INORGANIC REPLICATION IN ELECTRON MICROSCOPY 811 



ward, thus tilting the repHca to various angles. In the extreme upper left, 

 this tilt produces negative shadows, regions where the electrons pass through 

 three thicknesses of replica. In none of the more usual positive shadows 

 observed in other parts of the rephca was any film observed, or in a replica 

 twice as thick also studied. The conclusion is that aluminum does not 

 diffuse. It is tempting to speculate that the short-range forces responsible 



Fig. 10— Gold-manganin replica (ie = 150A, da = 20°). Resolution about lOOA. Rep- 

 lica is much thicker than optimum. A film in the shadow, clearly evident in the original 

 micrograph, does not show except for two whitish areas where it has torn and curled. 



for recrystallization do not permit diffusion and, that when diffusion does 

 occur with materials such as chromium, it is due to some other component 

 such as an oxide. Resolution, although compHcated by the granular structure, 

 appears to be about 100 A. 



The micrograph of Fig. 10 is from a gold-manganin replica, produced by 

 simultaneous evaporation of two volumes of gold and one of manganin 

 (alloy, 84% Cu, 12% Mn, 4% Ni) at da = 20°. The total thickness was 

 about 150 A, so mass thickness, about 24 Aig/cm^, was much greater than the 



