The star Orionis was used to calibrate the H+[NII] observations; Lyrae and Hydrae were used for the
[OIII] observations. After cosmetic cleaning and applying the
whitelight correction described above, the counts in each stellar
image were summed, and then scaled by the telescope aperture size and
exposure time. To obtain pixel values in units of counts cm^{-2}
sec^{-1} Å^{-1} we must then divide by the effective spectral
bandpass of the etalon at each pixel. We emphasize that this value is
not necessarily the same as either the spectral sampling of the cube
or the etalon resolution. Rather, it is a measure of the amount of
flux transmitted at a given wavelength for a sequence of etalon
spacings; it depends primarily on the finesse (*N*_{R}) of the etalon.
This ``effective bandpass'' was calculated by summing the flux under a
synthetic monochromatic spectrum spanning an entire order of the
appropriate theoretical Airy function and dividing by the peak value
of the Airy function. We derived an effective bandpass of 3.0Å for
our data sets. The scaled values of observed stellar counts were then
compared with published flux values ([Hayes 1970]; [Hayes & Latham 1975]),
corrected for atmospheric extinction. Each pixel in the final spectra
has units of ergs cm^{-2} sec^{-1} pixel^{-1} frame^{-1},
yielding total fluxes from the line fits in units of ergs cm^{-2}
sec^{-1} pixel^{-1}. The final estimated systematic error in the
flux calibration was 7.5%.