There is a clear difference between the direct images obtained through the various filters. The I band images display a smooth distribution of mainly evolved red stars, whereas the B and V images reveal the presence of absorbing dust, H II regions and young stars. In the B-I maps of most galaxies, we detect clumpy, ring-like or elongated blue or red structures. Next we present as b/w figures (Plates 1-22) for each galaxy a blue (B band, except V band for NGC 7469) direct image for morphological information and various colour maps.
The Sb type Seyfert 2 galaxy NGC 1068 is the nearest object
displaying both AGN and starburst activity, with 1''
corresponding to a projected size of 110 pc. In Plate 1, we show
its B band image. The nucleus exhibits a Seyfert 2 optical
spectrum in direct light, but spectropolarimetry has revealed a
broad, polarized component to the narrow emission lines (e.g.\
H
), indicating the presence of an obscured Seyfert 1
nucleus (Antonucci & Miller 1985; Miller et al. 1991). UV,
optical and near-IR (NIR) observations of the bright inner
(
15'') disk indicate a young stellar spectrum with
strong Balmer absorption and emission lines, suggesting a large
population of B type stars and their associated H II regions,
typical of a starburst (e.g. Bruhweiler et al. 1991). NGC 1068 is
extraordinarily rich in molecular gas, with an estimated total
H
mass of 
(Planesas et al. 1991).
The MIR and FIR emission is extended (Braatz et al.\
1993; Cameron
et al. 1993) indicating that
the IR emission originates in dust not solely heated by
the AGN. The complex linear radio source of 
7'' in
extent has two compact jets on opposite sides of the nucleus
(, Wilson & Ulvestad 1983, 1987). The NLR is cospatial with the
radio emission and displays an outflow in the direction of the
jets, driven by interaction with radio plasma (Baldwin et al.
1987), characteristic of AGN activity. Both the NLR (Pogge 1989)
and the ENLR (Unger et al. 1992) share a conical shape aligned
with the radio axis, indicative of photoionization by a nuclear
radiation field collimated by the torus. A ROSAT X-ray
image of NGC 1068 (Wilson et al. 1992) reveals, in addition to a
compact central source, larger scale emission, which is
coincident with the outer edges of the optical disk and is
probably due to SNRs, X-ray binaries or a starburst driven
wind, and not to electron scattering of the hidden Seyfert
1 nucleus.
The primary aim of this work is to use the colour maps to investigate the properties of the nucleus, and circumnuclear regions on kiloparsec scales. However, most of the host galaxy is also imaged, so we shall begin by briefly commenting on the large scale structures.
There are two previous studies of the large scale colour
properties of NGC 1068. Schild et al. (1985) obtained CCD
images in B, V and R. Ichikawa et al. (1987) published
ratios of photographic images, based on plates taken using B,
V, and i filters. Our large scale B-I (
;
) map is shown in Plate 2. Our map agrees
well with the B-R image of Schild et al. excluding regions
within 5'' of the nucleus. The oval distribution of H II
regions, centred on the nucleus and with major axis about
30'', is clearly seen in Plate 2. We see much more detail in
the structure of the H II regions than Schild et al. because of
our deeper imaging and better seeing.
There is an arc of blue resolved features NE-SW through the nucleus roughly parallel to the radio jet axis. These knots (labelled by Schild et al. 1985) are known to be related to sites containing luminous young stars (e.g. Bruhweiler et al. 1991; Truong & Bruhweiler 1991). All the knots have similar spectra, so they should have comparable star formation ages. Our images cover the knots 2 to 4 (see Schild et al.). The colours of these knots within a 3'' aperture are listed in Table 3. As for all the galaxies in Table 3, the colours include contribution both from the knots and the underlying old redder population. Therefore, the real colours of the knots must be even bluer than those reported here. The colours are in good agreement with Schild et al. (who used a 7'' aperture), and confirm that they are sites of recent star formation. Knot 4 is extremely blue and could be a very recent star formation site, as supported by its Wolf-Rayet type optical spectrum (Bruhweiler et al. 1991).
