2026-07-072026-07-07http://salesiana.dossiersoluciones.com/handle/123456789/97195Two classes of Gamma Ray Bursts have been identified so far, characterized by T_90 durations shorter and longer than approximately 2 seconds. We show here that the BATSE 3B data allow a good fit with three Gaussian distributions in log T_90. The chi^2 statistic indicates a 40 % probability for two Gaussians, whereas the three-Gaussian fit probability is 98 %. Using another statistical method, it is argued that the probability that the third class is a random fluctuation is less than 0.02 %.10 pages, 5 Postscript figures. Submitted to The Astrophysical Journal, October 1997AstrophysicsA 3rd Class of Gamma Ray Bursts?text