Contrast by Scattering 41 



When fast electrons collide with atoms, it happens very rarely 

 that they lose their whole energy in a single collision. Most of 

 the energy is lost gradually, in steps of 10-30 volts at a time. 

 Some of the electrons are scattered elastically, i.e., without ap- 

 preciable loss of energy. Most of our knowledge of these 

 processes is derived from experiments in gases and from theo- 

 retical investigations of the collision of electrons with single 

 atoms and molecules.^^- "^^ Our knowledge of the scattering of 

 fast electrons in solids, however, is still rather rudimentary. The 

 theoretical investigations have not yet gone very far, and the 

 older experimental material, up to the year 1938, is of entirely 

 inadequate accuracy. It will be shown that in electron micro- 

 scopy a loss of about 10 ev by an electron of 60,000 ev velocity 

 has observable effects, whereas, until 1938, no energy loss less 

 than about 200 ev could be observed with certainty. The reason 

 is that until the advent of the electron microscope, experimenters 

 had no high voltage source at their disposal with a voltage 

 fluctuation less than about 0.5 per cent. High voltage accumu- 

 lator batteries were available in very few laboratories only. The 

 development of negative feedback stabilizers, which was under- 

 taken for the purposes of electron microscopy, at once enabled 

 physicists to improve the accuracy of their measurements by a 

 factor of about a hundred. G. Ruthemann,^^ in 1941, was the first 

 to measure the velocity losses of fast electrons in thin metal foils 

 with an error limit of less than a volt. He. demonstrated that 

 in solid bodies, as in gases, fast electrons lose their energy in 

 well-defined discrete steps. This result is of considerable im- 

 portance, as it justifies treating metals and other solids at least 

 approximately as highly compressed gases, and using for them 

 the very large amount of knowledge collected in the investigation 

 of electrons in gases. Another important result is that, similarly 

 as in gases, the number of elastically scattered fast electrons is 

 rather small in comparison with the number of those which have 

 suffered energy losses. 



These results are suflicient for a discussion of the origin of 

 contrast. Figure 14 shows the simplified ray-diagrams of an 

 electron microscope of the transmission type, for elastically and 



