4 8 



L. B. Tuckcrman 



(assumed monochromatic) light and other incidental factors. If 

 a perfect match were obtained Al would equal zero. 



If the light incident on the system is ellipticallv polarized. A/ 

 and I m will in general be functions of its initial ellipticity e o and its 

 initial azimuth o . 



Starting with a perfect match, 



A/ 



and varying e o within the limits of an observed match, equation 

 (87) becomes: 



o I in J 



) 



(88) 



Again, varying the azimuth 6 o within the limits of an observed 

 match : 



£ it] «.*#/('..«./>.* . . 



o I m J 



) 



(89) 



The maximum possible variations or errors of e o and o compatible 

 with an observed match are then : 



and 



8^„ 



SO 



:I/£-^ 



) 



del I I 



f(/,a,/?, y , ..'••) 



: I/,^± 



(90) 



(90 



Although the form of f (f m , a, /?, y, ..'..) is not definitely 

 known, certain general conclusions about the variation of sensi- 

 bility with varying experimental conditions may be deduced for 

 special halfshade systems. 



Type A. 



Lippich Half Nicol. 



Brace Sensitive Strip, etc. 

 Here /— 4 and I'=(i — *)-I\- Then from (11) 



204 



