Using the bright galaxy counts from the 2 Micron All‐sky Survey (2MASS) second incremental release, two techniques for probing large‐scale structure at distances of ∼150 h−1 Mpc are investigated. First, we study the counts from two sets of six 5°×∼80° strips in the two galactic caps. In the six southern strips a deficit of ∼30 per cent was found relative to a predicted homogeneous distribution at Ks∼ 11. These strips were also in good agreement with a model incorporating the radial density function of the southern 2dF Galaxy Redshift Survey (2dFGRS), which shows a deep underdensity between ∼90 and 180 h−1 Mpc. These results indicate a very large ‘local hole’ in the Southern Galactic Cap (SGC) to ≳150 h−1 Mpc with a linear size across the sky of ∼200 h−1 Mpc, but with a significantly lower mean underdensity of ∼30 per cent compared with that suggested by the bright APM SGC counts. The counts in the northern set of strips are low overall but indicate a more varied pattern. When all the available 2MASS data with |b| > 30° were aggregated, they indicated underdensities of ∼18 and ∼30 per cent at Ks∼ 11 for the northern and southern areas, respectively. Our second method uses the ratio of the counts between 11.38 < Ks < 12.38 and 12.88 < Ks < 13.38 to the ratio of the corresponding counts of the homogeneous model binned in 25 deg2 areas; the counts from these areas provide a smoothed map over the sky of the slope in the counts between Ks= 11.38 and 13.38. Visually, the resulting map shows the expected complex form of the cosmic web and picks out known rich clusters, demonstrating the usefulness of this ‘slope statistic’ as a probe of large‐scale structure at distances of ≲150 h−1 Mpc. Most interestingly, the map also shows large regions, ∼100° across, of steep counts in both hemispheres. Thus, the present 2MASS data suggest the presence of a potentially huge contiguous void stretching from south to north. Not only would this delineate further the limits for the cosmological principle but it would also show the possible presence of significant power on scales of ≳300 h−1 Mpc in the galaxy power spectrum.

We present a new CCD survey of bright galaxies predominantly within the Northern and Southern strips of the 2dF Galaxy Redshift Survey (2dFGRS) areas. We use the new CCD data to check the photographic photometry scales of the 2dFGRS 100‐k release, APM Bright Galaxy Catalogue, APM–Stromlo Redshift Survey, Durham–UKST (DUKST) Survey, Millennium Galaxy Catalogue (MGC) and Sloan Digital Sky Survey (SDSS). We find evidence for scale and zero‐point errors in the 2dFGRS 100‐k release Northern field, DUKST and APM data of 0.10, 0.24 and 0.31 mag, respectively; we find excellent agreement with the MGC and SDSS photometry. We use our new CCD data to correct the photographic photometry, we then compare the CCD number counts in both the Northern and Southern survey areas. We find conclusive evidence that the Southern counts with B < 17 mag are down by ≈30 per cent relative to both the Northern counts and to the models of Metcalfe et al. in the same magnitude range. We further compare the number redshift distributions from the B < 17 mag Durham–UKST and B < 19.5 2dFGRS redshift surveys using the corrected photometry. While the Northern n(z) from 2dFGRS appears relatively homogeneous over its whole range, the Southern n(z) shows a 30 per cent deficiency out to z= 0.1; at higher redshifts it agrees much better with the Northern n(z) and the homogeneous model n(z). The Durham–UKST n(z) shows that the Southern ‘hole’ extends over a 20 × 75 deg2 area. The troughs with z < 0.1 in the Durham–UKST n(z) appear deeper than for the fainter 2dFGRS data. This effect appears to be real since the troughs also appear to deepen in the 2dFGRS data when magnitude‐limited at B < 17 mag and so this may be evidence that the local galaxy distribution is biased on ≳50 h−1 Mpc scales which is unexpected in a ΛCDM cosmology. Finally, since the Southern local void may persist over the full area of the APM and APM Bright Galaxy Catalogue with a ≈25 per cent deficiency in the counts below B≈ 17, this would mean that its extent is approximately 300 h−1 Mpc × 300 h−1 Mpc on the sky as well as ≈300 h−1 Mpc in the redshift direction. Such a 25 per cent deficiency extending over ≈107h−3 Mpc3 may imply that the power‐law behaviour of the galaxy correlation function extends to ≈150 h−1 Mpc with no break and shows more excess large‐scale power than detected in the 2dFGRS correlation function or expected in the ΛCDM cosmology.

Observations in the submillimetre (submm) waveband have recently revealed a new population of luminous sources. These are proposed to lie at high redshift and to be optically faint because of their high intrinsic dust obscuration. The presence of dust has been previously invoked in optical galaxy count models which use the Bruzual & Charlot evolution models with an exponential τ=9 Gyr star formation rate (SFR) for spirals, and these fit the count data well from U to K. We now show that by using either a 1/λ or Calzetti absorption law for the dust and re‐distributing the evolved spiral galaxy ultraviolet (UV) radiation into the far‐infrared (FIR), these models can account for all of the ‘faint’ (1 mJy) 850‐μm galaxy counts, but fail to fit ‘bright’ (2 mJy) sources, indicating that another explanation for the submm counts may apply at brighter fluxes, e.g., quasi‐stellar objects (QSOs) or ultraluminous infrared galaxies (ULIRGs). We find that the main contribution to the faint, submm number counts is in the redshift range 0.5