APPLICATION OF ELECTRON DIFFRACTION 597 



which we recognize as a completely clean condition of the crystal 

 surface, could be produced by a comparatively mild heating, to a 

 temperature roughtly estimated to be about 900° C. The conclusion 

 that such a mild heating was sufficient to clean the surface completely 

 is of interest, as it appears to be in disagreement with similar evidence 

 concerning gas on surfaces obtained from photoelectric and thermionic 

 measurements. 



Transient electron diffraction beams of another type appeared some 

 time after the crystal had become cool following a heating. These 

 beams constitute what I shall call the electron diffraction pattern of the 



30 40 50 60 70 



MINUTES AFTER HEATING THE CRYSTAL 



Fig. 5 — Change in intensity of a typical "Type-3" beam and of a typical "Type- 

 4" beam as gas settled upon a clean crystal surface. Measurements of the intensities 

 of these two beams were made alternately over a period of an hour. 



fourth type. They could not be detected until the diffraction pattern 

 of the third type had become appreciably weakened, and did not at- 

 tain their maximum intensities until the pattern of the third type had 

 almost disappeared. The curve marked "Type-4" in Fig. 5 shows the 

 life history of a typical beam of this type following a heat treatment of 

 the crystal. In this particular experiment intensity measurements 

 were not made after sixty-one minutes. The dashed continuation of 

 the curve represents its general course as known from other experi- 

 ments. 



The two curves of Fig. 5 represent a single experiment which was 

 carried out at a time when the vacuum condition of our experimental 

 apparatus was intermediate between the best and the worst conditions 

 39 



