In the order of the conversion of the elements it appears to me, that water is to air what air is to fire, and that all the transformations of nature depend on them. Air, like the food of fire, assimilates with it, and is transformed into this first element. Water, rarefied by heat, is transformed into a kind of air capable of feeding the fire like common air. Thus fire has a double fund of certain subsistence; if it consume much air it can also produce much by the rarefaction of water, and thus repair, in the mass of atmosphere, all the quantity it destroyed, while ulteriorly it converts itself with air into fixed matter in the terrestrial substances which it penetrates by its heat or by its light. And so, likewise, as water is converted into air, or into vapours, as volatile as air, by its rarefaction, it is also converted into a solid substance by a kind of condensation. Every fluid is rarefied by heat and condensed by cold. Water follows this common law, and condenses as it grows cold. Let a glass tube be filled three parts full and it will descend in proportion as the cold increases, but some time before congelation it will ascend above the point of three fourths of the height of the tube, and increase still more considerably by being frozen. But if the tube be well stopped, and perfectly at rest, the water will continue to descend, and will not freeze, although the degree of cold be six, eight, or ten degrees below the freezing point; congelation, therefore, presents, in an inverted manner, the same phenomena as inflammation. A heat, however great, shut up in a well-closed vessel, will not produce inflammation unless touched with an inflamed matter; so, likewise, to whatsoever degree a fluid is cooled, it will not freeze unless it touch something already frozen, and this is what happens when the tube is shaken or uncorked; the particles of water, which are frozen in the external air, or in the air contained in the tube, strike the surface of the water, and communicate their ice to it. In inflammation, the air, at first very much rarefied by heat, loses its volume, and fixes itself suddenly. In congelation, water, at first condensed by the cold, takes a larger volume, and fixes itself likewise, for ice is a solid substance, lighter than water, and would preserve its solidity if the cold continued the same; and I am inclined to believe that we may attain the point of fixing mercury at a less degree of cold, by sublimating it into vapours in a very cold air; and also that water, which only owes its liquidity to heat, would become a substance much more solid and fusible, as it would endure a stronger and a longer time the rigour of the cold.