fuel 700 grains of white lead, which are well mixed with it, and 300 grains of pure white lead to cover the mixture. When the whole is heated, the carbonate of lead decomposes, forming pure oxide of lead, which is then reduced, as in the former case. By this process the results correspond to 1 grain in the quantity of lead produced from a given sample of fuel. Of course great care must be taken that the white lead is genuine.

Commercial samples are frequently adulterated with sulphate of lead and sulphate of baryta, oxychloride of lead, oxide of zinc, &c. This is a serious drawback to this otherwise excellent modification.

The following is the method of ascertaining the calorific power of fuel, employed by Dr. Ure, and described in his 'Supplement.'

The following calorimeter, founded upon the same principle as that of Count Rumford, but with certain improvements, may be considered as an equally correct instrument for measuring heat with any of the preceding (Lavoisier, Meyers, and others), but one of much more general application, since it can determine the quantity of heat disengaged in combustion, as well as the latent heat of steam and other vapours.

'It consists of a large copper bath capable of holding 100 gallons of water. It is traversed four times backwards and forwards in four different vessels, by a zigzag horizontal flue or flat pipe, nine inches broad and one deep, ending below in a round pipe, which passes through the bottom of the bath, and receives there into it the top of a small black-lead furnace, the innermost crucible of which contains the fuel. It is surrounded at the distance of an inch by a second crucible, which is enclosed at the same time by the sides of the outermost furnace, the strata of stagnant air between the crucibles serving to prevent the heat being dissipated into the atmosphere by the body of the furnace. A pipe from a double pair of bellows enters the ashpit of the furnace at one side, and supplies a steady but gentle heat to carry on the combustion kindled at first by half an ounce of burning charcoal. So completely is the heat which is disengaged by

L

the burning fuel absorbed by the water in the bath, that the air discharged at the top pipe is generally of the same temperature as the atmosphere. The vessel is made of copper, weighing 2 lbs. per square foot; it is 5 feet long, 1 wide, 2 deep, with a bottom 5 feet long, and 13 broad upon an average. Including the zigzag tin-plate flue, and a rim of wrought iron, it weighs altogether 85 lbs. Since the specific heat of copper is to that of water as 94 to 1000, the specific heat of the vessel is equal to that of 8 lbs. of water; for which, therefore, the exact correction is made by leaving 8 lbs. of water out of the 600 or 1000 lbs. used in the experiment.

In the experiments made with former calorimeters of this kind, the combustion was maintained by a current or draught of a chimney open at bottom, which carried off at the top orifice of the flue a variable quantity of heat, very difficult to estimate.

'The heating power of the fuel is measured by the number of degrees of temperature, which the combustion of 1 lb. of it raises 600 or 1000 of water in the bath, the copper substance of the vessel being taken into account. 1 lb. of dry wood charcoal, by its combustion, causes 6000 lbs. of water to become 20° hotter. For the sake of brevity, we shall call this calorific energy 12,000 unities. In like circumstance, 1 lb. of Llangennock coal will yield by combustion 11,500 unities of caloric.'

This form of calorimeter of Dr. Ure's seems to possess many advantages over Laplace's and others, and is, no doubt, very convenient in use, although rather bulky.

The instrument known as Wright's calorimeter gives very accurate results, and is the one most generally used now in experiments on the heating power of fuel, in all but the most refined investigations. It is shown in the accompanying figure.

The copper cylinder A B is filled with a mixture of 20 grains of the combustible, and 240 of the deflagrating compound, which is composed of three parts of chlorate of potash, and one of nitrate of potash. A little piece of cotton soaked in chlorate of potash is placed partly in the mixture, the other

end projecting above the top of the cylinder; this is ignited, quickly covered with the bell-shaped part of the apparatus, and immersed in a measured quantity of water. As constructed,

the whole metallic apparatus weighs 6642-7 grains, and with this weight 290-1 grains of water are used. The temperature is recorded before and after making the experiment. During the deflagration the stop-cock is closed, which is, however, opened before taking the temperature the second time. A

tenth of the temperature that the water is raised by the combustion, is added for errors that are incidental to the use of the instrument.

If the instrument is made of the weight above given, the result is obtained by a very simple calculation. Each Fahrenheit degree by which the temperature of water has been augmented, corresponds to a pound of water converted into

20 grains of the coal will convert into steam (maximum effect) 9.9 lbs. of water

5. DETERMINATION OF THE SPECIFIC HEATING POWER.This represents the heat produced from a certain volume. of fuel. It may be ascertained by multiplying the absolute heating power by the specific gravity.

6. DETERMINATION OF THE PYROMETRIC HEATING POWER.— By pyrometric heating power is meant the degree of temperature which may be obtained by completely burning the fuel. This heating power not only depends upon the composition of the fuel, but chiefly on the time required for its combustion, and this again depends on the looseness and inflammability of the fuel. The absolute heating power of hydrogen is greater than that of carbon, but with regard to the pyrometric heating power it will be found that reverse is the case.

Carbon burned in contact with the air to carbonic acid will produce a heat of 2558° C.; if burned to carbonic oxide it only produces 1310°; hydrogen burning to water will produce a heat of 2080°. From this we learn that fuel rich in carbon, such as anthracite, coal, and coke, will produce a greater pyrometric effect than fuel rich in hydrogen, as wood, &c.

Density is an essential quality of fuel required to pro

duce great pyrometric effect. This is proved in the following way.

When atmospheric air first acts on the carbon contained in fuel, carbonic acid is formed, and the temperature rises to a certain degree, but on passing over glowing coal, carbonic acid becomes converted into carbonic oxide, and this causes a portion of the heat at first produced to become latent. This conversion of carbonic acid into carbonic oxide is more easy and complete, as the fuel used is more inflammable; and as a greater quantity of heat is thereby. rendered latent, it follows that the heating power of such a fuel is inferior. This accords with general experience; for it is well known that coke is able to produce a greater heat than charcoal.

Several good methods for determining pyrometric heating power were given in the last chapter.

7. DETERMINATION OF THE VOLATILE PRODUCTS OF CARBONISATION.—The amount of volatile matter yield on carbonising a fuel depends partly on the composition of the fuel, and partly on the temperature employed. If a fuel rich in oxygen and hydrogen is quickly heated, it will yield the greatest amount of volatile products. These are partly liquid (tar, naphtha, and acetic or ammoniacal water), and partly gaseous (carbonic oxide, carbonic acid, and light and heavy carburetted hydrogen). The more oxygen a fuel contains, the more carbonic acid and carbonic oxide it will produce; the more hydrogen it contains, the more illuminating gas it yields. The applicability of a sample of coal to the production of illuminating gas depends on these conditions.

Coal distilled at a low temperature yields much tar and comparatively little gas, and when a very high temperature has been used, less tar and more gas is produced, but the great heat will have reacted on the gas and injured its illuminating qualities. If the coal contains pyrites, the gas will contain sulphur compounds. The amount of water produced is generally larger than that of the tar.

In order to estimate the amount of volatile matter given off from any particular sample of coal, proceed in the