Investigation of Polarized Light. 221 



the Nicol with two maxima and two minima. The maxima 

 occur when the principal section is coincident with the plane 

 of incidence, and the minima in the alternate positions at right 

 angles to the former. If we reckon the angle of rotation a of 

 the prism from the position of the maxima, for which the 

 plane of polarization of the ray is at right angles to the plane 

 of incidence, and if we denote by J the intensity of the current 

 for the angle a as found from the reading of the galvanometer, 

 then, within the limits of experimental error, the relation 

 between these is expressed by the formula 

 J = A cos 2 a + B sin 2 a, 

 where A denotes the maximum current-intensity (for a = 0°), 

 and B the minimum intensity (for « = 90°). In order that 

 this connexion shall be verified without doubt, the positions 

 of maximum and minimum position must be determined with 

 as much sharpness as possible. The changeability of J is, as 

 the nature of the function shows, least in the neighbourhood 

 of a = 0° and a. = 90°: consequently, it is only possible to 

 determine these principal positions by direct experiment with 

 an uncertainty of more than a degree. We therefore pre- 

 ferred to determine the position of the greatest changeability 

 of J, namely, when a, = 45°. Whilst one of us slowly 

 rotated the NicoFs prism, the other observed at the gal- 

 vanometer the intensity of the current, and read off the 

 maximum and minimum values A and B. At the same time 

 the corresponding positions of the Nicol were read off on the 

 divided circle. For # = 45°, the formula leads us to expect 

 the value A + B 



If the Nicol be turned from one of the observed positions 

 through 45° we shall always obtain a value of J nearly equal 

 to the calculated value (A + B)/2. By a slow rotation of the 

 Nicol we brought it about that this number was actually read 

 off, and regarded the position so obtained as that actually 

 corresponding accurately to the azimuth 45°. Then, by 

 turning on, or back, through 45°, we obtain the corrected 

 principal positions, and obtain for these the old maximum 

 and minimum values A and B again. We must further 

 mention that on rotation of the prism the direction of the 

 emergent beam was not absolutely constant, but there were 

 slight displacements of the positions of entrance and emergence 

 of the beam to be observed on the glass wall (see fig. 1). As 

 there were but very slight changes of direction, and the dis- 

 placements of the spot of light thus produced fell pretty well 

 within the limits of accuracy of the method of measurement 

 described, we have neglected them. The attempt to eliminate 



