The brightest 200 GRBs with duration T-90 > 2 s, selected from the BATSE 4B
catalog; have been analysed, using a wavelet technique. The characteristic
s of statistically significant peaks including the rise time: fall time, FW
HM, peak height, peak area, time intervals between peaks and the number of
peaks per burst have been analysed. The results presented show that the dis
tribution of the areas of the peaks is compatible with a lognormal distribu
tion.
The properties of the peaks before and after the strongest peak have also b
een compared, and we have analysed the similarities and differences between
the pre-peak and post-peak sections of the burst.
The overall properties of peaks in GRBs provide considerable constraints on
models that attribute the variability to interactions with an external med
ium or to the activity of a central engine.