2005ApJ...621..246B

There are two main theories for the formation of S0 galaxies: that S0 and elliptical galaxies are formed in the same manner ab initio, or that S0s are mainly spiral galaxies from which the gas has been stripped by interactions with either hot cluster gas or other galaxies. Perhaps the best way to test these formation ideas is to calculate the luminosities of S0 and spiral galaxies in the K band, where dust effects for the spiral galaxies are minimal. If S0s are stripped spirals whose star formation has been quenched, we expect from existing models that S0s should be ∼0.75 mag less luminous in K absolute magnitudes than early-type spirals. Using the nearly 25,000 galaxies in the Two Micron All Sky Survey that have total K magnitudes ≤11.25, we are able to select over 19,000 of these galaxies for analysis. Of these, 10,977 are in the Third Reference Catalogue of Bright Galaxies and 8230 are not. We separate galaxies by both Hubble type and whether they are in or out of nearby clusters and near or far from us. K-band absolute magnitudes are predicted using the velocity field model of Faber & Burstein for galaxies with a velocity of 6000 km/s or less with respect to the cosmic background radiation. We present histograms, medians, and means for galaxies of all Hubble types in absolute K magnitudes. In all cases, we find that giant ellipticals are more luminous than S0s, and that S0s are at least as luminous, if not more luminous, in the K band than are spiral galaxies. The K absolute magnitudes of the S0s compared to those of the spirals show that S0s are not gas-stripped spiral galaxies. While a few S0s can be gas-stripped spirals, as a class they are not. We suggest that the E+A galaxies seen in clusters at moderate redshift are S0s onto which hot gas has cooled, producing a temporary sprinkling of star formation.