272 NOTICE OF A VISIT TO PITMASTON. 



These passages are 5 by 3 inches ; the two lower ones in 

 each arch extend across, on the level of the floor, under both 

 flues, and communicate with the 4-inch cavity between the front 

 wall and flue next it. Thus, fourteen air-passages are in com- 

 munication with this cavity, and the other seven air-passages 

 communicate with the 6-inch cavity between the two flues. 

 By this arrangement the cold air was rapidly drawn from its 

 lodgment on the floor of the back-part of the house, brought 

 into the cavities by the sides of the warm flues, there heated, 

 and consequently rarefied ; it could not rest, but must of ne- 

 cessity continually ascend, and give place to a successive influx 

 from below through the respective air-passages. The flues, or 

 hot-air chambers, are 18 inches deep inside, and the cavity 

 between them is 23 inches deep ; this arrangement is said to 

 have afforded a complete circulation. It will readily be per- 

 ceived that it is a modification of the Polmaise system : in fact, 

 the same principle acted upon more than twenty years ago. 



In detailing the construction of the flues, both in this house 

 and in the Vinery, it was mentioned that the brick portions were 

 terminated by hollow iron cylinders ; the lengths of these were 

 equal to the heights of the respective back-walls above the 

 level of the brick portions of the flues. Metals being much 

 better conductors of heat than the materials of the flue, the iron 

 tube would, of course, extract more of the heat whicli the gases 

 still possessed after passing through the brick portions than 

 would otherwise be the case. In point of economy of fuel, no 

 mode of heating excels the flue, and its power exceeds that of 

 all others occupying the same space, if we except steam ap- 

 paratus, and this, for obvious reasons, is now rarely employed. 

 An immense assemblage of hot-water pipes is required to keep 

 up a house to 70°, when the external air is 20" below freezing ; 

 to heat the water sufficiently for all these pipes a strong fire 

 must be maintained ; and hence a large amount of air and car- 

 bonic acid at a high temperature must pass bj"^ the chimney, by 

 which, it is indeed probable, as much heat is dissipated in the 

 open air as is radiated from the pipes inside the house. That 

 brick flues have their faults no one will deny ; but surely im- 

 provement in flues will not remain almost stationary, as at 

 present? Or, will it be aflRrmed that they must ever be con- 

 structed nearly in the same way as they were when first intro- 

 duced by the Romans? A safety-valve, as previously noticed, 

 may be adopted so as to effectually prevent tlieir bursting ; and 

 Mr. Williams's idea of employing metal doubtless admits of 

 extended application. My opinion in regard to tliis is founded 

 on the circumstance of having actually seen cast-iron tubes em- 

 ployed for flues in France, in 1847, at two establishments, one 



