Upon these assumptions, Dr. Siemens condemns steam-pipes and stoves, hot-air pipes, and all other methods of directly heating the _air_ of apartments, and thereby making it warmer than were the walls, the ceiling, and furniture when the process of warming commenced. It is quite true that stoves, stove-pipes, hot-air pipes, steam-pipes, etc., do this; they raise the temperature of the air directly by _convection_, _i.e._, by warming the film of air in contact with their surfaces, which film, thus heated and expanded, rises towards the ceiling, and, on its way, warms the air around it, and then is followed by other similarly-heated ascending films. When we make a hole in the wall, and burn our coals within such cavity, this convection proceeds up the chimney in company with the smoke. But is Dr. Siemens right in saying that the air of a room, raised by convection above its original temperature, and above that of the walls, deposits any of its moisture on these walls? I have no hesitation in saying very positively that he is clearly and demonstrably wrong; that no such condensation can possibly take place under the circumstances. Suppose, for illustration sake, that we start with a room of which the air and walls are at the freezing point, 32° F., before artificial heating (any other temperature will do), and, to give Dr. Siemens every advantage, we will further suppose that the air is fully saturated with aqueous vapor, _i.e._, just in the condition at which some of its water might be condensed. Such condensation, however, can only take place by cooling the air _below_ 32°, and unless the walls or ceiling or furniture are capable of doing this they cannot receive any moisture due to such condensation, or, in other words, they must fall below 32° in order to obtain it by cooling the film in contact with them. Of course Dr. Siemens will not assert that the stoves or steam-pipes (enclosing the steam, of course), or the hot-air or hot-water pipes, will lower the _absolute_ temperature of the walls by heating the air in the room. But if the air is heated more rapidly than are the walls, etc., the _relative_ temperature of these will be lower. Will condensation of moisture _then_ follow, as Dr. Siemens affirms? Let us suppose that the air of the room is raised from 30° to 50° _by convection purely_; reference to tables based on the researches of Regnault, shows that at 32° the quantity of vapor required to saturate the air is sufficient to support a column of 0·182 inch of mercury, while at 50° it amounts to 0·361, or nearly double. Thus the air, instead of being in a condition of giving away its moisture to the walls, has become thirsty, or in a condition to _take moisture away from them_ if they are at all damp. This is the case whether the walls remain at 32° or are raised to any higher temperature short of that of the air.