ELECTRON TRANSMISSION THROUGH THIN METAL SECTIONS 877 



as seen in Fig. 7a. Figure 7b illustrates a case in which the contour family 

 can be identified on either side of a grain boundary with the displacement 

 very much in evidence at the boundary. This situation was anticipated in 

 Fig. 4. 



Proof that the dark contour lines seen in Fig. 6 are due to diffraction from 

 those regions was given in reference (1) where both the transmitted and 

 diffracted beams were imaged in an electron shadow microscope. The usual 

 transmission electron diffraction patterns from annealed sections generally 

 exhibit an array of spots characteristic of a single crystal. Sometimes 

 weak, broad Kikuchi lines are obtained but generally the bending of the 

 section and the area of the incident electron beam are such as not to favor 

 Kikuchi lines. 



The effect of cold working on the images of the sections was studied by 

 lightly pounding the center region of a f diameter disc (0.005'' thick) with 

 a small, rounded and poHshed steel rod against an anvil. The disc was then 

 electro-thinned and examined in the electron microscope. Originally it was 

 hoped that further information regarding laminar slip^ might be obtained 

 in this manner. However, no details of slip have been observed in the cold 

 worked sections, the general appearance being that seen in Fig. 8(a). Figure 

 8 was obtained from a section cold worked by pounding at room tempera- 

 ture and shows the recovery domains or early stage of polygonization.^ 

 These domains are not made visible by etching a polished surface and are 

 observed only by electron transmission. The domains are slightly dis- 

 oriented one with respect to the other and are made visible by the differences 

 in diffracted intensity. Since the extinction contours are absent in Fig. 8a 

 it is concluded that there is considerable internal strain in the domains as 

 previously mentioned. Figure 8b is a transmission electron diffraction pat- 

 tern of this section and shows arced rings made up of discrete spots. It 

 is concluded that each spot on an arc corresponds to a domain. Insufficient 

 domains are included in the primary beam to produce continuous rings. 

 The electron diffraction pattern of Fig. 8b is very similar to microbeam 

 x-ray patterns published by Kellar'^ et al and the domain size of about 2/x 

 from the electron micrographs is in excellent agreement with the results of 

 Kellar for pure aluminum. 



That domains sufficiently free of strain to yield extinction contours can be 

 obtained is illustrated in Fig. 9. Figure 9a is from a section prepared 36 

 hours after a block of the high purity, aluminum had been rolled to 0.005" 

 sheet with some annealing between passes. The contours are very much in 



* Recovery domains are not found in aluminum deformed by simple extension. Appar- 

 ently inhomogeneous strain is necessary. Extinction contours are observed in specimens 

 deformed in tension but the slip bands are not in evidence. 



8 J. N. Kellar, P. H. Hirsch and J. S. Thorp, Nature 165, 554 (1950). 



