3 Spectroscopic observations

Two-dimensional (long-slit) spectra were obtained using the OMR (Optomechanics Research Inc.) spectrograph mounted at the Cassegrain focus of the 2.16 m telescope at the XingLong Station of the Beijing Astronomical Observatory (BAO) in China, which also belongs to the National Astronomical Observatories, Chinese Academy of Sciences. The data were acquired during 26 predominantly moonless nights between March 1997 and January 2002. The vast majority of the observations were obtained during transparent or photometric conditions, the remainder through thin clouds.

The OMR spectrograph was equipped with a TEKTRONIX TEK1024, AR-coated back-illuminated CCD with $1024 \times 1024$, 24 $\mu$m pixels. A 300 line mm^-1 grating blazed at 5500 Å in first order was used, giving a dispersion of 4.8 Å pixel^-1 and a total spectral coverage of $\sim$4500 Å. According to the condition of seeing, we adjusted the slit width in between 2 $^{\prime\prime}$ and 3 $^{\prime\prime}$ on each night and used the selected configuration for all target spectra. Our spectra were centered on 5500 Å or 6000 Å (see Table 2), at a resolution (FWHM) of $\sim$10 Å, thus covering many interesting spectral features ranging from [O  II]$\lambda$3727 in the blue to [S  II]$\lambda$6731 in the red.

The slit was generally placed across the nucleus of each galaxy. In some cases where the location of the nucleus was not obvious, the slit was aligned to intersect the brightest part of the galaxy in order to maximize the chance of detecting the nucleus. Otherwise, the spectrograph was rotated about the optical axis of the telescope as required to keep the slit approximately perpendicular to the horizon, to minimize loss of light due to differential atmospheric refraction (e.g., Filippenko 1982).

Dome flats, illuminated by a hot, spectroscopically featureless arc lamp, were taken at the beginning and end of each night. He-Ar or Fe-Ar (before 1998, see Col. 6 of Table 2) lamps were observed immediately after observing each target object at the same position as the object for wavelength calibration.

Two or more of the KPNO spectrophotometric standards from Massey et al. (1988), such as G191B2B, Hiltner 600, Feige 34, HZ 44, BD +28$^{\circ}$4211, were observed each night for absolute photometric calibration. In good conditions, a standard was observed about every 3 or 4 hours. Table 2 lists the observations log and instrumental parameters for each run. Column 1 lists the observation run number, Col. 2 the observation period, and Col. 3 the number of nights actually observed, i.e. that were not lost because of poor weather conditions. Column 4 lists the centering wavelength used for each run. Columns 5-8 list the slit width, the arc lamp used, the number of spectral obtained, and the standard stars for each run. The individual integration times in each of the observations varied from 1200 s to about an hour, depending on the magnitude of the object and the seeing. For 39 objects, observations were performed twice in order to get a higher signal-to-noise ratio spectrum, so that we have observed a total 136 BCGs spectra. 107 of the total 136 spectra were observed at airmass less than 1.2, only 4 spectra were observed at airmass larger than 1.5. Table 3 lists these observation parameters. Column 1 lists the object name, while Cols. 2-4 and 5-7 list the observation date, exposure time, and airmass.