68 



POLMAISE METHOD OF HEATING HOT-HOUSES. 



was a quiet, clear, frosty night, the external air at 31° Fahr.* I 

 find the uniform range of the differential thermometer, from six 

 o'clock in the evening till eight in the morning, is from 62° to 

 57° or 58°. It must be remembered that this is legitimate 

 atmospheric temperature. There are no means within the house 

 of affecting the thermometer by radiation, either from pipes, or 

 from bricks heated by pipes, or by hot flues, and then attributing 

 the thermometric point to atmospheric temperature, which is 

 partly owing to radiation, a very common error; but it is the 

 temperature of the air taken by a thermometer suspended 

 against the upright side sashes of the house. The house itself is 

 a span roof, containing about 4000 cubic feet of air ; it is glazed 

 with sheet-glass, the ends are both glazed, and the upright side- 

 sashes are three feet in height, standing upon stone walls four 

 feet high. It is situated in a kitchen-garden, partly walled, on a 

 very considerable elevation, as will be evident when I state that, 

 though only 20 miles from London, I can from my own grounds, 

 on the same level, distinctly see the Downs in clear weather. I 

 have purposely abstained from using any covering material, 

 such as asphalte shutters, with a view of testing the heating 

 powers. These facts entirely corroborate those obtained at Pol- 

 maise by Mr. Murray, which were published in the Gardener's 

 Chronicle of the present year. 



Another inquiry of great practical importance is, How far the 

 success of this mode of heating depends upon the conditions 

 under which the principle is carried out ? Principles are all- 

 important, but experience can alone teach us the best mode of 

 applying them. When we reflect on the gradual advancement of 

 hot-water heating, the experience that has been slowly brought 



* Another experiment was made by closing four of the cold-air orifices, 

 reducing the exit of the cold air to two one-foot openings, and I could not 

 detect that the temperature of the house was always affected, while it was 

 quite evident that the velocity of the currents in these two was greatly in- 

 creased ; neither could I ascertain that the uniformity of the temperature 

 was at all affected ; the drains left open were those at the extreme end of 

 the house : had these been closed and those near the chamber opened, 1 

 think it probable the uniformity would suffer. I also closed all the hot-air 

 ventilation, allowing only the escape of hot-air at the opening close by the 

 hot chamber ; neither did this affect the general uniformity of the distribu- 

 tion. These evidently lead to the conclusion, that a small amount of cold- 

 air drain will prove sufficient: and also that any chamber or means of 

 diffusing the hot air is entirely unnecessary, except where required, as at 

 Nutfield, for the bottom heat. This experiment has not always been uniform 

 in its results : sometimes when only two cold drains are open the tempera- 

 ture of the house is less by two degrees than when all are open, but I can 

 never detect that it affects the uniformity of the distribution ; it must also 

 be understood that the temperatures named are not always the same ; but 

 the most important and best ascertained point is the equal diffusion of the 

 heat. 



