794 THE BELL SYSTEM TECHNICAL JOURNAL, OCTOBER 1951 



crystal lattice. It is easy to remember that 150- volt electrons have a wave- 

 length of one Angstrom unit, while the spacings between atoms in a solid 

 are of the order of several Angstroms. This fact of course did not escape 

 Elsasser, and it figures in his letter. 



From the quotation it is clear that the earliest patterns obtained from the 

 complex of large crystals were obscure, and the definitive proof of Elsasser 's 

 theory was obtained only when Davisson instituted his "program of thor- 

 ough search" and simultaneously in England G. P. Thomson instituted 

 his own. Two other items in the quotation require to be explained. The 

 hypothesis of "transparent directions" I will consider to be explained by 

 its name. Were it correct, the directions of the beams would be independent 

 of the speed of the electrons; since they are not, the hypothesis falls. The 

 reference to the "line-gratings of the surface atoms" induces me to proceed 

 at once to one of the principal contrasts between diffraction of electrons 

 and diffraction of X-rays. 



An optical grating is a sequence of parallel equidistant grooves or rulings 

 on a surface of metal or glass. The atoms on a crystalline surface are arranged 

 in parallel equidistant lines, and one might expect X-rays or electrons to be 

 diffracted from them as visible light is diffracted from an optical grating. 

 This expectation is frustrated in the case of X-rays, because their power of 

 penetration is so great that a single layer of atoms, be it the surface-layer 

 or any other, diffracts but an inappreciable part of the incident X-ray 

 beam; only the cumulative effect of many layers is detectable. Electrons 

 as slow as those that Davisson used are not nearly so penetrating. With 

 these indeed it is possible, as he was the first to show, to get diffraction- 

 beams produced by the surface-layer only. Such beams, however, are detect- 

 able only when the incident (or the emerging) beam of electrons almost 

 grazes the surface; and nearly always, when a beam is observed, it is due 

 to the cumulative effect of many atom-layers as is the rule with X-rays. 

 But the cumulative effect requires more specific conditions than does diffrac- 

 tion by the surface-layer: if the incident beam falls at a given angle upon 

 the surface, the momentum of the electrons and the wavelength of their 

 waves must be adjusted until it is just right, and, reversely, if the momen- 

 tum of the electrons has a given value the angle of incidence must be ad- 

 justed until it is just right. This also Davisson verified. 



As soon as Davisson made known his demonstration of electron-waves, 

 he was bombarded by entreaties for speeches on his work and for descrip- 

 tions to be published in periodicals less advanced and specialized than 

 Physical Review. To a number of these he yielded, and I recommend espe- 

 cially the talk which in the autumn of 1929 he gave before the Michelson 

 Meeting of the Optical Society of America; one finds it in print in volume 



