Abstract
The factors affecting the precision of measurement of the position of spectrum lines by visual and microphotometric methods are discussed. A statistical study of an extensive set of data shows that as a useful working rule the standard deviation of a single reading of the position of a line by means of a measuring microscope may be taken as 0-3 times the square root of the line width, if all lengths are measured in microns, whereas the standard deviation of a micro-photometer reading is about half this. The systematic errors of such measurements and methods for their determination are also discussed. The validity of these estimates of systematic and random error has been confirmed by a comparison of the radii of electron-diffraction circles with theory.