X-RAY MICROSCOPY 



ElECTBOSIAliC 00^- 

 MAGNEIIC LENS ~ 



TRANSVERSE 



MAGNETIC Field 



PRIMARY IMAGE 



PRIMARY ELECT 



ELASTICALLV REF 

 ELECTRONS 



SECONDARY IMA 



ELECTRON GUN 



MAGNETIC LENS 



Fig. 4. Principle of the focusing method, using 

 reflected electrons. 



of the x-ray fluorescent screen under normal 

 focusing condition. So, when using this 

 focusing method, the Hmit will be set by 

 insufficient stability during long exposure. 

 An important advantage of this focusing 

 method is that it works with the real speci- 

 men in position and that the focus can be 

 checked during exposure. 



The resolution, obtained by using this 

 focusing method is demonstrated in figs. 5 

 and 6. 



Stability. While the problem of focusing 

 is solved, the intensity problem still remains. 

 Thus long exposure times are required and 

 during this time, the microscope must be 

 stable. The stability concerns: 



(a) Electrical stability. Changes in both 

 anode voltage and lens current cause a 

 broadening of the electron spot. If the elec- 

 tron optical system is not well aligned, there 

 will be an associated image shift. Although 

 electrical stability can be achieved to a very 

 high degree for a short time, it becomes more 

 diflficult to obtain the same stability over 

 long periods. The use of electrostatic lenses 

 will not solve this problem because they give 

 more spherical aberration, resulting in a 



Fig. 5. Gold shadowed bull sperms, ca 2000X. 

 (Experimental Delft microscope.) 



site direction (see Fig. 4) and give an en- 

 larged electron image of the spot. For sym- 

 metry reasons it can easily be recognized 

 that this image lies in the electron source 

 and has the same size and shape when the 

 target is in focus. By using a transverse 

 magnetic field or by slightly tilting the lens, 

 if it is of the magnetic type, this secondary 

 image can be separated from the primary 

 electron path and caught on a fluorescent 

 screen. 



Calculations (14) show that the brightness 

 of this image is some 10* higher than that perimental Delft microscope.) 



Fig. 6. 1500 mesh per inch silver grid, showing 

 the resolution of the projection microscope. (Ex- 



664 



