in a tank or otherwise any extra amount of heat which may be carelessly applied in the furnace.
Let b represent a common boiler, having a feed cistern (c); let e and f be the ordinary flow and return pipes for heating the greenhouse, under the floor of which is placed the rainwater tank (h). All this will be seen to consist of what is found in an ordinary greenhouse only. The rest can be added to any common house and heating apparatus, and consists of a layer or coil of pipes (i) in the bottom of the tank, and communicating with the return pipe and boiler at k; while from the other end of this coil is fixed the pipe (l) terminating in a cistern (m), the bottom of which is on the same level as that of the feed cistern (c). An upright pipe (n), with a sliding tube (o), marked with degrees, is attached to the flowpipe (e), and also enters the aforementioned cistern (m). The tube (o) regulates the whole affair, according to its position; e. g. if this tube be raised one inch above the level of the water in the cistern and return-pipe (m) no circulation can take place through the coil unless the water be heated some 40°. If the orifice of the tube (l) be ten feet above the bottom of the boiler, and that of the tube (o) ten feet one inch, the circulation will commence only through the coil when the temperature of the water in the pipes exceeds 100° (the ordinary circulation will at all times proceed through the pipes in the greenhouse). If the temperature of the water be required to be 180°, the sliding tube would have to be raised about four inches above m; this is reckoning the water there to be about 40°; and it will be seen that these calculations are based on the table given at the beginning, and that no deduction is made for the water becoming slightly heated in the return-pipe (m); this being done for clearness. I find this is easily got over by the graduating of the tube (o) accordingly (the cistern (m) ought to be kept cold by a stream of air). The reader will see that no water can circulate only in the ordinary manner, unless the heat of the same exceed what is required; in that case it will, instead of flowing only through the pipes in the greenhouse, leave them and flow through the coil in the cold-water tank, and entering the boiler almost as cold as that in the tank, and continue to circulate through these until it becomes sufficiently cold to again flow through the pipes in the greenhouse.
In small greenhouses a very large fire may be made up
