For a TT supply I presume you will have a 100mA or 300mA slow tripping RCD immediately after the meter, or at least before the first earthed box...
or maybe even a 30mA normal RCD, if the circuit if occasional nuisance trips are acceptable.
What you need to ensure is that under any fault which does not trip the RCD, the touch voltage (between the earth terminal, which will be made slightly live by the fault current) and true earth (terra firma, and any pipes, fence posts stuck in it cannot exceed 50V, or for an 'agricultural installation 25V). The fault current is taken as
For a 30mA RCD, this implies an earth resistance of less than 1 kilohm for agriculture, and 2K for an all indoor installation, (if it was 25mA and 25V it would be 1K exactly) but one would try to get much lower than this to allow for soil drying out in summer etc. Of course for a 100ma then instead of 1K read 250 ohms, and for a 300mA trip read 100 ohms. Regardless of RCD type, I'd worry if it wasn't less than 100 ohms when new and damp, personally, and consider adding another rod, a rod length or two away connected in paralell, even though that might not strictly be needed to 'squeak a pass'.
There are 2 ways to test, one involving another earth electrode, and passing a fault current down the earth under test and looking at the rise in voltage, and the simpler one which is the basis of 'earth loop' type testers. The simpler method measures the phase to earth terminal voltage, and then (for a very short time !) connects a low value resistor (usually either 10 ohms or 100 ohms and wound like a fire element, to allow it to swea off the heat) between them, and monitors how much that voltage droops. The change in voltage divided by the current through the resistor is the impedance of the 'fault loop', i.e. the resitance of the live feed and the earth path together.
In a TT system, it is usually the earth rod resistance that dominates, as the live feed resistance will be sub ohm, otherwise the full load drop would be noticable.
While the test is underway, no-one should be near the earth electrode, as the 'step voltage' (the voltage between two nominal wet feet on the ground one pace apart ) could easily be dangerous.
Also for this test the live must be taken from the supply side of any RCD, as of course otherwise it will trip it.
actually, now I've typed all this guff I realise its all here... with nice diagrams.
http://www.tlc-direct.co.uk/Book/8.6.1.htm
oh well, happy reading.
regards M.