It can be seen from the the section above that four galaxies in our sample
have a knee in the radial disk profile, NGC 6946, NGC 3938,
NGC 3184, NGC 3423, and two
of them have also a Freeman II profile, NGC 3184, NGC 3423.
For these two galaxies, we have tried to fit the disk trend using Eq. (2),
, with 
, and n=3.
For the other two galaxies, we
have noticed that a single exponential fit is not adequate for the
whole disk, and that the calculated values for the radial scalelengths
vary with the radial range considered, as well as the central disk
surface brightness. If only the disk is considered, these results
seem to agree with the prediction of Phillips et al. (DDP model). In
their paper these authors present a model of dust distribution, called
"Triplex model", with the following assumptions:
The physical consequences of this model are threefold, (Davies 1990):
), because light is absorbed in the optical
region and is re-emitted in the infrared (this value has yet been
found by IRAS for nearby galaxies, see Davies 1990 and references
therein).
In this view the difference between Freeman I and
Freeman II profiles is explained in terms of a dimming in disk
luminosity, caused by dust and by a strong emission from a massive bulge
that compensates for this dimming (Freeman I), or from a little bulge,
unable to compensate for the dimming, so that the radial surface brightness
profile shows a drop in
the bulge-disk transition zone (Freeman II).
The observed profiles, however,
disagree with model predictions when the bulges are taken into account,
and a total fit bulge + disk is superimposed to data, as it can be
seen from Figs. 5 (click here) and 7 (click here). We have also noticed that radial disk
scalelengths vary with wavelength; in fact it is larger in
blue than in infrared for NGC 1232, NGC 6946, while the opposite trend
takes place for the other galaxies. The first result agrees with the
predictions of the DDP model, but it is also in good agreement with the
interpretation of RD. However it can be seen from the tables that a
trend with 
similar to the RD one is observed only for
two peculiar galaxies, and that the smoothest increase of the radial disk
scalelength with the wavelength is found for two other peculiar
galaxies, NGC 4321 and NGC 5194, while a pronounced increase of 
with 
is observed for NGC 3184, NGC 3423, NGC 3938.
This result for NGC 3423 is in disagreement with that found by RD.
The radial profile of the color index
(B-V) of NGC 3184 shows a trend toward blue for small values of radius; this
has been interpreted by Larson & Tinsley (L&T78) as an infall of gas
toward the galactic nuclei. As this phenomenon is not present in all
galaxies, they assume that a galactic wind blows up the infall in some
cases. When this occurs, the (B-V) index is red toward the galaxy centre,
i.e. like the one of NGC 1232.
Another interpretation is given by RD; according to these authors there is a
radial inflow of gas toward the galaxy centre, driven by a bar. This
inflow can also be responsible for the Freeman II profile. Radial
gradients have been found in the colour indexes (B-I) and (B-V);
according to the Larson model (Larson 1976), these are due to a
decrease in age and metallicity toward the edge of galaxy. A peculiar
behaviour is observed in the colour indexes (B-I) and (B-V) of NGC 4321
(M 100) and of NGC 5194 (M 51); they show a strong red peak corresponding
to a drop in luminosity, which occurs in the radial surface
brightness profiles, in particular in the B one. This reddening,
corresponding to the dimming in luminosity profiles, confirms,
according to Prieto, the presence of a huge amount of dust in the
bulge to disk transition zone. It has not been possible to apply the
same consideration to the same zone in NGC 3184, where the colour
index is flat in correspondence of the flat zone in the luminosity
profile. The two peculiar galaxies, M 51 and NGC 4321, in particular M 51,
show a real irregular luminosity profile, so the decomposition into bulge
and disk
components is less reliable than that obtained for the other galaxies.
There is a 1 : 2 scaling ratio between the (B-V) and the (B-I) index for
all the galaxies in our sample.
We also want to point out that our choice for the fitting formulas for
bulge and disk is not unique; in the work by Vila Costas & Edmunds
(1992) the definition of the disk
scalelengths is given as that radius at which the disk luminosity is
one half of its central value. This means that Eq. (1) must be
rewritten as: 
.
If one does not specify which fit is used in the work, some
misunderstanding can
arise, and there is the possibility to find values for the disk
scalelengths which differ, from author to author, by a factor of about
two. This phenomenon has already been considered in a work by Knapen
& Van der Kruit (1991).
These authors, starting from these discrepancies, question about the
usefulness of the bulge to disk decomposition and its physical
meaning, but, perhaps all the differences can be explained in term of a
different choice of the fitting formula.
Acknowledgements
The authors wish to thank Dr. F. Pedichini for his precious help in the acquisition of data, Dr. Francesca Natali for the helpful discussion during the data reduction, Dr. P. Monaco for reading the first version of the paper and A. Di Clemente for his help during the observations