262 GARDEN MANAGE3IENT. 



This house is 51 feet 9 inches in length, llHeetwide, and 6 feet 9 inches 

 high under the ridge. 



708. Much as it has been approved by the best practical horticulturists, the 

 tank system has its objections, which we find stated in the Ga-rdener's Chro- 

 nicle. " It is well known," the writer remarks, " that by means of a flow and 

 return-tank, the degree of bottom-heat in the pits can be very steadily main- 

 tained. Once the mass of soil on the beds is heated to the required pitch, very 

 little is required to keep it up, and sudden changes of temperature do not much 

 affect it : even such extremes of external temperature as 55° one night and 

 25° the next will only occasion a few degrees lower temperature in the beds. 

 But the case is different as regards the aii' of the house itself ; for, under the 

 above conditions, it would certainly be affected to a much greater extent,— 

 perhaps as much as 20°. Presuming that the temperature of the beds is just 

 what it ought to be, any attempt to counteract the coldness of the air in the 

 pit on a cold night would cause an excess of bottom-heat, which, by repetition, 

 must prove highly iDJ\xrious to the plants. If the communication of heat from 

 the tanks to the surface is only through the soil, the conduction of heat will 

 be very slow, whilst its escape through the glass is rapid." The remedy for 

 this, the writer goes on to say, is a separate command of heat for heating 

 the house, which he proposes to supply by a 2-inch steam-pipe, running from 

 the top of the boiler and ranging along the front of the house, immediately 

 under the glass, terminating at the further end of the return-tank. Why 

 steam, he does not say : perhaps a water-pipe would be sufl&cient for the pur- 

 pose ; but such is certainly an objection to which the system is open ; and this 

 is met in the range of houses proposed by Mr. Messenger, by hot-air cham- 

 bers \inder the tanks, the air being carried up into the body of the house in 

 the dij-ection of the arrows ascending towards the ridge. 



->£^(;^=g^55^i?3=*'^'«^ 



