Conclusions important to the practice of electron microscopy are drawn from experiments. Particles of Dow Latex 580G are unstable under electron beams and best micrographed at low intensity. Two different effects of radiation are noted. At low intensity the mean diameter in the present work was 2890A, the standard deviation of the spherical particles was 60A. 580G is improved as an internal standard if recalibrated by some other suitable standard (not the reverse) each time before use. Fresh replicas of diffraction gratings are still the most practical standards, reproducing gratings to two percent. By an independent optical method (spectroscopy) one can measure the grating space of a replica when it is on an eighth‐inch, 200‐mesh specimen screen under exact calibration conditions for electron microscopes. Specific replicas, and from them specific preparations of 580G are checked easily and accurately by the spectroscopic method. Replica age or other factors which may affect accurate reproduction of an original grating are then no longer important in calibration. A reliable method of calibration using 580G is suggested.