44 The Electron Microscope 



There is also another objection. If the physical aperture 

 plays such an essential part in producing contrast, how was it 

 possible to obtain micrographs, prior to 1939, before the physical 

 objective aperture was introduced? It could be said, and this 

 seems to have been the opinion of the authors on this subject, 

 that there always was an aperture, though in the older micro- 

 scopes it was of the order 0.1-0.5 radians rather than 0.002. 

 Good contrast could be produced in such instruments therefore 

 only if a great part of the electrons is deflected by more than 

 0.1-0.5 radians. There is strong evidence, to be discussed in a 

 later chapter, that in a single collision only a very small fraction 

 of the electrons is deflected by more than 0.1 radians, therefore, 

 contrast could be produced only by thick objects. Marton and 

 von Ardenne wxre aware of this and built up their theories on 

 the basis of multiple collisions. Multiple collisions mean very 

 large energy losses, in fact, it is easy to see that the mean energy 

 loss will increase with the square of the average deflection angle 

 produced by a great number of successive collisions. This pre- 

 sents the dilemma in an even sharper form : If an electron suffers 

 say ten collisions in the object, it will lose energy of the order 

 of 100-300 volts. How was it, then, that the resolution of the 

 electron microscopes was very noticeably improved when the 

 fluctuation of the driving voltage was reduced from 10 volts to 

 about 1 volt, although no steps were taken, nor could be taken, 

 to reduce the chromatic error arising from energy losses of the 

 order of several hundred electron volts ? 



The explanation is very simple, but rather surprising: There 

 are two sorts of chromatic error, one arising from fluctuations in 

 the driving voltage, the other from energy losses in the object. 

 Although the first is harmful, the second is beneficial. This docs 

 not affect the resolution, but increases the contrast to such an 

 extent that it can be said without exaggeration that microscopes 

 without small physical apertures produce good pictures mainly 

 hy the chromatic defect of their lenses. It may be added that the 

 spherical aberration has also a partly beneficial effect. 



The explanation is contained in the right half of figure 14 

 which illustrates inelastic scattering, i.e., assumes that the elec- 



