ELECTRON TRANSMISSION THROUGH THIN METAL SECTIONS 881 



evidence indicating that the internal strain is less than in Fig. 8. Another 

 section was prepared after the rolled sheet had stood for about one year 

 with the result shown in Fig. 9b. The domain boundaries in this section 

 are made visible chiefly through the discontinuities in the extinction con- 

 tours rather than overall contrast between domains. Apparently the do- 

 mains slowly reUeve their internal strains and become less distinct as re- 

 covery proceeds at room temperature. 



A possible complication in the study of thin sections in the electron 

 microscope is the effect of rather intense electron bombardment. There are 

 several possible ways in which the sections might be changed by bombard- 

 ment. One of these is simply annealing due to heating by bombardment. 

 However, the metal is a good conductor of heat and is in contact with the 

 heavy brass specimen holder so that high local temperatures would not be 

 expected as with thermal insulators or isolated particles. Another phenome- 

 non that is quite common is the deposition of a carbonaceous layer on the 

 areas exposed to the electron beam. This is generally due to the residual 

 hydrocarbon atmosphere in the vacuum system and is visible to the naked 

 eye as a black deposit. The remaining possibility is that of producing lattice 

 defects or vacancies by colUsion with the incident electrons. The cross- 

 section for this process is not known but it would be expected that for 50 

 KV electrons it would be quite small. 



Many thin sections of aluminum have been examined in the RCA EMU 

 instrument at moderate intensities using the biased electron gun with little 

 evidence for any changes occurring over the normal times required for ob- 

 taining pictures. However, if the peak intensity attainable with the biased 

 is used, quite significant changes occur as illustrated in Fig. 10. These im- 

 ages are taken from a sequence and show the effect of time of bombardment 

 on the recovery domains. The loss of contrast and irreversible changes in 

 details with time of bombardment are evident. Part of the effect is due to 

 heating and part to deposition but in general the behavior is not under- 

 stood. 



An outstanding feature of the domains in cold worked, high purity alu- 

 minum has been the relatively uniform size exhibited over a great many 

 samples prepared at room temperature. The deformations have ranged from 

 the order of about 30% to several hundred percent with the domain size 

 consistently in the neighborhood of 2/x. At low deformations of the order of 

 a few percent the domains are not found. No growth or change in size of 

 any consequence has been found after months at room temperature. The 

 relief of internal strains seems to be the only significant change with time. 

 A short anneal at or above the recrystallization temperature removes the 

 domains and gives rise to new crystals which exhibit extinction contours. 

 Observations such as this tie the domain structure quite firmly to recovery. 



