1884.] Some Experiments on Metallic Reflection. 195 



Owing to the unequal reflection of light polarised in and perpen- 

 dicularly to the plane of incidence, the experimental results contained 

 in the paper published in the " Proc. Roy. Soc." vol. 35, p. 26, 

 cannot be the true values of the amount of light reflected by the 

 mirrors. 



The lamp light which was incident upon the mirrors, being equivalent 

 to two beams of half its intensity polarised in and perpendicularly to 

 the plane of incidence, and light polarised in and perpendicularly to 

 this plane being unequally diffused by the paper, as well as unequally 

 reflected by the mirrors, the observed intensity must, as Professor 



Stokes pointed out, be r ^ and not iiL. 

 r r+1 2 



Assuming the values of J 2 and I 2 found from the determinations 



with polarised light, and the value of r from the measurements 



r J 2 4- 1 2 



contained in Table IX, the value of for the two mirrors at 



r + 1 



different incidences were calculated out,* and the numbers thus 

 obtained are given in the fourth and seventh columns of Table 

 VIII. 



In the case of the steel mirror the observed and calculated numbers 

 agree tolerably, as well, perhaps, as could have been expected, 

 recollecting the nature of the determinations, but with the speculum 

 metal mirror the results are discordant, the calculated results being 

 in all cases too low. The observations with unpolarised light were 

 made immediately after the mirror had been polished, whilst an 

 interval of several months elapsed before those with polarised light 

 were finished. Although the mirror was kept in a dry warm room, 

 and in a closed case containing lime, its surface was usually found to 

 be covered with a slight film ; this was readily removed by rubbing it 

 gently for a few seconds with a piece of wash-leather, and the surface 

 then appeared perfectly bright. 



After the conclusion of the experiments with polarised light, the 

 photometer was rearranged in its original form, and three observations 

 were made of the amount of unpolarised light reflected by the mirror 

 at an angle of 30° in order to ascertain whether the reflective power 



that the illumination was so different that was produced by light polarised in the 

 two ways, even though the polarisation of the light coming from the paper was 

 very feeble when the light incident was common light. I overlooked this when I 

 proposed (" Proc. Roy. Soc," vol. 35, p. 39) to measure r by measuring the 

 polarisation of the light coming in this case from the paper, and regarded r as only 

 " a little " greater than J on the strength of the author's assurance that the polari- 

 sation was so slight. — Or. Gr. S.] 



* The actual calculations were made with the equivalent formula - + r ^ , in 



* 100 + r 



which r = 63-38. 



