GALAXY NUMBER COUNTS .3. DEEP CCD OBSERVATIONS TO B=27.5 MAG

By combining a 26-h exposure taken with the prime-focus RCA CCD camera on the 2.5-m Isaac Newton Telescope and a 13-h exposure taken with the Tektronix CCD camera at the Cassegrain auxiliary focus of the 4.2-m William Herschel Telescope we have extended our determination of the form of the galaxy number-magnitude count relation on one CCD field to a blue magnitude limit of B-ccd similar to 27.5 mag. These data are deeper than any previously published B-band counts. We investigate the reliability and completeness of our data by subjecting simulated CCD frames to the same data analysis procedures as the real data. We see a clear evidence that the slope of the log N-m relation is decreasing faintward of B similar to 25. However, the counts are still rising at the limit of our data. We show that, for high q(0), it is not possible to reconcile the slope and numbers of galaxies at B > 25 with the slope of the local faint galaxy luminosity function unless this was steeper in the past, or density evolution has taken place. The required luminosity function slope is close to the steep slope detected for low-redshift, late-type galaxies. The presence of a change in the number-count slope is consistent with the idea that a redshift limit, or at least a redshift where the cosmological volume element ceases to rise sharply with redshift, has been reached in the galaxy distribution. We show that this hypothesis is supported by an observed flattening seen in the galaxy clustering amplitude at similar magnitudes. Low-q(0), pure luminosity models are still consistent with the count data, provided that the count models have a high normalization appropriate to the bright data at B similar to 18.