We have analyzed the UV properties of a sample of the 14 NLS1 galaxies observed with the IUE satellite. This sample comprises some unpublished observations of our own together with archival data. When the available number of spectra and their S/N allowed for it, we have studied the UV variability properties of the objects. Two objects (Mrk 1044 and IRAS13224-3809) have been found to experience significant variations in the continuum while the emission line wings remain constant.
We have studied the continuum SED of the NLS1 sample, from the FIR to the soft X-rays, comparing it with the SED of a sample of "normal" Seyfert 1 galaxies from Walter & Fink (1993). The FIR properties of NLS1 are not different from the FIR properties of "normal" Seyfert 1 galaxies with broad optical permitted lines. Only the UV luminosities tend to be smaller in NLS1, and the soft X rays spectra tend to be steeper, although a number of NLS1 are found to have spectral indexes very similar to those found in "normal" Seyfert 1 galaxies.
The analysis of the UV emission line spectra of the sample, has revealed that all NLS1 galaxies detected with IUE at sufficiently high S/N show broad wings in their emission lines (Ly , CIV and HeII). The positive detection of these broad wings demonstrates the existence of gas with a large velocity component in our line of sight, ruling out the hypotheses that (a) there is no BLR in NLS1, or (b) there is a BLR, but we cannot see it due to orientation effects. It is also difficult to accept that the broad wings we see in the UV emission lines result from a scattering process, since the optical counterparts are not detected even in polarized light and since the NLS1 continuum SED does not suggest a hidden AGN as in Seyfert 2 galaxies.
We have discussed how the absence of broad optical lines can be explained if the partially ionized zone in the high velocity line emitting gas is missing. The BLR would be made of fully ionized hydrogen gas in optically thin clouds similar to those proposed by Shields et al. (1995). In this way only the high ionization lines would be produced. Some of the implications of a BLR formed by highly ionized gas are discussed. In particular, the covering factor needed to account for the observed equivalent width of the emission lines should be very large.
© European Southern Observatory (ESO) 1997
Online publication: April 8, 1998