The inelasticity distribution and energy loss of leading particles, which are essential ingredients in statistical models for multiparticle production processes and the phase transition from hadronic to quark-gluon matter, have been calculated in an approach in which the dominant role is played by interacting gluons. The interpretation of existing experimental data indicates an appreciable reduction of the number of soft gluons in comparison with that deduced from a naive extrapolation of deep-inelastic-scattering experiments.PACS numbers: 13.85.-t, 12.35.EqThe theoretical and experimental study of the phase transition from hadronic to quark-gluon matter is a subject of high current interest. 1 The main physical quantity which controls the phase transition is the energy density e. Moreover, since the transition is expected to occur in (nuclear or hadronic) reactions, e = W/ Fis a fluctuating quantity, which will vary from event to event. Here Wis the (invariant) energy effectively used for hadronization and Fthe interaction volume. Wis related to the total available energy Vs by the inelasticity coefficientAn experimental investigation of the phase transition is thus possible only if the distribution of the inelasticity, X(K), is known. Another important field of applications of X(K) is represented by multiparticle production processes when treated from a statistical-hydrodynamical point of view. Such a point of view is adopted at present in most theoretical approaches to relativistic heavy-ion physics, 1 and has proven to be quite successful in hadronic reactions too. 2 The relevance of X(K) for statistical models lies in the fact that, 3 as opposed, e.g., to Regge models, the important physical variable is W rather than Vs. A typical physical observable <9(Vs ) appears as a convolution of O(W) and X(K), where O(W) is the corresponding quantity as a function of W\
d(^)=J0(W=Kj~s)x(K)dK.(
2)This second type of applications of X(K) has been stressed recently 4,5 after it became clear 6 that X(K) is not an immutable quantity but changes with s. In Refs. 4 and 5, relevant physical quantities like multiplicity distributions and average transverse momenta and their correlation were calculated by use of an empirical form of X(K), and their comparison with data brought new evidence for this new physical effect, i.e, X = X(K,^).Finally, the similarities observed between e*e~ and pp reactions, 7 when compared at the same W rather than -Vs, lead naturally to the consideration of X(K).There exists so far no satisfactory theoretical derivation of X(K) in hadronic reactions, and the same is a fortiori true for nuclear processes. The only existing calculations 8,9 relate X(K) to the empirically known leading-particle spectrum at 14-28 GeV(c.m.). 10 (This spectrum was calculated in a bremsstrahlung model where pions were radiated by the colliding hadrons 11 and was also addressed in the gluon bremsstrahlung picture. 12 ) These calculations are unsatisfactory because they do not account for the observed s dependence of X(K), and...