PROPERTIES OF LIGHT. 1^3 



yond the angles I have mentioned, the phenomenon will ' 

 cease to take place; and the farther we go from these 

 limits, on either side, the greater will be the quantity of 

 light reflected. 



This faculty of entirely penetrating transparent bodies, This property 

 which the liffht has acquired by its first reflection, it will ""^f^'"^^ o"" left. 

 , .^ . • • , I,- u I I . under different 



lose or retain in various circumstances, which 1 have stu- circum- 



died ; and thus I have been led to the following law, ac- stances. 

 cording to which this singular phenomenon is effected. 



If a second glass be made to turn round the first re- Lawof this 

 fleeted ray, a, constantly making with it an angle of 35° phenomenon, 

 25' ; and if in a plane perpendicular to this ray we conceive 

 two lines, one, 6, parallel to the first glass, and the other, 

 c, parallel to the second; the quantity of light reflected by 

 the second glass will be proportional to the square of the 

 cosine of the angle included between the lines be: it is at 

 its maximum when these lines are parallel, and null when 

 they are perpendicular. So that the limits of the pheno- 

 menon are relative to three rectangular axes, a, b, c, one of 

 which is parallel to the direction of the ray, another to the 

 first reflecting surface, and the third is perpendicular to the 

 two former. 



For the second glass let us substitute a metallic mirror, Metallic mir- 

 and call the rectangular axes of the second ray, analogous ""^^ substituted 

 to the axes a, b, c of the first, a', b\ c. If this ray be re- JIass. 

 ceived on a polished but unsilvered glass, which makes with 

 it an angle of 35° 25', we shall perceive the following phe- 

 nomena, which are independant of the angle of incidence 

 on the metallic mirror. If 6' be parallel to 6, that is, if the 

 metallic mirror be parallel to the axis 6, the ray it reflects 

 will retain its properties with respect to a glass situate pa- 

 rallel to the axis c'; it will penetrate it entirely. If b' be 

 parallel to c, the reflected ray will retain its properties for a 

 glass parallel to the axis h'. 



In the intermediate positions, the quantity of lio-ht, that 

 will have retained its property for a glass parallel to the 

 axis 6', i^ proportional to the square of the sine of the angl^ 

 comprised between the axes 6' A; an^ that which has re- 

 taiued its property with respect to ft glsM pftj-allel to th« 

 M2 »x„ 



