Low-grade pyrolusite manganese ore from the Paint Bank deposits of Virginia was prepared for flotation by grinding to different degrees of fineness. The coarsest size was a 28 Tyler Mesh and the finest was 100 Tyler mesh. Under size products of the grinding were as fine as minus 200 Tyler Mesh. Flotation tests were run on the ore in a Denver Sub-A Flotation Cell of 500 gram capacity. Each run was made with an amount of ore closely approximating this capacity. The flotation reagents were added in a definite order allowing three minutes between each addition. A conditioning period of fifteen minutes followed the addition of the reagents in each case. Crude tallol was tested as a collecting agent for the manganese dioxide in the ore, with auxiliary reagents. Runs were made varying the amounts of tallol, pine oil, frother, sodium silicate silica depressor, and the hydrogen ion concentration was varied with either sulphuric acid or hydrated lime. Crude oleic acid from tallol was also tested as a collecting agent for the manganese dioxide in the ore, with auxiliary reagents. Runs were made varying the amounts of crude oleic acid, sodium silicate depressor, and the hydrogen ion concentration was varied with either sulphuric acid or hydrated lime. Sulphonated tallol was also tested as a collecting agent, but the preliminary results did not seem to justify continuing with an extensive investigation. The grade of concentrate was fairly high but the yield of concentrate was less than two per cent of the original head sample. The crude oleic acid was found to be a better collector than the crude tallol, although neither one had much effect on the grade of the concentrate obtained. Sodium silicate had no beneficial action on the grade of concentrate, and with the crude oleic acid, it had a definitely deterrent effect. It was indicated that this was due to its coating the silica with a film of manganese dioxide, thus causing the silica to be collected as a particle of manganese dioxide. The sodium silicate was found to be useful in increasing the per cent recovery of concentrate. Pine oil, while not having any affect on the grade of concentrate, was effective in increasing the yield of concentrate. In general, an acid medium was found conducive to obtaining high grade concentrates, but this was more than offset by the low yields encountered. An alkaline medium was found to give the best overall recovery of the manganese dioxide. It was also indicated that the ore should be crushed through at least 35 mesh, and preferably 65 mesh, in order to obtain the greatest recovery. / Master of Science