Table 3: Colours of the maxima
The excellent correspondence of the colour properties of large scale structures, gives us confidence concerning the procedures used to produce our maps. We now turn to interpretation of the nuclear features.
Unlike the previous work mentioned above, our data can be used
to study details on scales of a few arcsecs. For this purpose we
show the B-I map of the innermost 
(
) in Plate 3. In the nuclear region on
scales less than 10'', there is an elongated blue structure,
extending along PA 40
for about 6'' (660 pc). It peaks
NE of the nucleus and has similar colours to the
average of the H II region knots (Table 3), except B-V, which
is redder. These colours are much bluer than a typical B-I
index for a spiral bulge of T = 2 (
), but redder
than the average B-I index for Seyfert 1 galaxies (
). It is likely that the blue elongation is caused by
scattered light from the nucleus of NGC 1068, which is
superimposed on the stellar population of the nuclear region,
which is redder. Therefore the observed colour index is an
upper limit to the blueness of the nuclear source whose
radiation is scattered.
Pogge & De Robertis (1993) presented narrow band emission line
and continuum imaging observations of NGC 1068. In the innermost
region of their ``UV/Red'' (3600/6100 Å) colour map, they find
a number of blue features. The bluest region is offset from the
nucleus by 
0.8'' in PA 
33
. The blue
elongation is along the radio axis and roughly coincident with
the peak of the [OIII] cone structure (Evans et al. 1991), radio
continuum NE peak (Wilson & Ulvestad 1983) and the 10 
m
emission peak (Tresch-Fienberg
et al. 1987). Additionally, there is an extended, diffuse UV excess continuum
component with a conical morphology. More complex situation is
revealed by the recent HST UV and optical imaging of NGC 1068
(Macchetto et al. 1994) who confirm the presence of the conical
blue structure in their UV/U continuum ratio map extending over
4'' with quite uniform colour. The UV peak is displaced by
0.05'' to north of the [OIII] peak and the UV continuum region
envelopes the [OIII] emitting region. Intriguingly, they found a
``twin-crescent'' object SW of the cone. This object is highly
polarized (
45%) and lies at the center of the overall
polarisation pattern of the circumnuclear region and at the apex
of the cones (, Capetti et al. 1995a,c). This structure probably
straddles the real nucleus, but is not the torus, because
it is misoriented with respect to the cones.
The morphology in Plate 3 is in good agreement with the structure observed by Pogge & De Robertis (1993) and Macchetto et al. (1994) and with the structure in the imaging polarimetry of NGC 1068 by Miller et al. (1991) and , Capetti et al. (1995a,c). We shall discuss the origin of this elongation in more detail, in conjunction with other galaxies, in the context of scattering of the nuclear light, in Sect. 6.2.
The SAB pec galaxy NGC 3227 (a B band image is shown in Plate
4) is interacting with a spheroidal companion NGC 3226. NGC 3227
has Seyfert type 1 spectrum (Salamanca et al. 1994). but NGC 3226
shows no signs of activity, either AGN or star formation. The
system is probably in the postcollision phase (Taniguchi et al.\
1990). The X-ray luminosity of NGC 3227 is low for a Seyfert 1
galaxy (Ward et al. 1987). It has a substantial amount of
molecular hydrogen emission, based on NIR spectroscopy (Fischer
et al. 1987) as well as
strong millimeter CO emission (Blitz et al. 1986; Meixner et al.\
1990). The resolved CO distribution has two intensity peaks
straddling the nucleus, aligned roughly E-W and separated by
2'' (Meixner et al.\). A more extended component (with a
diameter 17'') lies SE-NW along a 30
angle to the
galactic major axis, the NW part is the site of most of the
emission. In the maps of NGC 3227, 1'' corresponds to 100 pc.
Our full scale (
; 
) B-I map
(Plate 5) shows a wealth of structure. Especially, there is a
band of light (blue) colour, roughly in a shape of Z delineating
the nucleus, indicating the geometry of the dust absorption in
the galaxy. This geometry agrees well with that seen in the H I
21 cm absorption map by Mundell et al. (1995).
