In an article on the use of pulverized fuel we recently stated the conditions necessary to rapid and complete combustion. When those conditions—namely, fine division of fuel, intimate mixture of the proper proportion of oxygen as it exists in the mixed gases of the atmosphere, and sufficiently elevated temperature are maintained—there would be no smoke p/oduced. What would then pass off through a chimney would be carbonic acid, the nitrogen of the air, the sulphurous gases, and volatilized impurities of the coal or other fuel employed. The constitution of smoke in the ordinary acceptation of the term, is a mixture of unconsumed gases with particles of solid carbon, called soot. Some of the gases, carbonic oxide, etc., are combustible, as is also the soot. These combustible materials may be utilized, and it is the object of all smoke-consuming furnaces, stoves, etc., to consume them. It is the object of this article to investigate the general principles which must be observed to obtain perfect combustion of smoke. We have already shown that the admission of a proper quantity of oxygen is an essential condition, but as the chemical character of smoke varies according to the kind of fuel used, and the more or less perfect character of the combustion of the fuel which produces the smoke, it is obvious that this must be taken into account. It would by no means be correct to infer that because a furnace will consume the smoke and gases from a prime quality of anthracite coal, that the same furnace similarly adjusted would burn completely the dense smoke of bituminous coal. A perfect smoke-consuming apparatus ought, therefore, to take into account the amount of oxygen needed to consume different kinds of smoke, and be made adjustable thereto, accrding as circumstances may require. Although a small excess of air would not be likely to interfere with perfect combustion it will be plain, from the following considerations, that there must be both an economical limit to such excess, and another limit beyond which success would be defeated. We have seen that a proper temperature must be attained in order that combustion may commence, and that it must be maintained in order that combustion may continue. The capacity of gases for heat is much greater than that of solids; hence every pound of air not necessary to combustion admitted to contact with the combustible gases and floating carbon in smoke, will subtract heat and carry it away through the chimney. The economical limit is then reached as soon as complete combustion is attained. But if cold air in large quantity is mingled with smoke, the gases may become so cooled by the contact as to fall below the temperature of ignition, and thus prevent combustion, condensation of the gases taking place after a short time. This is illustrated by the smoking of a gas flame when placed in strong and irregular currents of air, and the instant deposit of soot when a cold body is placed in such a flame. Let the body thus placed in contact with the flame of an ordinary gas burner be of sufficient size and of sufficient conducting power to maintain a temperature at the point in contact with the flame below the point of ignition, and a continuous deposit and accumulation of unburned carbon will take place. The same is true wlien the body is a poor conductor of heat, but a rapid conveyer. This is the case with a thin glass vessel filled with water, suspended over a gas flame and in contact with it. The particles of water here rapidly convey away heat from the bottom of the vessel, while a deposit of carbon on the bottom immediately takes place. It is evident, then, that when it is possible to supply heated air to smoke in precisely the quantities needed, the best conditions for perfect combustion are obtained. It is further evident that because a furnace completely consumes its smoke, it is not therefore necessarily an economical furnace, since it is possible to waste more heat through a chimney by excessive draft than the amount gained by the consumption of smoke. Many important attempts have been made to secure perfect combustion of smoke, and so far as these apply to small furnaces where a complicated apparatus is out of the question, or to illy-constructed furnaces, to supplement imperfect action, they are many of them complete successes. Until the beau ideal of a furnace is attained, cheap, uncomplicated, and efficient, that shall burn fuel so completely that no smoke shall be generated, they will be valuable. The use of pulverized fuel, the only way in our opinion whereby the generation of smoke can be obviated, has hitherto entailed expense only possible to the larger kinds of establishments. It would seem, ' however, that the limits of possibility do not exclude a cheaper method of using pulverized fuel capable of more gonerla application, and of being employed on a smaller scale than has yet been accomplished.