Table 7 shows the efficiencies for finding at least one truth track in an event for the different event types. The absolute efficiency is the ratio of the number of events with at least one truth track over the total number of events. The relative efficiency is the ratio of the number of events with at least one truth track over the number of events with at least one truth track found by the truth segment finding algorithm. The errors are statistical. We obtain a 99% efficiency for suppling enough information to find at least one track, within the CTD geometrical acceptance, for NC and CC events. A similar absolute efficiency is also obtained for the NC events. The truth algorithm indicates that the best we can do in the CTD for CC events is approximately 92%. Clearly the FTD is needed for the linking of segments in a second level tracking trigger.
Table 7: Efficiency for finding at least one
truth track per event using the found segments.
Figure 7 show the absolute track efficiency using
the found segments versus minimum 
cut.
The absolute efficiencies for CC and BG events decrease
with increasing cut, while the efficiency for NC
remains relatively constant.
The differences for the different event types reflects the
difference in distributions of the different events.
The NC events have an electron with greater than
and hence should always produce a track within
the fiducial cuts for 
greater than 100 GeV 
.
CC and BG events have multiplicity distributions peaking at
low and some of these events will lie outside the
CTD-SLT acceptance, and hence the efficiency drops as a
function of .
The relative efficiencies as a function of cut are
constant for each event type with values of 100% for NC
events, and 98% for CC and BG events.
The SLT track finding efficiency has the potential to be
large.
Although many segments are outside the fiducial space,
we
still have a greater than 98% chance of finding at least
one track if it is within the CTD-SLT acceptance.
Loss in CC events is presumably due to tracks contained in
dense jets and hence contaminated segments.
Figure 7: Absolute track efficiency using found
segments versus minimum cut.
A study of the minimum number of hits required to define a segment has been made. The efficiency for finding a truth track, given the found segments, falls rapidly above a five hit definition (average in cell for a track) for each event type. Figure 8 shows the relative track efficiency for different segment minimum hit definitions. The figure indicates that four hits is an optimal choice. The number of nasty segments is constant with the minimum hit definition.
The maximum number of hits in a mask can range from 8 to 50 with little change in the efficiency for finding at least one track. Presumably the time required to find the segments is the important consideration in this case.
The track efficiency for the different event types is flat
for different road widths until the road becomes less than
(500 
m), at which point a rapid decrease
is observed; see figure 9, for which greater
than 0.5 GeV/c has been used.
A road change when predicted hits are not found is not
particularly important and as long as the increase is less
than two times the initial road width the efficiency is
maintained.
Figure 8: Track efficiency using found segments
for different segment hit definitions.
Figure 9: Track efficiency using found segments
for different road widths.