166 Messrs. Waidner and Mallory on a Comparison 



as possible similar to those under which they were used by him 

 in his mechanical equivalent determinations. On account of 

 the many advantages offered by a Griffiths platinum ther- 

 mometer for the standardization of calorimetric thermometers, 

 this instrument was selected for the comparison. In the 

 comparison the platinum and the mercurial thermometers 

 were placed side by side in a calorimeter and the two 

 read simultaneously while the temperature was slowly rising. 

 Readings on mercurial thermometers were taken by means 

 of a micrometer eyepiece. 



The stem corrections of the mercurial thermometers were 

 made, as in Rowland's experiment, by surrounding the ther- 

 mometer stem for a short distance above the calorimeter with 

 a water jacket, the remaining portion of the exposed stem 

 being assumed to be at the temperature of the surrounding 

 air. The mercury thermometers were placed in ice and dis- 

 tilled water for several hours and the zero determined before 

 each comparison. The corrected stem readings obtained in 

 our comparison were then reduced to Rowland's zero and the 

 corresponding temperatures of his air thermometer taken from 

 his tables of comparison of this thermometer with his Baudin 

 thermometers (Nos. 6163, 6166). The platinum thermometer 

 used in this experiment was one constructed by Mr. E. H. 

 Griffiths. It was provided with compensating leads so that 

 the temperature was independent of stem immersion, provided 

 this was great enough to prevent conduction down the leads 

 to the platinum coil. 



The resistance measurements were made with a Griffiths 

 resistance box (No. 7) especially designed for the measure- 

 ment of platinum temperature. We are under great obliga- 

 tions to the University of Chicago for the use of this box, 

 which was kindly loaned us for this work. The construction 

 and calibration of a resistance box similar to the one used by 

 us has been described by Mr. E. H. Griffiths in 'Nature,' 

 November 14th, 1895. The method there described was 

 employed by us. Two independent calibrations of the coils 

 and bridge wire gave practically identical results. 



Temperature on the platinum scale is defined by the 

 equation 



pt = rTZr x 100 > 



where pt denotes platinum temperature, R the resistance of 

 the thermometer at 0° C, Ri Ihe resistance at 100° C, and 

 R the resistance at the temperature pt. Platinum temperature 

 is therefore independent of the unit of resistance employed, 



