COLOURS IN METAL GLASSES AND IN METALLIC FILMS. 

 were calculated for gold and for silver for the following values of p : 



/* = !, p='5, /A ='6, ft = 7, /t = '8, /A = '9, /* =1-0. 

 Equations (19) and (20) may be written 



405 



(19'), 

 4} ...... (20'), 



whence the values of u and * for gold and for silver were calculated for the above 

 values of p. The values of n 2 K thence obtained were checked against those obtained 

 by means of equation (18), namely, 



i 8 K = 3 ........... (18'). 



In the case of silver with /JL less than '8 it was, however, seen to be better to obtain 

 HK as the quotient of the value of n z n got from (18'), by the value of u got from (19') 

 and (20'), owing to the large probable error when HK was determined directly from 

 (19') and (20'). 



From equation (13') we find 



which are the same as equations (19) and (20) with p. = 1. Consequently, as should 

 be the case, the medium of spheres is equivalent to the solid metal wherein the 

 spheres are of such varied sizes that they fill the whole space. Another check on the 

 tabulated numbers is afforded therefore by a 

 comparison of the calculated and observed 

 values of n l 2 K l , Mj/q and n } . 



I believe that nearly all the numbers here 

 given for silver and for gold are subject to an ^^ 

 error of less than 1 per cent. 



The values of n~K = 3 and of 7? for the 

 potassium-sodium amalgam of DKUDE'S table, 

 ' Phys. Zeitschriffc,' January, 1900, are less 

 carefully calculated. 



9. Consider now the incidence of plane 



Fig. 10. 



polarised light on a plate of this medium. We shall first suppose the plate to be 

 very thin and therefore optically crystalline. 



Suppose the two surfaces of the film are 2 = and z = d, and that zx is the plane 

 of incidence. 



