I know that I can illuminate(to some extent) a light bulb by connecting it
to the hot side of the mains and grounding it to the earth indicating that
there can be significant current flow in the earth at 50/60 Hz.

1) Can there also be significant flow at harmonic frequencies of the 50/60
Hz?  If so does this mean that an antenna low to the ground like mine,

could

be picking up some of that energy that may be in the LF bands?
Would raising it to 5 feet help eliminate some of that noise?

2) Does raising a loop have any effect on the received signal strength

other

moving the antennas resonance point?

3) Is it possible to shield one side of a loop antenna  with say wire
fencing a few feet away to  elimiinate noise coming from 1 direction and
thus making the unidirectional instead of bi-directional.

After all are not we playing the game of signal against noise.  To extend
the range of a given station should much consideration be given to
minimizing the undesirable noise, either man-made or natural ?   Isn;t the
noise the limiting factor here?

Please tell me if I am all wet...  I will go get a towel

A dry washcloth should suffice, Paul.

You surmise correctly on your first question.  Earth currents can induce
noise into a loop.  Sometimes there's not too much effect, sometimes quite a
bit.  Moreover, the proximity of the partially-conducting earth affects such
things as the depth of null you can obtain with the loop.

Second question:  Yes, height of the loop can make a difference in signal
strength.  It's not a linear relationship, but a bit of height may either
get you away from sources of noise, or improve signal strength, or both.

Third question, though:  It is not effective to "shield" a loop in the
manner you describe.  Currents in the fence will couple either signal or
noise--or both--into the loop, and also adversely affect its directional
properties.  It could eliminate your ability to achieve nulls in crucial
directions.

There are other ways, though, to utilize a loop in a unidirectional
configuration rather than bidirectional.  If you combine the output of a
loop with the output of a whip, with provision for matching the signal
levels of the two antennas, you can achieve quite a nice cardioid pattern in
situations where that would be more useful than the two nulls of a
figure-eight.  To learn more about steerable nulls and phasing and matching
of two antennas, I recommend visiting www.lwca.org and scrolling down to the
lower half of the home page.  Click the link for "On The Art of NDB DXing,"
and go to Chapters 4 and 5 for some very useful ideas.

73,
John Davis