Beavis et al. ReplyIn a recent publication [1], we discussed experimental measurements related to the centrality dependence of p production in Au 1 Au collisions. We compared our data with two of the most successful transport models available to describe nucleus-nucleus collisions, namely, RQMD and ARC. The accompanying Comment [2] attempts to "correct and complete some of the theoretical analysis" presented by us on behalf of the authors of the ARC model and to "modify" and "enhance" the conclusions in our paper. Its authors state that using a modification of their implementation of classical screening they are better able to describe our data. They also add useful plots presenting the "pure production" and the "no shielding" results of the ARC model. Similar calculations using the RQMD model were published several years ago [3,4].We are delighted to see that our data have prompted the authors of the ARC and RQMD models to refine their calculations to better describe our results [2,5]. We note that it is still possible to use very different theoretical approaches to produce results which now agree with our measurements made at low transverse momenta. Hence, the comparisons of the models with each other, and with data measured over a broader kinematic range, and for different collision systems now become necessary to understand the physics of p production and annihilation.Antiproton production is far from being completely understood. For example, in the year since the publication of our paper, there have been new measurements of p production in Au 1 Au͑Pb͒ [6-9] collisions in kinematic regions that are complementary to our measurement. Some of us have been involved in these measurements and plan to report on them elsewhere. The data suggest that there is a large yield of L particles which could contribute to the measured yield of antiprotons. Direct measurements of the L in light systems results in a L͞p yield that is ϳ3 [10], well in excess of unity. The preliminary ratio extracted from a comparison of E878 and E864 data for Au 1 Au͑Pb͒ collisions is ϳ4 [7,8]. Our collaboration has investigated the extent of L contribution in our p data [11] and concurs with the view that our data taken with the measurements of the other experiments are consistent with a L͞p ratio being much greater than 1. It is also noteworthy that even a simple thermal model of nucleus-nucleus collisions predicts a L͞p ratio in excess of 1 [8,12].In light of the above information, and the fact that the ARC model does not explicitly implement L production and annihilation, we expect the model will have to undergo further refinement to be able to simultaneously describe the E878 data, and the data from the other experiments.