242 



Professor Sir James Dewar 



[Jan. 23, 



deflection is observed, in one direction with the Hquid air vessel and 

 to'Jjhe opposite side when the vessel of hot water is substituted. 



Another aid to the heat insulation of vacuum vessels is the intro- 

 duction of finely divided powders to the exhausted space : the results 

 of the use of some of these are given in my discourse on " Liquid Air 

 as an Analytic Agent."* 



Fig. 2. 



Proc. Roy. Inst., xv. p. 823. 



