In the first moment of combustion the air, being heated, rarifies, and the mercury descends; but when, as in combustions of phosphorus and iron, no elastic fluid is formed, absorption becomes presently very sensible, and the mercury rises high into the jar. Great attention must be used not to burn too large a quantity of any substance in a given quantity of gas, otherwise, towards the end of the experiment, the cup would approach so near the top of the jar as to endanger breaking it by the great heat produced, and the sudden refrigeration from the cold mercury. For the methods of measuring the volume of the gasses, and for correcting the measures according to the heighth of the barometer and thermometer, &c. see Chap. II. Sect. V. and VI. of this part. The above process answers very well for burning all the concrete substances, and even for the fixed oils: These last are burnt in lamps under the jar, and are readily set on fire by means of tinder, phosphorus, and hot iron. But it is dangerous for substances susceptible of evaporating in a moderate heat, such as ether, alkohol, and the essential oils; these substances dissolve in considerable quantity in oxygen gas; and, when set on fire, a dangerous and sudden explosion takes place, which carries up the jar to a great height, and dashes it in a thousand pieces. From two such explosions some of the members of the Academy and myself escaped very narrowly. Besides, though this manner of operating is sufficient for determining pretty accurately the quantity of oxygen gas absorbed, and of carbonic acid produced, as water is likewise formed in all experiments upon vegetable and animal matters which contain an excess of hydrogen, this apparatus can neither collect it nor determine its quantity. The experiment with phosphorus is even incomplete in this way, as it is impossible to demonstrate that the weight of the phosphoric acid produced is equal to the sum of the weights of the phosphorus burnt and oxygen gas absorbed during the process. I have been therefore obliged to vary the instruments according to circumstances, and to employ several of different kinds, which I shall describe in their order, beginning with that used for burning phosphorus.