4 Observational results

4.1 General description

The examination of the data in various 1D or 2D displays, showed that the colour distributions in the studied S0's are determined by several effects:

1.

To a first (rough) approximation the colours are constant upon a given isphote and decrease monotonuously with increasing isophotal radius, as occurs in E-type galaxies;

2.

Obvious dust patterns occur in a number of objects, giving rise to high contrast colour features;

3.

Diffuse dust may be concentrated in disks, leading to reddening of the disk as compared to the adjoining spheroid, or sometimes to the minor axis asymmetry in light and colour, previously discussed by Michard & Simien (1993) (MS93);

4.

Since, from population effects alone, bulge and disk may have somewhat different mean colours, and possibly have different radial colour gradients, deviations of the isochromes from the isophotes will occur. It turns out that the disk dominated major axis often becomes redder, far from center, than the minor axis for a given isophote: this may be tentatively ascribed to a larger colour gradient in the bulge, this component blueing outwards more than does the disk. On the other hand a stretch of bluer major axis (bluer disk!) appears in two of the studied objects.

 

Table 7: Reference colours and gradients, separately for mean isophotes and bulge dominated regions respectively. Very uncertain values are labelled with a colon. The sign of gradients $\Delta C/\log r$ has been changed. Obs: source of observations. $r_{\rm m}$ mean isophotal radius in arcsec of colour measurement. r₁,r₂: range for estimating gradients. Com: "component", as Iso for mean isophotal colours, Bul for bulge. In the first case the measurement radius is $r_{\rm e}$. B-V, etc.: mean local colours. $\Delta$: colour gradients. Notes: Scarce calibration data for NGC 3098, 5308, 5422. NGC 2549: bulge not measured, as not prominent enough against disk. NGC 3098: small bulge not measurable from OHP frames. NGC 3245: bulge not measured, as not prominent enough against disk. NGC 4026: bulge measured on W side, as E side is dusty. NGC 4036: bulge measured on N side, as S side is very dusty; isophotal gradients probably affected by the strong inner dust features. NGC 4111: bulge not measured, as affected by dust pattern. NGC 5308: bulge not measurable from OHP frames. NGC 5422: bulge measured on W side, as E side is dusty; isophotal gradients probably affected by the strong inner dust features. NGC 7332: large bulge measured on majA to avoid dust pattern. NGC 7457: B-R data taken at $r_{\rm e}/4$ from low fields CFHT frames; bulge not seen

