POLARIZED LIGHT AT THE SURFACE OF A UNT AX A L CRYSTAL. 
607 
pretty clearly too big, as also possibly is that in line 14. A very small displacement 
in the position of the sugar cell, so that the light traversed it somewhat obliquely, 
would give rise to an error of the kind here considered. The cell was usually adjusted 
by observing the beam of light reflected from its first face. This could be made with 
a little care to travel back through the biquartz, and in that case the light clearly fell 
normally on the cell. 
. The first twelve sets of observations recorded in Table II. were made on October 
28, 1880, the last two a few days later. 
Thus, unless there is some regular source of error in the experiments, we must 
conclude that the formula connecting the plane of polarization and the angle of 
incidence in the case in which only the ordinary ray traverses a crystal of Iceland 
spar, as given by the electro-magnetic theory, is only true approximately. 
The method does not enable ns to determine accurately by experiment the position 
of the plane of polarization of the incident light with reference to the face of the 
crystal on which it falls; we can however compare the rate of change of the position of 
this plane, as the angle of incidence varies, found from experiment with its value 
deduced from theory. 
We arrive at the conclusion that this rate of change as given by theory is too rapid 
when the angles of incidence are large, that for angles of incidence of from 30° to 60° 
the theoretical and experimental rates agree, while for lower angles of incidence the 
theoretical rate is again possibly too great; the last inference however being a little 
uncertain. 
Two other shorter sets of observations made about the same time confirm these 
statements exactly, and it does not seem worth while to print a table of the numbers 
actually arrived at. 
4 I 
MDCCCLXXXII. 
