Tectonic drift.

"From memory, the British monitoring point is Herstmonceux Castle (Royal Observatory) ... and its moving NE at about a cm per year ..."

"Away from the mid atlantic ridge. About 2 cm P.A. from memory. I suggest a search engine trawl on keywords +plate tectonics+ or +plate movement+ or +seafloor spreading rates+ and +atlantic+ might be a start if you want this data... There we go, a quick google search comes up with a dandy automatic calculator. Plug in Lat/Long and choose the correct oceanic ridge and Bobs your parents brother..."

Results:

Cardiff (North Mid-Atlantic Ridge) 51° 28' N, 3° 10' W, NUVEL-1
Full spreading rate : 2.2 [cm/yr]
Spreading direction: 110.2 - 290.2 [deg. from North]
Parameters used for calculation:

Ridge: North MAR
Plate Model: NUVEL-1
Pair of plates: eu-na
Latitude of Euler pole: 62.4 [deg.]
Longitude of Euler pole: 135.8 [deg.]
Angular velocity: 0.22 [deg./m.y.]
Latitude inputted: 51° 28' [deg.]
Longitude inputted: -3° 10' [deg.]

Cardiff (North Mid-Atlantic Ridge) 51° 28' N, 3° 10' W, NUVEL-1A
Full spreading rate : 2.1 [cm/yr]
Spreading direction: 110.2 - 290.2 [deg. from North]
Parameters used for calculation:

Ridge: North MAR
Plate Model: NUVEL-1A
Pair of plates: eu-na
Latitude of Euler pole: 62.4 [deg.]
Longitude of Euler pole: 135.8 [deg.]
Angular velocity: 0.21 [deg./m.y.]
Latitude inputted: 51° 28' [deg.]
Longitude inputted: -3° 10' [deg.]

"The unit of [deg./my] means degree per million years.

I suppose that you tried to calculate the rate of the MAR or other north Atlantic-Eurasian plate boundary, because 0.22 is the angular velocity based on NUVEL-1 plate model for North America-Eurasia plates.

It means that the Eurasia plate rotates 0.22 degree per million year relative to North America Plate around the Euler pole. Therefore the angular velocity is same for any place when you calculate for the MAR. For dating for old alignment, you can use the calculated velocity [cm/year].

Kyoko OKINO"

"You can do a google search of John Dewey and you will get numerous hits regarding plate tectonics. John Dewey was once a professor at Oxford and is now at some university in California with Walter Pittman."

"Probably several over a complete precessional cycle. You could cut down the search time by narrowing the date range. You would also need to know the altitude of the horizon, both at 31.6 deg and 211.6 deg (if a reverse alignment is possible)."

"Even with modern engineering techniques, an error in the alignment of a structure of less than 1 arc minute is utterly insignificant. Why assume that an alignment error of a trivial 30 arc *seconds* is anything more than human error?"
Because error is cumulative and occurs over several miles and reference points indicating more intention than error.