228 Mr Griffiths, On Thermo metric "Fixed Points:' [Nov. 23, 



Section II. Boiling Point. 



The following is, I believe, our definition of standard boiling 

 point : " the temperature of the steam issuing from water boiling 

 under a pressure of 760 mm. of mercury at temperature 0° and at 

 sea-level 45°." 



From a practical point of view this definition is, I now 

 believe, sufficient. Until recently I have been doubtful as to the 

 rapidity with which, if the barometer changes, the steam acquires 

 the temperature corresponding to the new pressure, and I anti- 

 cipated that experiment would show there was a considerable lag. 

 I had hitherto failed to satisfy myself on this point since the un- 

 certainty arising from want of knowledge as to the actual tem- 

 perature of the box coils was greater than the temperature 

 changes consequent on the changes of pressure usually occurring 

 during an experiment, and results obtained from observation on 

 different dates would convey no information concerning the lag, if 

 any. As previously indicated, I now can measure the changes in 

 temperature of the box coils (I do not here refer to the absolute 

 temperature) certainly to within o- 01 C, and the resulting observa- 

 tions are sufficient to settle this lag question, at all events to my 

 own satisfaction. On October 7th Mr Whetham and I were 

 engaged in standardising a thermometer for the purposes of a 

 joint investigation. The pressure on this occasion proved to be 

 so variable that it appeared hopeless to attempt an accurate deter- 

 mination of the resistance at 100°. We, however, continued the 

 observations for a period of nearly 3 hours in the hope of throw- 

 ing some light on the matter to which I have referred. The 

 results are shown in the following table, but are much better 

 exhibited when plotted. The actual numbers, as also the varia- 

 tions, are so dissimilar that I have adopted the following method 

 for showing the relative movements. The mean pressure through- 

 out the observations was 755"00 mm. ; the mean resistance (after 

 correction for temperature of coils) was 1792 # 546 of my box unit 

 (1 box unit = "01 ohm approximately). In columns 2 and 3 I give 

 the differences from these means respectively at the given times *, 

 the barometer differences are then plotted on 10 times the scale 

 of the resistance differences, and thus the paths of the two 

 irregular curves are brought comparatively near together. It is 

 evident that hardly any signs of lag are observable ; in fact, if 

 there is any lag, it would appear to be rather in the barometer 

 than in the temperature of the steam. In column 4 I give the 

 boiling point reduced from the pressure according to Regnault's 



"' The times were only recorded to the nearest minute, further accuracy being 

 unnecessary. 



