The PHENIX CDR[25] and updates establish in great detail the feasibility of detecting
dimuons from the DY process (
) and quarkonium production into the first muon
end cap for rapidities 
at 
. The mirror kinematic coverage of
the second arm should be very similar. There may be some differences in the hadron absorption
characteristics either due to a desire to have a lower muon threshold or to
a physical difference in the configuration of the second arm.
The requirements on triggering, muon identification, and pattern recognition are
considerably less stringent for pp physics than for the quark-gluon plasma
(QGP) search program for which the first end cap was optimized.
Because of the large effort which has gone into the design of the first end cap it is worth
comparing the scale of the signal-to-background for the QGP and pp physics
programs for the DY process. The comparison is made at equivalent luminosities, 
, for
AuAu and 
collisions. Backgrounds are taken to be dominated by random
coincidences of either pion punch-through or muons from pion and kaon decay which emulate
high-mass DY muon pairs. Compared to the pp background, the rate of soft hadron
production from central
AuAu collisions is, 
, where ``
'' indicates that higher 
production may well go as a higher power of A[40]. The background rate from random
coincidences of such events is
. Again with reference to pp collisions, the rate
of production of DY pairs in central heavy-ion collisions is 
. Thus
the overall signal-to-noise factor, 
, is
times better for the 
program at equivalent
luminosities.
The principal operational differences between dimuon detection as described in the
Conceptual Design Report[25] and what is
proposed here are connected with detection of rather low-energy muons near central rapidity. Here the
opening angle between the muons is large enough that they are accepted into different arms (hence the
increase in acceptance). Also, one or more muons may be detected in the arms of the central magnet.
Optimization of the configuration of the second arm requires further simulation for polarized proton
physics as well as investigation of its performance for the QGP program. As an example, it may be
desirable to have less total hadronic absorption in the second arm in order to lower the minimum mass
threshold for central 
and 
events. Figure 15 shows that operation of the second arm without
the ``nose cone'' hadron absorber of the first arm permits recording of 
events well above the
pion decay continuum. Clearly 
production in low multiplicity 
collisions is feasible
even without maximal suppression of hadron decays.
Figure 15: Simulation of 
events and pion decay background at
. The upper points are events accepted in the
second muon arm with no hadron absorbing ``nose cone'' placed before
the central magnet pole tip. The lower curve shows the
corresponding decay background in the first muon arm in the
configuration of the PHENIX CDR[25] (which includes the
``nose cone'', neutron shield, and Pb photon/electron shield). The
simulations were performed with PISA[41], using pion
production cross sections from Alper, et al. [42].