of the Metals. 163 



the transparent metallic leaf is bounded on both sides by air, or 

 on one side by air and on the other by a highly refracting me- 

 dium such as glass. 



When the components of an incident ray, polarized parallel 

 and perpendicular to the plane of incidence, have an equal in- 

 tensity, then in the transmitted light the intensity of the com- 

 ponent polarized perpendicular to the plane of incidence prepon- 

 derates. 



As yet it is only in the case of gold that the author has suc- 

 ceeded in obtaining leaves which were bounded on both sides 

 byair, and at the same time sufficiently uniform and even. For 

 this purpose gold-leaf of a convenient shape was placed upon 

 distilled water and allowed to straighten itself out. It was then 

 lifted off by means of a small frame of brass which had been 

 previously burnt black, and of which the inside measure was 

 4 miilims. in breadth and 28 millims. in length. After the 

 evaporation of the water there thus remained in the frame a 

 transparent gold plate which could be fastened by means of the 

 frame to the axis of a goniometer and examined. Such a plate 

 is never quite flat, and consequently there is always an error 

 in the measurement of the angle of incidence, and this error may 

 amount to several degrees. 



The goniometer had a perfectly similar disposition to that 

 which Jamin* employed in his researches. Two brass tubes 

 blackened internally could be turned on a graduated circle about 

 the axis of the goniometer, to which they were directed perpen- 

 dicularly, and carried at their extremities two circles with NicoFs 

 prisms, in order to allow of the determination of the plane of 

 polarization of the incident and emergent rays. The rays re- 

 flected from or passing through metal were received by a Babi- 

 net's compensator previously to their entry into the NicoFs prism 

 used as the analyzer, in order to neutralize the difference of phase 

 of the components polarized parallel and perpendicular to the 

 plane of incidence. The rays emerging from the compensator 

 were then polarized in a straight line, and the azimuth /3 of then- 

 plane of polarization could be measured by means of the analy- 

 zing NicoFs prism, by turning it until a black band again ap- 

 peared between the parallel lines of the compensator ; at this spot 

 the light emerging from the compensator was fully extinguished. 



The brass tube which carried the polarizing NicoFs prisms 

 was provided with two diaphragms with cross lines, and the en- 

 tire apparatus was so placed that the shadows of both cross lines, 

 received on a plate of ground glass, covered one another when 

 a heliostat projected sunlight upon the NicoFs prism and then 



* Ann. de Chim. et de Phys. vol. xix. p. 370 et seq. ; vol. xxxi. p. 1/0. 



E2 



