544 PROCEEDINGS OP THE AMERICAN ACADEMY. 



of electrolyte, size of electrodes, etc.) remain constant, and there is every 

 reason to believe that the maximum temperature reached is the same in 

 the two cases. Any criterion of the temperature corresponding to either 

 of these spectra would appear to be wanting unless they correspond to 

 the same temperature. Nor does the fact of the absence of some of the 

 strong lines of a metal appear to offer any measure, however qualitative, 

 of the temperature, since other strong lines usually belonging to the same 

 temperature are present. 



It is certainly very easy to be led into reasoning in a circle in a case 

 where much of our knowledge of temperatures has been deduced from 

 observations on spectra, and much of our theory as to the correspondence 

 of certain spectra with certain temperatures has been drawn ultimately 

 from the same source. 



Jefferson Physical Laboratory, 

 Harvard University, Jan. 15, 1904. 



