138 Experimental Investigations 



when it is proposed to confine the heat in a body, of 

 which the temperature has been raised, or to hinder its 

 dissipation as much as possible, care is taken to sur- 

 round the heated body with substances known to be 

 bad conductors of heat. 



The results of many of my experiments having led 

 me to suspect that the cooling of bodies is not effected 

 in the manner which is generally supposed, I made the 

 following experiment, with the intention of clearing up 

 this interesting part of the science. 



I procured two bottles, nearly cylindrical, of the same 

 form and the same dimensions when measured exter- 

 nally, one being of glass, and very thick, and the 

 other of tin or tinned iron, which was very thin. Each 

 of them is three inches ten lines in diameter, very 

 nearly, and five inches in height ; and each has a neck 

 one inch three lines in diameter, and one inch two lines 

 in height. The glass bottle weighs 13 ounces, i gros, 

 and 1 8 grains poids de marc; and the other thin me- 

 tallic vessel weighs only 5 ounces, I gros, and 65 grains. 



Having very exactly weighed the bottle of tinned 

 iron, I found its exterior surface to be 54.462 inches, 

 which give 0.21142 of a line for the thickness of 

 its sides, taking the specific gravity of the metal at 

 7.8404. 



The mean thickness of the sides of the glass bottle 

 is more than six times as great, as may be easily de- 

 duced from a calculation founded on the weight of the 

 bottle, the quantity of its surface, and the specific grav- 

 ity of glass. 



Having filled these two bottles with boiling water, 

 I hung them up by slender strings in the midst of 

 the tranquil air of a large chamber, at the height of 



