The collapse of structure effects observed in hot nuclei is interprete
d in terms of a dynamic lattice model which describes the process of n
ucleon (clusters) evaporation from a hot nucleus, predicting the Final
mass distribution. Results are compared with experimental data for th
e B-10+Be-9 and B-10+B-10 reactions, and indicate that thr structures
observed in the low-energy mass distributions in both simulation and e
xperiment are a consequence of the competition between the residual in
teractions and the thermalization dissipative process. As a characteri
stic feature of complex evolving systems, this competition lends to lo
ng term memory during the dissipative path, the observables becoming t
hus insensitive to the actual microscopic interactions.