1. Center for Space Research, Center for Theoretical Physics and Department of Physics, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA

Abstract

The basic nature of cosmic -ray burst sources¹ has remained a mystery. Schaefer² and Schaefer et al.³ recently discovered three images on archival photographic plates probably due to optical flashes from the error boxes of three -ray burst sources. This discovery opens up the possibility of ground-based optical studies of these objects (see, for example, ref. 4) and provides an important new set of data that can constrain models significantly (see, for example, ref. 5 and references therein) for these intriguing sources. London and Cominsky⁶ have considered a model for the optical flashes, wherein -ray bursts are emitted by a collapsed object in a close-binary stellar system and a small fraction of the radiation is reprocessed into optical radiation in the surface layers of the companion star (see refs 7 and 8 for possible alternative explanations). Based on their own estimates of the time, , required to reprocess the radiation into optical light, London and Cominsky concluded that a binary model would not account for the observed optical flashes. They considered, however, only main-sequence stars and hydrogen-depleted degenerate dwarfs as possible companion stars. We reconsider here the binary model and conclude that it is viable. In particular, under the assumption that the optical flashes were produced by -ray bursts of about the same intensity as those observed, we find that nearby (100 pc) binary systems with secondaries whose masses are less than 0.06 M can fit all the observational constraints for the three optical/-ray pair events (hereafter 1928/1978, 1944/1979 and 1901/1979; see Table 1).