2 Observations

The observations used in this work have been gathered at three different instruments:

1.

High resolution frames were obtained for 5 objects at the CFHT, in 1989-90 by J.L. Nieto and 1992 by one of us. The so called HRCam was used, with a scale of 0.107 arcsec/pixel. The total field was only 641 $\times$ 1011 pixels. Note however that the data for NGC 3115 were obtained in 1989 with the Cassegrain camera at the same scale and field. With the HRCam, a guiding star has to be located at a specific place in the field, so that the studied object is often far from centred, and the usable galaxian range may be rather limited: the range of measurable majA values is 30- 50 arcsec. Unfortunately, only the B and R colours are available, with R according to Cousins. Stellar images give a mean FWHM of $0.66\hbox{$^{\prime\prime}$ }$ in B and $0.57\hbox{$^{\prime\prime}$ }$ in R.

2.

Obervations of fair resolution were obtained by one of us for 6 more objects, at the 2 m Bernard Lyot telescope of the Pic du Midi (TBL). With the catadioptric focal reducer, the scale was 0.242 arcsec/pixel, and the usable field 951 $\times$ 1001 pixels. Frames in B, V, R and I were generally obtained. Some problems with the optical adjustements of the camera occured in 1994, so that the PSF is seen to vary significantly in the field, while the PSF's in I are rather poor, as compared to the other bands. Several frames from this run are affected by geometrical errors (of electronic origin), involving some limitations in their possible uses. The mean FWHM's from stellar images are $1.15\hbox{$^{\prime\prime}$ }$ in B, $1.05\hbox{$^{\prime\prime}$ }$ in V, $1.03\hbox{$^{\prime\prime}$ }$ in R and $1.47\hbox{$^{\prime\prime}$ }$ in I for the main run of 1994, while the observations of NGC 7332 are less good.

3.

The above material was completed by frames from the series obtained at the 120 cm of the Observatoire de Haute-Provence (OHP) by J. Marchal and one of us. In 1990-91, the available scale was 0.84 arcsec/pixel, and the usable field $321 \times 511$ pixels, or about 4 $\times$ 7 arcmin. Later a camera with 0.76 arcsec/pixel and a field of about $7\times 7$ arcmin became available, and was used in our 1994 observing run. Frames in U, B, V, R and I were obtained, but part of the U images were discarded due to unsufficient S/N. The seeing at the OHP is generally in the 2-3 arcsec but may become still worse!