A stringent new limit on bursts from nearby galaxies

If typical gamma-ray bursts GRBs] have X-ray counterparts similar to those detected byGinga, then sensitive focusing X-ray telescopes will be able to detect GRBs three orders of magnitude fainter than the detection limit of the Burst and Transient Spectroscopy Experiment BATSE]. If a substantial portion of the burst population detected byBATSE originates in a Galactic halo at distances greater than or equal to 150 kpc, existing X-ray telescopes will be able to detect GRBs in external galaxies out to a distance of at least 4.5 Mpc. As reported in Hamilton, Gotthelf and Helfand (1995) the Imaging Proportional counter IPC] on board theEINSTEIN Observatory detected 42 transient events with pointlike spatial characteristics and timescales of less than 10 seconds. These events are distributed isotropically on the sky; in particular, they are not concentrated in the directions of nearby external galaxies. For halo models of theBATSE bursts with radii of 150 kpc or greater, we would expect to see several burst events in observations pointed towards nearby galaxies. We see none. We therefore conclude that if theGinga detections are representative of the population of GRBs sampled byBATSE, GRBs cannot originate in a Galactic halo population with limiting radii between 150 kpc and 400 kpc. Inasmuch as halos with limiting radii outside of this range have been excluded by theBATSE isotropy measurements, our result indicates that all halo models are excluded. This result is independent of whether the flashes we do detect have an astronomical origin.