The fluid to be examined must be acidulated with acetic or hydrochloric acid. If the fluid be neutral or alkaline, the acid may be added at once. If on the other hand the fluid redden litmus, and the acid be either unknown or a mineral acid, potash must first be added in a slight excess, and then the alkali must be supersaturated with acetic or hydrochloric acid. The reasons for these precautions are stated under hydrosulphuric acid as a liquid reagent. The fluid being thus prepared, it is subjected to a stream of hydrosulphuric acid gas for ten or fifteen minutes. The first portions of the gas turn the arsenical solution to a bright lemon-yellow colour, and the subsequent portions throw down a yellow flocculent sulphuret of arsenic. If the proportion of oxide in solution is small, a yellowness or yellow milkiness only is caused, owing to the sulphuret being soluble in an excess of hydrosulphuric acid. But on expelling that excess by boiling, a distinct precipitate and colourless fluid are produced. The precipitate is then to be collected thus. The precipitate is allowed to subside, and the supernatant fluid being withdrawn, the remainder is poured into a filter. When all the fluid has passed through, the portions of precipitate on the upper part of the filter are washed down to the bottom. The filter is then gently compressed between folds of bibulous paper, and the sulphuret removed with the point of a knife before it dries, and dried in little masses on a watch-glass by the side of a chamber-fire, or still better in a vapour-bath. In this way it is very easy to collect a twenty-fifth part of a grain of the sulphuret. Another method which takes more time, but will enable the least skilful person to collect extremely small quantities, is to allow the sulphuret to subside in the original fluid in which it is formed, to pour off the supernatant liquid, and pour the remainder into a small glass tube, Fig. 7. After the precipitate has thoroughly subsided, the supernatant liquid is to be withdrawn, and its place filled up with boiling water. The operation of alternate subsidence and affusion being repeated a sufficient number of times, the last portions of water should be gently driven off by heat, and wiped off the inside of the tube as the drops condense on it. Finally, the bottom of the tube, with the precipitate attached, is to be cut away with the file, and broken into small fragments with the view of preserving the whole sulphuret for the process of reduction. The sulphuret having been collected in either of these ways, it is now to be dropt into the tube, Fig. 3, and covered by means of the funnel, Fig. 4, with soda-flux. The process in other particulars is the same with that for reducing solid oxide of arsenic.