To illustrate the order of magnitude of the effects :

An ideal shape of ground electrode is a hemisphere, of diameter D
metres, pressed into the soil, flush with the soil surface..

Its resistance to the soil is easily proved and calculated :

R = S / Pi / D ohms,

where S is the soil's resistivity in ohm-metres and D is the diameter
in metres. Pi = 3.14. The metric system is by far the most simple.

The resistivity of salt sea water is 0.22 ohm-metres, constant
wherever you may sample and test. Unaffected by the melting of the
glaciers.

So with a diameter of 0.22 metres = 9 inches, the electrode resistance
= 1 ohm. Low enough?

If the ground electrode is a ball with diameter = 9 inches, immersed
in sea water at a sensible depth, then the electrode resistance will
be halved.

At radio frequencies the impedance of the connection to ground will be
that of the connecting wire only, even before the resistance of the
connecting wire to the water is taken into account. The high
permittivity of water will also tend to decrease impedance at RF.

Another illustration, following Lord Kelvin :

The resistance of a ground rod to soil is given by :

R = S / 2 / Pi / L * ArcSinh( 2 * L / D ) ohms,

where S = soil resistivity, L = rod length in metres and D is rod
diameter. ArcSinh is the inverse hyperbolic Sine function you will
find on your pocket scientific calculators.

So in sea water, at low frequencies, a rod 12 inches in length and a
diameter of 1 inch will have a resistance of 1.2 ohms. At HF, because
of the very low propagation velocity in water, propagation effects
predominate and the rod must be considered as a very lossy
transmission line. But its impedance to ground is still very low
because Zo is very small.

So the hull of a metal boat makes an excellent ground. Just connect
to it with an alligator clip at the end of a length of wire and stop
worrying about it.

By the way, the practical units of resistivity in ohm-metres should be
much preferred to the academic units of milli-Siemens. When dealing
with milli-Siemens I find I have to stand on my head and look
backwards.

1 milli-Siemens = 1000 ohm-metres.

The clock tells me it's 7.30 in the morning in Birmingham, the idle,
depressed ex-industrial city of the Midlands, where there used to be
10,000 factories, now superceded by the hardworking Chinese, and I'm
already half way down a bottle of Spanish Campaneo red. Hic!
----
Reg, G4FGQ.

An arithmetical correction. I forgot to divide by Pi.

The resistance of a hemispherical electrode, 9 inches diameter, in
salt sea water, is even smaller. It is only 0.32 ohms.

Incidentally. the resistance of a flat circular disk of diameter D
metres, in contact with the soil surface, is given by :

R = S / 2 / D ohms,

where S = soil resistivity in ohm-metres.

In sea water, a disk of 12 inches diameter has a resistance of 0.37
ohms. Which is negligible in comparison with the radiation resistance
of a 1/4-wave vertical antenna of 36 ohms.

Careful readers should make a note of these hints and tips, free to
USA citizens, in their notebooks.

My own notebooks extend from volumes A to letter S. I'm wondering who
to leave them to in my Will & Last Testament.
----
Reg, G4FGQ.

"Reg Edwards" wrote:
The clock tells me it's 7.30 in the morning in Birmingham, the idle,
depressed ex-industrial city of the Midlands, where there used to be
10,000 factories, now superceded by the hardworking Chinese, and I'm
already half way down a bottle of Spanish Campaneo red. Hic!

Reg, FYI: "NAPA, Calif. - French and California winemakers marked the 30th
anniversary of the storied Paris tasting with another sip-and-spit showdown.
California won - and by more than a nose. Native wines took the top five of
10 spots, with a 1971 Ridge Monte Bello cabernet sauvignon from the Santa
Cruz mountains coming out on top Wednesday."
--
73, Cecil, W5DXP

Reg, G4FGQ wrote:
"At radio fgrequencies the impedance of the connection to ground will be
that of the connecting wire only, even before the resistance of the
connecting wire to the water is taken into account."

In general, Reg is correct.

I`ve installed a marine radio on a yacht, engine powered not the sailing
kind. It worked well as expected and the owner paid promptly. I`ve also
installed radios of various types in many vessels including large ones
used on the high seas and small work boats serving the near offshore.
But, I never installed an antenna using radials on a boat or ship.
Radials can be hazardous to eyes and body parts at sea. For VHF, a
1/2-wave coaxial antenna has the same gain as a ground plane but needs
no radials. For HF, it`s easy and effective to use a loaded vertical
antenna against the sea as a return path.

Best regards, Richard Harrison, K5WZI