arsenates only give a precipitate on long-continued boiling. Arsenic is usually estimated either in the form of magnesium pyroarsenate or as arsenic sulphide. For the pyroarsenate method it is necessary that the arsenic should be in the _arsenic_ condition, if necessary this can be effected by heating with nitric acid; the acid solution is then mixed with "magnesia mixture" and made strongly alkaline by the addition of ammonia. It is then allowed to stand twenty-four hours, filtered, washed with dilute ammonia, dried, ignited to constant weight and weighed, the filter paper being incinerated separately after moistening with nitric acid. From the weight of magnesium pyroarsenate obtained the weight of arsenic can be calculated. In the sulphide method, the arsenic should be in the _arsenious_ form. Sulphuretted hydrogen is passed through the liquid until it is thoroughly saturated, the excess of sulphuretted hydrogen is expelled from the solution by a brisk stream of carbon dioxide, and the precipitate is filtered on a Gooch crucible and washed with water containing a little sulphuretted hydrogen and dried at 100 deg. C.; it is then well washed with small quantities of pure carbon disulphide to remove any free sulphur, again dried and weighed. Arsenic can also be estimated by volumetric methods; for this purpose it must be in the _arsenious_ condition, and the method of estimation consists in converting it into the _arsenic_ condition by means of a standard solution of iodine, in the presence of a cold saturated solution of sodium bicarbonate. The atomic weight of arsenic has been determined by many different chemists. J. Berzelius, in 1818, by heating arsenious oxide with excess of sulphur obtained the value 74.3; J. Pelouze (_Comptes rendus_, 1845, 20, p. 1047) titrated arsenic chloride with silver solution and obtained 75.0; and F. Kessler (_Pogg. Ann._ 1861, 113, p. 134) by converting arsenic trisulphide in hydrochloric acid solution into arsenic pentasulphide also obtained 75.0.