Introduction

Many authors have found evidence of correlations between galaxies (or clusters of galaxies) and background QSOs (e.g. Tyson 1986; Fugmann 1988; Bartelmann & Schneider 1993; Rodrigues-Williams & Hogan 1994; Wu & Han 1995). A natural explanation is that these associations are due to gravitational lensing, which can provide either a positive or negative correlation between foreground lenses and a flux limited background population, depending on the form of the background number count faintward of the sample flux limit. For a number count of the form $N(<m)\propto10^{\alpha m}$ a steep ( $\alpha>0.4$ ) slope causes a positive correlation, while a flat ( $\alpha<0.4$ ) slope causes a negative correlation. QSOs show both steep (at $B\lower .75ex \hbox{$\sim$} \llap{\raise .27ex \hbox{$<$}}19.5$ ) and flat (at $B\lower .75ex \hbox{$\sim$} \llap{\raise .27ex \hbox{$>$}}19.5$ ) number count slopes. A problem with most QSO-galaxy correlation measurements is that the amplitude of the correlation (both positive and negative) is a factor of $\sim2-5$ greater than expected from lensing in an $\Omega_0 =1$ Universe (e.g. Williams & Irwin 1998; Croom & Shanks 1999). An alternative is that patchy dust extinction in our own galaxy could cause some of the observed positive correlation between QSOs and galaxies. However, negative correlations require inter-galactic dust associated with the foreground lenses, making a dust explanation appear rather contrived.

Here we use data from the Issac Newton Telescope Wide-Field Camera (INTWFC) to cross-correlate galaxies with faint QSOs. A more in depth discussion of this work will be given in Croom & Shanks (2001), which will also include further modelling of the cross-correlation results. The QSOs in this study were taken from a number of deep optical and X-ray surveys (Boyle, Jones & Shanks 1991; Almaini 1996; Crampton, Cowley & Hartwick 1989; Koo, Kron & Cudworth 1986; Boyle et al. 1990; McHardy et al. 1998), the main aim being to have a large number of QSOs within the field of view of the INTWFC. The redshift distribution of the QSOs is shown in Fig. 1.

**Figure 1:** The redshift distribution of QSOs used in our analysis (solid line), compared to model redshift distributions of galaxies with B<23 mag (dotted line) and B<26 mag (dashed line). The normalization of the galaxy N(z) distributions is arbitrary.
$\begin{figure} \centering\centerline{\psfig{file=fig1.ps,width=7.0cm}}\end{figure}$

We observed a total area of 2.5deg² in the B-band with the INTWFC, containing a total of 192 QSOs. Galaxy catalogues were produced using SEXTRACTOR (Bertin & Arnouts 1996). Typical $5\sigma$ detection limits were equivalent to an isophotal magnitude of $B_{\rm ccd}=27-27.5$ . The total magnitude at which all objects are detected at $\geq3\sigma$ is typically $B_{\rm ccd}=25.5-26$ , which forms our completeness limit.