Energy Dependence of Multiplicity Fluctuations in Central Pb+Pb Collisions

At high energy densities a new phase of matter, the quark-gluon-plasma (QGP), is expected to be formed. The energy dependences of various observables show anoma- lies at low SPS energies which might be related to the on- set of deconfinement (1). Both the onset of deconfinement and an approach to the critical point of strongly interacting matter might lead to an extraordinaryincrease of multiplic- ity fluctuations. In addition to the effects mentioned above also the normal statistical fluctuations are interesting to study, since the size of the multiplicityfluctuationsdepends on the nature of the particle ensemble (micro-canonical, canonical, grand-canonical) (5). This motivates a detailed study of the energy (2, 3) and centrality (4) dependence of multiplicity fluctuations measured by the scaled variance ω = Va r(n) / . For the study of the energy dependence of multiplic- ity fluctuations, very central Pb + Pb collisions were se- lected, for which the fluctuations in the number of partic- ipants can be neglected. The multiplicity distribution of negatively charged hadrons at top SPS energy in the most central Pb+Pb collisions at 158A GeV is shown in figure 1 in comparison with a Poisson distribution, which is close to the prediction of the grand canonical model. The mea- sured distributions are significantly narrower than the Pois- sonian ones at 20A, 30A, 40A, 80A and 158A GeV(2). The energy dependence of scaled variance for negatively charged hadrons is shown in figure 2, left. The detector ac- ceptance changes with energy so that a direct quantitative comparison of scaled variance at different energies is not possible. A comparison to string hadronic models show that UrQMD reproduces the data on scaled variance, al- though the mean multiplicity is overpredicted(see figure 2, right). The Venus model underpredictsthe mean multiplic- ity and overpredicts ω at higher energies. In (5) it is shown that the measuredfluctuationsare smaller than predictedby the canonicaland grand-canonicalmodel, but seem to be in agreement with the micro-canonical model.