From: Razmic Gregorian <pascal@biochem.wustl.edu>
  To  : rasmb@bbri.harvard.edu
  Date: Mon, 02 Jun 1997 10:34:06 -0500

sed v.

I'm using the Origin package to fit sedV data, and until this point, I'd
had good agreements using all three methods presented, transport, second
moment, and dc/dt.  AND, within dc/dt good agreement between the
apparent s value whether I calculated it by integration or by fitting a
gaussian to the g(s) curve.  Ok. so, all this fell apart last week. I
collected a series of 6 runs, and analyzed them as always.  Using
Walter's dc/dt (which works by integration, right?) and integration
using origin, I get a DECREASING s value with time, even though there is
no apparent gross loss of total concentration and my boundary is still
where it should be, while the result I get by fitting a gaussian to the
g(s) curve has a constant center.  What AM i doing wrong here?
Or, alternatively, what IS going on here?

Thanks,
Razmic