Testing the Gamma-Ray Burst Pulse Start Conjecture

We test the hypothesis that prompt gamma-ray burst pulse emission starts simultaneously at all energies (the Pulse Start Conjecture). Our analysis, using a sample of BATSE bursts observed with four channel, 64-ms data and performed using a pulse fit model, generally supports this hypothesis for the Long GRB class, although a few discrepant pulses belong to bursts observed during times characterized by low signal-to-noise, hidden pulses, and/or significant pulse overlap. The typical uncertainty in making this statement is < 0.4 s for pulses in Long GRBs (and < 0.2 s for 40% of the pulses) and perhaps < 0.1 s for pulses in Short GRBs. When considered along with the Epk decline found in GRB pulse evolution, this result implies that energy is injected at the beginning of each and every GRB pulse, and the subsequent spectral evolution, including the pulse peak intensity, represents radiated energy losses from this initial injection.