I shall not enter into a more detailed description of the construction of this apparatus, which is still capable of being altered and modified in many respects, but shall only add, that when it is to be used in experiment, the lamp and reservoir with the contained oil must be accurately weighed, after which it is placed as before directed, and lighted; having then formed the connection between the air in the gazometer and the lamp, the external jar A, Pl. XI. is fixed over all, and secured by means of the board BC and two rods of iron which connect this board with the lid, and are screwed to it. A small quantity of oil is burnt while the jar is adjusting to the lid, and the product of that combustion is lost; there is likewise a small portion of air from the gazometer lost at the same time. Both of these are of very inconsiderable consequence in extensive experiments, and they are even capable of being valued in our calculation of the results. In a particular memoir, I shall give an account to the Academy of the difficulties inseparable from this kind of experiments: These are so insurmountable and troublesome, that I have not hitherto been able to obtain any rigorous determination of the quantities of the products. I have sufficient proof, however, that the fixed oils are entirely resolved during combustion into water and carbonic acid gas, and consequently that they are composed of hydrogen and charcoal; but I have no certain knowledge respecting the proportions of these ingredients. SECT. V. _Of the Combustion of Alkohol._ The combustion of alkohol may be very readily performed in the apparatus already described for the combustion of charcoal and phosphorus. A lamp filled with alkohol is placed under the jar A, Pl. IV. Fig. 3. a small morsel of phosphorus is placed upon the wick of the lamp, which is set on fire by means of the hot iron, as before directed. This process is, however, liable to considerable inconveniency; it is dangerous to make use of oxygen gas at the beginning of the experiment for fear of deflagration, which is even liable to happen when common air is employed. An instance of this had very near proved fatal to myself, in presence of some members of the Academy. Instead of preparing the experiment, as usual, at the time it was to be performed, I had disposed every thing in order the evening before; the atmospheric air of the jar had thereby sufficient time to dissolve a good deal of the alkohol; and this evaporation had even been considerably promoted by the height of the column of mercury, which I had raised to EF, Pl. IV. Fig. 3. The moment I attempted to set the little morsel of phosphorus on fire by means of the red hot iron, a violent explosion took place, which threw the jar with great violence against the floor of the laboratory, and dashed it in a thousand pieces.