224 BELL SYSTEM TECHNICAL JOURNAL 



Those of us who are studying electron diffraction are most fortunate 

 in having before us a perfect model for our experiments and a fund of 

 valuable data in the vast amount of work that has been done in the 

 last fifteen years on the diffraction of X-rays. It is for this reason 

 that, in spite of a rather difficult technique, so many and such varied 

 results have been obtained in less than two years. Already we have 

 passed on from crystal diffraction to diffraction by optical gratings. 

 The first results of this sort were reported a month or so ago by Rupp 

 and are in agreement with our expectations. Electrons are diffracted 

 by an optical grating as if they were waves of length hjmv. 



I have still to mention the beautiful but puzzling results which 

 have been obtained in Japan by Nishikawa and Kikuchi. It is too 

 bad to have to conclude my remarks with mention of the only results 

 so far obtained which are distinctly puzzling. Nishikawa and Kikuchi 

 have been studying the scattering of high-speed electrons by thin 

 sheets of mica and calcite. The method of their experiment is identical 

 with that of the original Laue experiment except that the heterogeneous 

 beam of X-rays is replaced by a homogeneous beam of electrons. 

 The results, as I have mentioned, are puzzling. If the incident 

 beam is homogeneous, as stated, it is equivalent to a beam of mono- 

 chromatic waves, and no diffraction pattern — or at most a very simple 

 one — should be observed; and yet, when extremely thin sheets of 

 mica are employed, elaborate and beautiful patterns of sharply defined 

 spots are obtained — and patterns which cannot be readily explained 

 even on the assumption that the incident beam contains a large 

 range of wave-lengths, instead of a single wave-length only. When the 

 speed of the incident beam is changed, the form of the pattern remains 

 the same but its scale factor is altered. This also is unlike anything 

 observed with X-rays. The results are such as might be expected 

 if the diffracting system were a two-dimensional mesh rather than a 

 three-dimensional lattice. 



When somewhat thicker sheets of mica are used, the pattern of 

 sharply defined spots is replaced by an array of rather fuzzy rings and 

 lines. Again the observations are contrary to our expectations, and 

 their explanation is far from obvious. 



It may be significant that these are the only experiments, so far 

 reported, in which the diffracting material is an insulator. But 

 whether the clue lies here or elsewhere, it is highly unlikely, I think, 

 that the explanation of these results will conflict with the conception 

 we now have of electrons which are sometimes particles and sometimes 

 waves. 



