Quasi-stellar objects, ultraluminous infrared galaxies, and mergers

We test the hypothesis that QSOs are formed via strong tidal interactions o r mergers, initially going through an ultraluminous infrared phase. Our app roach is to look for traces of this phase in the host galaxies of QSOs. We select a sample of low-redshift bona Dde QSOs that may be in a transitionar y stage between ultraluminous infrared galaxies (ULIGs) and QSOs. These obj ects, which we call transition QSOs, have an intermediate position in the f ar-infrared color-color diagram between the regions occupied by the two cla sses of objects. We carry out a systematic spectroscopic and imaging study of these objects in order to determine their interaction and star-forming h istories. By modeling the spectra, we obtain ages for the recent starburst events in the host galaxies and interacting companions. We have discussed t he first five objects in the sample in detail in previous publications; her e we present results for the remaining four objects, and discuss the sample as a whole. We find that all nine transition QSOs are undergoing tidal int eractions and that eight are major mergers. Every object also shows strong recent star-forming activity, and in at least eight cases this activity is directly related to the tidal interaction. The ages we derive for the starb urst populations range from currently active star formation in some objects to poststarburst ages less than or similar to 300 Myr in others. There is also a clear connection between interactions, starbursts, and QSO activity. Seven of the QSOs in the sample are also ULIGs; statistical considerations show that the two phenomena are necessarily physically related in these ob jects. Our results imply one of two scenarios : (1) at least some ULIGs evo lve to become classical QSOs, and the transition stage lasts less than or s imilar to 300 Myr; or (2) at least some QSOs are born under the same condit ions as ULIGs, and their lifetime as QSOs lasts less than or similar to 300 Myr. We discuss other properties and trends found in the sample, and propo se a model that accounts for all of them, as well as the youth of these sys tems.