THE MICROSCOPE BY MEANS OF THEIR OPTICAL CHARACTERS. 603 



If the illumination in the two cases be the same, the supposed 

 position of extinction is correct. If not, the nicol is restored 

 to its original position, and the stage is rotated slightly towards 

 the direction in which the darkness was the greater. The 

 same test is then again applied, and, if necessary, the process 

 is repeated, till the rotation of one nicol through equal angles 

 in both directions produces the same result. 



The angle through which the nicol must be rotated is that 

 which will produce a faint illumination for rotation in one direc- 

 tion. It is usually between half a degree and two degrees. As 

 a rule it would be sufficient if a graduation were provided showing 

 a rotation of a nicol through J, 1 and 1J degrees. 



Where the position of extinction is the same for all colours, 

 this method may be applied either with monochromatic or 

 white light, the latter being preferable, not only because the 

 illumination is greater, but also because, when the true position 

 of extinction has not been obtained, the two directions of rota- 

 tion of a nicol give different interference colours. 



Wright has devised a bi-nicol ocular in which the results of 

 the rotation of two upper nicols in opposite directions may be 

 observed simultaneously.* A similar effect is obtained by the 

 insertion of plates of right- and left-handed quartz, which rotate 

 the nicol through equal angles in opposite directions. This is the 

 principle of the Bertrand eye-piece, but in its usual form the 

 plate is so thick, 2'5 mm., that it rotates the light through a 

 large angle, about 60 for sodium light and greater or less amounts 

 for light with shorter or longer wave length. If it be reduced 

 to a thickness of forty microns corresponding to a rotation of 

 1 for sodium light, much greater accuracy is obtained, both 

 with monochromatic and white light, f 



Wright's bi-quartz ivedge plate, a combination of wedges and 

 plates of quartz, enables a rotation of any convenient amount 

 in opposite directions to be obtained.! 



In all these determinations greater accuracy can be secured 

 by increasing the illumination, but care must be taken that 

 the lower nicol is not injured by over-heating (see p. 600). 



It is unnecessary to dwell here on the other methods which 



* Loc. cit., pp. 374-376, 379. 



f S. Nakamura,Centr. f. Min., 1905, pp. 267-279. 



% Am. Journ. Set., Series IV., vol. xxvi., 1908, pp. 377-380. 



