And, Fred, the effort and sharing is appreciated.

When dealing with George's data you might try the parallel experiment that
Arch has just documented .

TomNie

I'm
doing the test for the people who might wish to use the process, not to
prove
anything to myself.