100 The Electron Microscope 



C is the constant of spherical aberration, which depends only on 

 the shape of the field, not on its linear dimensions or intensity. 

 The focal length, /, and the accelerating potential, V, are prefer- 

 ably used in the combination ( t^ j as in a magnetic lens field of 



given strength / is proportional to V, i.e., ( V7 ) is a characteristic 



constant of the lens. In electrostatic lenses ( vW 1 should be 



1 



W 



chosen as a characteristic quantity, leaving V^ in the final factor 



-1 



instead of V ^. But this varies so slowly with V that we can 

 leave it out of account in the following discussion, noting that 

 we are free to chose V in such a way as to suit the thickness and 

 scattering or absorbing properties of the object best. For very 

 thin objects it may be best to make V only 10,000 or even 5,000 

 volts. Electrons with energies of this order have photographic 

 efficiencies considerably below the optimum, which for most 

 emulsions is somewhere between 50 and 100 kev.^^ But, at 

 least in principle, it is easy to accelerate the electrons up to the 

 optimum velocity before they reach the photographic plate. In 

 practice, this introduces some complications as compared with 

 present instruments in which all parts behind the object can be 

 earthed. But the complication may be w^ell worth while, as 

 considerable improvement could be expected in the contrast and 

 detection limit of small particles of low atomic weight, especially 

 organic ones. 



There are, therefore, only two effective possibilities for reduc- 

 ing the resolution limit : reducing the aberration constant C, or 



V . . V 



increasing the lens strength -r, in electrostatic lenses -p. Let 



us first consider the second possibility. The simplest method 

 to increase the lens strength appears to be to scale down the 

 lens dimensions, and scale down / in the same ratio by keeping 

 constant the magnetic potentials between the poles, or the elec- 

 tric potential differences between the electrodes of the lens. In 

 the case of magnetic lenses, this is of no avail, as the pole pieces 