NGC	Obs.	Com.	$r_{\rm m}$	r1,r2	B-V	$\Delta$	V-I	$\Delta$	B-R	$\Delta$	U-V	$\Delta$
2549	-	Iso	11	3,30	0.92 $\pm ~.02$	.06	1.20 $\pm ~.02$	.04	1.48 $\pm ~.02$	.09	1.45 $\pm ~.02$	.24
3098	-	Iso	8	3,20	0.80 $\pm ~.03$	.10:	1.16 $\pm ~.03$	.01	1.38 $\pm ~.03$	.07	1.36 $\pm ~.03$	.20
3098	CFH	Bul	3	1,5	-	-	-	-	1.42 $\pm ~.02$	.06	-	-
3115	-	Iso	30	3,40	0.95 $\pm ~.01$	.05	1.21 $\pm ~.01$	.00	1.52 $\pm ~.01$	.08	1.43 $\pm ~.02$	.17
3115	CFH	Bul	5	3,8	-	-	-	-	1.52 $\pm ~.02$	.08	-	-
3115	OHP	Bul	10	8,20	0.96 $\pm ~.03$	.07	1.23 $\pm ~.03$	-	1.52 $\pm ~.03$	.08	1.49 $\pm ~.03$	.18
3245	-	Iso	21	3,30	0.87 $\pm ~.01$	.05	1.14 $\pm ~.01$	.05	1.39 $\pm ~.01$	.07	1.38 $\pm ~.02$	.22
4026	-	Iso	16	3,30	0.90 $\pm ~.01$	.04	1.18 $\pm ~.01$	.05	1.47 $\pm ~.01$	.06	1.42 $\pm ~.01$	.21
4026	TBL	Bul	10	4,15	0.88 $\pm ~.02$	.06	1.18 $\pm ~.02$	.05	1.44 $\pm ~.02$	.09	-	-
4026	OHP	Bul	10	4,15	0.89 $\pm ~.02$	.06	1.20 $\pm ~.02$	.04	1.46 $\pm ~.02$	.08	1.40 $\pm ~.02$	.26
4036	-	Iso	20	3,30	0.90 $\pm ~.01$	.12	1.16 $\pm ~.01$	.07:	1.45 $\pm ~.01$	.15	1.46 $\pm ~.01$	.26
4036	TBL	Bul	10	5,20	0.89 $\pm ~.02$	.11	1.11 $\pm ~.02$	.05	1.43 $\pm ~.02$	.11	-	-
4036	OHP	Bul	10	5,20	0.90 $\pm ~.02$	.07	1.17 $\pm ~.02$	.04	1.46 $\pm ~.02$	.09	1.47 $\pm ~.02$	.22
4111	-	Iso	10	4,25	0.83 $\pm ~.02$	.08	1.12 $\pm ~.02$	.03	1.40 $\pm ~.02$	.10	1.23 $\pm ~.03$	.23
4350	-	Iso	15	3,30	0.91 $\pm ~.01$	.10	1.21 $\pm ~.01$	.08	1.48 $\pm ~.01$	.13	1.40 $\pm ~.03$	.26
4350	TBL	Bul	10	3,15	0.85 $\pm ~.02$	.12	-	-	1.49 $\pm ~.02$	.18	-	-
4350	OHP	Bul	10	5,15	0.92 $\pm ~.02$	.13	1.14 $\pm ~.02$	.05	1.47 $\pm ~.03$	.17	1.43 $\pm ~.03$	.33
5308	-	Iso	13	3,30	0.88 $\pm ~.01$	.08	1.17 $\pm ~.01$	.03	1.43 $\pm ~.01$	.08	-	-
5308	TBL	Bul	3	2,5	0.90 $\pm ~.03$	.17	-	-	1.43 $\pm ~.03$	.17	-	-
5422	-	Iso	13	3,30	0.94 $\pm ~.01$	.15	1.20 $\pm ~.01$	.09	1.54 $\pm ~.01$	.21	-	-
5422	TBL	Bul	5	3,10	-	-	1.15 $\pm ~.03$	.10	-	-	-	-
5422	OHP	Bul	8	5,12	0.93 $\pm ~.04$	.14	1.19 $\pm ~.04$	.08	1.50 $\pm ~.04$	.14	-	-
7332	-	Iso	14	3,30	0.85 $\pm ~.02$	.02	-	-	1.37 $\pm ~.02$	.09	-	-
7332	TBL	Bul	8	3,12	-	-	-	-	1.38 $\pm ~.03$	.09:	-	-
7457	-	Iso	33	3,30	0.84 $\pm ~.02$	.01	-	-	1.39 $\pm ~.02$	.07	-	-

 

Table 8: Selected measurements of disk colours and colour gradients. The sign of gradients $\Delta C/\log l$ has been changed. Obs: source of observations. $l_{\rm m}$: Mean majA of bulge colour estimates in arcsec. l₁,l₂: range for estimating gradients. B-V, etc.: mean local colours. $\Delta$: disk colour gradients against majA l. Notes: NGC 3115: Disk not measured, as nowhere dominant enough. NGC 4036: Disk not distinguished from "mean". NGC 7457: Disk not distinguished from "mean"

