A problem which frequently presents itself to the spectroscopist is that of comparing photographs with a view to detecting differences which are not sufficiently conspicuous to make them obvious by simply superposing the plates. Differences of the kind in question may arise in the spectrum of a particular substance as the result of variations in the conditions of excitation of the spectrum which may produce changes in the distribution of intensities, changes in the nature of the lines, or the emission of new lines.