We present a preliminary measurement of the angular clustering of faint ( R \leq 25 ) field galaxies in which we concentrate on the behavior of \omega ( \theta ) on small angular scales ( \theta \mathrel { \raise 1.505 pt \hbox { $ \scriptstyle < $ } \kern - 6.0 pt \lower 1.72 pt% \hbox { { $ \scriptstyle \sim$ } } } 10 ^ { \prime \prime } ) .
The galaxies are strongly clustered and \omega ( \theta ) is well-characterized by a power law of the form A _ { \omega } \theta ^ { - \delta } .
The best-fitting value of the power law index , \delta , is , however , steeper than the fiducial value of \delta = 0.8 , indicating that there are more pairs of galaxies separated by \theta \mathrel { \raise 1.505 pt \hbox { $ \scriptstyle < $ } \kern - 6.0 pt \lower 1.72 pt% \hbox { { $ \scriptstyle \sim$ } } } 10 ^ { \prime \prime } in our sample than would be otherwise expected .
Using the best-fitting form of \omega ( \theta ) , we estimate that \sim 10 \% of the galaxies are in physically close pairs ( separations \mathrel { \raise 1.505 pt \hbox { $ \scriptstyle < $ } \kern - 6.0 pt \lower 1.72 pt \hbox { { $% \scriptstyle \sim$ } } } 21 h ^ { -1 } kpc ) .
This is a factor of order 2 larger than local galaxy samples but comparable to galaxy samples with \left < z \right > \sim 0.4 .
The mean redshift of our galaxies is of order 0.95 , and , therefore , our result suggests that there was little or no evolution in the merger rate of galaxies between z \sim 1 and z \sim 0.4 .