230 On curvilinear Iron Roofs for Hot-houses. 



each end of the house, and I had the satisfaction to observe, 

 that the temperature of that part of the house, in which the 

 thermometer stood, was raised between two and three de- 

 grees, when the external air was at 40°. This effect was, I 

 conclude, produced by the heated air being impelled into the 

 body of the house amongst the plants, instead of being per- 

 mitted to rise, as it had previously done, and to come in- 

 stantly into contact with the roof: and by suspending light 

 bodies amongst the plants, I ascertained that the previously 

 confined air was thus constantly kept in a state of rapid motion. 

 The air is suffered to escape through passages of four inches 

 wide and two inches and a half high, at E, which passages are 

 placed at the same equal distances, as those in the front 

 wall, and like those are opened or closed, as circumstances 

 require. The trouble of opening or closing such passages, 

 after substances of proper form are prepared and suspended, 

 for the purpose, is very small, much less, I think, than that 

 of moving the lights of any house of ordinary construc- 

 tion ; and the effect of the kind of ventilation obtained upon 

 the growth of my plants and fruit, is every thing I wish 

 it to be. 



I have stated that my whole house is heated by a single 

 flue : this enters at the west end of it, and thence passes 

 along the whole front within sixteen inches of the wall. It 

 then returns twenty feet towards the middle of the house 

 and back again, the smoke escaping at the end opposite to 

 that at which it enters. The flue is consequently single at 

 the end of the house, which adjoins the fire place, and triple 

 in the last twenty feet of the opposite end ; by which means 

 a nearly equal temperature is every where given. 



