DEEP HUBBLE-SPACE-TELESCOPE GALAXY AND PAIR COUNTS AS TESTS OF MERGERHISTORY

We consider the use of counts of faint galaxy pairs as tests of the ro le of merging in galaxy evolution, using both simulations and Hubble S pace Telescope (HST) data in comparison with models incorporating a ra nge in the evolution of merger rate. Artificial data, including simula tion of the surface-brightness selection effects implicit in HST image ry, indicate that simple pairing statistics ale robust enough for this test. However, the differences between pair fractions for various pla usible scenarios of merger-rate evolution are small enough to be easil y hidden by statistical errors or environmental differences. With a Po isson-based algorithm, we identify pairs with physical distances betwe en components less than or similar to 60 h(-1) kpc. Deep HST WFPC2 ima ges of the held around the radio galaxy 53W002 at z = 2.4 in F814W (I) and F606W (V) show that 12%-20% (similar to 10 pairs) of galaxies at m(I) less than or similar to 24 or m(V) less than or similar to 24.6 a re paired. This supports a mild enhancement of pair fraction toward hi gh redshifts (less than or similar to 1-2), as concluded in other rece nt studies of faint galaxy pairs. The magnitude limits eliminate most possible members of the grouping associated with 53W002 itself. We com pare the counts of galaxies and galaxy pairs in the 53W002 held and ot her recent studies to a detailed model. incorporating overall ''tradit ional'' galaxy spectral evolution and various merger rates, in conjunc tion with our previous evaluation of the local merger sate. The data s upport evolution of the merger rate, such that the comoving density of merging systems varies as (1 + z)(alpha), with 0 less than or equal t o alpha less than or equal to 2. This is consistent with the constrain t alpha < 3, beyond which local elliptical galaxies (E/S0 types) are o verproduced from merging, and includes the case alpha approximate to 2 required by our model to support the ''merger hypothesis,'' whereby a lmost all elliptical galaxies are merger remnants. The alpha approxima te to 2 case also expects a large fraction of early galaxies undergoin g merging. Therefore, we support the merger-rate evolution required to support the ''merger hypothesis'' and the expectation that early merg ing is important in the overall evolution of galaxies. Assuming merger remnants to be elliptical galaxies subsequently undergoing passive ev olution, our model can reproduce at once counts of faint galaxies and galaxy pairs, as well as the present-day luminosity function of ellipt icals, for the Hubble constant H-0 approximate to 50 km s(-1) Mpc(-1) and a cosmological constant Lambda = 0. These lines of evidence do not appear to offer compelling arguments for more exotic cosmologies or s tar formation histories. A preliminary study shows the fraction of pai red galaxies in the 53W002 field to be much larger than that in the Hu bble Deep Field (5%-6% at most in the HDF), suggesting physical held-t o-held fluctuations in pair statistics. Further progress to constrain merging history requires not only larger samples, preferably coupled t o at least partial redshift information, but better indications as to which faint objects in multiple systems should be considered as indepe ndent galaxies and which are parts of single dynamical units.