NGC	Obs.	Com.	$l_{\rm m}$	l1,l2	B-V	$\Delta$	V-I	$\Delta$	B-R	$\Delta$	U-V	$\Delta$
2549	OHP	Dis	40	20,50	0.92 $\pm ~.02$	.05	1.18 $\pm ~.02$	.00	1.47 $\pm ~.02$	.06	1.44 $\pm ~.02$	.18
3098	CFH	Dis	20	7,25	-	-	-	-	1.38 $\pm ~.01$	.04	-	-
3098	OHP	Dis	20	7,25	0.80 $\pm ~.02$	.02	1.16 $\pm ~.02$	.02	1.37 $\pm ~.02$	.04	1.35 $\pm ~.03$	.06
3245	OHP	Dis	30	10,40	0.87 $\pm ~.02$	.04	1.14 $\pm ~.02$	.01	1.40 $\pm ~.02$	.06	1.40 $\pm ~.03$	.09
4026	TBL	Dis	40	25,60	0.91 $\pm ~.02$	.05	1.18 $\pm ~.02$	.00	1.47 $\pm ~.02$	.07	-	-
4026	OHP	Dis	40	25,60	0.90 $\pm ~.02$	.08	-	-	1.47 $\pm ~.02$	.09	1.46 $\pm ~.02$	.12
4111	TBL	Dis	40	25,50	0.82 $\pm ~.02$	.04	-	-	1.31 $\pm ~.02$	.07	-	-
4111	OHP	Dis	40	25,50	0.80 $\pm ~.02$	.02	1.09 $\pm ~.02$	-.03	1.32 $\pm ~.02$	.02	1.15 $\pm ~.02$	.11
4350	TBL	Dis	20	13,34	0.90 $\pm ~.02$	.04	-	-	1.55 $\pm ~.02$	.09	-	-
4350	OHP	Dis	20	13,34	0.91 $\pm ~.02$	.08	1.20 $\pm ~.02$	.03	1.50 $\pm ~.02$	.04	1.53 $\pm ~.02$	.09
5308	TBL	Dis	25	15,40	0.92 $\pm ~.02$	.05	1.17 $\pm ~.02$	.05	1.46 $\pm ~.02$	.05	-	-
5308	OHP	Dis	25	15,40	0.89 $\pm ~.02$	.03	1.14 $\pm ~.02$	.03	1.43 $\pm ~.02$	.03	1.44 $\pm ~.03$	.10
5422	TBL	Dis	30	25,40	-	-	1.21 $\pm ~.02$	.12	-	-	-	-
5422	OHP	Dis	30	25,40	0.92 $\pm ~.02$	.08	1.20 $\pm ~.02$	.13	1.50 $\pm ~.02$	.07	-	-
7332	TBL	Dis	40	20,50	-	-	-	-	1.33 $\pm ~.02$	.07	-	-
7332	OHP	Dis	40	20,50	0.85 $\pm ~.02$	.03	-	-	-	-	-	-

4.2 Radial colour profiles

4.2.1 Mean isophotal colours

In the Appendix (Figs. A1 to A12), the distribution of mean colours for consecutive isophotes are shown against the isophotal radius r in a logarithmic scale. The 4 colours B-V, V-I, B-R and U-V are available for 8 objects; U-V is missing in the data for NGC 5308, 5422, 7332 and 7457, and V-I also for the two latter. Both the OHP low resolution data and the high resolution data from CFHT and TBL are shown. The examination of the graphs suggests the following comments:

1.

The agreement between colour profiles from the different instruments is generally fair. The deviations resulting from different resolutions are lessened by introducing adequate inner cutoffs for various observations, but may remain large in cases where a minute red feature occurs near the galaxian center as in NGC 4111. Disagreements at large r are obviously due to errors in the sky background corrections. Some differences between measurements from wide field and low field exposures may come from the calibrations, because the set of aperture photometry data will not necessarily be the same in such a case: an example of this situation is the B-R graph for NGC 3115;

2.

The radial colour profiles are rather similar for the various objects, but there are important systematic changes of the gradients in the various colours. V-I is always the flatter profile, then B-V, B-R, U-V have stronger and stronger gradients. The comparison of colour gradients for different colours is of interest in view of the uncertainties in the origin of this property;

3.

The inner colour profiles, as measured with our higher resolution vary widely from object to object: for instance, NGC 2549 shows a rather flat colour run, while NGC 3245, 4026 or 4111 have a central red peak. This is essentially due to the presence of various dust concentrations in the central galaxian regions. Such features may be largely washed out by lack of resolution.

In Table 7 are given the mean colours at $r_{\rm e}$ and the logarithmic gradients $\Delta C/\Delta \log r$, where C is any of the colours under study. As expected these parameters show significant variations from galaxy to galaxy. It is perhaps suggestive to consider the "statistics" of gradients for each colour:

in B-V, 12 cases, mean = 0.072, $\sigma = .041$

in V-I, 10 cases, mean = 0.045, $\sigma = .029$

in B-R, 11 cases, mean = 0.104, $\sigma = .047$

in U-V, 8 cases, mean = 0.224, $\sigma = .031$.

