398 JAMES HILLIER 



the fundamental ambiguity that always exists in the electron micro- 

 graphs between the density and thickness of a specific specimen 

 point, and for showing surface characteristics of larger objects Avhere 

 the thickness of the underlying material has reduced the contrast 

 or that underlying material has an internal structure that masks the 

 structure of the surface. See Figure 9. 



The last point indicates a major limitation of the shadow-casting 

 technique, particularly for biological work. Information obtained 

 through the method of shadow-casting is confined to the infinitely 

 thin outer surface of the specimen and usually completely submerges 

 the image of any existing internal structure. It is only in the very 

 specially prepared biological specimens that the surface carries sig- 

 nificant information. For this reason, the writer considers the present 

 practice in some laboratories of shadow-casting all specimens a dan- 

 gerous procedure leading both to erroneous interpretations and to 

 an apathy toward the basic problems of structure determination. 



4. Examination of Compact Solids 



Coincident with development of the electron microscope as a com- 

 mercial instrument, its inability to examine surfaces was presented 

 as one of its major limitations. This arose from the fact that suffi- 

 ciently thin sections could not be cut of most materials and that 

 electronic images by reflection from a surface could not be made 

 successfully owing to the wide range of velocities present in the 

 reflected electron beam. However, this apparent difficulty was over- 

 come almost immediately by the development of what is known as the 

 replica technique by H. Mahl in Germany and V. K. Zworykin and 

 E. G. Ramberg in the United States. Basically, this technique in- 

 volves the production of a thin plastic replica of the surface to be 

 examined. The replica is made thin enough to be suitable for exam- 

 ination in the electron microscope. The geometrical details of the 

 surface structure are made visible by the fact that at different points 

 the replica has difi'erent thicknesses and, hence, a different scattering 

 power for the electrons. At first it seemed unlikely that such a 

 replica could reproduce the surface structure on a sufficiently fine 

 scale. At present, however, it has been well established that the 

 chief limitation in the technique is the molecular size of the material 

 used in making the replicas. Since this is usually considerably less 

 than 10 m/x, the method is quite satisfactor}'- for resolving powers of 



