172 



Holborn and Day — Gas Thermometer at 



hj a simple method, and finally a second platin-iridium bulb 

 with walls of double the thickness of the first was used in a 

 somewhat different oven in verification of the earlier results. 

 The use of the same thermoelement throughout enabled the 

 observations to be compared with those of the previous series. 

 The thermo-electric force of this element (platinum — platin- 

 rhodinm), like the others involving the use of combinations of 

 metals of the platinum group, may be represented by an 

 equation of the second degree within the limits of temperature 

 covered by the investigation (250-1150°). The form of the 

 curve is therefore known when such elements are calibrated at 

 three known temperatures. 



To render such calibration independent of the gas-thermom- 

 eter, a series of melting points of pure metals was determined 

 in which we took advantage of the opportunity to investigate 

 rather carefully the conditions which may affect the melting 

 temperature in certain cases. 



To this is further added a chapter on the measurement with 

 thermo-elements, the accuracy obtainable and their lasting 

 qualities. 



]. Determination of the Expansion. 



To determine the expansion of the platin-iridium bulb, a bar 

 of the same alloy (80 Pt, 20 Ir) was obtained, 500 mm in length 

 and 5 mm in diameter and its linear expansion studied up to 

 1000°. In connection with this we also undertook the measure- 

 ment of the expansion of several other substances at high tern- 



Natural size. 



peratures and hope to communicate in detail the full results of 

 the investigation at another time, restricting ourselves for the 

 moment to such as are necessary to the discussion of the gas 

 thermometer. Slight changes which the figures here commu- 



