Large-scale fluctuations in the distribution of galaxies

The null hypothesis that the three-dimensional power spectrum measured from the APM Survey is consistent with the one-dimensional power spectrum measured from the pencil-beams surveys of Broadhurst et al. and of Szalay et al. is tested. The external estimates of the mean power that we make are sensitive to details of the model for the survey geometry and to the assumed level of the distortion of the pattern of galaxy clustering caused by peculiar motions. We find that the measured 3D clustering of galaxies can account for the presence of peaks in the one-dimensional power spectrum but is less successful in recovering the detailed appearance of the observations. We find no strong evidence for any additional large-scale structure in the deep pencil beams beyond that recovered from the APM Survey. We conclude that it is unlikely that large-scale structure can be responsible for the steep local number counts of bright galaxies.