12H6 



the detlections became less than without heating, showing the etïect 

 to be reversible. 



In order to find out whether the deflectiou for unequal teiiipera- 

 tures of the jets — maxiuiuni 3.2 inni. — could be ascribed to a 

 chemical reaction between the mercury and the glass, we heated 

 tube B for an hour with the pressure off, so as to give some of 



