588 PROCEEDINGS OF THE AMERICAN ACADEMY. 



show fluorescence in solution often show it also when dried in gelatine, 

 and sometimes even in the pure solid state; but in the latter case the 

 fluorescence color is usually quite different from that of the same 

 substance in solution. 



Our present methods of investigation seem inadequate. From the 

 mass of fact which has been gathered it seems possible that fluorescence 

 in solids, and especially in inorganic solids, is determined by the presence 

 of small quantities of impurities, but in the uranium salts and the platino- 

 cyanides we have substances which are, according to the usual standards, 

 pure, and which are nevertheless capable of exhibiting strong fluorescence, 

 and organic substances which are pure in the same sense often exhibit 

 this property in the strongest degree. The conclusions which may be 

 drawn from the chemical attacks on the question are few and uncertain, 

 and there are as yet no general principles which can be stated. 



The only methods of investigation which have been applied to the 

 problem with the exception of chemical analysis and synthesis are those 

 of spectroscopy. 3 In spectroscopy again we have a multitude of facts 

 with only an occasional connection of narrow scope, and the results 

 which may be hoped for are immediately limited by our limited knowl- 

 edge of the method itself. There is satisfaction in knowing the spectrum 

 of the light emitted by a fluorescing substance under definite conditions 

 of excitation, and such knowledge has of course scientific value, but 

 any extension of our knowledge of fluorescence based on spectroscopic 

 methods can go but very little farther, since we know practically nothing 

 of the mechanism of spectrum production even in the simplest case. 

 The fundamental problem of the appearance of a fluorescence spectrum 

 goes back to the question how any spectrum is produced, and the complete 

 answer to one of these questions includes the answer to the other. 



One general fact appears to hold for every case of fluorescence, and 

 that is selective absorption by the fluorescing substance. This substance 

 is able to select from the variety of light vibrations which enter it those 

 of certain definite periods, transforming their energy in part into heat, 

 about the period of which we know very little, and in part into light, 

 which is emitted and which determines the spectrum of the fluorescence. 

 This fact of selective absorption brings with it the conclusion that there 

 is something in the substance which has the power of taking up periodic 

 disturbances of definite period, and with their aid setting up within 

 itself disturbances of the same general nature but usually of different 



3 For a possible variation see Nichols and Merritt, Pliys. Rev., 19, 396 (1904). 



