ON LIGHT. 377 



non j and to study it in its simplicity, we must in idea 

 break it up into its component elements, and examine 

 their phenomena per se. Now it results, from a series 

 of experiments too extensive and refined to be here de- 

 tailed, and from reasonings upon them which the gene- 

 rality of our readers could hardly be expected to follow, 

 that when a ray, polarized in any plane, undergoes re- 

 flexion in a different plane, the reflected portion comes 

 off in all cases more or less elliptically polarized that 

 is to say, that it consists of, or can be resolved into, two 

 rays, the one polarized in the plane of incidence, the 

 other in a plane at right angles to it that both these 

 portions have undergone a change of phase at the mo- 

 ment of reflexion, but not the same for both, so that 

 arriving at the surface in the same phase, they quit it in 

 different, and therefore constitute by their superposition 

 an elliptically polarized ray. The amount of ellipticity 

 varies, for each reflecting medium (according to the 

 nature of its material) with the angle of incidence at 

 which the reflexion takes place, and also with the inclina- 

 tion of the plane of incidence to that of the primitive 

 polarization of the incident ray. If the reflexion take 

 place on ordinary transparent media of not very high 

 refractive power, as glass, or water, and at the polariz- 

 ing angle, the degree of ellipticity is so slight that the 

 vibration may be considered as rectilinear, and the re- 

 flected ray as completely polarized in the plane of 

 incidence. As the refractive power of the surface in- 

 creases, the ellipticity impressed is greater, and in some 

 substances, of very high refractive power, such as dia- 



