754 TRUEBLOOD. 



4. Thermometers and the Measurement of Temperature, 

 a. Description of thermometers. 



All temperature measurements in this work have been made by 

 means of platinum resistance thermometers. A large amount of 

 experimenting, the details of which are not of much interest, was done 

 before a satisfactory type of thermometer was evolved. The require- 

 ments of a thermometer suited to exact calorimetry at fairly high 

 temperatures and pressures are rather severe. The chief ones may 

 be said to be: (a) mechanical durability, (6) permanence of calibration, 

 (c) ability to follow fluctuations of the temperature of the sm-roundings 

 with reasonable rapidity, {d) similarity in all respects of the two 

 members of a pair of thermometers. The last-named is particularly 

 important where thermometers are to be used differentially, because 

 the advantage of the differential method obviously decreases as the 

 difference in the resistance of the two thermometers at the same 

 temperatiu-e increases. 



The type of thermometer which was finally adopted and used is 

 illustrated in Fig. 8. It is essentially an adaptation of the calorimetric 



mil 



Figure 8. One of a pair of similar platinum resistance 

 thermometers used for the temperature measurements. 



thermometer of Dickinson and Mueller,* but is naturally much more 

 substantial mechanically, and much less sensitive. It is of the Callen- 

 dar type, the leads being of No. 26 copper wire, silver soldered at one 

 end to the platinum, at the other end to copper binding posts. In 

 the instrument shown, the binding posts are silver-plated. It was 

 found desirable to do this and also to silver-plate the lugs at the 



4 Bulletin of the Bureau of Standards 3, 641 (1907). 



