ON HEATING GREENHOUSES BY MEANS OF HOT WATER. 87 



first, and the water there, being expanded by the steam, is forced up 

 through the pipe at E, circulates in the direction of the arrows, enter- 

 ing the boiler again at F, and this circulation is constant as long as 

 the water is heated, owing to the water at the^top of the boiler being 

 hotter than that at the bottom ; and the expansion at the top con- 

 stantly forces the water up through the pipe at E, because, in con- 

 sequence of the whole pipe being at a higher level than the boiler, 

 the boiler is always full, and the water has no other means of escape. 

 It seems hardly necessary to urge that the pipe should be perfectly 

 horizontal, it being generally admitted that water has no particular 

 fancy for running up hill. It will be obvious enough that if the pipe 

 is below the level of the top of the boiler, the boiler will not be always 

 full ; and if the top of the boiler is empty, there would be merely a 

 steam pressure on it, which might induce a blow-up, and would never 

 produce circulation. C. C.'s plan of placing the boiler in any con- 

 venient corner, and carrying the pipes along the floor, is well calcu- 

 lated to produce an explosion, since the whole of his pipe would be 

 below the level of his boiler, and he would have no circulation at all. 

 The water may circulate in a gutter made water-tight with Roman 

 cement, and this is the cheapest plan, and answers well if the gutter 

 be laid truly horizontally. The fireplace must have an air-tight door, 

 and there must be a door to the ash-pit, with a regulating valve, or 

 some similar contrivance, to admit more or less air, as required. 



My object in giving a sketch of the boiler, as above, is, that it can 

 be cast and fitted up at any foundry, as mine was, as it is not every 

 person who can afford to pay 6/. 6s. for one of Stephenson's boilers, 

 which, however, are excellently made. The whole of my apparatus, 

 including boiler, forty-five feet of gas-pipe (three-inch diameter 

 inside), and fitting and fixing, cost 10/. only, j The boiler is sunk in 

 the floor of my back kitchen, which adjoins the greenhouse. My 

 fire will remain lighted for six hours without trouble, and frequently 

 remains in for nine hours. The quantity of "culm" (query, 

 anthricites ?) used is quite insignificant, and culm gives far more 

 heat and burns much longer than coke, and I could keep up a high 

 temperature if I wished. 



