In order to approximately ascertain the relative reducing action of water gas, carbon monoxide, and superheated steam on iron ore, the author decided to have carried out the following experiments, which were conducted by Mr. Carl J. Sandahl, of Stockholm, who also carried out the analyses. The ore used was from Bilbao, and known as the Ruby Mine, and was a good average hematite. The carbonaceous material was the Trimsaran South Wales anthracite, and contained about 90 per cent. of carbon. A small experimental furnace was constructed of the form shown by illustration, about 4 ft. 6 in. high and 2 ft. 3 in. wide at the base, and gradually swelling to 2 ft. 9 in. at the top, built entirely of fireclay bricks. Two refractory tubes, 2 in. square internally, and the height of the furnace, were used for the double purpose of producing the gas and reducing the ore. The end of the lower tube rested on a fireclay ladle nozzle, and was properly jointed with fireclay; through this nozzle the steam or air was supplied to the inside of the refractory tubes. In each experiment the ore and fuel were raised to the temperature "of from 1,800 to 2,200 deg. Fahr." by means of an external fire of anthracite. Great care was taken to prevent the contact of the solid carbonaceous fuel with the ore. In each experiment in which steam was used, the latter was supplied at a temperature equivalent to 35 lb. to the square inch. The air for producing the carbon monoxide (CO) gas was used at the temperature of the atmosphere. As near as possible, the same conditions were obtained in each experiment, and the equivalent weight of air was sent through the carbon to generate the same weight of CO as that generated when steam was used for the production of water gas. [Illustration] _First Experiment, Steam (per se)_.--Both tubes, A and B, were filled with ore broken to the size of nuts. The tube, A, was heated to about 2,000 deg. Fahr., the upper one to about 1,500 deg. NOTE.--In this experiment, part of the steam was dissociated in passing through the turned-up end of the steam supply pipe, which became very hot, and the steam would form with the iron the magnetic oxide (Fe_{3}O_{4}). The reduction would doubtless be due to this dissociation. The pieces of ore found on lowest end of the tube, A, were dark colored and semi-fused; part of one of these pieces was crushed fine, and tested; see column I. The remainder of these black pieces was mixed with the rest of the ore contained in tube, A, and ground and tested; see column II. The ore in upper tube was all broken up together and tested; see column III. When finely crushed, the color of No. I. was bluish black; No. II., a shade darker red; No. III., a little darker than the natural color of the ore. The analyses gave: