388 On the CONTRACTION 



below the furface. The jar was freely expofed to the air of the 

 room, the temperature of which was from 6o° to 62 °. 



The experiment commenced at noon : 



Top Thermom. Bottom do. 



3 2 ° - 3 2 ° 



In 10 minutes, - 33-f- - 34-f- 



— 3° - 3S-S ' 37 



_ 5 o . 37 . 3 g + 



— an hour, 38 38-f- 



and 10 minutes, 42 - 38.25 



— 30 44 - 40 



50 46+ - 41+ 



— 2 hours and 10 minutes, 48 - 42.5 



3° 5° " 44 



5° 50.5 - 45 



— 4 hours, - 54-49 



Confiding in the indications of the thermometers, from this 

 experiment we learn, that when heat flows on all fides from the 

 ambient air into a column of ice-cold water, the warmer por- 

 tions of the fluid actually defcend, and take pofTeflion of the bot- 

 tom of the vefTel. 



This downward courfe proclaims an increafed denfity, and 

 teftifies that the cold water is contracted by heat. As foon, 

 however, as the fluid at the bottom exhibits a-temperature of 

 38 , this courfe is retarded and foon flopped, and with the rife 

 of temperature beyond 40 ° is totally changed : for when the 

 mafs attains this degree, the experiment equally fliows, that the 

 warmer fluid afcends and occupies the fummit, by its route an- 

 nouncing its diminifhed denfity, and proving that water is now 

 expanded by heat. 



Expe- 



