The first thing I dowsed was a 914 Xerox copier. A colleague had lost a
pocket watch in his garden and asked me to help him find it. I made a
dowsing rig from office stuff and held it out. It pointed right to the
Xerox. He took the rig home and found his pocket watch.
 
Dowsing has a proper place among remote sensing methods. These methods
include resistivity, metal detection, soil compaction measurement, infrared
photography, oblique photography, and so on and so on.  Not a single one of
these methods is "replicable" or totally effective.  Remote sensing simply
is neither consistent nor reliable, in an absolute sense. There are too
many variables in nature for consistency to be assured.
 
The only way to ensure that you have found all the cultural resources on a
site is to dig the site, all of it.
 
I don't want to impoverish my clients, so I will use any clue and every
method that might reduce the amount of expensive dirt-moving. In the case
presented at the onset of this thread, the problem was to find a cemetery
that was "known" to exist in a tract of about an acre. My approach would
begin with the local locational model, coupled with historical research.
 
Then I would visit the site to grok its fullness.
 
Then I would walk the site with a dowsing rig to see if my suspicions were
correct.
 
Then I would hire a Gradall and begin stripping in the area where my
hunches, my dowsing, and the models predicted the cemetery should be.
 
My client is paying for the use of my brain, my experience, and my
intuition. Whatever gimmickry I choose to employ is immaterial if I find
the cemetery.
 
Ned Heite