388 Prof. E. Wiedemann on the 



in a rotating mirror, which were separated by perfectly dark 

 intervals. The duration of the oscillations was 100 x 10 ~ 7 sec. 

 The dark interval which corresponded to the decrease from a 

 high intensity to a very small intensity, let us say to about 

 r _i_ f the original value, amounted to about -j 1 ^ of the 

 oscillation period, i.e. 10 -6 sec. 

 Hence 



T — ° — T -&X10 -5 /> — 10 7 



The true value is no doubt greater. 



(c) Electric discharges in Geissler's tubes often appear so 

 close together in a rotating mirror that they follow each other 

 at intervals of T oooo second, and yet they are separated by 

 almost absolutely dark intervals. We may here assume that 

 during yooWo second the brightness has not sunk more than 

 to ^ Q , and then we obtain 



J< = ^=J *-*x'- 5 , 6 = 2-3 xlO 5 . 



(d) In the phosphoroscope fluorescent liquids do not shine 

 after a longer interval than ^/qq seconds. We may, there- 

 fore, put here 



J ( = f^Jo«- ixl0 "", or 6 = 5x10*. 



b is, however, certainly greater. Its value cannot, as yet, 

 be determined more accurately by this method, since we 

 are not as yet acquainted with any phosphorescent liquids, 

 and can therefore only give an upper limit for t. In the 

 meantime the magnitude of b may also be determined for 

 pure solutions of fluorescent substances in the following 

 manner: — We determine it for solutions of any fluorescent 

 substance, eosin, in solid gelatine, and then in gelatine con- 

 taining increasing quantities of glycerine or water, which, 

 as I have shown, phosphoresce distinctly. By extrapolation 

 we obtain the value for an eosin solution in glycerine with- 

 out addition of gelatine. Since in winter the sun's rays, 

 necessary for these measurements, hardly fall upon the 

 Institute in Erlangen, I hope to report later as to nume- 

 rical data in continuing the investigations upon fluorescence 

 and phosphorescence. The interferences observed with 

 fluorescent bodies with large differences of path cannot be 

 used, like the similar phenomena in gases, for the determina- 

 tion of b. In gases each impact produces luminous motions, 

 which die away upon the free path, but with constantly illumi- 