4.2.2 Bulge colours and gradients

There are two ways to distinguish the colours of the different components in S0's, say the bulge and the disk: on the one hand, it is possible to analyse the images in terms of these two components, independently for the two pass bands; on the other hand one can look at such parts of the images where one of the component strongly dominates the other. The first type of analysis will give model dependent results; in the second technique some "pollution" of the bulge light by the disk may be feared (or conversely). Since the bulge stands out very clearly along the minA of the near edge-on galaxies of the present sample, we have measured the corresponding colours and gradients, in such cases where the bulge appeared sufficienly clean of dust patterns. An analysis in terms of bulge and disk components is available for nearly all objects studied here (Michard 1998 and unpublished) (M 98), and the results were used to check that the disk contribution, along the minA, in the range selected for defining the bulge, was smaller than about 20%. For NGC 4026, 4036 and 5422, only this side of the bulge apparently not affected by dust was used. For NGC 7332, the minA is affected on both sides by the dust pattern: we tried to measure the bulge in a strech of the majA where the disk remains negligible, judging from the e₄ Carter coefficient. The bulge is not dominant enough for measurement in NGC 2549, 3245 and 7457; it is badly affected by dust in NGC 4111.

Since the axis ratio of bulges is certainly not constant, according to such bulge+disk decompositions where this parameter was not a priori constrained (Seyfert & Scorza 1996; Michard 1998), the logarithmic colour gradient of a bulge is not the same if measured in terms of minA or majA abscissae. We have here chosen to give bulge colour gradients $\Delta C/\log r$ in terms of the mean radius r. The results of the present attempt to measure bulge gradients, together with reference colours in the corresponding r range, are given in Table 7.

The measured reference colour for bulges are not much different of the isophotal colours at $r_{\rm e}$ given in the same table, because for this kind of objects, the $r_{\rm e}$ values do not fall much farther outside than the radii adopted here to define the bulge colours. The estimated bulge gradients are also similar to the isophotal radial gradients, because the measurement ranges often overlap, so that the gradients are bulge dominated. In a few favourable cases, i.e. NGC 3115, 4350, 5308, the bulge gradient is larger than the average one.

4.2.3 Disk colours and gradients

For most of the galaxies in the sample, there are extended stretches of the majA where the light is essentially due to the disk: such regions can be located from the bulge+disk analysis quoted above. An exception is NGC 3115 where the disk is nowhere much brighter than the spheroid. Disk colours and gradients have been measured wherever it seemed feasible, and the results are given in Table 8. The gradients are calculated against the majA l, so that they might be compared to various radial gradients of interest for disk galaxies. They cannot be compared to the mean isophotal gradients above without precautions, because of the variations of axis ratio in the range of measurement. It has been checked however that $\Delta C/\log l$ and $\Delta C/\log r$ are the same within errors, except in a few cases, where the later is larger in connection with a strong decrease of q in the domain of measurement. It should also be stressed that the run of majA colours may show significant fluctuations associated with ring-lens structures (see below), making the evaluation of gradients somewhat unsecure.

We have compared the disk colour gradients(DCG) of Table 8 to the bulge colour gradients (BCG) of Table 7 (sometimes replaced by the isophotal gradients if deemed equivalent): it is found that disk gradients are systematically smaller. A numerical comparison give the following results:

. B-V 8 cases, mean BCG: .10, mean BCG-DCG: .05, $\sigma$: .04;

. V-I 7 cases, mean BCG: .035, mean BCG-DCG: .00, $\sigma$: .02;

. B-R 8 cases, mean BCG: .11, mean BCG-DCG: .05, $\sigma$: .04;

. U-V 6 cases, mean BCG: .24, mean BCG-DCG: .13, $\sigma$: .04.
Note that measurement errors provide a large part of the $\sigma$ values for the differences in gradient between the two components.

On the other hand, the reference colours quoted for bulge and disk in a given galaxy are very much alike, without systematic differences. Further examination will show that, if taken upon a given isophote, majA colours (disk dominated) may differ slightly from minA colours (bulge or envelope dominated).

