Introduction

Timing solutions have given the position[2] and flux[1] of 15 pulsars in 47 Tuc (all millisecond pulsars), out of the 20 pulsars listed in those papers, and this represents the current level of sensitivity, with all of the pulsars being only visible during periods of interstellar scintillation (5 more are occasionally visible but are not yet located because of their rare appearance or binary orbit). We can expect perhaps 200 more pulsars[1] in 47 Tuc, if the luminosity distribution is similar to the galactic distribution of millisecond pulsars, which is given by

$$dN=L^{-1} d\, \log L$$ (1)

From this, for every lower decade of flux, there will be 10 times more pulsars, up to a cutoff (since the integral of pulsar count diverges for no cutoff). We should expect that they are spatially distributed in the same way, independent of luminosity. Camilo et al[1] gave mean fluxes for the 15 located pulsars, using timing measurements from the 64m Parkes telescope, and the flux totals to 2.06 mJy. We assume that the flux of the 5 other unlocated pulsars are going to be small compared to this, and so if there are 200 pulsars, there will be about 4 mJy in the extended component, and 2 mJy from the as yet undetected pulsars. We suggest that imaging the core of the cluster with the ATCA, with a suitably compact array (for a large synthesised beam), we should be able to see the group of pulsars as a single extended component, and be able to remove the effects of background sources and those bright pulsars that may be in a process of scintillating.