In Plate 6, we show the B-I image of the innermost 
(
) of NGC 3227. The colour map shows
two blue maxima on either side of the nucleus (NW and SE) with a
separation 
2''.
Star forming regions are known to be associated with dense molecular clouds and dust. CO mapping indicates the presence of molecular clouds and hence the material available for star formation (Heckman et al. 1989). Therefore, a correlation between the CO distribution and the colour maps would be expected if the blue maxima are caused by recent star formation. Indeed, we find a good correspondence between the CO map (Meixner et al. 1990) and our B-I map (Plate 6), where two blue maxima are coincident with the CO emission peaks. The agreement is especially good for the SE maximum.
However, there is no direct evidence of ongoing star
formation in the nuclear region of NGC 3227. On the other hand,
the nucleus of NGC 3227 is strongly polarised, with a degree of
polarisation rising smoothly from 
in the red to
nearly 3% in the blue (Schmidt & Miller 1985), and essentially
constant PA, both for the continuum and the broad lines. This
polarisation can be best explained by interstellar dust
scattering, where the dust is intermixed with the narrow line
region (NLR) gas (Schmidt & Miller). The PA of the
polarisation is, interestingly, roughly parallel to the nuclear
radio source axis (Ulvestad et al. 1981) and the optical major
axis, suggesting that the dust clouds are distributed in the
same plane as the overall galactic structure. Indeed, the
double maxima are orientated along the major axis, and it is
plausible that they represent scattering of the nuclear light in
unobstructed directions.
The colours of the blobs in a 1'' aperture are given in Table 3. They are quite red, and not typical of young hot stars or scattered light, but the real colours must be bluer, since they include also the underlying structure, which is redder than the actual blue maximum. Note that the aperture used is very small (to avoid contamination from the other maximum and the nucleus) and the colours therefore prone to photon statistics errors.
NGC 4151 is an SABab galaxy (de Vaucouleurs et al. 1991) and it
is the best studied bright Seyfert 1 galaxy. A B band image of
NGC 4151 is shown in Plate 7. Its redshift is 0.0033 and 1''
corresponds to 100 pc on the sky. The BLR is point-like but the
NLR extends in a biconical structure 
4'' at PA
60
(Evans et al. 1993) and the ENLR is resolved into
a jet-like feature with a series of high-ionization blobs
covering 
15'' SW and a fainter plume 5'' NE at PA
50
(e.g. Heckman & Balick 1983;
Perez et al. 1989).
Extended X-ray emission at the brightest parts of the ENLR was
discovered by Elvis et al. (1983). The radio structure is a
collection of resolved sources (Harrison et al. 1986). The
nuclear jets are at arcsec scale at 
and
misaligned with the ENLR by 
30
, whereas at smaller
scales, the radio component is elongated at PA
57
,
similar to the NLR and ENLR (Heckman & Balick 1983; Pedlar
et al. 1993).
In Plate 8, we plot the B-I map of the innermost 
(
) of NGC 4151. There is a strong
colour gradient with the circumnuclear region being bluer than
the rest of the bulge. This elongated blue structure is at
PA
60
extending over 7'' in diameter. The
elongation is parallel to the small scale radio components and
the NLR (Evans et al. 1993). In the nuclear region we find two
unresolved blue components symmetrically located each side of
the nucleus (at PA
80
) separated by 
2''. They
are misaligned by 
20
with respect to the ionisation
cones (Evans
et al. 1993) and the small scale radio axis. This structure is in good
correspondence with the knots in the similar resolution radio
map by Johnston
et al. (1982).
The colours of the two maxima are given in Table 3. The B-V colours are surprisingly red; the other colours are consistent with normal spiral bulges. Since the real colours (without the underlying structure) must be bluer, these colours could be caused by a young stellar population or by scattered nuclear light. Again, with no firm evidence of ongoing star formation in the nucleus of NGC 4151, we prefer the latter possibility, but discuss the problems associated with this interpretation in Sect. 6.2.
