THE SCIENTIFIC WORK OF C. J. DAVISSON 795 



18 of the Journal of that Society. It is written with such clarity, grace and 

 humor as to make one regret that Davisson was not oftener tempted to 

 employ his talents for the benefit not of laymen precisely, but of scientists 

 who were laymen in respect to the field of his researches. I quote the first 

 two sentences: "When I discovered on looking over the announcement of 

 this meeting that Arthur Compton is to speak on 'X-rays as a Branch of 

 Optics' I realized that I had not made the most of my opportunities. I 

 should have made a similar appeal to the attention of the Society by choos- 

 ing as my subject 'Electrons as a Branch of Optics.' " 



Though in this period his duties as expositor took a good deal of his time, 

 Davisson found opportunity to prosecute his work and to begin on certain 

 applications. One obvious development may be dismissed rather curtly, as 

 being less important than it might reasonably seem. One might have ex- 

 pected Davisson to strive to verify de Broglie's law X = h/p to five or six 

 significant figures. This would have been difficult if not impossible, since 

 the diffraction-beams of electrons are much less sharp than those of X-rays; 

 this is a consequence of the fact that the diffraction is performed by only a 

 few layers of atoms, the primary beam being absorbed before it can pene- 

 trate deeply into the crystal structure. But even if it had been easy the 

 enterprise would probably have been considered futile, for de Broglie's law 

 quickly achieved the status of being regarded as self-evidently true. Such a 

 belief is sometimes dangerous, but in this case it is almost certainly sound: 

 the law is involved in the theories of so many phenomena, that, if it were 

 in error by only a small fraction of a per cent, the discrepancy would have 

 been noted by now in more ways than one. Davisson established the law 

 within one per cent, and there are few who would not regard this as amply 

 satisfactory. 



The greatest of the uses of electron-diffraction lies in the study of the 

 arrangement of atoms in crystals and in non-crystalline bodies. Here it 

 supplements the similar use of X-ray diffraction, for it serves where X-ray 

 diffraction does not, and vice versa. Once more I quote from a lecture of 

 Davisson's: "Electrons are no more suitable for examining sheets of metal 

 by transmission than metal sheets are suitable for replacing glass in windows. 

 To be suitable for examination by electrons by transmission, a specimen 

 must be no more than a few hundred angstroms in thickness. It must be 

 just the sort of specimen which cannot be examined by X-rays. Massive 

 specimens can be examined by electrons by reflection. The beam is directed 

 onto the surface at near-grazing incidence, and the half-pattern which is 

 produced reveals the crystalline state of a surface-layer of excessive thin- 

 ness. . . . Invisible films of material, different chemically from the bulk of 

 the specimen, are frequently discovered by this method." Many experi- 



