The impact of GHRS on atomic physics - A fruitful collaboration between laboratory and stellar spectroscopy

High-quality spectra of various cosmic sources obtained with the GHRS require accurate atomic data for a detailed analysis. They have also shown the need for data for nearly all elements in the periodic table, at least for the lower ionization stages. These requirements have initiated new experimental and theoretical work in laboratory astrophysics in order to improve the atomic data base for stellar spectroscopy in the satellite-ultraviolet region. Analyses of astrophysical emission-line spectra have demonstrated the importance of a detailed knowledge about the atomic structure of complex atoms or ions, represented, e.g., by the iron-group elements. We give in this paper examples of old and new spectroscopic problems in the ultraviolet, e.g. the B II resonance line, and explain in view of the atomic structure why we may still find numerous unidentified features in a GHRS spectrum. We also show how stellar spectra can be used to obtain information about highly excited energy levels in complex atoms. We discuss possible excitation mechanisms in emission-line spectra, in particular a pair of strong lines in eta Carinae, where fluorescence might be enhanced by stimulated emission. This implies the discovery of a natural ultraviolet cosmic laser.