2 New observations

Narrow band imaging in the H$\alpha$ emission line ($\lambda$ = 6562.8 Å) of 13 galaxies was obtained in March 20, 2002, using the 2.1 m telescope at San Pedro Martir Observatory (SPM) (Baja California, Mexico).

The target galaxies are listed in Table 3 as follows:

• Column 1: VCC (Binggeli et al. 1985) or CGCG (Zwicky et al. 1961-68) designation.
• Column 2: NGC/IC name.
• Column 3: UGC name.
• Columns 4 and 5: J2000 celestial coordinates.
• Column 6: photographic magnitude as given in the VCC or in the CGCG.
• Column 7: heliocentric velocity (km s^-1) from the VCC or from Gavazzi et al. (1999a).
• Column 8: exposure times in minutes for the ON-band filter.
• Column 9: transmissivity ( $R_{{\rm ON}}$) of the ON-band filter at the redshifted H$\alpha$ line.

We used a Site 1024 $\times$ 1024 pixels CCD detector with pixel size of 0.31 arcsec. Each galaxy was observed through a narrow band interferometric filter ($\sim$90 Å width) centered at $\lambda$ 6603, for the galaxies at the redshift of Virgo ( $350<V<3000~{\rm km~s}^{-1}$), and at $\lambda$ 6723 Å, for galaxies in the Coma supercluster. These observations provided us with the ON-band images and required 15-20 min integration time. The OFF-band images were obtained through the r-Gunn filter and were exposed one fifth of the ON-band ones. The observations were obtained with the seeing ranging from 1.2 to 3 arcsec, but in photometric conditions. They were flux calibrated using the standard stars Feige 34 and Hz44 from the catalogue of Massey et al. (1988), observed every 2 hours. Repeated measurements gave <0.05 mag differences, which we assume as the typical uncertainty ($1 \sigma$) of the photometric results given in this work.

The reduction of the CCD frames follows a procedure identical to the one described in previous papers of this series (e.g. Gavazzi et al. 2002), based on the IRAF STSDAS reduction packages, and it will be briefly summarized here. To remove the detector response each image was bias subtracted and divided by the median of several flat field exposures obtained on empty regions of the twilight sky. Cosmic rays were removed either using the task COSMICRAY in IRAF or manually by direct inspection of the frames. The sky background was determined in each frame in concentric object-free regions around the galaxies and then subtracted from the flat-fielded images. The typical uncertainty on the mean background is estimated 10% of the rms in the individual pixels. This represents the dominant source of error in low S/N regions.

H$\alpha$ fluxes and equivalent widths are estimated subtracting the contribution of the continuum from the ON-band measurements. As the continuum was estimated using the broad band r filter, which in fact includes the H$\alpha$ and [NII] lines, the corrected fluxes and equivalent widths are computed according to Eqs. (1) and (2) of Paper III, and their uncertainties are given by:

$$% {\sigma_{F_{\rm c}} = \sigma_{F}\bigg(1+{\frac{\int R_{{\rm... ...d}\lambda}{\int R_{{\rm OFF}}(\lambda){\rm d}\lambda}}\bigg) }$$ (1)

$$% {\sigma_{EW_{\rm c}} = \sigma_{EW} \bigg(1+{\frac{\int R_{{... ...{\int R_{{\rm OFF}}(\lambda){\rm d}\lambda}\right)^2}\right)}.$$ (2)

Galaxies with substantial $\rm H\alpha +[NII]$ structure are given in Fig. 14. The contours of the OFF frames are superposed to the NET (ON-OFF) frames (grey-scale).