The element manganese having many peculiarities in its reactions with the other elements, is now extensively used in the arts, its combinations entering into and are used in many of the important processes; it is consequently often brought before the chemist in his analysis, and has to be determined in most cases with considerable accuracy. Many methods have been proposed for this, all of them of more or less value; those yielding the best results, however, requiring a considerable length of time for their execution, and involving so large an amount of manipulatory skill as to render them fairly impracticable to a chemist at all pressed for time, and receiving but a mere trifle for the results. As I have had to make numerous estimations of manganese in various compounds, as a public analyst, I have been induced to investigate the volumetric methods at present in use to find their comparative values, and if possible to work out a new one, setting aside one or more of the difficulties met with in the use of the older ones. This paper is a part summary of the results. First, I will detail my process of estimation, then on the separation. From all compounds of manganese, excepting those containing cobalt and nickel, the manganese is precipitated as binoxide; those containing these two elements are treated with phosphoric acid, or as noted under Separation. A.--The Estimation. The binoxide of commerce, as taken from the mine, is well sampled, powdered, and dried at 100°C. 0.5 gramme of this is taken and placed in a 250 c.c. flask; in analysis the binoxide on the filter, from the treatments noted under separation is thoroughly washed with warm water; it is then washed down in a flask, as above, after breaking the filter paper; sufficient water is added to one-third fill the flask, and about twice the approximate weight of the binoxide in the flask of oxalate of potassa; these are agitated together. A twice perforated stopper is fitted to this flask, carrying through one opening a 25 c c. pipette nearly filled with sulphuric acid, sp. gr. 1.4, the lower point of which just dips below the mixture in the flask, and the upper end, carrying a rubber tube and pinch cock to control the flow of acid. Through the other opening passes a glass tube bent at an acute angle and connected by a short rubber tube to an adjoining flask, two-thirds filled with decinormal baryta solutions. These connections are all made air tight. Sulphuric acid is allowed in small portions at a time to flow into the mixture. Carbonic acid is evolved, and, passing into the adjoining flask, is absorbed by the baryta, precipitating it as carbonate. To prevent the precipitate forming around or choking up the entrance tube, the flask must be agitated at short intervals to break it off. The reaction so familiar to us in other determinations is expressed thus: MnO_{2}+KO,C_{2}O_{3}+2SO_{3} = MnO,SO_{3}+KO.SO_{3}+2CO_{2},