Fairall-9 (F-9 = ESO 113-IG45) is a Seyfert 1 galaxy () in the Southern Hemisphere (
and
) at
z=0.0461 (Ward & Wilson
1978). It was discovered spectroscopically as a Seyfert 1 galaxy by
Fairall (1977), who noted its high luminosity of
and classified it as an extreme Seyfert 1 galaxy, almost a QSO
(Fairall 1977; Ward & Wilson 1978; West et al.\
1978). Its morphology shows an unresolved nucleus (
at
, taking
) with an envelope with several spiral arms extending until
from the center and an elliptical companion galaxy at 28'' SE.
The diameter of the outer envelope is 64'' (40 kpc), and a jet is suggested
to exist out to
SE of the nucleus (West et al.\
1978). Griersmith & Visvanathan (1979) classified it as an
early type spiral galaxy, but Véron-Cetty et al.\
(1991) consider it an elliptical giant galaxy. Its optical linear polarization
is low,
, as in most Seyfert galaxies, consistent with
interstellar polarization by dust from our galaxy only (Martin et al.\
1983).
F-9 has displayed an unusually large amplitude in its variability: it decreased in
brightness from 1978 to 1984 with and in the UV
with
, corresponding to maximum to minimum
brightness ratio
(Wamsteker et al. 1985) and
increased afterwards, but never recovered its discovery brightness. Similar
variations have been observed in the NIR (Glass 1986; Clavel
et al. 1989) and X-rays (Morini et al. 1986), but with smaller
amplitude than that seen in the UV. The optical variations between 1979 and 1990
have been extensively documented by Lub & De Ruiter (1992).
More recent X-ray observations have been made with GINGA (Makino, private communication)
and ROSAT (Walter et al. 1995). FIR observations with IRAS have been
reported by Edelson & Malkan (1987) and radio observations by
Véron-Cetty et al. (1991).
The emission lines have been studied in considerable detail. The narrow lines of
, have been suggested to show a blue asymmetry in the profile
(Whittle 1985; Busko & Steiner 1988, 1989) and no
flux variability (Kollatschny & Fricke 1985; Wamsteker et al.\
1985; Lub & De Ruiter 1992). On the other hand, the
intensity and profiles of the broad lines are highly variable. The broad UV lines
(
,
and
) have been extensively studied by Chapman et al.\
(1985), Wamsteker & Colina (1986) and Clavel et al.\
(1989), and the optical lines by Wamsteker et al. (1985) and
Lub & De Ruiter (1992). The complex structure of the UV line
profiles has been discussed by O'Brien & Zheng (1990) and
Zheng & O'Brien (1990). A detailed profile decomposition by means
of gaussian components has been made for
\
by Wamsteker et al. (1985). Cross correlation studies of the total
intensity of
,
and
showed a delay in the line response to UV
continuum variation of
for all three lines (Clavel
et al. 1989). Koratkar & Gaskell (1989) suggest that CIV
seems to show evidence of infall motion, while
suggests random motion.
Photoionization model calculations on the total line intensities by
Clavel & Santos-Lleó (1990) indicate a Broad-Line-Region
(BLR) at 100 light-days from the ionizing source, which needs a cut-off energy
of 0.83 ryd in the big blue bump, a density of
and a
column density of
to be able to explain the Wamsteker-Colina effect, the apparent
saturation of CIV at high luminosities as defined by Shields et al.\
(1995). Binette et al.\
(1989) require a similar cut-off energy at 0.73 ryd using a different
photoionization code.
We discuss here all observations made in the UV (Sect. 2) and compare these with the observations made at other wavelengths to derive the Spectral Energy Distribution (SED) in Sect. 3. In Sect. 4 we give a detailed analysis of the emission line variability, including line decomposition. Section 5 combines the results of Sects. 3 and 4 to compare the data with photoionization calculations and derive the nature of the Broad-Line-Region (BLR) in Sects. 6 and 7. A summary of the results is given in Sect. 8.