It may be taken for granted that air which has been once breathed is unfit for further respiration. Such air, we have seen, contains about 4 1/2 per cent of carbon dioxide, and experiment has shown that a far smaller proportion is poisonous, and it is generally admitted that a disagreeable or stuffy sensation is perceived by man when the air in a room contains about 1 per cent of carbon dioxide in 1000 of pure air. It is probable, however, that the feeling of closeness is not altogether due to the carbon-dioxide, but is in part owing to the exhalation from the skin and to the volatile products resulting from chemical changes in the body that are discharged from the lungs. M. Boussingault has estimated that a horse of average size eliminates 4800 litres, or between 8000 and 9000 pints, of carbon dioxide per diem, from which it is clear that a very large quantity of pure air must be supplied to prevent the proportion of the C02 from rising above 1 per cent. The object of good ventilation is to effect this renewal of the air without creating a draught, or, in cold periods of the year, causing a sudden inrush of air at a low temperature. In the case of the human subject it is held that to preserve the air in a fit state of purity, with ordinary or natural ventilation, at least 2000 cubic feet of .space should be allowed for each individual, and a much larger space, as might be expected, is required for the horse; but as a set-off to this it may be remarked that it has been found that the horse can bear without injury a larger percentage of C02 than man, and secondly, the construction and arrangements of stables are more favourable to natural ventilation than are the majority of human habitations. So if the space occupied by stables be somewhat smaller than those built for man, the renewal of the air is more effectually accomplished.

The term " natural ventilation" has been suggested by Pettenkofer, who has shown that, besides the interchange of air and gases that takes place through open windows and doors, or through the cracks and interspaces of boards and ill-fitting frames of windows and doors when these are closed, the walls of most dwelling-houses, when composed of brick or sandstone and mortar, allow air to pass through them with much greater facility than is usually believed. Thus it is found by experiment that the natural pores of such walls allow of the passage of 2000 cubic feet of air through -

42 square feet of a free wall of sandstone.

30 ,, „ „ quarried limestone.

25 „ „ „ brick.

19 „ „ „ tufaceous limestone.

14 „ • „ „ mud.

Two circumstances are of great importance in regard to the activity of natural ventilation - the relative temperature of the air within and without, and the presence of moisture in the walls. Moisture, by filling up the pores, greatly obstructs, if it does not altogether prevent, the passage of air through walls. It is present in large quantities in new buildings, in which, therefore, no natural ventilation takes place, and which are proverbially unhealthy.

Pettenkofer found that in a room 14 feet square and 14 feet high, with a difference of temperature of 34° F. (66° F. inside, and 32° F. outside), the contents changed once in an hour; with a good fire in the stove it increased 25 per cent, and even when all apparent apertures and crevices were closed it only diminished 28 per cent. ' A stable built of mud is capable of permitting the entrance of a considerable volume of fresh air by natural ventilation, and can thus afford shelter to more animals than one built of sandstone. The activity with which the exchange of air is secured by natural and wholesome ventilation in a stable does not depend upon its cubic capacity, but upon the extent and nature of its ventilating walls and appliances, and hence a small stable built of porous material may secure better ventilation than a large one, partly because for each animal there is more ventilating surface with equal cubic space, and partly because the air permeates the walls more readily.