ignited, the blast is turned on, and the mixture of nitrogen and carbon monoxide passes over the iron, heating it to a red heat, while the fuel in contact with the retort does the same thing. When the fuel and retort full of iron are at a cherry-red heat, the air blast is cut off, and the pipe connecting the furnace and retort, together with the pipe in connection with the bottom of the retort, are closed, and steam, superheated by passing through a pipe led round the retort or interior wall of the furnace, is injected at the bottom of the red-hot mass of iron, which decomposes it, forming magnetic oxide of iron and hydrogen, which escapes by the second tube at the top of the retort, and is led away either to a carbureting chamber if required for illumination, or direct to the gasholder if wanted as a fuel. The mass of incandescent fuel in the furnace chamber, surrounding the retort, keeping up the temperature of retort and iron sufficiently long to enable the decomposition to be completed. The hydrogen and steam valves are now closed and the air blast turned on. The hot carbon monoxide passing over the hot magnetic oxide quickly reduces it down to metallic iron, which, being in a spongy condition, acts more freely on the steam during later makes than it did at first, and being infusible at the temperature employed, may be used for a practically unlimited period. What more simple method than this could be desired? Here we have the formation of the most valuable of all fuel gases at the cost of the coke and steam used, a gas also which has double the carrying power for hydrocarbon vapors possessed by coal gas, while its combustion gives rise to nothing but water vapor. In this course of lectures I have left much unsaid and undone which I should have liked to have had time to accomplish, and if I have been obliged to leave out of consideration many important points, it is the time at my disposal and not my will which is to blame. And now, in conclusion, I wish to express my thanks to my assistants, Messrs. J.A. Foster and J.B. Warden, who have heartily co-operated with me in much of the work embodied in these lectures. * * * * * STEREOSCOPIC PROJECTIONS.