2 Follow-up spectroscopy

We observed all the candidates from the "List 2" during one observing run, between 20-24 May 1996. The observations were carried out with the CAFOS (The Calar Alto Faint Object Spectrograph) at the 2.2m telescope at the German-Spanish Observatory at Calar Alto (Almeria, Spain). The details of the observations are given in Table 1. The observations from May 1996 were done in excellent weather conditions (the photometry being better than 5$\%$) and therefore the campaign was not dedicated to the follow-up spectroscopy of the new candidates but mainly to the spectrophotometry of a bigger sample of emission-line objects, including objects from Paper I. Because the main goal was accurate flux determination, we used a 4$^{\prime\prime}$ slit widths, which on the the other hand increased the error in the redshift determination of the new candidates. Thus the error in redshift of the objects presented in this paper is bigger than the error of the objects from Paper I.

  

Table 1: The details of the spectroscopic observations from May 1996

The data were reduced using the MIDAS routines. The frames were biased and flatfield-corrected. For the extraction of the 1-dimensional spectra from the 2-dimensional data, the optimal extraction algorithm of Horne (1986) was used. The spectra were rebinned to a linear wavelength scale using a third or fifth order polynomial fitted to the dispersion curve of the comparison spectra. A flux calibration was applied, including the correction for the atmospheric extinction using the mean extinction values for Calar Alto. Finally the wavelength scale was checked by comparison with the night-sky lines. A more detailed description of the reduction procedure is given by Stickel et al. (1993).

Once fully reduced, the emission lines in each spectrum were measured by fitting a Gaussian Curve. The quoted redshifts were derived as means of the redshifts determined from the individual strong lines, and the errors of the redshifts were calculated as error of the mean. The observed redshifts were further corrected for the motion of the Earth and transformed in heliocentric redshifts. The internal errors from the May 1996 campaign are around $\Delta z=0.0002$. In order to estimate the external errors due to the 4$^{\prime\prime}$ slit width we compared the redshifts of some galaxies observed also in a previous campaign (see Paper I) with a 2$^{\prime\prime}$ slit width. We obtained a total error of $\Delta z=0.00035$. For the few galaxies with absorption lines the errors are larger, up to $\Delta z=0.0007$.