Figure 1: Azimuthal B-R colour profiles for NGC 4026 from TBL frames. Abscissae are polar angles in degrees, with the majA at 0 and 180 and the E minA at 90. The colours are measured within ranges of isophotal radii given as labels to each curve. Interesting features are the very red narrow inner disk; the blueing of the minA regions in the outer isophotal "rings"; the colour asymmetry in the minA regions, the eastern one being redder, with definite red peak due to dust, near PA of 30 and 150$^\circ$

Figure 2: Azimuthal colour profiles in NGC 4026 from TBL frames. This completes Fig. 11 by showing the profiles for the outermost measured range in three different colours. The colour contrast between the majA and the minA is minimum in V-I

Figure 3: Radial B-R colour profiles in NGC 4026 and 5308 from TBL frames. Abscissae: Mean isophotal radius r in arcsec and logarithmic scale. The colours of the 4 halves of each axes are shown separately. Upper frame: NGC 4026, PA = 179$^\circ$. Full dots: South majA. Small circles: North majA. Black squares: East minA. White squares: West minA. The minA becomes bluer outwards than the majA, much more so for the western side, as the eastern one is affected by dust (see also Fig. 14). Lower frame: NGC 5308, PA = 59$^\circ$. The various symbols then relate to orientations 120$^\circ$ clockwise from the ones above. The majA is redder everywhere, due to a thin red disk at small r and a blue bulge at large r. There is no minA asymmetry in this case

Figure 4: Azimuthal colour profiles through the inner part of NGC 3098 (upper curves), and the medium part of NGC 3115 (lower curves). Abscissae are polar angles in degrees, with the majA at 0 and 180. The colours are measured within ranges of isophotal radii given as labels to each curve. A narrow bluer region occurs along the majA

Figure 5: Asymmetries in the SuBr distributions along the two axis of NGC 4026 from TBL frames. Abscissae: Radial distances in arcsec. Ordinates: Differences in magnitudes between the two axes, with circles for the majA and crosses for the minA. There is no colour effect in the asymmetry along the majA, while the asymmetries along the minA change regularly with the colour: the E side is less bright and redder, due to dust in the disk of the object

 

Table 9: Measurements of the colour contrast of the dust lane in NGC 4036. Obs: Source of the data. Colour contrasts of the dust lane are averages of several points selected in 2D colour maps against an ad hoc background (see text). They are in magnitudes

NGC	Obs.	B-V	V-I	B-R	U-V
4036	TBL	0.18	0.19	0.25	-
id	OHP	0.15	0.11	0.21	0.24

 

Table 10: Selected measurements of mean colours and colour differences upon major and minor axes. A colon is appended to more uncertain values. OBS: source of observations. r: Mean isophotal radius. l and s: true semi-major and -minor axes. B-V, etc.: mean isophotal colours. $\Delta C$: colour difference, i.e. colour at the majA minus colour at the minA. Not: Notes. a) Inner disk of NGC 4026. b) NGC 4036: possible effect of dust pattern. c) NGC 4036: well outside dust pattern. d) NGC 4350: dusty disk? e) Inner disk of NGC 5308

NGC	OBS	r	l	s	B-V	$\Delta C$	V-I	$\Delta C$	B-R	$\Delta C$	U-V	$\Delta C$	Not
3098	CFH	16.	37.	6.9	-	-	-	-	1.38	0.10	-	-
4026	TBL	4.0	5.0	3.0	0.93	.085	1.18	.015	1.49	.07	-	-	a
id	id	6.3	7.5	4.8	0.91	.04	1.16	.01	1.47	.06	-	-	-
id	id	10.	13.	7.1	0.91	.05	1.15	.01	1.46	.055	-	-	-
id	id	16.	25.	10.	0.89	.08	1.15	.02	1.43	.085	-	-	-
id	id	25.	55.	14.	0.86	.09	1.14	.04	1.42	.125	-	-	-
4026	OHP	10.	13.	7.1	0.92	.02	-	-	1.48	.03	1.45	.09	-
id	id	16.	25.	10.	0.91	.03	-	-	1.47	.045	1.42	.105	-
id	id	25.	55.	14.	0.87	.05	-	-	1.45	.06	1.36	.155	-
id	id	40.	91.	20.	0.84	.09	-	-	1.45	.05	1.32	.21	-
4036	TBL	16.	25.	10.	0.93	.11	1.17	.07	1.48	.13	-	-	b
id	id	25.	40.	15.	0.91	.085	1.13	.015	1.43	.085	-	-	c
4036	OHP	16.	25.	10.	0.93	.06	1.19	.05	1.49	.08	1.51	.14	b
id	id	25.	40.	15.	0.90	.045	1.15	.03	1.44	.04	1.43	.130	c
id	id	40.	65.	25.	0.92	.015	1.13	.01	1.43	-.01	1.36	.08	-
4350	OHP	10.	16.	6.	0.93	.03	1.22	.04	1.50	.04	1.43:	.22:	d
id	id	16.	29.	9.	0.92	.05	1.19	.08	1.47	.06	1.39:	.21:	d
id	id	25.	46.	14.	0.89	.04	1.17	.09	1.44	.10	1.28:	.21:	d
id	id	40.	64.	25.	0.83	.06:	1.18	.10:	1.42	.10:	-	-
5308	TBL	4.0	5.1	2.8	0.90	.070	1.18	.040	1.45	.095	-	-	e
id	id	6.3	8.5	4.2	0.90	.060	1.18	.035	1.44	.080	-	-	-
id	id	10.	20.	5.8	0.89	.085	1.18	.040	1.43	.095	-	-	-
id	id	16.	36.	8.3	0.85	.110	1.16	.055	1.39	.130	-	-	-
id	id	25.	56.	12.	0.81	.160	1.13	.055	1.35	.170	-	-	-

