2 Sosmati — Platinum- Rhodixhm Thermoelement. 



therinoeleineiit. It was shown that the relation of thermal 

 e.ni.f. to temperature can be represented over the i-ange 

 studied (4-00° to 1550°) by two parabolic curves, by means of 

 which temperatures between the standard uielting points can 

 be interpolated with an avei*age accuracy of one degree. 



Since this thermoelement can be conveniently used for the 

 measurement of temperatures beyond 1550°, continuing up to 

 the melting point of the platinum wire at 1755°, it was thought 

 desirable to extend its interpolation curve to cover the region 

 between palladium and platinum (151:9°-1755°). In this region 

 the thermoelement is the most accurate means we have for 

 measuring temperatures. Below 300°, on the other hand, 

 the sensitiveness of the platinum-rhodium element is very 

 low compared with the platinum resistance thermometer, the 

 copper-constantan thermoelement, or the mercury thermometer. 

 Nevertheless, it is often convenient to use this element for 

 measurements in the lower range; hence we have in addition 

 determined the course of the thermoelement curve from 0° to 

 300°. The present paper contains the results of these exten- 

 sions in the form of a table, which has been found very 

 convenient in this laboratory for the thermoelectric measure- 

 ment of all temperatures from 0° to 1775°. 



In connection with the work on the nitrogen thermometer 

 referred to above, some data were obtained on the thermal 

 electromotive force of rhodium alloys of various compositions, 

 which have been brought together with the few earlier deter- 

 minations that are available, to show the relation of the 

 thermal e.m.f. to the percentage of rhodium in the alloy. 



2. Scale of Temperatures. 



The melting and boiling points of pure substances form the 

 best basis for the calibration of secondary measuring devices, 

 such as the t]iermoelement, after these points have once been 

 determined ou the gas thermometer. The 0° and 100° points 

 are familiar. In the neighborhood of 200° and 300° the boil • 

 ing points of pure naphthalene and benzophenon were used. 

 The best gas thermometer determinations of these two points, 

 by Callendar and Griffiths,* and by Jaquerod and Wassmer,t 

 difEer by 0-26° at 218° and 0-4° at 305°. We have adopted 

 the data of Jaquerod and Wassmer, until further gas thermom- 

 eter nieasiirements shall have brought out the reasons for these 

 differences. The values are 217-68° for naphthalene and 

 305'4:4:° for benzophenon, at 760"™ pressure. :}: 



*PliiL Trans. Roy. Soc, elxxxii, A, 43-72, 119-157, 1892. 



t Jour. Chim. Phys., ii, 52-78, 1904. 



JThe results of Jaquerod and Wassmer have been used as the standard 

 since 1904 by the Eeseareh Laboratory of Physical Chemistry at the Massa- 

 chusetts Institute of Technology, in their work on electrical conductivity 

 at high temptratures. 



