coimexioii "mth Physical Optics. 



19 



sage, with its ambiguous sign prefixed in the " unforeseen 

 difficulties," shows that the methods are not likely to furnish 

 results accurate enough for testing important laws of nature. 

 It is not to be wondered at that the Professor, in much varia- 

 tion of results, should find some which were ^satisfactory.^ 

 As a pi'actical answer I here give some comparisons between 

 my own results and those of others. The most striking and 

 most essential difference between Fresnel's theoretical formula 

 and my experimental one is, that according to the former the 

 whole incident light should be reflected at 90° incidence, or 

 rather, that this is the limit to which the intensity of reflection 

 tends at very high incidences ; my formula gives only a quan- 

 tity varying from about 72 to 76 rays reflected out of every 

 100, as this limit for the ordinary kinds of crown plate and 

 flint tjlass. Hence we have a difference which cannot be 

 hidden in errors of experiment. This was one of Bouguer's 

 particular points. In speaking of the light reflected by water, 

 he says (p. 135), "dans I'incidence pour ainsi dire infiniment 

 petite la lumiere reflechie est les trois quarts de I'incidente ou 

 de la directe." We must note here that he measured his 

 angle of incidence from the surface and not from the normal, 

 and that a part only, instead of the whole, was reflected at the 

 highest possible incidences. Below I give Dr. Faraday's re- 

 sults with crown, plate, and flint glass, at 45° incidence (the 

 only angle he used) with my own at 40° and 50°, from the 

 before-named papers, as well as some calculated results. 



From Dr. Faradai/s results. 

 Crown glass, No. 3, reflected ^ = 4*13 of every 100 rays. 



Plate glass, No. 2, 

 Flint glass, No. 9; 



7, 





.-Fa = ^-29 



6, 





2^1 = 4-52 



2, 





26 ~ ^'^^ 



8, 





Ws — 3*88 



9, 





rU = 5-13 



Incidence. 



40° 

 50° 



From my own results. 



Crown glass, p, 

 reflects, of every 100. 



. 4-94 . . . 4-78 



. 5-68 . . . 5-92 



Calculations by my own formula : 



Incidence. Crown. Plate. 



40° ... 4-767 . . . 4-778 



45° . . . 5-f05 . . . 5-243 



50° ... 5=810 . . . 5-882 

 C2 



Flint. 



5-29 

 6-73 



Flint. 



5-320 

 5-884 

 6-656 



