592 BELL SYSTEM TECHNICAL JOURNAL 



found to be predominant under all conditions, was the normal diffrac- 

 tion from the space lattice of the nickel crystal. The dififraction pat- 

 tern of this type is analogous to a Laue X-ray pattern produced by the 

 interaction of a beam of heterogeneous X-rays with a single crystal. 

 The second type of diffraction pattern consisted of a curious but very 

 simple assemblage of twelve electron diffraction beams which, at the 

 time of their discovery, we designated "anomalous beams." 



At the time of publication of the note in "Nature" the nickel 

 crystal, with which we were e.^erimenting, had not been heated after 

 the bulb containing it was sealed from the pumps. Subsequently the 

 crystal was heated many times by electron bombardment. The effect 

 of these heatings was to destroy the diffraction pattern of the second 

 type, made up of the so-called anomalous beams, and to intensify 

 greatly the first type of diffraction pattern. Simultaneously patterns 

 of two entirely new types appeared for the first time. These new 

 diffraction patterns were comparatively short-lived. Both of them 

 disappeared completely within a few hours after heating the crystal. 

 After its initial increase in intensity, following such a heating, the 

 diffraction pattern of the first type also became weaker with the 

 passage of time, but at a rate very much less rapid than the rate at 

 which these new patterns changed. After the lapse of some days the 

 diffraction pattern of the second type was again found. At first the 

 beams of this pattern could barely be detected, but they became 

 progressively stronger as the diffraction pattern of the first type became 

 weaker. 



Experiments more or less similar to the experiment just described 

 were performed many times. After the initial heating of the crystal 

 the rapidity of occurrence of the changes in the diffraction patterns 

 varied greatly, from one experiment to another. These changes oc- 

 curred very slowly when the experimental conditions were such as to 

 warrant the belief that the vacuum in the experimental tube was unusu- 

 ally high. The changes took place much more rapidly when the 

 vacuum was known to be comparatively poor. We concluded that the 

 changes in the diffraction patterns which were observed following a 

 heat treatment of the crystal were caused by gas being gradually 

 adsorbed upon its surface. The two new types of diffraction pattern 

 must arise from a clean or nearly clean surface, and their complete 

 disappearance seemed conclusive evidence that the surface had become 

 covered by at least one layer of gas atoms."* The diffraction pattern of 

 the first type was weakened only slightly at the time of their disap- 



* We have not obtained any information regarding the nature of the adsorbed gas. 

 In referring to this gas the word "atom" has been used to mean either atom or 

 molecule. 



I 



