ll if the air at the ceiling line is much hotter. Hence it occurs that,
in warmed rooms of such size as I have mentioned, where one or two
petroleum lamps are used for lighting them, after two or three hours of
occupation by a family of three or four persons in winter weather, the
air at the ceiling line has become so poisonous that a bird dies if
allowed to breathe it for a very short time--sometimes, indeed, for only
a few minutes. With candles, if the illumination of the room is
maintained at the same degree as in the case of lamps, the contamination
of the air is very much worse. It is doubtless the case that poisonous
germs are rapidly developed in atmospheres which are called "stuffy;"
and although, in a healthy state of the body, we are able to breathe
them without perceptible harm, yet even then the slight headache and
uneasiness we feel is a symptom which does not suffer itself to be
lightly regarded, whenever, from some cause or other, the general
condition is weak.

The products of combustion from coal gas (which are steam and carbonic
acid mixed with an infinitesimal quantity of sulphur) are,
proportionately, far less injurious to animal life than the products
from an equal illuminating power derived from either oil or candles.
They are, however, it is certain, destructive to germ life; and
therefore, if taken off from the ceiling level, where they always
collect if allowed to do so, no possible inconvenience or danger to
health can be felt by any one in the room. But in our endeavors to take
off the foul air at the ceiling, we encounter our first serious check in
all schemes of ventilation. We draw the elevation and section of the
room, and put in our flues with pretty little black arrows flying out of
the outlets for vitiated air, and other pretty little red arrows flying
in at the inlets; but when we see our scheme in practice, the black
arrows will persist in putting their wings where their points ought to
be; in other words, flying into instead of out of the room.

One of the best ways of finding the true course of all the hot and cold
currents in a room is to make use of a small balloon, such as used to be
employed for ascertaining the specific gravity of gases; and, having
filled it with ordinary coal gas, balance it by weights tied on to the
car till it will rest without going up or down in a part of the room
where the air can be felt to be at about the mean temperature, and free
from draught. Then