Measurements of e+e- pair production at the Bevalac

A complete kinematic description of a lepton pair requires 6 independent variables, in contrast to the case of a real photon, where there are only 3 independent variables. The mass and momentum of the virtual photon may be specified independently, with the remaining variables being the two angles specifying its production direction, and the two angles of the decay in its rest frame. It has become customary to use as kinetics variables: the mass, M; the transverse momentum, Pt; and a longitudinal variable, either the rapidity -- Y, or the scaling variable -- X. Data are almost always presented as a function of only one of these variables. This may be due to the difficulty inherent in presenting multi-dimensional data, to poor statistics, to limited experimental acceptance, or to some combination of these reasons. It should be noted, however, that these one-dimensional projections always represent integrations over the remaining variables -- integrations within which the shape and limits of the acceptance of the experimental apparatus must be taken into account. The effects of the acceptance must be understood in comparing experimental data either to theory or to results from experiments with different coverage of the total phase space. Although we have a strong interest in studying nucleus-nucleus collisions using the tool, most of the discussion presented here will be about our measurements in p+Be collisions. It was necessary for us to make these measurements because, in the relevant range of incident energies, the yield of electron pairs from nucleon-nucleon collisions had not been measured, nor was the hadronic production mechanism understood. A brief discussion of lepton production in hadronic collision will help put out our results in a proper framework. 15 refs., 9 figs., 2 tabs