1 Introduction

The Italian-Dutch satellite, Beppo-SAX, transformed the field of gamma-ray burst studies by discovering the X-ray afterglow from the gamma-ray burst and determining the position of the burst to arcminute accuracy. The ensuing optical observations then showed conclusively that the bursts originated at large redshifts, and were of extreme luminosity. The predictions by Paczynski & Rhoads (1993); Katz (1994); Mészáros & Rees (1998); and Vietri (1997); have largely been born out by observations. The general decrease in the X-ray intensity has been found to follow a powerlaw of the form $t^{-\alpha}$ where the value of $\alpha$ ranges between 1.1 and 1.9 for the few bursts that have been observed (Costa et al. 1997; Costa et al. 1998; Yoshida et al. 1998). So far the bursts in X-rays have been followed for about 10 days before they fade below the current instrument sensitivities. By following the burst for several months using AXAF, the powerlaw decay can be tested to greater limits. The great advantage of AXAF is the ability to position the gamma-ray burst to arcsecond accuracy within a thousand seconds. This makes the association with a galaxy or any other object virtually a hundred percent clear, since the mean spacing between the faintest galaxies is of the order of several arcseconds. Since AXAF is so sensitive, it is feasible to wait until an optical identification has been made or not. If no optical identification has been made, then AXAF can position the gamma-ray burst and still determine if it is associated with a galaxy. There could be a class of gamma-ray bursts that fade very rapidly in the optical band or are intrinsically fainter because of the surrounding medium (Panaitescu et al. 1998).