Construct a monochromator from a single interference filter.

Fig. I Lomb i incidenco incidenc irchers in the behavioral and visual sciences The upper panel of Figure 1 shows the output of a tly need a source of monochromatic light whose Bausch and Lomb 641-nm interference filter at five Lgth can be varied. The purpose of this note is different angles of incidence. The peak wavelength of ribe principles and construction of a reliable the distribution moves toward shorter wavelengths by expensive monochromator. increasing the incident angle, and the band pass monochromator produces monochromatic light broadens. For the shortest peak wavelengths, the band ring collimated light with an interference filter; pass is shown to be bimodal. One mode results from igth is varied by changing the angle of incidence the interference of light polarized in the plane of infilter relative to the collimated light beam. Incidence and the other from light polarized perpence filters of this type are constructed of many dicular to the plane of incidence (Heavens, 1965). When spaced metal films (termed metal-dielectricplane polarized light is used with the same filter (lower nterference filters). Interference between light panel of Figure 1) there is no band-pass distortion, flected and transmitted by the metal films even at the largest angle of incidence (50 degrees). is to isolate a narrow band of monochromatic Imperfect collimation and polarization will proVavelength of the monochromatic light depends duce a slight broadening of the band pass, which pacing between the metal films. If the filter is may occur to only one side of the peak wavelength. from a position perpendicular to the light beam Within limits, however, a slight broadening will be Iincidence), the effective spacing as traversed by inconsequential for researchers in the behavioral t ray is increased and the wavelength changes sciences. shorter values. Wavelength calibrations were made with an Edgerton, Germeshausen, and Grier (EG&G) 580-585 spectroradiometer. The interference filter was positioned perpendicular to the collimated light beam. At normal WITHOUT POLARIZER incidence, peak wavelength of the interference filter is longer than at any other incident angle; this was verified. Slit widths of the EG&G spectroradiometer were set for a 5-nm band pass and the light energy passed by the interference filter was calibrated every 5 nm, 25 nm on either side of the peak wavelength. Without moving the interference filter or the spectroradiometer, the polarizer (Pola Coat Inc.) was removed from the light beam and the calibrations repeated. The angle of incidence was changed, and the procedure WITrH POlAR IZER repeated. Following these calibrations, the interference