This difference, which is by no means very considerable, may arise from errors which are unavoidable in experiments of this nature, or it may be owing to the composition of oil not being as yet exactly ascertained. It proves, however, that there is a great agreement between the results of our experiments, respecting the combination of caloric, and those which regard its disengagement. The following desiderata still remain to be determined, viz. What quantity of caloric is retained by oxygen, after combining with metals, so as to convert them into oxyds; What quantity is contained by hydrogen, in its different states of existence; and to ascertain, with more precision than is hitherto attained, how much caloric is disengaged during the formation of water, as there still remain considerable doubts with respect to our present determination of this point, which can only be removed by farther experiments. We are at present occupied with this inquiry; and, when once these several points are well ascertained, which we hope they will soon be, we shall probably be under the necessity of making considerable corrections upon most of the results of the experiments and calculations in this Chapter. I did not, however, consider this as a sufficient reason for withholding so much as is already known from such as may be inclined to labour upon the same subject. It is difficult, in our endeavours to discover the principles of a new science, to avoid beginning by guess-work; and it is rarely possible to arrive at perfection from the first setting out. CHAP. X. _Of the Combination of Combustible Substances with each other._