concerning Heat. 139 



five feet from the floor, and at the distance of four feet 

 asunder. 



The temperature of the air of the chamber, which 

 did not vary a quarter of a degree during the .whole 

 time of the experiment, was 9! degrees of Reaumur's 

 scale. 



An excellent mercurial thermometer, with a cylindri- 

 cal bulb, of four inches long and two lines and a half 

 in diameter, suspended in the axis of each of these 

 bottles, indicated the temperature of the contained 

 water ; and the time employed in its cooling for every 

 five degrees of Fahrenheit's thermometer was carefully 

 observed, during eight hours. 



The glass being considered as a very bad conductor 

 of heat, and the sides of the bottle being so thick, who 

 would not have expected that the water in this bottle 

 would have been more slowly cooled than that in the 

 very thin bottle of tin ? 



The contrary, however, was the event ; the bottle 

 of glass was cooled almost twice as quickly as that of 

 tin. 



While the water included in the bottle of tinned 

 iron employed 56 minutes to pass through a cer- 

 tain interval of cooling, namely, through ten degrees, 

 between the 5oth and 4Oth degree of the thermometer 

 of Fahrenheit above the temperature of the air of the 

 chamber, the water in the glass bottle employed only 

 30 minutes for the same change. 



It appears to me that the result of this experiment 

 throws great light on the mysterious operation of the 

 communication of heat. 



If we admit the hypothesis that hot bodies are 

 cooled, not by losing or acquiring some material sub- 



