PLANTS, WITH GLASS ROOFS. 173 



communicate by means of stop-oocks ; so that whenever there was 

 more heat in the pipes than was wanted for heating the air of the 

 house, it could be transferred to the reservoir tank. To F . 14g 

 save the expense of stop-cocks where the cisterns could be 

 wholly or partially uncovered, the orifices of the con- 

 necting pipes might be stopped by plugs ; and when the 

 reservoir tank is above the level of the heating-pipes, the 

 connexion between them might be made by means of j 

 siphons with stop- cocks. 



The pipes employed are generally of cast-iron, and Section of re- 

 round, as being more conveniently cast; but any other nerve cust* 

 , ~ ... 11 i ana tiot-wa- 



metal and form will answer; and when there is no t er pip&r. 



great pressure on the pipes, earthenware may be used, 

 the joints being made good with cement ; and at the angles, where 

 elbow-joints would be necessary, small cisterns could be employed, 

 or elbows of earthenware might be made on purpose. For ob- 

 taining a large heating surface, flat cast-iron pipes have been used, 

 placed vertically in some cases, and in others horizontally; but 

 round pipes of four inches in diameter are in most general use. When 

 the object is to obtain a supply of heat in the shortest time, then the 

 boiler and pipes should be of small capacity ; and this is generally 

 desirable in the case of greenhouses, where heat is occasionally wanted 

 for a few hours in damp weather, not for the sake of raising the tem- 

 perature, but for drying up cold damp : nevertheless, even in green- 

 houses it is desirable to have a reservoir of heat for supplies in very 

 severe weather. In stoves in which fire heat is employed the greater 

 part of the year, both boiler and pipes may be of large capacity ; and 

 this should also be the case in early forcing-houses. Whatever 

 mode of heating or kind of pipes may be adopted, the pipes should 

 always have a gradual ascent from the place where they enter the 

 house, or are intended first to give out heat, towards the farther 

 extremity ; otherwise, the circulation will be less rapid, and conse- 

 quently the heat less equally distributed. The quantity of pipe 

 required to heat any house depends on various circumstances ; such 

 as the form and construction of the house, the temperature that is to 

 be kept up in it, and the temperature of the external air. Various 

 calculations have been made on the subject by different engineers, and 

 more especially by Mr. Hood, who says : " It may be taken as an inva- 

 riable rule, that in no case should pipes of a greater diameter than 

 four inches be used, because, when they are of a larger size than this, 

 the quantity of water they contain is so considerable, that it makes a 

 great difference in the cost of fuel, in consequence of the increased 

 length of time it will require to heat them, which is four and a half 

 hours for four-inch pipes, three and a quarter hours for three-inch 

 pipes, and two and a quarter hours for two-inch pipes, supposing the 

 water to be at 40 before lighting the fire, and the temperature to 

 which the water was raised 200. Pipes of two or three inches 

 diameter therefore are to be preferred for greenhouses and conserva- 

 tories which only require fire-heat to be applied occasionally." After 



