my fit suggestion

["David W. Hertzog" <hertzog@uiuc.edu>]

Steve,

If you fit A exp (-Lt) + B in the range 10 us -> 24 us and then take the
result for B and FIX it for subsequent fits where you walk the start
time from 10 us toward 0 us, you will have a fit where B is fixed in a
situation where it is dominant, and then you are extending it under the
exponential.

Then, you should do the procedure in reverse, namely fit as early as you
can with B floating, then fix B again and walk the fit start time.

Do the 2 procedures give the same result?  If not, then our answer is
dependent on this sequence -- clearly not a good thing.  We will have to
then understand the level or accommodate it in a systematic.

Dave

Fred,
	Since you are an expert, I cc:'d you on this too in case
Berkeley might find it of interest.  The start time (and stop time)
scans are rather important.

	The other suggestion I made -- for your mu+ world -- is to
isolate those events that are pulling your Rate down some much for
gondolas 5/12 when you relax mu conditions.

This is really interesting !!