About 110 emission lines are present in the SW spectrum of RR Tel, shortward
of 1900 Å (several of them being present in the 1983 long exposure
spectrum SWP 20246 only).
For all the unsaturated lines the FWHM values (in Å) and the reddening
corrected line emission intensities (in erg 
) are given in
Table 1.
Apparently there is no indication of any short-term modulation in the emission
intensities, which show only a secular decline (see also Sect. 6). For this
reason in Table 1 we have listed only one value per year for the emission
intensity.
Most identifications are from the paper by Penston et al. (1983) but they have been checked also against the lists of ADB and DF, and improved using recent lists of emission lines such as those of Kelly (1987), Morton (1991), Pradhan & Peng (1995), and Williams (1995). We have changed some previous identifications and provided a tentative identification for some previously unidentified lines using our personal judgment but keeping in mind that, to a large extent, line identifications is more of an art than a science. In Table 1 we have marked with a bold character the previously unidentified lines that have been given an identification.
We recall that in RR Tel some information on the identification of a specific line comes from the line width itself because, as pointed out by Thackeray (1977) and Penston et al. (1983), there is a positive correlation between the line ionization level and the line width, as measured from the FWHM; this has helped us in some dubious cases. In this respect we point out that no appreciable variations in the line FWHMs are evident from our data for the 1978-1983 time interval. It is also worth pointing out that the doublets of NIV] near 1485 Å and of SiIII] near 1888 Å both show a rather wide additional component in the emission profile with FWZI larger than 1.5 Å. This component is not evident in other lines of comparable ionization, e.g. in the SIV] and OIV] emissions near 1400 Å.
Incidentally, we want also to point out that, in our opinion, the very wide
and shallow emission features (FWZI about 12 Å) underlying the strong lines of
CIV 
1550 and HeII 1640 (and partially NV
1240) in the SWP 20246 long exposure
spectrum are real and
not an artifact associated with the severe overexposure in the line center.
Such profiles are common
in the high resolution spectra of CVs (Robinson et al. 1995;
Friedjung et al. 1997) and their origin is commonly associated
with material rotating at high velocity in the inner part of an accretion disk.
In the SWP 20246 spectrum we have found some additional (rather weak) features with respect to those reported by DB and by ADB. Since these features are not present in other spectra we have listed them separately in Table 3 together with tentative identifications and some notes.
Figure 1: The emission intensity variations from 1978 to 1993 for lines of
different excitation
Figure 2: The changes in the MgV line at 1805 Å between 1978 and 1993. Units
in the y-axis are in 