178 



Prof. G. G. Stokes. On a remarkable [Feb. 26, 



of the spectrum. The particular part varies from one crystal to 

 another ; it may be anywhere from the extreme red to the extreme 

 violet. It stands out by its greatly superior brightness on the 

 general ground of the continuous spectrum, and when it is fully 

 formed the reflection over the greater part of it appears to be total. 

 The appearance recalls that of a bright band such as the green band 

 seen when a calcium salt, or the orange band seen when a strontium 

 salt, is put into a Bunsen flame. The bright band is frequently 

 accompanied right and left by maxima and minima of illumination, 

 forming bands of altogether subordinate importance as regards their 

 illumination. Sometimes these seem to be absent, and 1 cannot say 

 whether they are an essential feature of the phenomenon, which some- 

 times fail to be seen because the structure on which the bands depend 

 is not quite regularly formed, or whether, on the other hand, they are 

 something depending on a different cause. 



Disregarding these altogether subordinate bands, and taking 

 account of the mean illumination, it seems as if the brightness of 

 the spectrum for a little way right and left of the bright band were 

 somewhat less thau that at a greater distance. 



When the main band occurs at either of the faint ends of the 

 spectrum, it is visible, by its superior brightness, in a region which, 

 as regards the continuous spectrum, is too faint to be seen, and thus 

 it appears separated from the continuous spectrum by a dark 

 interval. 



When the angle of incidence is increased, the band 1 moves in the 

 direction of increasing refrangibility, and at the same time increases 

 rapidly in breadth. The increase of breadth is far too rapid to be 

 accounted for merely as the result of a different law of separation of 

 the colours, whieh in a diffraction spectrum would be separated 

 approximately according to the squared reciprocal of the wave-length, 

 while in bands depending on direct interference the phase of illumina- 

 tion would change according to the wave-length. 



8. The transmitted light being complementary to the incident, we 

 have a dark band in the transmitted answering to the bright band in 

 the reflected. In those crystals in which the band is best formed, it 

 appears as a narrow black band even in bright light. When the band 

 first appears as we recede from a normal incidence it is extremely 

 narrow, but it rapidly increases in breadth as the angle of incidence 

 is increased. 



9. Some of the general features of the phenomenon were prettily 

 shown in the following experiment : — 



Choosing a crystal in which the bright band in the reflected light 

 began to appear, as the incidence was increased, on the red side of the 

 line D, so that on continuing to increase the incidence it passed 

 through the place of the line D before it had become of any great 



