284 



On the Means of supplying 



of procuring a constant supply, proportional to the expenditure. 

 This end seems to be most effectually attained by cisterns on 

 the pipes, whose temperature varying with that of the pipes 

 themselves, the evaporation from them is greatest, when the pipes 

 are hottest, i. e. when the greatest degree of artificial temperature 

 is being maintained, when the drain upon the atmosphere by 

 condensation, &c., is also greatest. 



After several experiments, I believe that zinc cisterns fixed 

 on the pipes will be found the most effective and commodious 

 contrivance for this purpose. The troughs occasionally provided 

 are objectionable only in as far 

 as they are much too small, and 

 yield a brief and scanty supply ; 

 and if their size were greatly 

 increased, they would render 

 the pipes very heavy and cum- 

 bersome. The annexed sketch 

 {Jig, 42.) represents the troughs 

 or cisterns which 1 have em- 

 ployed, which may be made at 

 small expense. 



My pipes are placed side by side, and the cistern, which is 

 1 ft. wide, embraces both of them. It is 6 in. deep to the top 

 of the pipes, and is fitted to their curvature to the depth of 

 l^in. or 2 in. more. These cisterns being properly fitted, and 

 luted on the pipes with wet sand, the water in them will reach a 

 temperature of 140° — 145°, when the pipes are at 200° ; but 

 the smallest interstice makes a great difference in the heat trans- 

 mitted. With respect to the surface of cisterns required in any 

 given house, I am not prepared to speak with certainty ; but I 

 find that cisterns fixed in the manner here described, whose 

 temperature varies from 120° to 145°, evaporate about an inch 

 and a quarter, or rather more, per 24 hours, or about | of a gal- 

 lon per square foot of surface ; so that a computation may easily 

 be made of the number required in any given house to produce 

 a given effect. The proportion which I employ in an orchi- 

 deous stove is about 1 square foot of evaporating surface, to 10 

 square feet of glass ; but innumerable causes will affect the result 

 in such a degree, as to make the quantity requisite for any house 

 a matter rather of experiment than calculation, although calcu- 

 lation may afford some approximate estimate. Tlie length of the 

 cisterns is limited only by the extent of smooth surface in the 

 pipes. It would be difficult to fit a cistern to the rings on the 

 pipe; and, as these occur at every 3 ft., 2 ft. 8 in. or 2 ft. 10 in. 

 is the utmost convenient length for a cistern. It will also be per- 

 ceived, that in this arrangement the pipes should lie side by side, 

 which is always the best where practicable ; where they are one 



