CORRELATIONS OF GALAXIES AT Z=0.3

The angular correlation functions for galaxies with magnitudes 20 < b(J) < 21.5 and 21 < b(J) < 22.5 are measured from photographic images of nine distinct fields. New expressions for the variance of correlation function estimates are used, which show that the nine fields are consistent with a single angular correlation function w(theta) and with previous estimates of its amplitude and slope-though the errors on previously published w(theta) values are probably underestimated. Assuming a spatial correlation function of the form xi(r, z) = (r/r0)-gamma(1 + z)-3-epsilon, and using the measured N(z) for these magnitude ranges, we derive xi(r = 250 h-1 kpc, z = 0.18) = 48 +/- 10, and xi(r = 250 h-1 kpc, z = 0.27) = 55(+11; -9) for galaxies in these samples, largely independent of gamma and epsilon (1 sigma errors). This is approximately 2 times lower than the small-scale clustering seen in nearby galaxies at b(J) < 18 (after allowing for modest growth in clustering since z = 0.27), yet very similar to the clustering of the nearby IRAS-selected galaxy population. The correlations of the reddest third of the b(J) approximately 22 galaxies, however, are larger and consistent with the b(J) < 18 clustering, not the IRAS clustering, for a clustering growth rate of epsilon greater-than-or-equal-to -1. Thus to b(J) < 22.5, most of the apparent evolution in clustering of blue-selected galaxies appears to be due to an increasing fraction of late-type or star-forming galaxies at fainter magnitudes, combined with weaker small-scale clustering for late types. These conclusions are entirely empirical and independent of galaxy evolution models.