Spectroscopic observations were made with the Canada-France-Hawaii Telescope (CFHT) using the Subarcsecond Imaging Spectrograph (SIS; Le Févre et al. 1994) in August 1997. The Stis2
CCD with 21
m pixel and the R150 grating were used and the resultant dispersion is 2.88 Å per pixel. The wavelength coverage extends from 4500 to 9000 Å. Two multi-slit masks were used to obtain spectra of candidate cluster galaxies in the field. Each slit has 7.8-arcsec-long and 0.78-arcsec-wide aperture, which gives
18 Å instrumental resolution. MG 2016+112 B, B1, and D were observed within one slit of the first mask (Mask1) and C with a slit in the second one (Mask2). In total, 6 and 5.8 hours exposures were obtained for Mask1 and Mask2, respectively.
The data was pre-reduced using standard IRAF tasks. We then follow the reduction procedure of multi-slit data described in Le Févre et al. (1995). Wavelength calibration was done with the arc-line spectra taken at the observation. Typical internal error of wavelength determination is
0.3 Å. The flux calibration was done using the spectroscopic standard star GD248 (Oke 1990). Note that the flux calibration becomes more uncertain above 7500-8000 Å because no order-separating filter was used and the flux is contaminated by the UV and blue contribution of the second order filter. We tried to carefully take into account this effect but flux calibration may remain uncertain above 8000 Å.
Figures 1 and 2 show the obtained spectra of image B and C smoothed with a 5
pixel Gaussian filter. Strong UV emission lines of Ly,
NV
Å, CIV
Å, HeII
Å, and CIII]
Å are seen in both of the spectra. SiIV and OV lines at 1400 Å are clearly seen in the B spectrum but only marginal in C. The redshift of C is the same as B within the uncertainties. There is no sign of contamination by a system at a different redshift.
Copyright ESO 2001