As stated, the need and extent of ventilation is based on the presence of carbonic acid gas and its comparative volume. The freshest air, however, always carries some of this impurity, probably 2 parts in 10,000 volumes, while it rises to about 4 parts in towns ; this latter being, for towns, considered as pure air. In manufacturing centres this figure is exceeded, yet by good ventilation the standard of allowed impurity (about 8 parts to 10,000 in large centres) can be kept even in schools and rooms containing a full number of persons. Outside those districts having, naturally, a more or less poluted atmosphere the allowed degree of impurity is about 6 parts, or, more accurately, 2 parts per 10,000 in excess of the outer air.

There is no practical way of dealing with the carbonic acid gas so as to eliminate it from the air of rooms and so leave the contained air pure ; nor would this be desirable, for many reasons. Instead of this, ventilation is resorted to, so as to displace the whole air contents of the rooms, and replace these with quite new air from outside ; and ventilation may therefore be described as the substitution of fresh air for vitiated air. This, as will be seen, must be done to a sufficient extent to keep the proportion of carbonic acid gas (which is being constantly produced by the occupants of the rooms, gas-burners, etc.) down to the allowed limit of impurity; therefore the greater number of people there are in a room, or place, the greater must be the volume of air continually entering and passing through it. On this account the volume of air to be dealt with is always calculated at per person, also, where it appears desirable, per gas-burner.

An adult person gives off 0.5 to 0.6 cubic feet of carbonic acid per hour, and to maintain a standard of purity of 2 parts of C02 to 10,000 of air (in excess of the outside air) it is necessary that each adult, in health and in repose, have 2500 cubic feet of air per hour. If undergoing moderate exertion, an adult needs one-fourth more air, owing to his producing more C02, and this has to be remembered in factory and similar buildings. A gas-burner yields a greater volume of C02 than a man, and is calculated as being equal to three adults, at least, and should have fresh air accordingly. In schools the infants are commonly allowed 1500 feet of air per head, with 2000 feet for scholars of full age. In sickness the figures given should be increased 50 per cent.

Although scarcely within the province of the ventilating engineer, it may be stated that the effects of ventilation may be interfered with by the space allowed each person in a fully occupied place like a school. In board schools the minimum space is 100 cubic feet per head ; in factories and workshops it should be 250 feet ; in barracks 600 feet is allowed, while in hospitals 1500 to 2000 feet is usual. It follows also that floor area should bear some relation to space ; for, as a writer has said, people could become suffocated at the bottom of a well though the space above be unlimited. There is no gain in calculating space as being higher than 12 feet; and, for schools, the least floor area should be 8 square feet per scholar, while in hospitals each patient should have 100 square feet as a minimum. For ordinary purposes the floor area should be one-twelfth of the cubic space allowed.

In providing the needed supply of fresh air to rooms, regard has to be made to the velocity with which it enters and flows through. Cool air is perceivable as a disagreeable draught when travelling at a low rate compared with warmed air. It is usually considered that a movement of 3 feet per second is ordinarily too fast, though, should the air be raised to 620 or 65° Fahr., this rate of movement will not be noticed, at least not disagreeably. There are no objections to the air moving rapidly in the delivery ducts or tubes leading to a room, provided the movement is slow in the room itself. This is effected by making the delivery openings, with their gratings (or "registers"), of greater area than the ducts, usually in square trumpet-mouthed form, ensuring a gentle delivery and causing the air to spread out as it enters. For ordinary living apartments the velocity can be kept low at all points, as the air of such rooms does not require to be changed oftener than about three times per hour; but in fully occupied places, like schools, concert halls, etc., the question of velocity has to be considered in deciding the sizes of ducts and their delivery openings.

In dealing with the practical side of ventilating works, the subject must be divided into two parts, namely, that which is called Natural Ventilation and Mechanical Ventilation. In the former the air movement is obtained by flues, which might be called chimneys, from the fact that they have a regular active upward movement of air within them, and as such they are considered as extractors. They do not have such a powerful draught within them as chimneys have, as their temperature is so much lower ; but, broadly speaking, the air movement is sufficient to make this mode of obtaining the needed change of air occur satisfactorily. There are very strong exponents of natural ventilation, though in some cases reliance is put in the extractive action which takes place when wind blows over the upper end of a flue or duct. A very brief experiment with a piece of tube (glass, if convenient) open at both ends, but having a loose fitting ball of cotton-wool in it, will show that quite a light breeze over the upper end will cause the wool to rise. This movement is not due to any difference in temperature, but wholly to the exhaustion of air within the tube, by the horizontal movement of air over its top opening. It will be found that, by blowing slightly downward towards the top opening of the tube, there will still be a rise of air occur in it, but if the angle is too acute there will be a downward movement, like the familiar down-blow in chimneys.

The experiment will clearly show that the wind is an active, and sometimes very powerful, cause of an upward movement of air in a ventilating duct or flue, provided its upper end terminates high like a chimney, and does not, by terminating low, have down-blow occur in it. This latter detail has to be borne in mind, for while architects take every care to terminate smoke flues (chimneys) at sufficient height to prevent down-blow, they not infrequently give no consideration to this in the termination of air flues. It is necessary, as will be understood, if a natural up-draught is to be the means of effecting the desired air-change in the rooms below.