Estimation of section thickness,etc. by quantitative electron microscopy

Methods have been developed for the estimation of section thickness, and of the mass-thicknesses, and hence the masses, of material embedded in sections. They are based on the current densities measured with a Faraday-cage or on a new approximation for the relationship between exposure and optical density (OD) of the electron microscopical plate, which is valid for OD < 5. This relationship is quite distinct from the response to light, and also from an approximation for electrons which was only valid for OD < 1.2, and which was the basis for an earlier method. The old methods and the two groups of new ones were tested, and also compared with the interference microscope. It was found that all but one of the new methods, in both groups, were homogeneous; the older methods and the interference microscope differed very significantly between themselves and with the homogeneous methods. The errors in the interference microscopical estimations were almost certainly due to the presence of embedded material in the sections, which could not be detected with the interference microscope because of its low resolution and the poor contrast in the thin sections. (A second set of experiments using only pure resin gave a good agreement between the homogeneous methods and the interference microscope.) The orders of accuracy of the methods are quite high. A measurement repeated twelve to thirty-five times will yield a value which has a 95% chance of being within 5% of the true value, while a single observation has a 95% chance of being within ～ 25% of the true value with most methods, while some place it within ～ 10%. Simplified, but less accurate, applications of the methods are also suggested.