High resolution electron microscopic study of negatively stained macromolecules and thin sections of tissue embedded in a variety of media are difficult to interpret because of the superimposed phase image granularity. Although all of the information concerning the biological structure of interest may be present in a defocused electron micrograph, the high contrast of large phase image granules produced by the substrate makes it impossible to distinguish the phase ‘points’ from discrete structures of the same dimensions. Theory predicts the findings; however, it does not allow an appreciation of the actual appearance of the image under various conditions. Therefore, though perhaps trivial, training of the cheapest computer produced by mass labor has been undertaken in order to learn to appreciate the factors which affect the appearance of the background in high resolution electron micrographs.
Deviation probabilities for arithmetic progressions and other regular discrete structures
