Light is polarized at the Surface of the Metals, 301 



and it is only bv refracting them in peculiar directions that theory 

 indicates, that we may show plenty of evidence for the law of 

 their tints. In short, this difficulty almost entirely disappeared 

 in the laminae of silver polished by the hammer. Then the por- 

 tion of light, which takes the fixed polarization at each reflec- 

 tion, becomes extremelv feeble comparatively to that which pre- 

 ser\'es the moveable polarization, at least when we do not pre- 

 sent the laminae to the incident rays under an extreme obliquity; 

 for we know that in this case all the plane surfaces, even those 

 which have been purposely unpolished, assume the specular po- 

 lish. Thus by avoiding gr^at inclinations, and confining our- 

 selves to few reflections, the laws of moveable polarization are 

 alone perceived, and the tints of the fasciculi which nothing al- 

 ters are developed with the greatest regularity according to the 

 series of Newton's rings. This case is fortunately that which 

 first offered itself to my observations, and it served me as a guide 

 to pass on to the more complicated case, in which the moveable 

 polarization becomes less sensible, and the fixed polarization 

 more considerable. Now, since the only difference of a polish 

 more or less smooth determines more abundantly the passage 

 of the light reflected from one of those states to the other ; 

 ought we not to conclude, that here, as in the crystals endowed 

 with double refraction, the moveable polarization is also the first 

 which is exercised, when the luminous molecules are sufficiently 

 removed from the reflecting surface in order that the asperities 

 of the latter should be insensible at the distance at which they 

 are ? But the distance always diminishing, and the effect of 

 the inequalities of the surface becoming more sensible, it hap- 

 pens, if they are very little, that the reflecting force becomes 

 energetic enough to give polarization to a great part of the lu- 

 minous molecules; whereas, if the asperities are stronger, and 

 consequentlv the reflecting power weaker, a greater number of 

 these particles continue their oscillations without being fixed. 

 We have here, therefore, in the action of the bodies on light, 

 the exam])le of an effect analogous to those of cajjillarity ; for if, 

 as M. Laplace has demonstrated, the latter arc produced by the 

 attraction more or less strong which a bodv exercises at its sur- 

 face, according as that surface is plane, or concave, or convex, 

 in the same way in the new phaenomena wiiich I announce, the 

 different configuration of the reflecting surfaces exercises on the 

 luminous molecules a different mode of polarization. But the 

 phaenomena of capillarity are produced by differences of curva- 

 ture appreciable to our senses, and even to our measures, instead 

 of our being obliged, in order to change the mode of action of 

 bodies upon light, to produce undulations almost imperceptible, 

 such as the unequal nature of the polish gives them. Further, 



we 



