of Energy in the Spectra of Solids, 433 



mode of action between the ponderable atoms and the sur- 

 rounding- gether. 



This relation, which does not require a knowledge of the 

 absolute temperature, is better fitted for experimental verifica- 

 tion than the preceding relations. For the moment we do 

 not possess direct data for such a verification. It is only in 

 a very imperfect manner that I have been able, by measuring 

 the ordinates and areas of the curves given by Prof. Langley, 

 to obtain the following values of I mas and E : — 





'maximal 



E 

 451 .. . 4-90 5-2 0-140 



603 



798 

 1088 



4-05 21-5 532 0*163 



3-63 39-7 1074 0-134 



3-27 64-9 1730 0-123 



Bearing in mind the imperfection of the method of verifi- 

 cation, the numbers of the last column, which according to 

 our theory should be equal, do not seem unfavourable to it. 



6. Tracing the Curves. — To be able to trace the curve of 

 energy it is necessary to give special values to p and/(0). If, 

 for example, we admit with M. Stefan* that the radiant energy 

 emitted by a body bears a direct ratio to the fourth power of 

 its absolute temperature, we shall be able to put 



p=l, f(d) = W; 



from which equation (10) will take the form 



I^Bjh- dh\-6 ( 16 ) 



Having determined the two constants B x and c of this equation 

 from a diagram of the normal invisible spectrum of lampblack 

 at 178° C, given by Prof. Langley, I have calculated and traced 

 the theoretical curves for this temperature and for 327° C. = 

 600° abs. (fig. 3). We see that these curves have all the 

 qualities that Prof. Langley attributes to the curves determined 

 experimentally by him ; that is to say : — 



(1) The radiant energy represented by the ordinates dimi- 

 nishes on both sides, yet extends indefinitely towards the 

 side of the less refrangible rays. • 



(2) When the temperature increases all the ordinates in- 

 crease, but in unequal proportion. Those which represent the 

 energy of the more refrangible waves increase always more 

 rapidly than those which correspond to the longer undulations. 



(3) Hence the fact, independently observed, that the maxi- 



* Stefan, Sitzber. Wien. Akad. vol. lxxix. p. 391 (1879). 