NGC 7469 (a V band image is shown in Plate 9) is a well
studied Seyfert 1 galaxy of Hubble type SAB(rs)a and redshift
0.0167 (1'' corresponds to 490 pc on the sky) exhibiting a
circumnuclear starburst ring. It shows strong 3.3 
m
emission feature from dust heated to 
, in a
region between 1 and 3'' from the nucleus, with the heating of
the dust grains probably due to radiation from hot young stars
in the circumnuclear H II regions (Cutri et al. 1984). NGC 7469
also exhibits the PAH emission features in the MIR spectrum
(Roche et al. 1991), common for starbursts. The 12.5 
m
emission is elongated 3'' along PA = 78
probably
representing unresolved integrated emission from the nucleus and
two brightest starburst ring features (Keto et al. 1992).
NGC 7469 has strong emission in the 
CO J= 1-0 (Meixner
et al. 1990) and H
= 1-0 S(1) (Genzel
et al. 1995) lines within a few arcsec from the nucleus, providing evidence
for a concentration of molecular gas.
A high resolution CO map of NGC 7469 (Meixner et al. 1990) shows,
in addition to a CO peak coincident with the nucleus, a SW
component 
3.5'' from the nucleus and an extension to the
NW. In Plate 10, we present our V-I map of the innermost
(
). The V-I map shows
an irregular ring of blue emission around the nucleus with a
diameter 
3''. There are three notable blue maxima, in
the S and in the NE and NW quadrants. The overall dimensions
agree quite well with the CO emission (Meixner et al.\) although
there is not a one-to-one correspondence. The SW CO plume is
not reproduced, neither the NE maximum in the colour map.
A high resolution (0.4'' FWHM) radio map at 6 cm of NGC 7469
(Wilson et al. 1991) is dominated by a strong compact source
centered on the Seyfert nucleus. Much of the remaining radio
halo (
8'' diameter) is concentrated into an annular
structure of diameter 
3'', with bright maxima to the NE
and SW. An [OIII]/H
image of Wilson et al. shows that
H
is extended around a bright peak coincident with the
continuum nucleus. The [OIII] emission is much more centrally
concentrated with a plume to the N.
Mauder et al. (1994) presented high resolution, speckle masking
images at three optical wavelengths (5400 - 6700 Å) with the
central source subtracted. The images show a ring shaped
emission region with a radius of 1.3'' and a lumpy structure
with three main maxima at PA 
40, 245 and 320
. The
maxima in our V-I image (Plate 10) correspond well with those
in the radio, emission line and red continuum images. There is
an especially good coincidence for the NE peak. The agreement
for the SW peak is less good, but we have to remember the much
higher resolution in the radio map. Also, the overall dimensions
are in exact agreement. This similarity between the radio and
optical indicates that the circumnuclear structure seen in the
continuum colour map represents direct emission from hot stars
in the ring,
i.e. a starburst.
The colours of the blue circumnuclear blobs in a 2'' aperture are given in Table 3. The V-I colours are remarkably uniform and close to normal old stellar population colours. Again, the real colours without the underlying structure must be bluer.
Mrk 3 is an S0 type galaxy (Jenkins 1981; Wagner 1987) with a
Seyfert 2 nucleus (Koski 1978) at a redshift of 0.0134; 1''
corresponds to 410 pc on the sky. A B band image is shown in
Plate 11. The X-ray spectrum of Mrk 3 (Awaki et al. 1990) shows a
heavy absorbing column (
cm 
) and a
strong Fe K line. This can be explained by absorption in and
scattering off an obscuring torus. Awaki et al. suggest Mrk 3
being an intermediate case between less obscured (e.g. Mrk 348)
and completely absorbed (e.g. NGC 1068) Seyfert 2s.
The H I mass is 
(Pedlar et al. 1984),
large for ellipticals. This may be due to interaction with
UGC 03422, located 6' NW along PA = 155
, although there
is no evidence for this in direct optical images (Wagner 1987).
