PLANTS, WITH GLASS ROOFS. 205 



determining the quantity of pipe as a sufficient approximation for ordinary 

 purposes : " In churches and very large public rooms, which have only 

 about an average number of doors and windows, and moderate ventilation, 

 by taking the cubic measurement of the room, and dividing the number thus 

 obtained by 200, the quotient will be the number of feet in length of pipe," 

 four inches in diameter, which will be required to obtain a temperature of 

 about 55 to 58. For smaller rooms, dwelling-houses, &c., the cubic mea- 

 surement should be divided by 150, which will give the number of feet of 

 four-inch pipe. For greenhouses, conservatories, and such-like buildings, 

 where the temperature is required to be kept at about 60, dividing the cubic 

 measurement of the building by 30, will give the required quantity of 

 pipe ; and for forcing-houses, where it is desired to keep the temperature 

 at 70 to 75, we must divide the cubic measurement of the house by 20 ; but 

 if the temperature be required as high as 75 to 80, then we must divide by 

 18 to obtain the number of feet of four- inch pipe. If the pipes are to be 

 three inches diameter, then we must add one-third to the quantity thus ob- 

 tained ; and if two-inch pipes are to be used, we must take double the length 

 of four-inch pipe. 



" The quantity of pipe estimated in this way will only suit for such places 

 as are built quite on the usual plan." (Treatise, <^c., p. 125.) The above 

 calculations for heating are made on the supposition that the lowest external 

 temperature will be 10 ; but in situations " exposed to high winds, it will 

 be prudent," Mr. Hood observes, " to calculate the external temperature 

 from zero, or even below that, according to circumstances ; and in very warm 

 and sheltered situations, a less range in the temperature will be sufficient." 

 Local circumstances, therefore, may require from 5 to 10 per cent, to be 

 added to, or deducted from, the length of pipe found according to the fore- 

 going rules. As a proof of the soundness of Mr. Hood's calculation, we may 

 state that the great stove at Chatsworth is heated at the rate of one super- 

 ficial foot of heated pipe to thirty cubic feet of air; and the temperature kept 

 up during the severest weather of the winter of 1840-41, was 60", though 

 there were frequently from 20 to 35 of frost during the night. This house 

 is sixty feet high, with glass on all sides, exposing a surface of 60,000 feet, 

 and enclosing 1,050,000 cubic feet of air. The quantity of coal consumed 

 was about two tons per night. (Gard. Chron. April 17, 1841, p. 243.) 



501. The situation in which the pipes are placed is, in general, what we have 

 stated to be that most suitable for smoke flues, (492), viz., along the front 

 and ends of houses placed against a back wall, and entirely round, detached, 

 or span-roofed houses. In the case of pits or frames with flat roofs, the pipes 

 may be either placed in front or in the middle, always bearing in mind 

 that heated air ascends, and that the quantity heated in a given time, will, 

 all other circumstances being alike, depend on a regular supply to the 

 heating body, by a current distinct from that by which the heated air 

 escapes. Such a current is formed by the cross drains adopted by Mr. Penn, 

 and exhibited in various other sections of plant-structures given in this work. 

 For the same reason it is desirable, when practicable, and under certain circum- 

 stances, to confine the pipes on each side, so that the air which passes up among 

 them may not escape without being heated. To illustrate the effect of this 

 arrangement, W T C may take Perkins's double boiler, and compare it with the 

 common boiler. It would not occur to any person who had not reflected on 

 the subject, that water could be boiled any sooner in one boiler than another, 



