[en] Expensive experiments to detect a deconfined parton phase have been done and are being planned. In these experiments it is hoped that nuclear collisions at relativistic energies will exhibit signals of this new phase. So far all the results may be interpreted in terms of independent nucleon-nucleon interactions. These elementary collisions at very high energies are therefore worth examination since each such collision produces a highly excited entity which emits a large number of hadrons. In the hadronic phase this results in the GS multiplicity distribution. In the parton phase, parton branching results in the popular negative binomial distribution. Though neither the GS nor the NB distribution alone agrees with the data beyond 200 GeV, it is fitted exceedingly well by a weighted sum of the two distributions. Since the negative binomial distribution arises from the branching of partons, we interpret the increase with energy of the negative binomial component in the weighted sum as the onset of a deconfined phase. The rising cross section for the negative binomial component parallels very closely the inclusive cross section for hadron jets which is also considered a consequence of partons branching. The consequences of this picture to nuclear collisions is discussed. (author). 8 refs., 9 figs., 3 tabs