4.3 Colour maps and dust patterns

Maps of isochromes and isophotes are shown in Figs. A1 to A12 in the Appendix. Such maps are specially useful to recognize "dust patterns" and their geometries. These are of course well seen in objects listed in the subclass S0₃ of the RSA, that is NGC 4036 and 5422, and less easily in the subclass S0₂ as NGC 7332. Besides this, dust patterns are detected in the colour maps of several S0's. Small size but high contrast patterns are seen in NGC 3245 and 4111. More extended low contrast dust formations appear in the colour maps of NGC 2549 and 4026. The Table 9 gives a few measurements of the contrast of dust formations in NGC 4036. Here the measured points were identified on 2D colour maps, while the background was chosen as the average colour upon an isophotal contour outside the pattern. It should be noted that colour contrasts are distinctly smaller for the OHP frames, as a result of their poor seeing. In view of seeing effects, relative colour contrasts are more or less compatible with expectations for the same type of dust as the one in our Galaxy.

Besides obvious dust patterns, colour maps show possible dust concentration in the inner disk of such objects as NGC 3115, 4026, 5308, where the inner isochromes are much flatter than the isophotes. This phenomenon is further discussed below.

4.4 Isochromes against isophotes

The deviations of isochromes from isophotes have been studied mostly from many "azimuthal colour profiles": examples of such diagrams are shown in Figs. 1, 2 and 4. If the isochromes follow the isophotes, the azimuthal profiles are horizontal lines except for noise fluctuations. This is so for NGC 2549 or 7457 (with rather poor S/N ratio). In other cases various kinds of fluctuations occur, regular or not, and located either near the center of the object or at large r values. These deviations are possibly due to diffferent physical causes, and are best described by simultaneously considering the 2D light and colour maps.

Features in the azimuthal profiles are classified as follows:

1.

"Dust patterns", already discussed in the previous subsection, show up as red peaks in the azimuthal profiles, where the "measurement ring" intercepts such a feature;

2.

"Diffuse dust in disks" may be present without forming well defined patterns, due to an unfavourable projection, or to a diffuse large scale distribution. The dust layer may then be detected by a reddening of the majA as compared to adjoining regions. For a favourable range of inclinations, it may also produce an asymmetry in light and colour along the minA, as the extinction and reddening are larger on one side of the bulge than the other. A discussion of this phenomenon may be found in MS93. From these criteria, dust is present in NGC 4026, 4350, 5422. From the first criterion alone, dust might also be present in the innermost range of the disk of NGC 3115, up to a majA length a=5 arcsec, and in part of the disk of NGC 5308, up to a=15. A red central disklet is also seen in NGC 4026, according to Fig. 1. This criterion is however not unambiguous, because the disk might well be locally redder than the adjoining regions from a difference in their stellar populations. Indeed we find that the inner "red disks" of NGC 4026 and 5308 have less colour contrast in V-I than B-V: this is not in line with the dust layer explanation... which however cannot be ruled out due to the effects of seeing upon these minute features;

3.

From azimuthal colour profiles a thin disk bluer than its surrounding occurs in NGC 3098, inwards of a radius of about $r_{\rm e}$, or 8 arcsec. This is also the case in NGC 3115, roughly in the range $15< r_{\rm e} < 30$. The blue disk of NGC 3098 is washed out in our poor resolution OHP data. The one of NGC 3115 is barely seen at the limit of our small field CFHT data. This phenomenon is presented in Fig. 4;

4.

