XII. ELECTRON MICROSCOPY 393 



2. General Types of Preparations Suitable for Eleetron 



Microscopy 



Specimens suitable for examination with the electron microscope 

 must satisfy three conditions: 



(fl) The specimen must be sufficiently thin. If a high resolving 

 power electron microscope image is to be obtained, it is necessary 

 that the electrons penetrate the specimen without appreciable loss 

 in energy. This means that the thickness of the specimen, plus its 

 support, must be less than some definite value. It is difficult to give 

 an exact value for this permissible thickness since it is determined by 

 the nature of the problem as well as the properties of the specimen 

 and the velocity of the electron beam. For instance, the problem 

 may be the determination of the shapes and sizes of metallic particles. 

 Since silhouettes only are of interest, the particles may have dimen- 

 sions from several microns down. If, on the other hand, the problem 

 is the internal structure of such particles, they must be less than a 

 hundred millimicrons in thickness if they are to be observed by trans- 

 mitted electrons. In biological material, where the density is seldom 

 very great, layers, including the supporting membrane, of less than 

 a quarter of a micron are usually quite satisfactory. 



(h) The specimen must provide sufficient contrast. Since the 

 electron microscope image exists only by virtue of the differences in 

 the scattering of electrons by different points of the specimen, it is 

 necessary that the structures to be observed produce such differences 

 in the scattering. For instance, isolated bacteria mounted on a thin 

 collodion membrane will give an image of ideal contrast as far as 

 the contours of the organisms are concerned, while bacteria mounted 

 on a collodion membrane but covered by a uniform organic film equal 

 in thickness, say, to the height of the organism will provide an image 

 in which the bacteria are hardly distinguishable, if at all. Similarly, 

 in the first case, structures of nearly equal densities within the pro- 

 toplasm of the individual cells will be indistinguishable, unless it is 

 possible to enhance the contrast l)y some technique of specimen prep- 

 aration or by some instrumental development (Fig. 5). 



(c) The specimens must be prepared in such a way that they are 

 unaffected either by the vacuum or by the electron bombardment. 

 Since the specimen is placed in a vacuum, it is obvious that it must 

 contain no volatile liquid. This eliminates the possibility of exam- 

 ining undried specimens of anj' type. In passing, it is to be noted that 



