230 PROGRESS IN MICROSCOPY 



C. R. Burch has built Schwarzschild-type objectives which are, 

 as yet, the best ever designed. He showed that, if both mirrors are 

 spherical, the objective can be corrected from both spherical aber- 

 ration and coma but that occlusion of the incident beam by the small 

 mirror cannot be less than 45 per cent. Nor can the numerical aperture 

 of an objective comprising two spherical mirrors exceed 0-50 approxi- 

 mately (although it can be increased to 060 in the lower infra-red up 

 to 1 i-i, and 80 above 2 5 //). 



The numerical aperture can be increased by retaining the spherical- 

 aberration correction but coma develops and the field becomes very 

 small. With an objective of numerical aperture 65, the field diameter 

 does not exceed lO/i at 2500 A. 



In Chapter 1, § 3, it was mentioned that the central occlusion 

 brought about by the small mirror altered the diffraction disk and 

 reduced the diameter of Airy's disk while retaining the zero-minimum 

 rings and making the bright rings definitely stronger. Such increased 

 intensity of the bright rings lessens image contrast. If the occluded area 

 does not exceed 10 per cent of the total wave area in the space image 

 (see Chapter I, § I and Fig. 1.7), the diffraction phenomenon is but 

 slightly altered, this corresponding to an occlusion ratio sinwo/sinw of 

 approximately 030. Using a non-spherical surface (the mirror M), 

 Burch reduced occlusion to 4 per cent. When both surfaces are non- 

 spherical, the numerical aperture 65 is increased to 095. The N.A. 

 can still be increased further by setting on the object a plano-convex 

 lens whose centre is in coincidence with the object. In this way the 

 objective of N.A. 065 (dry-front-lens) becomes an objective of 

 N.A. 0-98 and, with immersion, the dry-front-lens objective of 

 N.A. 0-95 becomes an objective of N.A. 1-4. 



Objectives comprising 2 spherical mirrors have been used by 

 E. R. Blout, G. R. Bird and D. S. Grey for infra-red micro-spectro- 

 scopy. The objectives were of N. A. 063. Setting on the object 

 a silver-chloride or a thallium-bromide-iodide lens increased the N.A. 

 to 1-25 and 1-50, respectively. This arrangement gives good resuhs 

 in the range 2 to 20 /f. 



S. Miyata, S. Yanagawa and S. Noma have evolved non-spherical 

 concave mirror objectives by evaporating in vacuo, on a spherical 

 mirror, a layer of zinc sulphide of variable thickness. 



Mirror objectives are also employed in the ultra-violet. W. Thorn- 

 burg has developed a reflecting objective, featuring a low occlusion- 



