The availability of increased resolution, dispersion, and luminosity from plane gratings at high angles of incidence and diffraction is discussed from the standpoints of theory and practice. Reduction in the resolution given by actual gratings at angles above a certain maximum for a given wavelength usually arises from close-lying line-satellites originating from ruling defects. Variation of satellite displacements and intensities with wavelength gives rise to such undesirable effects as error of coincidence. The bright 10-in. gratings now produced by the M.I.T. interferometrically controlled engine can be used effectively at very high angles (12th-order green from 7500 grooves per inch), but as in all gratings the angle above which resolution fails to increase further diminishes with decreasing wavelength. The pattern dimensions and intensities of satellites are here discussed qualitatively as they affect resolution in various spectral regions, and are quantitatively discussed by one of us elsewhere.
The use of gratings and echelles in series for increasing spectroscopic efficiency is discussed, and spectrograms made with two echelles thus used are shown. Two gratings used in series transmit only a narrow wavelength range at one setting because of the wide angular spread of the beam from the first disperser. An echelle beam, on the other hand, spreads but little, and can be caught on a second echelle to give broad spectral coverage. Two echelles used in series give high speed and resolution without the careful relative adjustment required to produce a satisfactory grating mosaic.
GEORGE R. HARRISON and GEORGE W. STROKE, "Attainment of High Resolution with Diffraction Gratings and Echelles," J. Opt. Soc. Am. 50, 1153-1157 (1960)