200 FRESNEL. 



spar. Whenever the separation of the two rays was so 

 small as to escape detection by the eye, the observer 

 remained in doubt and did not venture to pronounce it 

 doubly refractive. Now, however, by the aid of a 

 method which a member of the Academy has pointed 

 out,* the existence of double refraction manifests itself 

 by characteristics quite independent of the separation of 

 the two images. No substance, however thin it may be, 

 possessed of this property, can escape this new mode of 

 examination. But, if it were certain that double refrac- 

 tion could not exist without our perceiving the very 

 manifest phenomena on which this method is founded, it 

 would not appear equally incontestable that it ought 

 necessarily to accompany them ; and a doubt in regard 

 to this might seem the more natural since the author of 

 this method has himself found certain plates of glass 

 which, without separating the images in a perceptible 

 degree, yet give birth to all the phenomena in question : 

 since a distinguished philosopher of Berlin, M. See- 

 beck, afterwards proved that all glass rapidly cooled 

 enjoyed the same property ; and since, lastly, a very 

 able experimenter of Edinburgh produced the same 



* The author here alludes to his. own discovery of the polarized 

 colours, made also quite independently by Brewster about the same 

 time. These tints are now familiar to most persons by means of the 

 little instrument called the polariscope. By placing a plate of sele- 

 nite, mica, &c., far too thin to exhibit any separation of images, in 

 polarized light, and viewing it through an analyzer, these brilliant 

 tints convey distinct evidence of the existence of that property, since 

 they are shown theoretically to depend solely upon its existence, how- 

 ever insensibly small its amount may be. It therefore seems impor- 

 tant for the verification of theory, to show independently its existence 

 in any substances which exhibit the tints. Glass ordinarily possesses 

 no such power; but plates of unannealed glass exhibit the tints. 

 Hence the importance of the experiments mentioned to show its 

 existence directlv. Translator. 



