constituents by treatment before combustion, its value will be greatly increased. The hydrogen that is in excess of the quantity required to form water with the oxygen in the fuel is alone available for combustion, and is termed the disposable hydrogen. The remainder of the hydrogen is regarded as being already in combination as water, and may be viewed as an actual source of loss, because this water has to be vaporised when the fuel is burned. Cellulose (CHO) contains no disposable hydrogen, whilst cannel coal contains a considerable quantity.

The temperature at which fuels kindle varies considerably, since an initial temperature is required before combustion is effected. Slow oxidation may, it is true, take place at ordinary temperatures, but in metallurgical practice it is only comparatively rapid combustion that is really useful. Extreme density and, on the other hand, great tenuity equally hinder inflammability. For the former reason, anthracite, diamond, and graphite are ignited with great difficulty. Fuels richest in hydrogen are the easiest to light, the inflammability of resinous wood and of bituminous coal being due to this fact. Combustible gases, however, do not ignite below a cherry-red heat, on account of the extreme dispersion of their molecules.* The inflammability of vegetable charcoal depends upon the temperature at which it is prepared. Thus, when wood is carbonised at 300° to 400°, the CC:L. charcoal ignites at 360°; and when a temperature of 1200° to 1300° has been employed, ignition cannot be effected below 600° to 800. This is due to the fact that the hydrogen is driven off almost entirely at the higher temperatures. Peat ignites at 225; pine wood at 295°; ordinary coal at 325°, the meltingpoint of lead; coke, anthracite, hydrogen, and carbonic oxide at a dull-red heat.

The length of flame given by fuel in burning is so important that coals have advantageously been classed as " long-flaming" and "short-flaming." Certain fuels disengage at the moment of combustion a large proportion of volatile matter, and these constitute the long-flaming varieties. The manner in which the fuel is burned, however, has a great effect upon the flame. Charcoal, for example, if burned with free access of air, merely glows; but if burnt with a limited supply of air in a thick layer so arranged that the products of combustion from the lower portion pass through the upper, carbonic oxide will be formed, which will burn with a blue flame. The volume of the flame depends, amongst other conditions, upon the velocity of the current of air * Frankland, Phil. Mag., vol. xxxvi. (1868), p. 309. See also Ernst, Chemisches Repertorium, vol. xvii. p. 2.

by which the combustion is effected, the volume decreasing as the velocity rises.

Calorific Power.-The calorific power of a fuel is the total heat generated by the combustion of a unit of weight of the fuel. The heat is measured in two ways, either by the number of units of weight of water raised 1°, or by the number of units of water evaporated. The latter method, which was proposed by Professor Rankine, gives numerical results 537 times less than the former. In expressing the calorific power of a fuel, the amount of heat generated on the combustion of carbon to carbonic anhydride is taken, as the standard of comparison. This calorific power of carbon is expressed by the number of parts by weight of water capable of being heated from o° to 1° by the combustion of one part by weight of carbon. It is found by direct experiment to be 8080 units. The unit of heat varies with the thermometric scale and the unit of weight employed. The unit most largely adopted, the metric unit or calorie, is the quantity of heat required to raise 1 kilogram of water from o° to 1° C.; whilst the British thermal unit is the amount of heat required to raise 1 lb. of water one degree Fahr. Thus I calorie = 3.96832 British units, and 1 British unit 0.251996 calorie. Expressed in equivalent foot-pounds, 1 calorie = 1390.

I

=

For experimentally determining the calorific power of a fuel, a calorimeter is employed. Count Rumford's calorimeter consisted simply of a vessel, filled with water, containing a worm-pipe, through which the products of combustion passed from a funnel outside. In this way they imparted their heat to the water, whose rise in temperature was noted. All calorimeters are similar in principle to Rumford's. In the more modern instruments, however, the vessel in which the combustion takes place is entirely surrounded by water and by an air jacket.*

If a fuel consists only of carbon and hydrogen, its calorific power may be calculated by multiplying the weight of each of the elements in one part of the fuel by their respective calorific powers as found by experiment. For example, a fuel consists of 85.71 per cent. of carbon and 14.29 per cent. of hydrogen. What is its calorific power? The calorific powers of carbon and hydrogen are respectively 8080 and 34,462,and

85.71 × 8080 = 692536.80

14.29 × 34462 = 492461.98

1184998.78

*On calorimeters, consult Ganot's Physics; Watts' Dictionary of Chemistry; Zeit. anal. Chem., vol. xxiii. p. 453; Journ. Soc. Chem. Ind., 1886, p. 635; F. J. Rowan, ibid., 1888, p. 195.

which is the heat evolved on the combustion of 100 parts of the fuel. This, when divided by 100, gives 11,849.98 as the calorific power of the fuel. So simple a case as this is rarely met with, it being usually necessary to determine the amount of disposable hydrogen in the fuel, and to multiply the result by 34,462.

Scheurer-Kestner has shown that carbon in combination develops more heat than carbon in the form of charcoal. Consequently the calculated values of the calorific powers of coals are, as a rule, too low. In order to obtain results approximating more closely to the truth, it is advisable to employ for carbon the calorific power of 9000 instead of 8080, and for hydrogen the number 30,000 instead of 34,642. Even this correction does not apply to certain bituminous lignites, and actual experiment is the only safe guide for commercial purposes.

