Figure 1:  Black = all index 0 vertices and red = index 0 vertices in events with one vertex.  Left is the tofMult and right is the refMult.  Both look pretty smooth.

Figure 2:  Red and Black are identical to Figure 1.  Blue is from those events were tofMult > (0.00222*refMult0^2+5.758refMult0+100)  Basically these are what I've defined as split.  (though I had an additional cut of refMult > 50 and tofMult > 500 to stay away from all the nice correlation in the corner.  That removes the features though.   I guess I don't really see anything extremely obvious.  This is for the high tower stream, but the physics stream doesn't look too different.  The hump is a little more rounded.

Figure 3:  Index 0 refmult in black.  All others in red.  This is the physics stream for the files I've chosen.

Figures 4a and b:  vpdVz - Vz of index 0 versus refMult of index 0.  Graph on the right is simply a zoomed in version of the graph on the left.  There is a centrality dependence of the vpdvz - vz distribution, but it really only kicks in at very low resolutions.

Pick day 50 night shift
From when I was running so I know things were smooth!

Run 11050002 from beginning of fill
Daq rate ~430 Hz

7366 events:
st_physics_11050002_raw_5030001.daq

5252 events:
st_ht_11050002_raw_3150001.daq

894 events:
st_zerobias_adc_11050002_raw_5510001.daq
st_zerobias_adc_11050002_raw_2520001.daq
st_zerobias_adc_11050002_raw_3510001.daq
st_zerobias_adc_11050002_raw_1510001.daq
st_zerobias_adc_11050002_raw_4510001.daq

vpd-mb: 260021
zerobias: 9300
NPE_11: 260522
Central: 260123

Run 11050017 from end of fill
Daq rate~ 400 Hz

7639 events:
st_physics_11050017_raw_5040001.daq

3457 events:
st_ht_11050017_raw_1160001.daq

1070 events:
st_zerobias_adc_11050017_raw_2500001.daq
st_zerobias_adc_11050017_raw_1520001.daq
st_zerobias_adc_11050017_raw_4520001.daq
st_zerobias_adc_11050017_raw_5520001.daq
st_zerobias_adc_11050017_raw_3520001.daq

getDaqList2009.tcl 11050002 p
delete *.lis
edit *.hpss
hpss_user.pl -f file.lis