Transparent Inactive Crystal Plates. 193 



probable that the recorded differences between calculated and 

 ohserved values are in part due to the effects of such surface 

 films. 



Previous measxirements. — F. E. Neumann 1 measured the 

 nniradial azimuths of light waves, entering a calcite cleavage 

 plate at different angles, and found that his results of observa- 

 tion agreed closely with those of calculation. His observations 

 were made in white light and through a prism which separated 

 the two images E and O so that each could be studied alone. 

 Although Neumann's measurements were repeated 40 times 

 and should therefore be very accurate, the use of white light, 

 and also of intervening glass surfaces, introduced sources of 

 probable error which might, under certain conditions, prove 

 serious. 



Later, in 1852, R. T. Glazebrook 2 undertook a more elaborate 

 series of determinations of the uniradial azimuths of transmitted 

 waves. His apparatus and method for determining the 

 positions of total extinction were more accurate and refined 

 than those of Neumann, but the agreement between his 

 observed values and those of calculation was only approximate, 

 and less satisfactory than Neumann's values. Glazebrook's 

 measurements, however, were made for the most part on 

 polished plates on which surface films may have been especially 

 active and may have modified the results accordingly. As 

 Drude has shown, an exceedingly thin surface is sufficient to 

 produce marked elliptic polarization in the reflected waves. 



Extinction positions of an unmounted plate. — (Line No. 7, 

 Table II.) A series of readings in sodium light was taken to 

 determine, if possible, the positions of maximum extinction of 

 the central part (E and O overlapping) of a calcite cleavage 

 plate. The angles listed are the average of two sets of 10 

 readings each; the first for +a>, the second for — a>. It was 

 found that the individual settings often varied a degree or more ; 

 their average, moreover, was not midway between S and S e 4- 

 90° or vice versa. In short, accurate determinations under 

 such conditions are not possible, and the different positions of 

 extinction were rarely found to be precisely 90° apart. 



Extinction positions of mounted plate. — On lines No. 8 and 

 No. 9, Table II, the results of the determinations of the ex- 

 tinction position of a mounted cleavage plate of calcite are 

 listed, in which the measurements were made in Nernst light, 

 (1) with the aid of the bi-quartz wedge (line No. 8) and (2) 

 without it (line No. 9), each angle being the average of 10 

 settings. In the strong Nernst light, the individual settings 

 agreed more closely, but their averages agree neither with the 

 uniradial azimuths, observed or calculated, nor with the other 



1 Pogg. Ann., xlii, 11-12, 1837. 2 Proc. Roy. Soc, clxxiii, 617, 1882. 