"Blueing of the outer spheroidal component" of S0's is expected by analogy with ellipticals. Indeed in many of the azimuthal colour distribution, the outer measured isophotal rings show a large colour contrast between minor and major axes regions, the earlier ones being bluer. This is clear in our azimuthal colour profiles for NGC 3098, 4026, 4036, 4111, 4350, 5308, 5422, and is also marginally seen in others. It is easily checked that the minA regions become bluer with increasing radius, while the majA regions stay at nearly the same redder colour (see Fig. 3 for examples). It is therefore proposed that this feature of the azimuthal colour profiles is due to a colour gradient (similar to to the one observed in E-type galaxies that is a blueing towards larger radii), occuring in galaxian components other than the disk, that is the bulge itself and an outer envelope (sheroidal halo or thick disk), which may or not be considered as an extension of the bulge. It is suggestive that the colour contrast here discussed is minimum in V-I and maximum in U-V, exactly like the radial colour gradients discussed above.

In Table 10 are collected a number of results about the colour differences between the minor and major axes in our sample galaxies, as a function of isophotal radius and colour. As these differences are very sensitive to seeing, in the case of near edge-on galaxies, results for our OHP data are separated from the higher resolution CFHT and TBL data (and eventually discarded, if deemed too much affected by the poor PSF's). These data have been recovered from azimuthal colour profiles similar to Figs. 1 and 2. Depending upon the studied object and radius range, the majA minus minA colour difference may be due either to dust concentration in the disk or to blueing of the bulge component, or both. Our guesses as regards the relevant effect are indicated below, in the descriptions of each object.

As is the case for ellipticals, one may wonder if the suggested colour gradients in the bulge of S0's may be explained by the advocated population gradients (see a review in Wyse et al. 1997), or if a concentration of diffuse dust towards the center has to be invoked, as in the theoretical models of Witt et al. (1992), and in the applications by Goudfrooij & de Jong (1995), or Wise & Silva (1996). It may be added that, in the case of S0's, dust concentration in the disk is likely to play a role in colour distributions, not only in the inner part of several objects as discussed above, but also in their outer radial range. The situation is expected to be rather complicated, since we deal with galaxies built from at least two components, i.e. bulge and disk, and two sources of colour variations, i.e. populations changes and dust occurence.

 

Table 11: Measurements of the minor axis asymmetry. Obs: source of observations. s: distance along the minA where the asymmetry is measured in arcsec. For NGC 5422 the B measurement is from an OHP frame

NGC	Obs.	s	U	B	V	R	I
4026	TBL	12	-	0.18	0.15	0.11	0.08
id	OHP	id	0.20	0.11	0.09	0.08	-
4036	TBL	25	-	0.14	0.13	0.11	0.08
id	OHP	id	0.12	0.13	0.11	0.07	0.08:
4350	OHP	12	0.10	0.10	0.08	0.08	0.05
5422	TBL	15	-	0.20	0.13	0.09	0.03

4.5 The minor axis asymmetries

The phenomenon described in MS93, and tentatively attributed to rather thin, but extensive, layers of dust in S0 disks, has been reconsidered from the present higher resolution data. A clear cut case is NGC 4026, where the asymmetries are shown in Figs. 5, in 4 bands from TBL data, and display the expected colour effect.

Table 11 gives some new measurements of the minA asymmetry. It is interesting that the measured asymmetries are smaller for OHP frames than for TBL frames (cases of NGC 4026 and 4036) probably as a seeing effect. It has been found however that in some cases OHP frames of edge-on galaxies display large minA asymmetries due to an elongated and asymmetric PSF!

MS93 emphasized that the data at hand did not allow to distinghuish between extensive continuous layers of dust, as postulated in their attempted modeling, or systems of dust clouds. The high resolution colour maps obtained for NGC 4026 and 5422 suggest that some structure is present in the disk of these objects, notwithstanding the unfavourable projection.

Figure 6: Possible colour changes associated with rings in NGC 2549. Abscissae: majA central distances in arcsec and r1/4 scale. Ordinates: Upper frame: Differences in mag between a B majA SuBr profile and a least square fitted r1/4 law: this strongly emphasizes local ring features. Middle frame: B-R majA profile. Lower frame: U-V majA profile. There is some correspondence between the ring features and humps in the colour profiles, particularly in U-V