The "Natural" form of ventilation, while undoubtedly receiving aid by the passage of wind over its flues, does not rely on this action. Its chief reliance is on the difference in temperature of the air in the flues from that of the rooms and the outer air, this difference inducing a draught in just the same way as it happens in a chimney. A very useful table for ascertaining the volume of air that flues will deal with, and for deciding the sizes of flues, is as follows :-

Extraction of air in cubic feet per minute by a ventilating flue or shaft of 1 square foot area. These calculations make no allowance for friction, which will reduce the figures by one-third to one-half, according to the state of the interior surface of the flue. (From Hood on Warming Buildings, by Frederick Dye, M.R.I.)

 Height of the Flue inFeet. Difference in Temperature of Air entering the Flue and the outside Air. 15 o.F 20° F. 25° F. 30° F. 15 245 284 318 348 20 285 330 368 404 25 318 368 410 450 3° 347 403 450 493 35 376 436 486 531 40 403 465 518 570 45 427 493 551 605 50 45° 518 579 635

Architects have, apparently, a difference of opinion as to whether the flues or shafts carrying the warm vitiated air from rooms should go up as separate flues (as chimneys always do) or whether they should join at convenient points and ultimately terminate at top in one or more outlets of large area. The latter is the usual - and somewhat necessary - plan when the ventilating arrangements are introduced after the building is up; and when, of necessity, the flues have to be of sheet metal on the faces of the walls. If a natural draught is to be relied on there can be no doubt whatever that separate brick-built flues are best, but the other plan can be adopted when necessity makes it compulsory, or if it is proposed to use some auxiliary aid to assist the air movement. Every one having - to do with chimneys knows the peculiar, and commonly bad, results frequently obtained if two or more join one another, and the same must apply to air flues; but if something is introduced at the upper termination of the air flues to assist or induce the draught, then the joining of flues will give good results. The assistance referred to would be in the nature of a powerful gas ring or, more usually, a mechanically propelled air-extracting fan. Of course, the latter arrangement would come under the heading of mechanical ventilation.

Fig. 69,.

Assuming that a system of natural ventilation (and the same applies to mechanical ventilation) is accompanied by a means of warming the air, the methods adopted for this may be introduced here. No building, as a bank, a store, school, institution or large place of business, all of which would be systematically ventilated, is ever without a heating installation, and, in the majority of cases, this is arranged so as to warm the air as it enters. It would not be reasonable to warm such places without doing at least something to raise the temperature of the large volumes of air that the ventilating arrangements bring in. In imparting warmth to the air, draughts are not only avoided, but ventilation (natural) is aided, as will be understood.

One form of air-heating appliance which commonly appears is the ventilating radiator. This is a radiator which is visible in the room, but is directly connected with the fresh-air inlets, as Fig. 69 shows. This needs no further description than to say that, in finding the area of radiator surface to afford a certain temperature, 25 per cent, must be added, if the radiator is of the ventilating kind. These radiators have valves fitted in their bases to control the flow of air, and it is argued that by this means the radiator can be made to heat and reheat the air of the room, when there are few people in it and the vitiation trifling, thus effecting a saving in fuel compared with the heat units taken by full ventilation.

In Fig. 70 is afforded a view of an "indirect" radiator, not visible in the room and devoted wholly to warming the air before it enters. This is connected up as a radiator or stack of pipe would be, and is then encased soundly, and provided with inlet and outlet ducts or flues as shown. A radiator, so fixed, can be made to heat more than one room, if its size is suitable. It is only necessary to take two or more delivery ducts from the casing ; or to take one vertical duct of suitable size and branch it (with short branches) into rooms on floors above.

Fig. 70.

This illustration shows the warm air entering near the floor line. There is much difference of opinion regarding this point, and the subject will be found fully dealt with a little farther on. In no case should the delivery gratings or registers be in floors, as dirt gets swept into them. These radiators require to have 50 per cent, greater surface than the ordinary direct-heating non-ventilating kind, to afford a given temperature in the rooms.

The rule observed in calculating the sizes of ducts with indirect radiator work is to give them an area of 2 square inches for each square foot of radiator surface. Thus a 60 feet radiator would have ducts of 120 square inches, or say a tube or square form 17 by 7 inches. When delivery ducts go to upper floors, the proportion of 2 square inches per foot of surface is reduced. For a first floor 1 1/2 square inch, and a second floor 1 inch, will do. If required, the cold-air duct need not be of quite so large an area as the delivery tubes ; three-fourths to seven-eighths is sufficient.

In Fig. 71 is shown the plan of a warm-air heating installation in which a stove is used for affording the heat. The details are as follow :-

The heater has to be of a suitable make, having solid gills, or extended ribs and parts, to its exterior surface, the purpose of these extensions being to afford a largely increased heating area at a comparatively low temperature. In America these heaters are made in independent form with an iron air-casing around them, but in this country the stove is nearly always enclosed in a brick-built chamber. Fig. 72 affords a view of one of Musgrave's stoves with a longitudinal section of the chamber, showing some of the warm-air ducts leading from it. Although the chamber should be of confined area in order to prevent cold air getting through too freely, yet there must be an opening, and sufficient space within for cleaning the outer surface of the stove, as may be needed. All stoking is done, of course, from outside the air chamber.

Fig. 71.