Hence we can only operate upon very small quantities, such as ten or twelve grains of alkohol, in this manner; and the errors which may be committed in experiments upon such small quantities prevents our placing any confidence in their results. I endeavoured to prolong the combustion, in the experiments contained in the Memoirs of the Academy for 1784, p. 593. by lighting the alkohol first in common air, and furnishing oxygen gas afterwards to the jar, in proportion as it consumed; but the carbonic acid gas produced by the process became a great hinderance to the combustion, the more so that alkohol is but difficultly combustible, especially in worse than common air; so that even in this way very small quantities only could be burnt. Perhaps this combustion might succeed better in the oil apparatus, Pl. XI.; but I have not hitherto ventured to try it. The jar A in which the combustion is performed is near 1400 cubical inches in dimension; and, were an explosion to take place in such a vessel, its consequences would be very terrible, and very difficult to guard against. I have not, however, despaired of making the attempt. From all these difficulties, I have been hitherto obliged to confine myself to experiments upon very small quantities of alkohol, or at least to combustions made in open vessels, such as that represented in Pl. IX. Fig. 5. which will be described in Section VII. of this chapter. If I am ever able to remove these difficulties, I shall resume this investigation. SECT. VI. _Of the Combustion of Ether._ Tho' the combustion of ether in close vessels does not present the same difficulties as that of alkohol, yet it involves some of a different kind, not more easily overcome, and which still prevent the progress of my experiments. I endeavoured to profit by the property which ether possesses of dissolving in atmospheric air, and rendering it inflammable without explosion. For this purpose, I constructed the reservoir of ether a b c d, Plate XII. Fig. 8. to which air is brought from the gazometer by the tube 1, 2, 3, 4. This air spreads, in the first place, in the double lid ac of the reservoir, from which it passes through seven tubes ef, gh, ik, &c. which descend to the bottom of the ether, and it is forced by the pressure of the gazometer to boil up through the ether in the reservoir. We may replace the ether in this first reservoir, in proportion as it is dissolved and carried off by the air, by means of the supplementary reservoir E, connected by a brass tube fifteen or eighteen inches long, and shut by a stop-cock. This length of the connecting tube is to enable the descending ether to overcome the resistance occasioned by the pressure of the air from the gazometer.