Thus the products of the combustion of gas (which are principally steam) serve a useful purpose in lighting, by keeping at the ceiling level a certain stratum of heated vapor, which holds up, as it were, the carbonic acid and exhalation from the lungs given off by those using the room.  The obvious inference, therefore, is that if we take off these products from the level of the ceiling, we shall take off at the same time the impure and vitiated air.  On the other hand, if we make use of a system of artificial lighting, which does not produce any steam, then we shall have to adopt means to keep the air at the ceiling level warm, in order to prevent the heated impure air from descending in comparatively rapid currents, after having parted with its heat to the ceiling.  It may very frequently be observed on chilly days that a number of currents of cold air seem to travel about our rooms, although there may be no crevices in the doors and windows sufficient to account for them; and, further, that these currents of cold air are not noticed when the curtains are drawn and the gas is lighted.  The reason is that there is generally not enough heat at the ceiling level in a room unlighted with gas to keep these currents steady.  Hence the complaints of chilliness which we constantly hear when electric lights are used for the illumination of public buildings.  For example, at the annual dinner of the Institution of Civil Engineers, held at the end of April last in the Conservatory of the Horticultural Gardens, the heat from the five hundred guests, and from an almost equal number of waiters and attendants, displaced the cold air from the dome of the roof, and literally poured down on the assembly (who were in evening dress) in a manner to compel many of them to put on overcoats.  If the Conservatory had been lighted with gas suspended below the roof, this would not have been the case, because sufficient steam would have been generated to stop these cold douches, and keep them up in the roof.  In fact, if electric lights are to be used in such a building, it will be necessary to lay hot-water pipes in the roof, to keep warm the upper as well as the lower stratum of air, and thus steady the currents.

Having pointed out difficulties which arise under certain conditions of the atmosphere in rooms built with care, to make them comfortable when electric lighting is substituted for gas, I will lay before you some few particulars relative to the condition of small rooms of about 12 ft. by 15 ft. by 10 ft., or any ordinary room such as may be found in the usual run of houses in this country.  The cubical contents of such a room equals 1,700 cubic feet.  If the room is heated by means of a coal fire, we shall for the greatest part of the year have a quantity of air taken out of it at about 2 feet from the floor by the chimney draught, varying (according to atmospheric conditions and the state of the fire) from 600 to 2,000 or more cubic feet.  This quantity of air