108 



PROGRESS IN MICROSCOPY 



is no longer J. At (1) and (2) the path difference has a value depending 

 on the lateral shift and the slope of the wave surfaces. Intensity at (1) 

 and (2) ditfers from the remainder of the field surrounding the image. 

 The object is seen, therefore, according to its ""slopes"; it is the optical- 

 path gradients, e.g. the index gradients, which are evidenced. The 



A 



_L 



Fig. 3.14. The duplicated waves in the differential method. 



method is differential. Figures 3.15(a) and (b), showing outhned wave 

 surfaces, will help to clarify these phenomena. These two figures 

 reproduce the left section of Fig. 3.14 merely eliminating the rounded-off 

 contours. In Fig. 3.15(a), the path difterenceJ between the two waves O 

 and E is zero. Owing to the lateral shift d between O and E and the 

 slope a of the wave surface, it is readily seen that the path difference 

 between the two waves O and E in the area (1) equates a. (^ (assuming a 



(I) 



j=o 



(a) 



(b) 



(I) 



Fig. 3.15. Path difference between the twx^) wa\es O and E. 



to be not steep). If, as in Fig. 3.15(b), the path difference .1 between 

 the two waves O and E is not zero, the path difference in the area (1) 

 becomes \~nd. In any case, when changing over from an imageless 

 area to one where the object is featured by the surface-wave slope a, 

 the path difference steadily varies by a.d. 



If I =^ 0565 //, the field is purple outside the image (crossed 

 polarizers). The hue varies in the '/-slope area (I): when W = 001 5 u. 



