2001ApJ...559..201M

Imaging spectroscopy using the new Palomar Integral Field Spectrograph is presented for the Paα line in four ultraluminous infrared galaxies. The resulting integral field data cubes reveal line emission possessing a wide range of complex spatial morphologies, often quite different from the appearance of the continuum. The velocity fields are equally diverse in nature, often failing to resemble typical modes of galactic motion. We see a variety of interesting phenomena in the individual mergers including star formation rates of 2-5 M_☉ yr^–1 in young tidal tails; a postencounter disk that obeys the Tully-Fisher relation; a large-scale emission-line nebula possibly associated with a massive outflow; an apparently single merging system possessing two distinct kinematical axes belying the presence of a second galaxy, mostly obscured by its merging companion; and possible formation of tidal dwarf galaxies. In most cases, we are able to establish the geometry of the merger and thus estimate the time in the merger process at which we are viewing the system. The resulting range in estimated ages, some of which are very young encounters (∼5x10⁷ yr), is not predicted by merger models, which produce high rates of star formation either 1-2x10⁸ yr after the first encounter or very late (∼10⁹ yr) in the merger process. Even in the very young mergers, despite a sample selection based on extended line emission, the ultraluminous activity appears to be centrally concentrated on the nucleus of one of the progenitor galaxies–namely, the galaxy with a prograde orbital geometry. The inferred extinction to these concentrations is high, usually at least 1 mag at the wavelength of Paα. The presence of a significant population of very young ultraluminous mergers, together with the majority of ultraluminous infrared galaxies existing in the final stages of merger activity, indicates that the ultraluminous galaxy phase is at least bimodal in time. An evolutionary scenario is proposed for ultraluminous infrared galaxies, wherein the far-infrared luminosity may undergo multiple ultraluminous bursts during the course of the encounter. A substantial fraction of the merger lifetime may be spent in a phase identified with the less powerful luminous infrared galaxy class.