Polanzalion of Light at Surface of Metals. 297 



made to act first on the upper side of the small piston, the un- 

 der side of both being then open to the condenser. (N. B. The 

 upper side of the large piston is in this arrangement always kept 

 in open communication with the condenser) : of course the large 

 piston descends in equilibrium, both sides of the large being open 

 to the condenser. When both pistons reach the bottom, the 

 exhausting valve is shut, and the steam from above tlie small 

 piston is sent under both pistons, and both ascend ; — the upper 

 side of the large, as before observed, being open to the condenser, 

 and the small piston ascending in equilibrium. When they reach 

 the top, the under side is again o):)ened to the condenser, and a 

 fresh supply of steam admitted on the piston in the small cyhn- 

 der, and the stroke is again repeated. In the engraving the pipe 

 A a, which is the communication between the top of the large 

 cvlinder and the eduction pipe, is broken off at a, for the pur- 

 pose of showing the lower valves, but is actually continued to 

 and joins the eduction pine at B, 



Fig. 3 is the same in arrangement as fig. 2, only the action 

 of the power is reversed — the valves which in fig. 2 are at the 

 top of the small cylinder being here placed at its lower end. 



tiV. Dclerminatum of the Laws according to which Light is 

 polarized at the Surface of the Metals. By M. Riot. 

 Extracted from a Memoir read to the Institute on the \5th 

 of May 1815*. 



When Malus had discovered the polarization which light un- 

 dergoes on being reflected at the surface of diaphanous bodies, 

 he also ascertained that this phsenomenon is not produced, at 

 least in the same way, on the surface of the metals. Since that 

 memorable epoch in science, he returned twice to this singular 

 exception; but without doubt, if time had not been wanting, he 

 would have seen the propriety of modifying the first ideas which 

 he published, and the true lawsof the metallic polarization would 

 not now be a desideratum. 



Several facts have since been added to the results of Mains. 

 1 have shown in my work upon Light, p. 137, that in general two 

 kinds of reflections take place at the surface of coloured bodies! 

 one, which seems to take place outside of the bodies, acts indis-p 

 criminately on all the luminous molecules, and produces a white 

 ray, if the incident light be white ; the other, deeper seated, 



• Annales de C'/timie, tome xciv, p. '209. 



acts 



