After several trials of different materials, I have found marble the best substance for constructing the mercurial pneumato-chemical apparatus, as it is perfectly impenetrable by mercury, and is not liable, like wood, to separate at the junctures, or to allow the mercury to escape through chinks; neither does it run the risk of breaking, like glass, stone-ware, or porcelain. Take a block of marble BCDE, Plate V. Fig. 3. and 4. about two feet long, 15 or 18 inches broad, and ten inches thick, and cause it to be hollowed out as at m n Fig. 5. about four inches deep, as a reservoir for the mercury; and, to be able more conveniently to fill the jars, cut the gutter T V, Fig. 3. 4. and 5. at least four inches deeper; and, as this trench may sometimes prove troublesome, it is made capable of being covered at pleasure by thin boards, which slip into the grooves x y, Fig. 5. I have two marble cisterns upon this construction, of different sizes, by which I can always employ one of them as a reservoir of mercury, which it preserves with more safety than any other vessel, being neither subject to overturn, nor to any other accident. We operate with mercury in this apparatus exactly as with water in the one before described; but the bell-glasses must be of smaller diameter, and much stronger; or we may use glass tubes, having their mouths widened, as in Fig. 7.; these are called _eudiometers_ by the glass-men who sell them. One of the bell-glasses is represented Fig. 5. A. standing in its place, and what is called a jar is engraved Fig. 6. The mercurial pneumato-chemical apparatus is necessary in all experiments wherein the disengaged gasses are capable of being absorbed by water, as is frequently the case, especially in all combinations, excepting those of metals, in fermentation, &c. SECT. II. _Of the Gazometer._ I give the name of _gazometer_ to an instrument which I invented, and caused construct, for the purpose of a kind of bellows, which might furnish an uniform and continued stream of oxygen gas in experiments of fusion. Mr Meusnier and I have since made very considerable corrections and additions, having converted it into what may be called an _universal instrument_, without which it is hardly possible to perform most of the very exact experiments. The name we have given the instrument indicates its intention for measuring the volume or quantity of gas submitted to it for examination.