264 FRAGMENTS OF SCIENCE. 



cation lending its weight of proof to the undulatory theory 

 on which the predictions are founded. 



The selenite ring-system, already referred to, is a most 

 delicate reagent for the detection of polarized light. When 

 we look normally, or perpendicularly, at an incipient cloud, 

 the colors of the rings are most vividly developed, a dim- 

 inution of the color being immediately apparent when 

 the incipient cloud is regarded obliquely. But let us con- 

 tinue to look through the Nicol and selenite normally at 

 the cloud : the particles augment in size, the cloud becomes 

 coarser and whiter, the strength of the selenite colors be- 

 coming gradually feebler. At length the cloud ceases to 

 discharge polarized light along the normal, and then the 

 selenite colors entirely disappear. If, now, the cloud be re- 

 garded obliquely the colors are restored, very vividly, if not 

 with their first vividness and clearness. Thus the cloud 

 that has ceased to discharge polarized light at right angles 

 to the illuminating beam, pours out such light copiously in 

 oblique directions. The direction of maximum polariza- 

 tion changes with the texture of the cloud. 



But this is not all ; and to understand, even partially, 

 what remains, a word must be said regarding the appear- 

 ance of the colors of our plate of selenite. If, as before 

 stated, the plate be of uniform thickness, its hue in polar- 

 ized light is uniform. Suppose, then, that by arranging 

 the Nicol the color of the plate is raised to its maximum 

 brilliancy, and suppose the color produced to be green ; on 

 turning the Nicol round its axis the green becomes fainter. 

 When the angle of rotation amounts to 45° the color dis- 

 appears ; we then pass what may be called a neutral posi- 

 tion, where the selenite behaves, not as a crystal, but as a 

 bit of glass. Continuing the rotation, a color reappears, 

 but it is no longer green but red. This attains its maxi- 

 mum at a distance of 45° from the neutral position, or, in 

 other words, at a distance of 90° from the position which 



