A theoretical and experimental investigation of contour phenomena observed in electron microscope images near focus has been carried out. An explanation for the apparent reversal of the Fresnel fringes at exact focus is given and a method is described for the determination of the degree of asymmetry from the fringe patterns. A procedure for empirically correcting the asymmetries usually present in magnetic electron microscope objectives is outlined and some of the results obtained with a compensated lens are shown. A number of instrumental defects which may prevent the attainment of the ultimate resolving power of an instrument, together with methods for their elimination, are listed.