Measurement of High Temperatures. 43 
high temperatures. This method depends upon our ability to 
obtain, from results such as are recorded in Table IX of that 
paper, a general expression for the radiating power of any 
given body as a function of the temperature, or, what amounts 
to the same thing, to find the values of the quantities, A, E 
and I, in Kirchhoff’s equation 
WV, 
e= Zz oh | yo (1) 
These quantities being given, the temperature of a source of 
light could be determined by compar ring the intensity of por- 
tions of its spectrum—as for instance “those lying between & 
and A+da, X and 4’+d2’, 2”, and 4” +d", &e.—with the corres- 
ponding wave lengths of the spectrum of a body of known 
accurate measurements have been hitherto unattain 
. Crova, in Comptes Rendus, has faba a similar 
method. He gave, however, no measurements of glowing tem- 
peratures upon which to base his method, and ignored entirely 
the very serious difficulties to be overcome before it can be 
made Sees available. M. Crova proposes the following 
three modes of procedure : 
1. “Au moyen de la longueur d’onde de la radiation qui 
limite Je reg vers le violet.”+ 
2. “Par la Fata du maximum calorifique du spectre qui 
se fbstcks utant plus du violet que la ee or ek "émis- 
sion et -jeagg sestiion 
8, 
Spectres. 
. La mesure vigoureuse des températures pourra étre 
faite par voie spectrometrique des que l'on connaitra la Joi 
€xacte de l’émission pour toutes les radiations et des constantes 
numériques pour chaque longueur d’onde. . . 
n the first method, the visible spectrum is “falsely assumed 
to have a clearly defined boundary at its end nearest the violet. 
* Kirchhoff, 2 elaatecians a iiber das Sonnenspectrum, Anhang, § 2. 
+ Crova, Etude spectrometrique de quelques sources rearioh & Comptes 
Rendus, Ixxxvii, 322. 