At 5 GHz, Mrk 3 exhibits a pair of highly collimated jets
perpendicular to the major axis of the galaxy, 2'' in extent
along PA = 84
(Kukula et al. 1993). Embedded in the jets
are compact components, the western end terminating in a bright
lobe containing a hotspot. Mrk 3 has a dust-dominated steep
spectrum in the NIR (Edelson & Malkan 1986), high optical
polarisation (
for lines and continuum at PA =
130
; Schmidt & Miller 1985) and shows broad lines in
polarised light (Miller & Goodrich 1990).
Recent HST imaging by Capetti et al. (1995b) revealed a striking
spiral shape of the ionized gas emission, with a large number of
resolved knots in a fainter extended region. The brightest part
of the spiral is confined into a cone of 
40
full
opening angle, and its morphology is similar to the radio
(Kukula et al. 1993). The optical emission lines are blue
asymmetric, indicating absorption by dust surrounding the NLR
clouds. The nuclear emission thus seems to be heavily absorbed
in the line of sight, while the ionizing flux is intrinsically
not strongly collimated.
The B-I image of the central 
(
) region of Mrk 3 is shown in Plate 12. The central
region is redder than the surrounding bulge at PA 
100
with extent 
11''. Within this red elongation, there is
an arclike blue extension, with two main blue maxima, roughly
elongated at PA = 95
. The W maximum is bluer at
1.3'' from the nucleus, while the E maximum is at
0.8'' distance from the nucleus. In the E, the elongation
curves towards larger PA, fading away at 
4.6'' from
nucleus. The double structure is roughly aligned with the radio
axis (85
; Ulvestad & Wilson 1984). The opening angle of
the ``cone'' agrees well with that measured from the emission
lines (Capetti et al. 1995b), and the curvature of the blue
continuum is similar (S-shaped) to that seen in the emission
lines (Capetti et al. 1995b) and in the radio (Kukula et al.\
1993). Most likely, the blue structure is scattered light from
the nucleus following the biconical structure in Mrk 3. Outside
of the blue emission, the colour is typical of galactic bulges,
displaying no sign of star forming regions in its host galaxy,
rendering the possibility of jet-induced star formation
unlikely.
Pogge & De Robertis (1993) presented a ``UV/Red'' colour map of
Mrk 3, which revealed two distinct structures: a pair of UV
(blue) knots on either side of the nucleus along PA = 85
,
separated by 1.5'', and a diffuse blue component with a
distinct bi-conical morphology extending out to 5'' on either
side of the nucleus along PA = 114
. Our B-I map is in
good agreement with their map, with the PA's agreeing within the
errors. We discuss the origin of these blue maxima in Sect. 6.2.
The V-R map of Mrk 3 (Plate 13), which resembles an
[OIII]/H
excitation map, shows a spiral structure in the
outer regions with a radius 
4'' from the nucleus, in
agreement with Capetti et al. (1995b) results. Interestingly,
there is also a red double structure in the nuclear region,
which corresponds well to the blue maxima in the B-I map
(Plate 12). The colours of the maxima in a 2'' aperture are
given in Table 3. The BVRI colours are redder than for the
NGC 1068 elongation, but still in agreement with scattering,
being contaminated by the underlying red stellar population.
Mrk 78 is a Seyfert 2 galaxy at redshift 0.038 (1'' equals
1.15 kpc). Plate 14 shows its B band image. At 2 cm, Mrk 78
shows a triple structure, with a bright core and fainter
components to the E and W, parallel to the NLR axis (Pedlar
et al. 1989). There is no detailed morphological correlation,
however, between the radio and the ENLR (Unger et al. 1987), with
the axes misaligned by 
25
(Capetti et al. 1994). This
misalignment can be caused by the ENLR being in a different
plane from the rest of the galaxy, or by different covering
factor E and W of the nucleus (Pedlar et al. 1989).
The [OIII] emission is biconical with opening angle
40
extended along the radio axis at PA = 67
(
in extent; Capetti
et al. 1994). There is also a resolved obscuring zone close to the center,
elongated E-W, with 
0.8'' extent and displaced from the
center of the cone by 
0.2''. The lack of far-UV
emission, comparison with the FIR emission, and photon budget
arguments indicate a strongly anisotropic nuclear source.
