Figure 1: The blue (a) and red (b) optical spectra of the nuclear region of
MCG-6-30-15. Prominent emission lines and the CaII absorption
feature have been marked.
Optical spectra of MCG-6-30-15 were obtained with the AAT on the
night of 1995 April 10. The RGO spectrograph with the 300B grating
(grating angle 
) was used to obtain blue (3500-5500Å)
spectra and the Faint Object Red Spectrograph (FORS) provided red
(5700-10000Å) spectra. Red and blue spectra could be taken
simultaneously
via the use of a dichroic beam splitter.
The spectral (FWHM) resolutions are 
Å in the blue and 
Å in the red. Standard data reduction was performed using the FIGARO software package provided by STARLINK. Wavelength
calibration was performed using a Cu-Ar arc. We estimate that the
wavelength calibration is accurate to within 1Å in the blue spectrum and
3Å in the red spectrum. Absolute flux calibration utilized the standard
star L745-46A. Finally, the atmospheric water bands within the red
spectrum were corrected for using the extremely metal-deficient red giant
star HD126587 as a continuum reference source.
In order to be sensitive to intra-night variations in flux or spectrum, we took 18 separate red/blue spectra of MCG-6-30-15. A normal Galactic F-star is situated 6 arcsec to the south of the nucleus of MCG-6-30-15 (Pineda et al. 1980): the slit position was arranged so as to cover both the nucleus and the star. Assuming the star to have a constant flux during the night, this provides a good control against which we can search for flux variations in the nucleus. The exposure time of each spectrum was 600s in the blue and 540s in the red.
After reduction it was found that only 8 red/blue spectra were unaffected
by the variable weather conditions during the night. The spectrum of the
nuclear region of MCG-6-30-15 was extracted from each long-slit spectrum.
No significant difference in either overall flux or individual line fluxes
could be found between these spectra. Furthermore, no significant
variability could be detected when the overall nuclear flux was compared
with that of the nearby F-star. These conclusions remain robust even if we
include periods of data that were affected by poor weather. We can set
upper limits of 
per cent on intra-night variation of the optical
flux of the nucleus. Given the lack of any detectable variability, the 8
`good' spectra were combined to form a single co-added spectrum with
effective exposure times 4800s in the blue and 4360s in the red.
Figures 1a and 1b show the blue and red coadded spectra, respectively.
These spectra possess extremely high signal-to-noise and, therefore,
uncertainties are dominated by calibration effects and atmospheric
variations. Comparing the 8 separate good spectra (i.e. those obtained
prior to co-adding), we estimate that there is a 
per cent
uncertainty in the overall normalization of each spectrum. These are most
likely due to `grey' variations in atmospheric conditions. Treating these
uncertainties as independent, the coadded spectra should each have an
uncertainty of approximately 3 per cent in overall normalization. The
uncertainties in the red and blue spectra should be statistically
independent.
Several features are immediately apparent from these spectra. The emission
line spectrum is dominated by broad Balmer lines (H 
, H 
,
H 
and H 
) and narrow forbidden oxygen lines ([OII] 
and [OIII] 
). The emission line spectrum, which is
clearly related to the Seyfert activity, is examined in more detail below
(Section 2.1.2). The spectrum between 9000Å and 10000Å may be severely
affected by incorrect subtraction of atmospheric water features, and so the
line-like features in this region of the spectrum should be treated with
caution. The presence of the CaII doublet in absorption (near
4000Å) shows there to be a non-negligible fraction of starlight from
the host galaxy contributing to this spectrum. We now examine this stellar
component.