from the surface of Transparent Bodies. 83 



It is not difficult to deduce from Mr. Green's equations the 

 following : — 



cot e = Q(/x 2 cot i + cot r) ~\ ._. 



cot e t = — Q(/x 2 cot i — cot r) J 



Equations (2) belong to the case of light polarized perpen- 

 dicular to the plane of incidence ; and since light polarized in 

 the plane of incidence is proved by Mr. Green to undergo no 

 change of phase by reflexion, it is easy to see that e t —e is the 

 difference of phase produced by reflexion in the two components 

 of light originally polarized in any azimuth. 



The quantity which M. Jamin has selected for comparison 

 with theory is 



e t — e 

 ISO* 



A few trials convinced me that the coefficient Q, as defined by 

 equation (1), would never represent the observations; I there- 

 fore proposed to myself to examine whether it were not possible 

 to represent M. Jamm's observations by equations (2), consider- 

 ing Q as an unknown constant to be determined by experiment. 



The formulas from which M. Cauchy deduces the difference of 

 phase cannot be reduced to the form of equations (2) . In the 

 following table I have given both M. Cauchy' s results and those 

 deduced from the modification of Mr. Green's formulas, which I 

 have proposed. 



Table I. 



Sulphuret of Arsenic (transparent). 



Difference of phase. 



//,=2-454. Q=2-55. 



G2 