Capetti et al. suggest that obscuring dust (both on pc and kpc
scales) prevents UV radiation to escape along our line of sight,
while the stellar continuum remains largely unobscured.
We present our B-I map in Plate 15. It shows the nucleus as
redder than the bulge of the galaxy. The reddest region close to
the nucleus is slightly extended (
) along PA
330
. This extension is
almost perpendicular to the E-W [OIII] morphology (Capetti et al.\
1994), and may show the location of the obscuring dust in this
galaxy. There is also an arc-like blue region forming a broken
ring from E through N to SW at a distance 
3'' from the
nucleus. Again, although this could be a star forming region
around the nucleus, or show a gradient in the stellar
population, there is no direct evidence for this. A more
appealing possibility is that it represents a scattering mirror
which is less efficient or further away from the nucleus
than in the cases of NGC 1068, Mrk 3 and Mrk 573 (see Sect.
6.2). The colours of the blue ring are redder in V-I, but
bluer in B-V than for the NGC 1068 elongation (Table 3).
Mrk 348 is an almost face-on S0/a-sb type galaxy (Huchra 1977)
with a Seyfert 2 nuclear spectrum (Koski 1978). The B band
image is shown in Plate 16. The velocity field of the neutral
hydrogen and tidal plumes indicate an extensive tidal
perturbation by a companion galaxy, NGC 266 (Simkin et al. 1987).
Its small reddening (
both for lines and
continuum) is mainly produced in our Galaxy. At a redshift of
Mrk 348 (0.014), 1'' corresponds to 420 pc projected distance
on the sky.
Spectropolarimetry has revealed the presence of broad wings to
the permitted lines in Mrk 348 (Miller & Goodrich 1990). The
radio structure is linear on a scale of 0.2'' (Neff & de
Bruyn 1983; Ulvestad & Wilson 1984). There is also an extended
distribution of ionized gas in the nuclear region (Bergeron
& Durret 1987). The 
luminosity of Mrk 348 is a factor of 30
brighter (Warwick et al. 1989) and the Fe K
line an order
of magnitude fainter (Koyama et al. 1989) than in NGC 1068.
Probably Mrk 348 is seen directly through much less column
density than in the case of NGC 1068.
We present our B-I map of the central 
(
) of Mrk 348 in Plate 17. It shows the
nucleus as redder than the near-nuclear region. At larger
distances, the host galaxy appears again as red. Between these,
there is an intermediate blue ring-like region, 
3'' in
diameter elongated at PA 
15
. Again, although the map
of Mrk 348 is less clear than for the other galaxies, this blue
structure may be related to scattering of the nuclear light,
with less efficient ``mirrors'' than in the other galaxies
studied.
Mrk 573 is a Seyfert 2 galaxy at a redshift of 0.017 (1''
corresponds to 510 pc projected distance on the sky). Plate 18
shows its B band morphology. The extinction towards the
nucleus, 
, is probably mostly intrinsic to the
galaxy. It has a triple radio source at 6 cm with two lobes of
similar luminosity at distance 
1.5'' from the nucleus
along the radio axis at PA = 125
(Ulvestad & Wilson
1984). Narrow band [OIII] and H
images show an
elongated emission line nebulosity extending about 
along the radio axis, and two oppositely directed cones of
ionising radiation (Haniff et al. 1991; Tsvetanov
& Walsh 1992). The H
peak coincides with the continuum peak, but the
[OIII] maximum is shifted 4'' along the radio axis to SE
(Whittle et al. 1988; Tsvetanov & Walsh 1992). Perpendicular to
the line image elongation there is a low-ionisation valley
through nucleus at PA 
40
(Tsvetanov & Walsh 1992).
Kinney et al. (1991) compared the number of ionising photons seen
by the ENLR gas to the number of ionising photons seen in Earth
for a sample of Seyfert 2 galaxies. For Mrk 573, they found a
ratio 
1. However, corrected for extinction and biconical
covering factor, the ratio becomes 
5 (Tsvetanov & Walsh
1992), providing strong evidence for anisotropy of the radiation
field.
