Bill made a Root macro which strips off the wire occupancies from YMon
and puts them in a text file.  I took this data and incorporated them into
my calibration results, and have updated the October shutdown web page:

   http://penn01.fnal.gov/~cot/oct2001/

In particular, the problem list now has "Low/High Occupancy Wire" sections
which place cuts on the YMon occupancy (which I renormalize to 1.0 for each
superlayer).  See the "superlayer occupancy plots" PS file in the "Preliminary"
section of the web page.  I place a 3.5 sigma cut for low and high (although
I don't repeat wires for "high" which also have cell-based noise problems, to
keep the list small).  I added the normalized occupancy as a fourth value in
the () for N in the lists, for example:

   SL5-073 (A,156,12-10-0): #02/14 N = 1.000 (1.000,0.000,0.000,1.217) 

is a wire which passes calibration cuts, but has a 1.217 occupancy (roughly
7 sigma high).  Its neighbor wire is a "Noisy TDC" channel:

   SL5-073 (A,156,12-10-0): #03/15 N = 1.464 (1.667,0.000,0.000,1.093) 

that is (1.0 < N < 2.0) but no NULL noise.  In general, though, Noisy TDC
channels do not seem to have high YMon occupancies, and so do not register
double hits from chamber signals as a rule.

     I regenerated all of the problem cell plots in the "Preliminary" section
so that you can see occupancy (the black dots in the hit plot) versus 
calibration hit level (the red line is "same side" calibration, blue is
"opposite side" even or odd, and the green solid is "null" calibration, with
zero TREF value).  Hopefully, this can provide us with further clues 
as to the origin of dead/hot wires and noisy cells.  Bob has already done 
a lot of cross-referencing with YMon in his last message.