law that the maximum of polarization takes place in reflection from the joint surface of two media. The angle of incidence is complement to the angle of refraction, or the tangent of incidence (as the author expresses it,) is equal to the quotient of the indices of refraction of the media.
After describing in a series of propositions the various degrees in which light becomes polarized by reflection or refraction at different angles, and the number of reflections or refractions necessary to effect complete polarization at various angles remote from that which produces the maximum, the author investigates the origin of a certain quantity of unpolarized light which exists even at the maximum polarizing angle in reflection from substances of high refractive power; and he shows it to depend on the different refrangibility of differently coloured light. For when the incidence is such that the mean refrangible rays are completely polarized, it is evident that the incidence will not be such as to polarize completely either the red or the violet rays, and consequently a beam composed of these will appear as white light not polarized; and when the polarization is effected at the surface of substances of high refractive and dispersive power, this portion will form a large proportion of the whole reflected light. On the contrary, any pencil of homogeneous coloured light, though only once reflected, may be completely polarized, even at the surfaces of the densest substances, if incident at an angle correctly adapted to its refrangibility.
The author purposes, on some future occasion, to point out the laws which regulate the polarization of light under various other circumstances not noticed in the present communication.
It is not surprising, says the author, that neither Sir Isaac Newton, nor Dr. Herschel, nor any other writer who has followed on the same subject, have given any explanation which appears to him to be satisfactory, since they have not been in possession of the phenomena connected with this inquiry. After stating various objections which he considers as conclusive against the alternate disposition to be reflected or transmitted inherent in the rays of light, and recurring at certain equal intervals which are expressed under the name of fits by Sir Isaac Newton, Mr. Knox proceeds to describe those new phenomena which form the principal subject of bis paper, and which he has been enabled to make by the assistance of the method of observing such appearances employed by Dr. Herschel. This method consists in using the shadows of some opake substance held over thin plates of glass, for the purpose of distinguishing from each other the several effects produced by different surfaces employed at the same time. If a plate of unsilvered glass be laid upon a table before
