Results of an H-alpha + [N II] narrowband imaging survey of x-ray-selected E and SO galaxies are presented. A new technique is described for objectively optimizing removal of stellar continuum light while providing well-defined estimates of systematic errors. The procedure has the additional benefit of eliminating sky contamination specifically in image regions occupied by galaxy light. Detailed consideration of the measured spectral energy distributions is included in the flux calibration procedure, and emission-line luminosities (or upper limits), corrected for Galactic foreground extinction, are tabulated for metric apertures. The resulting luminosities for the sample members show no tendency to correlate with x-ray luminosities. Observed H-alpha + [N II] luminosity (L(H-alpha + [N II])) does correlate with far-infrared luminosity (L(FIR)), with typical ratios of L(H-alpha + [N II])/L(FIR) that are an order of magnitude below those for late type galaxies. Some tendency is found for L(H-alpha + [N II]) to correlate with nonthermal radio power, thought this may be due to eccentricities of the sample selection. No connection is apparent between the "boxiness" or "diskiness" of stellar isophotes and emission-line or far-infrared luminosity. Optical nebulosity in early type galaxies probably contains a significant multiparameter dependence on active galactic nuclei behavior, accretion from the hot interstellar medium, and mass injection from external sources. Emission associated with cooling flow phenomena on a galaxy scale may be overwhelmed in many instances by other sources of mass and ionization, suggesting that analogies between nebulosity in these sources and emission in cluster cooling-flow galaxies should be treated with caution.