Pogge & De Robertis (1993) presented a ``UV/Red'' map of
Mrk 573, where they find a biconical structure oriented at PA =
124
with an opening angle 
60
. They also find a
curved feature 3.5'' SE of the nucleus and a similar feature
4.5'' NW. Both are coincident with [OIII] emission line
features, but outside the radio continuum lobes. Recently, Pogge
& De
Robertis (1995) obtained narrow band images in five emission
lines and a continuum band. In addition to the features
mentioned above, which they found to be elongated perpendicular
to the radio axis and rotated following the spiral arms
rotation, they found two further arc-like structures at 1.7''
SE and 2.2'' NW of the nucleus at PA 124
, bracketing the
radio lobes. Based on the morphology and ionisation gradient,
Pogge & De Robertis (1995) suggest that the inner pair of
features is reminiscent of edge-brightened bow shocks where the
emission line gas is interacting with the radio plasma from the
AGN, while the outer pair may represent episodic ejection of
plasma from the nucleus.
We present our B-I map of the central 
(
) of Mrk 573 in Plate 19. It shows two blue
maxima on either side of the nucleus with a separation of
7'' at PA 
135
and a red ``bridge''
perpendicular to them crossing the nucleus. The double
structure is roughly parallel to the radio axis (PA = 124
;
Ulvestad & Wilson 1984) but the blue component is more
extended than the radio continuum emission. There is also a
tightly wound spiral arm structure tracing an elliptical ring
with major axis 
aligned at PA 
10
. Most
likely, the blue maxima define the geometry of the radiation
cones, with the red bridge indicating the large scale obscuring
material across the nucleus. While the blue maxima could
alternatively be understood as arising from star formation
triggered by the radio jet, there is no sign of young blue stars
in the continuum images. Furthermore, the ENLR line ratios
indicate an AGN-type ionising spectrum. Note, however, that the
10 
m flux of Mrk 573 is relatively larger than in typical
Seyfert 2s, suggesting that a starburst component may contribute
in the nucleus (Maiolino et al. 1995). There is a good
general correspondence between the maps of , Pogge & De Robertis
(1993, 1995) and our B-I map (Plate 19); except that we can
resolve the blue maxima, which are more diffuse and situated
further away from the nucleus than in NGC 1068 and Mrk 3. For
discussion of the blue maxima, see Sect. 6.2. The colours of
the blue maxima in a 3'' aperture are given in Table 3. They
are bluer in B-V and V-R than those of the NGC 1068
elongation, and in good agreement with arising from scattering.
Except for its supernova 1986N, NGC 1667 has not been studied in detail so far. It is a relatively isolated type Sc galaxy with a Seyfert 2 nucleus at z = 0.0153; 1'' corresponds to 470 pc on the sky. A B band image of NGC 1667 is shown in Plate 20. Evidence for star formation activity in its host galaxy comes from starburst-like FIR colours, slightly extended MIR emission, and optical morphology and emission line ratios (see references in Heckman et al. 1989). The radio emission is diffuse and single-component (Ulvestad & Wilson 1989), indicating that most of the radio emission in NGC 1667 does not arise from the AGN.
We present our large scale B-I map (
) of NGC 1667 in Plate 21. The image delineates
the spiral structure of this galaxy, the individual H II regions
in the spiral arms, and the geometry of dusty material, as in
the case of NGC 3227 (Plate 5). Note especially the red
``bubble'' extending up to 9'' N of the nucleus. Plate 22
shows the B-I map of the central 
(
) nuclear region. In addition to the
innermost spiral arm, there is a red double structure across the
nucleus with a separation 
3'' at PA 
45
. The
colours of these red maxima are given in Table 3. Although this
red structure could be related to the material obscuring a
Seyfert 1 nucleus, we defer further discussion, but encourage
finding direct evidence of anisotropy in this little studied
Seyfert 2.