Calorific Intensity. The calorific intensity, or pyrometric effect of a fuel is the highest temperature which the fuel is capable of producing when burnt in air. Measurements of calorific intensity are based on the fact that the heat produced by combustion is transferred to the product of combustion, and it may be determined by calculation on the assumption that the calorific intensity of a simple combustible body burnt in oxygen is equal to its calorific power divided by the product of the relative weight of its products of combustion and the specific heat of those products, or, expressed as an equation,

where W represents the weight

Wx C

W x c

of the substance, C the calorific power of the substance, w the weight of the product of combustion, and c its specific heat. For example, the calorific intensity of carbon burnt to carbonic anhydride in oxygen, is

The calorific intensity of carbonic oxide burnt to carbonic anhydride is

If the substance is a mixture, the weights and specific heats of

* Revue scientifique. 1888; translation in Journ. Soc. Chem. Ind., 1888, p. 615; Bull. de la Soc. industrielle de Mulhouse, 1875, p. 241; Comptes Rendus, vol. Ixvi. (1868), pp. 1047, 1220; vol. lxvii. (1868), pp. 659, 1002; vol. lxviii. (1869), p. 608; vol. lxix., p. 412.

the various products of combustion must be introduced. In determining the calorific power of hydrogen, the water obtained is assumed to be in the liquid state. In determinations of the calorific intensity, however, the water is in the form of gas. From the calorific power, therefore, must be subtracted the amount of heat which would be given out on cooling the steam to o°. For one part of water at 100° this would be 537 calories. Again, in raising the product of combustion from o to t°, a greater amount of heat will be needed to raise it the first 100° while that product is liquid than would have been required if it had been gaseous. These two amounts of heat are in the proportion of the specific heats of steam and water, that is, as 0.4805 is to I. Hence the extra quantity of heat to be added to the 537 calories is (1 − 0.4805) 100, or 51.95 calories. The calorific power must, therefore, be decreased by (51.95 + 537) 9, or 5300.5 calories, since I part of hydrogen yields 9 parts of water. Hence the calorific intensity of hydrogen is

The calorific intensity is merely a theoretical quantity, as it is based on the assumptions that the products of combustion of the fuel have constant specific heats for all temperatures, and that they absorb all the heat produced. Neither of these assumptions is true, inasmuch as the specific heats of gases generally increase with the temperature, and as there is also a considerable amount of heat lost by conduction and by radiation, and dissociation limits the temperature. The computation of calorific intensity has consequently but little commercial value.

The calorific intensity of a fuel may be found by direct experiment by means of pyrometers. These are thermometers so constructed as to measure high temperatures.

-

Pyrometry. This term is applied to the measurement of high temperatures. It is not possible, it is true, to attain as accurate measurements of high temperatures as it is of low ones by the aid of thermometers; nevertheless, very precise determinations have been made, and, as the measurement of high temperatures is of great importance in metallurgy, some attention must be devoted to the principles on which pyrometry is based.

The author has elsewhere pointed out that, notwithstanding the importance attached by early experimenters to the action of * Lecture at the Royal Institution. Nature, vol. xlv. 1892 p. 534.

heat on metals, they had but little definite information respecting the relative degrees of intensity of heat; and their views were not inadequately expressed in the eighth century by Geber, who stated that great difficulties arose in conducting operations with the aid of heat, because heat cannot be measured, "sed quoniam non est res ignis, quæ mensuari possit."* The name of Josiah Wedgewood is always associated with the early attempts to provide a practical method of pyrometry; but, although he wrote a thousand years after Geber, he seems to have merely changed the language of the latter, "heat cannot be measured," into a lament that there were no trustworthy instruments for effecting the measurement of "the higher degrees of heat, from a red heat up to the strongest that vessels made of clay can support." He therefore devised a pyrometer, which depended on the contraction that clay experiences when strongly heated. It is not necessary to give a history of pyrometry in this place; the author would merely point out that Wedgewood demonstrated the necessity for the accurate measurement of high temperatures, and that, from his time, the invention of more or less suitable instruments has proceeded rapidly.

In the long interval between the work of Wedgewood and that of the late Sir William Siemens, such pyrometric appliances as were actually in industrial use depended mainly on what is known as the "method of mixtures,"-that is, upon the employment of a body, the specific heat of which was known, to transfer or carry heat from a furnace, or source of heat, to a measured volume of water, the rise in temperature of which was indicated by an ordinary thermometer. Pyrometers depending on the expansion of metallic strips or rods are also employed; but they may all be set aside with the general statement that, although they are useful in affording rough approximate measurements, they are useless for accurate pyrometry. It may be well to add that a noteworthy advance in thermometry has recently been made in Professor Ramsay's laboratory at University College, by Messrs. Baly and Chorley, who employ the fluid alloy of sodium aud potassium, instead of the mercury of the ordinary thermometer; and, as temperatures of 600° Centigrade may thus be measured by making the thermometers of very hard glass, further details of their experiments will be looked for with much interest.

Weinhold has thus briefly set forth the principles on which pyrometers have been constructed:

From the edition of his Summa Perfectionis Magisterii, published in Venice, 1542, p. 28. (Additions have probably been made to Geber's original text.)

Phil. Trans. Royal Soc., vol. 1xxii. 1782, p. 305.

Pogg. Ann. vol. cxlix. (1873), p. 186.