346 BABCOCK. 



where 6 = temperature in degrees Centigrade 



dp = "platinum temperature" defined by (20) 

 R = resistance of thermometer at temperature 0. 

 Rm° = resistance at 100° 

 i?o° = resistance at 0° 

 5c = a third constant whose value depends on the purity of the 

 platinum. 



6. Calibration of Standard Thermometer. 



Resistance at 0° C. Thermometer was deeply immersed in a thermos 

 bottle containing commercial distilled water ice, finely shaved and 

 soaked with distilled water. Preliminary measurements showed no 

 appreciable difference between commercial ice and distilled water 

 frozen in the laboratorv. The average of six different determinatiors 

 was taken. 



Resistance at 100° C. Thermometer in a Regnault hypsometer 

 equipped with manometer; hypsometer filled with distilled water. 

 Alternate readings made on bridge and barometer. Barometer instru- 

 mental error determined by cornparison with a standard 25 mm. bore 

 barometer. Boiling points taken from Landolt and Bornstein's 

 Tabellen. The average of four different determinations was taken. 



Resistance at 218° C. Thermometer immersed in a naphthalene boil- 

 ing apparatus, similar to the usual sulphur boiling tube, but equipped 

 with two side condensing tubes. The purity of the naphthalene was 

 tested, as usual, by measuring its melting point. In these measure- 

 ments the thermometer tube was protected by an aluminium radia- 

 tion and splash shield. Alternate readings were made on the bridge 

 and the barometer. Boiling points were taken from Burgess and 

 LeChatelier "Measurement of High Temperatures," p. 450. The 

 average of three different determinations was taken. 



7. Calibration of Observing Thermometer. 



The observing thermometer in its calibration tube, and the stand- 

 ard thermometer, were both immersed in the constant temperature 

 bath, which was set to regulate at some definite temperature. This 

 bath temperature was not constant but fluctuated between narrow 

 limits due to the regulation. Both thermometers followed these 

 fluctuations but the observing thermometer lagged behind the stand- 

 ard. Comparisons were made by averaging the readings over at 

 least one complete temperature wave for both thermometers. Seven 

 comparisons for the two thermometers were thus made, thi'ee at 



