614 



Sir James Dewar 



[Jun. 17; 



surface of the fiask to be examined. A side tube in this cap con- 

 nected it to the vacuum circuit, while the neck of the fitting carried 

 a small steel drill which could be rotated to and fro while protected 

 bv an oil-cup against influx of air. After exhausting and testing the 

 arrangement for tightness, etc., connection was opened between the 

 annular space of the container and the exhausted circuit by a small 

 rotating of the drill. Subsequent examination revealed that a clean 

 2^ mm. hole was thus produced. The vacuum maintained by the 

 flask was thus directly measured. The subsequent response of the 

 flask to different degrees of vacuum was also measured. In the 

 result it was ascertained that not only were the polished surfaces at 

 fault, but also that the condition of the charcoal was improved by 

 further heating and exhaustion, so that the rate of loss was materially 

 reduced. Before piercing, the vessel had been losing 2*85 kilos, per 

 day, whereas after the re-exhaust the charcoal was so much improved 

 that the rate of loss was reduced to 1 • 64 kilos, per day. A pressure 

 of OMU<s mm. was registered when the vessel was opened as described. 

 There was no appreciable leak. With steam heating the pressure 

 rose as follows, due mainly to gases expelled from the charcoal : — 



Time 

 Press, in mm. Hg. 



10 mill. 

 1-02 



30 min. 

 1-72 



2 hr. 

 2-075 



2i hr. 

 2-107 



3hr. 

 2-24 



34 hr. 

 2-80 



The subsequent exhaust was then carried out as previously 

 described, with the followino- results :— 



On the vessel becoming cooled after the steam heating, liquid 

 air was poured in. The pressure fell in a few hours to a limit 

 of 0*000015 mm. Hg., and the rate of evaporation measured by 

 the meter was one litre in 68^ seconds, equal to 1630 grams per 

 day. It is interesting to note that from the Rankine equation, 



log. p. = A - * ap])lied to the observations marked {a^ and {h) above. 



between the tempei-atures 21')' and 100 0, tlie deduced value of the 

 B. (proportional to the latent heat of the occluded charcoal gases), 



