in the Neighbourhood of the Polarizing Angle. 9 



of behaviour of various parts of the same surface. A surface, 

 originally clean, and then greased with olive-oil carried upon 

 a previously ignited platinum wire, frequently showed streaki- 

 ness, when the eye, observing through the nicol, as usual, 

 was focused upon the surface. 



Except perhaps in the case of oleate, none of these experi- 

 ments, many times repeated, gave any evidence of a real 

 departure of properly skimmed surfaces from the laws of 

 Fresnel ; and it looked very much as if all the results enun- 

 ciated for liquids by Jamin were vitiated by the presence of 

 greasy films. That a film of extreme tenuity would suffice 

 was certain. The band from water was very obviously dete- 

 riorated by a film of olive-oil, which needed to be condensed 

 four or five times in order to stop the camphor movements. 



But it was impossible to rest here. It was necessary actu- 

 ally to measure, or, if that were not possible, to find limits for, 

 the ellipticity of the various surfaces. And for this purpose a 

 much more elaborate apparatus had to be installed. 



Sunlight, reflected horizontally from the heliostat, passed 

 through a diaphragm in the shutter of about \ inch diameter, 

 and thence to a collimating lens of 23 inches focus. It was 

 next reflected in the required oblique direction by an adjust- 

 able mirror, and caused to traverse the polarizing nicol, 

 mounted in a circle that allowed the orientation of the nicol 

 to be read to a minute of angle. After reflexion from the 

 surface under examination the light traversed in succession a 

 quarter- wave-plate and the analysing nicol, and was then 

 received into the eye, either directly, or with the intervention 

 of a small telescope magnifying about twice. In either case 

 the eye was focused upon the diaphragm, which was provided 

 with cross wires ; so that the rays which fell upon any part 

 of the retina constituted a parallel pencil, not only at the 

 surface of the liquid, but also in their passage through the 

 nicols and quarter- wave-plate. The latter was of mica, and 

 both it and the analysing nicol w T ere mounted so as to be 

 capable of rotation about the direction of the reflected ray. 



The adjustments were made as follows. The analyser and 

 quarter-wave- plate being removed, the mirror and polarizer 

 were adjusted until the dark spot was central in relation to 

 the cross wires. A rotation of the mirror, altering the anffle 

 of incidence, moves the spot vertically, while a rotation of the 

 polarizer moves it horizontally. The zero for the eye-nicol 

 could have been found by rotating the polarizer and then 

 recovering the dark spot ; but in order to avoid risks of dis- 

 placement, which might be fatal in such a delicate inquiry, I 

 preferred to leave the first nicol untouched, and to depolarize 



