TH E 



AMERICAN JOURNAL OF SCIENCE 



[FOURTH SERIES.] 



Art. XVIII. — On the Gas Thermometer at High Tempera- 

 tures ; by Ludwig Holborn and Arthur L. Day. 



Second Pape 



[Communication from the Physikalisch-Technische Reichsanstalt, Charlottenburg, 



Germany.] 



Lsr a former paper* we described our measurements with 

 the gas thermometer in some detail. The investigation was 

 concerned principally with establishing the conditions under 

 which this instrument may be used with certainty as a stand- 

 ard at high temperatures. We have shown that this is pos- 

 sible using pure nitrogen as the expanding gas in a bulb of 

 platin-iridium, this material being much superior to the porce- 

 lain so long in use for the purpose, both in the accuracy of the 

 results obtainable and the convenience in handling. 



Since then the work has been continued in the direction 

 indicated at the close of the previous paper. In order to 

 obtain the full advantages of the exactness which the use of 

 the platin-iridium bulb has rendered possible, the first step was 

 to measure the undetermined coefficient of expansion of this 

 metal at high temperatures. This was very necessary because 

 the correction which the observations with the gas thermome- 

 ter require on account of the expansion of the bulb increases 

 more rapidly than the temperature to be measured. With the 

 platin-iridium bulb, for instance, it amounts to 10° at 500°, 30° 

 at 1000° and 40° at 1150°, while the increase in the coefficient 

 of expansion with the temperature, which of course is not 

 taken into account in the earlier paper, affects the measurement 

 1°, 5°, and 7° at the above temperatures respectively. 



Secondly, the influence of the pressure which the expanding 

 gas exerts on the glowing walls of the bulb was investigated 



* Ludwig Holborn and Arthur L. Day, this Journal (IV), viii, 165, 1899. 



Am. Jour. Sci.— Fourth Series, Vol X, No. 57. — September, 1900. 

 12 



