in the Neighbourhood of the Polarizing Angle. 3 



cise, on peut remarquer une tendance du coefficient k a 

 diminuer avec i'indice, et quand celui-ci est approximative- 

 ment egal a 1*45, k est nul : l'indice continuant a decroitre, 

 le coefficient k reprend des valeurs sensibles et croissantes." 



Since the time of Jamin many valuable observations upon 

 reflexion have been made by Quincke and others, which it is 

 not necessary for my purpose further to refer to. In 1889* 

 Drude made the very important observation that the ellip- 

 ticity of a freshly split surface of rock-salt is very small, but 

 that it rapidly increases on standing. And he concludes 

 generally that solid bodies reflect from natural cleavage sur- 

 faces according to the laws of Fresnel. It is remarkable that 

 fluids are excluded from this statement. Indeed Drude ex- 

 pressly remarks that in fluids the natural surface-sheet, which 

 forms the transition from one medium to another, exercises 

 an essential influence upon the reflexion. 



Nearly contemporaneously with Drude's observation I made 

 a first attempt to examine whether the different behaviour 

 which Plateau found for water and for alcohol in respect of 

 superficial viscosity, and which I was disposed to attribute 

 to surface contamination, corresponded to anything in the 

 phenomena of reflexion near the polarizing angle, but at 

 that time 1 was misled by a faulty nicol, and failed to observe 

 anything distinctive. Subsequently, having proved that the 

 superficial viscosity of water was due to a greasy contamina- 

 tion, whose thickness might be much less than one millionth 

 of a millimetre, I too hastily concluded that films of such 

 extraordinary tenuity were unlikely to be of optical importance, 

 until, prompted by a remark of Sir Gr. Stokes, I made an 

 actual estimate of the effect to be expected. The thickness 

 required to stop camphor movements, viz., 2 x 10 -7 centim., 

 is about g-J-Q of X D . This will be one factor in the expression 

 for the amplitude of reflexion due to grease. There will be 

 another somewhat small factor expressive of the difference of 

 optical quality between grease and water ; so that the intensity 

 of reflexion at the polarizing angle might on this view be of 

 the order 10 ~ 6 , or 10 ~ 7 , a quantity quite appreciable, when 

 the incident light is from the sun. Thus encouraged, I re- 

 turned to the attack, and on October 2nd, 1890, examined the 

 image of the sun as reflected from water at the polarizing 

 angle. The arrangements were very simple. Sunlight re- 

 flected horizontally into the dark room from a heliostat out- 

 side was again reflected downwards by an adjustable mirror. 

 The water was contained in a large porcelain dish to the 



* Wied. Ann. xxxvi. p. 532. 

 132 



