PHYSICAL METABOLISM 235 



screen painted with some fluorescent substance, the 

 colour of the fluorescent light is the same through- 

 out the whole length of the visible spectrum which 

 excites the fluorescence, and is found in some cases 

 with quartz prisms to extend as far into the ultra- 

 violet as ten times the length of the visible 

 spectrum. By examining the light thus obtained 

 with a crossed prism whose edges are parallel to 

 the original spectrum two spectra are obtained ; 

 one, the fluorescent spectrum, which is merely 

 shifted some distance parallel to itself, and wider 

 than the original spectrum ; the other, the visible 

 solar spectrum, which now appears distorted in a 

 straight line at an angle to the original spectrum. 

 The fluorescent spectrum, which is parallel to the 

 latter, lies well within the space between the 

 original spectrum and its shifted image ; so that 

 every ray in it is less refracted than the correspond- 

 ing ray in the visible spectrum, at least as regards 

 those solar rays which are of greater refrangibility, 

 and to which the fluorescence owes its origin. 



A glance at the accompanying figure will make 

 the efiect quite clear. 



By this means it is found that the fluorescent 

 spectrum is always not merely of less refrangi- 

 bility than the exciting light, but also that its 

 period is independent of part of the spectrum 

 which gives rise to it.^ Thus, as Stokes first ob- 

 served, the homogeneous light of high refrangi- 



1 That is, speaking generally. See Moore, Astrophysical 

 Journal, March, 1905. 



