according to its degree of endothermicity, an endothermic substance is more or less unstable, for it is always ready to emit heat, or to do work, as soon as an opportunity is given to it to decompose. The theoretical and practical results of this circumstance will be elaborated in Chapter VI., when the endothermic nature of acetylene is more fully discussed. A very simple experiment will show that a notable quantity of heat is set free when calcium carbide is brought into contact with water, and by arranging the details of the apparatus in a suitable manner, the quantity of heat manifested may be measured with considerable accuracy. A lengthy description of the method of performing this operation, however, scarcely comes within the province of the present book, and it must be sufficient to say that the heat is estimated by decomposing a known weight of carbide by means of water in a small vessel surrounded on all sides by a carefully jacketed receptacle full of water and provided with a sensitive thermometer. The quantity of water contained in the outer vessel being known, and its temperature having been noted before the reaction commences, an observation of the thermometer after the decomposition is finished, and when the mercury has reached its highest point, gives data which show that the reaction between water and a known weight of calcium carbide produces heat sufficient in amount to raise a known weight of water through a known thermometric distance; and from these figures the corresponding number of large calories may easily be calculated. A determination of this quantity of heat has been made experimentally by several investigators, including Lewes, who has found that the heat evolved on decomposing 1 gramme of ordinary commercial carbide with water is 0.406 large calorie. [Footnote: Lewes returns his result as 406 calories, because he employs the "small calorie." The small calorie is the quantity of heat needed to raise 1 gramme of water 1° C.; but as there are 1000 grammes in 1 kilogramme, the large calorie is equal to 1000 small calories. In many respects the former unit is to be preferred.] As the material operated upon contained only 91.3 per cent. of true calcium carbide, he estimates the heat corresponding with the decomposition of 1 gramme of pure carbide to be 0.4446 large calorie. As, however, it is better, and more in accordance with modern practice, to quote such data in terms of the atomic or molecular weight of the substance concerned, and as the molecular weight of calcium carbide is 64, it is preferable to multiply these figures by 64, stating that, according to Lewes' researches, the heat of decomposition of "1 gramme- molecule" (_i.e._, 64 grammes) of a calcium carbide having a purity of 91.3 per cent. is just under 26 calories, or that of 1 gramme-molecule of pure carbide 28.454 calories. It is customary now to omit the phrase "one gramme-molecule" in giving similar figures, physicists saying simply