78 INSTRUMENTATION 



of a phase microscope, it may be impossible to avoid the occurrence of 

 parallax between the image of the opening in the condenser diaphragm 

 and the diffraction plate. Epstein (1950) has treated theoretically the 

 problem of the edge effect in phase contrast when the image of a point 

 source of light is projected on the edge of the conjugate area of the 

 diffraction plate but is defocused with respect to it. Excessive parallax 

 produces deterioration of contrast in the image of the object specimen. 

 It is evident that there is more overlap of the deviated and the un- 

 deviated light at the diffraction plate as the parallax increases. Curya- 

 ture of field of the group of lenses forming the image of the condenser 

 diaphragm may cause the parallax to become sufficiently greater as the 

 mean diameter of the opening in the condenser diaphragm increases 

 that it becomes necessary to limit the mean diameter of the opening in 

 the condenser diaphragm. 



The spherical aberration of the objective often restricts the outer 

 radius of the conjugate area. Spherical aberration produces non- 

 uniform phase changes in the wave front as it passes through the 

 objective. Such relative changes in phase are equivalent to changes in 

 optical path and will be different for different wavelengths of light. 

 Curves of chromatic lateral spherical aberration for most of the higher 

 power achromatic and apochromatic refracting objectives show a 

 rapid but smooth change near the outer edge of the aperture. Any 

 change in phase due to spherical aberration is superimposed on the 

 change in phase deliberately introduced by means of a diffraction plate. 

 Elementary theory shows that, to obtain optimum contrast with 

 objectives which are free from spherical aberration, the amplitude 

 transmission and the optical path across the diffraction plate should be 

 described by a step function; i.e., an abrupt discontinuity in optical 

 path or in amplitude transmission or in both should take place at the 

 boundary between the conjugate and complementary areas, but within 

 each area the optical path and the amplitude transmission should be 

 constant, as in Zernike's arrangement. When the conjugate area of a 

 step-type diffraction plate lies too far out in the aperture of an objective 

 with spherical aberration, the spherical aberration so modifies the region 

 of the conjugate area that changes in optical path can no longer take the 

 shape of a step. Instead of an optical path step at the conjugate area, 

 a region of rapidly changing optical path exists which is different in 

 magnitude, shape, and extent from what is required for good phase 

 contrast. A second rule of thumb is to select a value of the mean 

 diameter of the conjugate area in the range Yi to % the diameter of the 

 aperture of the objective in the plane of the diffraction plate. 



It is a familiar fact that the spherical aberration of an objective is 



