22 



Imprmements in heating hy Hot Water. 



4 in., but pressed or flattened until they became oblate, and 

 for this reason, that there might be sufficient fall or slope for 

 the small returning pipe. I erected the apparatus ni a two- 

 light frame {fig 6.), at the end of which I built a very small 



furnace, and inserted my boiler {fig. I-), passing the main 

 pipes along the lower end of the frame ; >j 

 from these proceeded the branches or la- 

 terals, amounting in all to near 40 ft., and 

 about 10 ft. or rather more of main pipe ; 

 these were properly levelled on old tan 

 and sawdust *, and filled with water. In 

 a short time the heated particles gradually 

 proceeded through mains and laterals, and into the tray. I 

 closed the sashes, and in a short time the bed was in a close 

 heat of 72°. Having placed a few plants in it, I soon found 

 they were covered with a soft dew, and having made up my 

 fire for the night, in the morning I was delighted to find the 

 bed had only lost 1°, it being ^\°. My next object was to find 

 the bottom heat, and by inserting a thermometer into several 

 parts of the bed, I found it from about 86° to 90°. These re- 

 sults seem to offer several advantages; indeed, I may say, you 

 may almost command every thing but light, through the 

 agency of hot water. Should you require a dry heat, by hav- 



* What I purpose using in my beds, is coarse sand ; the other materials 

 engender worms, fungi, &c. &c. 



The method for damping the sand on which the pots stand (as they would 

 be arranged in regular rows), would be, a pipe the length of the bed, or 

 half the length, perforated with small holes on each side, and at the bot- 

 tom about half an inch in diameter, with a piece 6 or 8 in. 

 long, attached at right angles, with a funnel mouth, 3 

 or 4 in. in diameter. ( j'%.8.) This perforated tube can be 

 slipped between the pots, and the water poured in at the 

 funnel, and when one side of the bed is watered to go 

 round to the other, of course, introducing the ]:)ipe between each row. I 

 think half the length better than the whole, as it would be more conve- 

 nient. 



