JRelations of Air, Heat, and Cold. . 453 



i? allowed to accumulate in that part of the screen. In this 

 latter case, the ciiBerence of temperature produced is very 

 nearly tbe double, and in the former it is no less ihan ten- 

 fold. But effects of the same kind, and which are alike 

 contrasted, though inferior in degree, will be perceived, if 

 a thin pellicle be spread over the compound surface of the 

 glass and tinfoil, or of the mica and silver leaf, the iTiere 

 proximity of the metallic surface repelling the atmosphere, 

 and consequently enfeebling the powers of absorption and 

 emission. 



" The very singular and unexpected f;icts now detailed 

 merit attention, and suggest a variety of improvements in 

 the practical managemeuc of heat. A vessel with a bright 

 metallic surface is^the best fitted to preserve liquors either 

 long warm, or as a conservatory to keep them cool. A 

 silver pot will emit scarcely half as much heat as one of 

 porcelain ; and even the very slightest varnishing of gold, 

 platina or silver, which communicates to the ware a certain 

 metallic gloss, renders this new kind of manufacture about 

 one-third' part more retentive of heat. The addition of a 

 covermg of flannel, though indeed a slow conductor, far 

 from checking the dissipation of heat, has directly the con- 

 trary tendency ; for it presents to the atmosphere a surface 

 of much greater propulsive energy, which it would require 

 a thickness of not fewer than three folds of this loose sub- 

 stance fully to counterbalance. The cylinder of the steani- 

 ensine has lately been most advantageously sheathed with 

 polished copper. 



" The progress of cooling is yet more retarded, by sur- 

 roundmsr the heated vessel on all sides, at the distance of 

 near an Tnch, with a case of planished tin ; and the addi- 

 tion of other cases, following at like intervals, augments 

 continually the eftext. With an obstruction of one case, 

 the rate of refrigeration is three tunes slower, with two 

 cases it is five times slower, with three cases it is seven times 

 slower, and so forch, as expressed by the succession of the 

 odd numbers. By multiplying the metallic cases, ihere- 

 ,fore, and disponing them like a nest at regular intervals, the 

 innermost coulJ be made to retain the sauic temperature 

 with little variation for many hours or even days. Such an 

 api^aratus would obviously be well calculated for various 

 culinary and domestic purposes. 



" In the conveyance of heat by means of steam, the sur- 

 face of the conducting tubes should have a metallic lustre. 

 On the contrary, if it be intended by that mode to warm an 

 apartment, they should be coated on the ouliide with soU 

 F f i paint, 



