60 



Mr. B„ IT. Poole on the Rate of 



i. e. 8*15 Xl0 -5 calorie per hour. This result being even 

 higher than the previous ones, the calorimeter and pitchblende 

 were put aside unopened for some months so that any chemical 

 action which might be occurring might cease. As a short 

 time previous to this experiment the wax had been melted, it 

 is possible that crystailographic changes may have been 

 occurring in the wax. 



During the interval, in the course of some experiments on 

 orangite which will be described in a later paper, a new 

 inner ice vessel was made with double walls and lid, the 

 space between the walls, about \ inch, being rilled with 

 cotton-wool. This vessel is watertight, and arranged so that 

 water from the outer ice cannot enter round the lid. In these 

 experiments the ice employed was specially frozen by an 

 ice-making firm out of water distilled in the laboratory. As, 

 however, the tempeiature variations in this ice seemed quite 

 as large as those in the ordinary commercial ice, the latter 

 was again employed in the subsequent experiments. The 

 distilled- water ice attacked the zinc vessel much more than 

 the ordinary ice, as the sides were found to be coated with 

 a thin layer of oxide. This was scrubbed off and the vessel 

 thoroughly dried and coated while warm with vaseline. 



Nine months after the insertion of the pitchblende in the 

 calorimeter it was again placed in the ice. The second 

 couple was not employed on 1 his occasion, as there seems to 

 be no doubt that variations in the ice temperature will be 

 detected by the couple used with the pitchblende, and it was 

 thought best to have as few leads entering the ice as possible. 

 The temperature of calorimeter is shown on the chart 

 (A, fig. 2). As the distance from galvanometer-mirror to 









Fig. 



2. 









TIME f?CCK0N£DFR0M oa 



1135 Sole Division 



r £ ur packing //v /c£ 

 f = \'c 



.. V 



B 







A^,^ 



A 





<5 



\ 





~ W 



<o 















i 



1 



D*rs 



scale had t ecu . reduced, to 101. cms.,. 1° C corresponds. to 

 about 1135 scale-divisions. As usual, the irregularities get 



i 



