they arrive at a part of it through which they cannot pass. The zero point is fixed at the temperature of low redness, 1077° F. The whole length of the groove or scale is divided into 240 degrees, each of which was supposed by Wedgwood equivalent to 130° F., the other end of the scale being assumed to represent 32.277 F. Wedgwood also assumed that the contraction of the clay was proportional to the degree of heat to which it might be exposed; but this assumption is not correct, for a long-continued moderate heat is found to cause the same amount of contraction as a more violent heat for a shorter period. Wedgwood's pyrometer is not employed by scientific men, because its indications cannot be relied upon for the reason just given, and also because the contraction of different clays under great heat is not always the

same.

In Daniell's pyrometer the temperature is measured by the expansion of a metal bar inclosed in a black-lead earthenware case, which is drilled out longitudinally to inch in diameter and 7 inches deep. A bar of platinum or soft iron, a little less in diameter, and an inch shorter than the bore, is placed in it and surmounted by a porcelain index 11⁄2 inches long, kept in its place by a strap of platinum and an earthenware wedge. When the instrument is heated, the bar, by its greater rate of expansion compared with the black-lead, presses forward the index, which is kept in its new situation by the strap and wedge until the instrument cools, when the observation can be taken by means of a scale.

The air-pyrometer. The principle and construction of the air-thermometer are directly applicable for pyrometric purposes, substituting a platinum. globe for the glass reservoir already described, for resisting great heat, and as large as possible. The chief cause of uncertainty is the expansion of the metal at high temperatures.

The second means of estimation is best represented by the "pyrometer' of Mr. Wilson, of St. Helen's. He heats a given weight of platinum in the fire of which the temperature is to be measured, and plunges it into a vessel containing twice the weight of water of a known temperature. Observing the rise of temperature in the water, he calculates the temperature to which the platinum was subjected, in terms of the rise of temperature of the water, the relative weights of the platinum and the water, and their specific heats. In fact, the elevation of the temperature of the water is to that of the platinum above the original temperature of the water in the compound ratio of the weights and specific heats inversely; that is say, that the weights of the platinum and the water being as 1 to 2, and

to

their specific heats as .0314 to 1, the rise of temperature of the water is to that of the platinum as 1 × .0314 to 2 × 1, or as 1 to 63.7, and the rule for finding the temperature of the fire is to multiply the rise of temperature of the water by 63.7, and add its original temperature to the product. The sum is the temperature of the fire, subject to correction for the heat absorbed by the thermometer in the water, and by the iron vessel containing the water, and the heat retained by the platinum. The correction is estimated by Mr. Wilson at th, taking the weight of water at 2000 grains, and that of the platinum 1000 grains, and it may be allowed for by increasing the above-named multiplier by 17th, to 67.45.

Mr. Wilson proposed that for general practical purposes a small piece of Stourbridge clay be substituted for platinum, to lessen the cost of the apparatus. With a piece of such clay, weighing 200 grains, and 2000 grains of water, he found that the correct multiplier was 46.

The third means of estimation, based on the melting points of metals and metallic alloys, is applied simply by suspending in the heated medium a piece of metal or alloy of which the melting point is known, and, if necessary, two or more pieces of different melting points, so as to ascertain, according to the pieces which are melted and those which continue in the solid state, within certain limits of temperature, the heat of the furnace. A list of melting points of metals and metallic alloys is given in a subsequent chapter.

LUMINOSITY AT HIGH TEMPERATURES.

The luminosity or shades of temperature have been observed by M. Pouillet by means of an air-pyrometer to be as follows:

A bright bar of iron, slowly heated in contact with air, assumes the following tints at annexed temperatures (Claudel):

MOVEMENTS OF HEAT.

When two bodies in the neighbourhood of each other have unequal temperatures, there exists between them a transfer of heat from the hotter of the two to the other. The tendency to an equalization, or towards an equilibrium, of temperatures in this way is universal, and the passage of heat takes place in three ways: by radiation, by conduction, and by convection or carriage from one place to another by heated currents.

RADIATION OF HEAT FROM COMBUSTIBLES.

It is a common assumption that the quantity of heat radiated from combustibles is very small in comparison with the total quantity of heat evolved. Holding the hand near the flame of a candle, laterally, the radiant heat, which is the only heat thus experienced, is much less than the heat experienced by the hand when held above the flame, which is the heat by convection of the hot current of air which rises from the flame. But it is to be noted that, whilst the radiant heat is dissipated all round the flame, the diameter of the upward current is little more than that of the flame, and the conveyed beat is therefore concentrated in a narrow compass.

M. Peclet, by means of a simple apparatus, consisting of a cage suspending the combustible within a hollow cylinder filled with water in an annular space, ascertained that the proportion of the total heat radiated from different combustibles was as follows:

These values serve to show that radiation of heat is considerable, and that dameless carbon radiates much more than flame, though the proportion of heat radiated from fuels depends very much upon the disposition of the material and the extent of radiating surface.

With respect to heated bodies, apart from combustibles as such, the radiation or emission of heat implies the reverse process of absorption, and the best radiators are likewise the best absorbents of heat. All bodies possess the property of radiating heat. The heat rays proceed in straight tres, and the intensity of the heat radiated from any one source of heat becomes less as the distance from the source of heat increases, in the Everse ratio of the square of the distance. That is to say, for example, that at any given distance from the source of radiation, the intensity of the radiant heat is four times as great as it is at twice the distance, and nine ames as great as it is at three times the distance.

The quantity of heat emitted by radiation increases in some proportion with the difference of temperatures of the radiating body and the surrounding medium, but more rapidly than the simple proportion for the greater differences; and the quantity of heat, greater or less, emitted by bodies by radiation under the same circumstances is the measure of their radiating power. Radiant heat traverses air without heating it.

When a polished body is struck by a ray of heat, it absorbs a part of the heat and reflects the rest. The greater or less proportion of heat absorbed by the body is the measure of its absorbing power, and the reflected heat is the measure of its reflecting power.

When the temperature of a body remains constant it indicates that th quantity of heat emitted is equal to the quantity of heat absorbed by th body. The reflecting power of a body is the complement of its absorbin power; that is to say, that the sum of the absorbing and reflecting powers ( all bodies is the same, which amounts to this, that a ray of heat striking body is disposed of by absorption and reflection together, that which is ne absorbed being necessarily reflected.

For example, the radiating power of a body being represented by 90, th reflecting power is also 90, and the absorbing power is 10, supposing tha

Table No. 106.-COMPARATIVE RADIATING OR ABSORBENT AND REFLECTING POWERS OF SUBSTANCES.