312 FORCING GARDEN. 
tions of this apparatus; but it may be proper to direct the 
attention of the reader to the close boiler represented in 
Fig. 32, in which is shown how the circulation may be 
conducted over a door or other obstacle. In this case the 
upper pipe must not ascend and descend twice: air-tubes 
ought also to be placed in the boiler, and on the highest 
part of the pipes; and the whole must be made consider- 
ably stronger than on common occasions. The annexed 
figure will give an idea of an isometrical elevation of a 
Fig. 33. 
hot-water apparatus for a vinery thirty feet long by eleven 
wide. A is the boiler, as in the figure on p. 308; B the 
upper or. delivering pipe; C the principal part of the upper 
pipe, of a flat form, presenting a greater radiating surface, 
in proportion to the quantity of heat; D the descending 
limb; E the returning pipe, of a cylindrical form. 
Mr. Fowler has employed the siphon asa part of the 
hot-water apparatus; and in his tract on the Thermo- 
siphon, as he calls it, has shown how its various modifica- 
tions may be employed in warming hot walls, as well as in 
heating glazed houses. The following statement of the 
principle is given in the Gardener's Magazine, vol. v. 
* Any one may prove that hot water will circulate in a 
