Comparisons of Mercury and Platinum Thermometers. 3 



a further series of simultaneous observations were taken, and 

 so on to the end of the scale. 



The mercury thermometers were used in a vertical position, 

 and all observations were taken while the meniscus was very 

 slowly rising, usually at a rate of about - 002 or '003° per 

 minute, so that the slight uncertainties of a falling meniscus 

 were avoided. At the point where the thermometer emerges 

 from the calorimeter, it is surrounded by a small water-jacket 

 in which a small thermometer and stirrer are placed. The 

 temperature of the mercury column above the water-jacket 

 is assumed to be that of the surrounding air as indicated by 

 a thermometer hung within an inch or two of the stem. 

 Thus the stem correction was applied in two parts by means 

 of the formula AN = -000156 n (t-t% where AN is the 

 correction to the observed stem-readin'g, t the observed 

 temperature, t' the temperature of that portion of the stem 

 emerging from the calorimeter, and N the number of stem- 

 divisions at the temperature t'. The stem correction was 

 applied to the observed stem-reading, and the corresponding 

 temperature on Rowland's air-scale was then taken from his 

 tables (Proc. Amer. Acad. xv. 1879, pp. 115 & 116) ; this 

 temperature was then reduced by the rise of the zero since 

 Rowland's experiments. 



The corresponding temperature on the Calendar-Griffiths' 

 air-scale was obtained from the observed resistance of the 

 platinum thermometer ; the way in which this was done will 

 be explained more fully later. 



Marnier in ivhich Baudin Thermometers were used by 

 Professor Roioland. 

 These thermometers were first calibrated by measuring the 

 length of a short column of mercury in different portions of 

 the stem, and in this way the relative volumes of different 

 parts of the tube were found. Temperature on the mercury- 

 in-glass thermometers was then defined as proportional to the 

 apparent volume of mercury in glass, when the thermometer 

 is vertical. It was then assumed that the apparent volume of 

 'mercury-in-glass could be expressed as a function of the second 

 degree of the temperature on the air thermometer, of the 

 following form : 



T = C / V-# / -jnT(40-T)|l-n(40 + T) \ 

 using the 0° and 40° points as fixed by the air thermometer; 

 where T is the temperature on the air thermometer, 



V is the volume of the stem of the mercurial ther- 

 mometer, as determined by the calibration, and 

 measured from any arbitrary point, 

 C, t ', m } n are constants to be determined. 

 B 2 



