The symptoms are: Car appears to warm up normally, taking about 5-10 minutes of driving in ambient temps of about 9C to warm up to normal temp of 90C. But then the temp gauge just carries on climbing. After half an hour or so it's nudging 140 C.

But: The fan doesn't automatically cut in, and on opening the bonnet there's none of the associated smell or sense of an overheating engine.  Cheers, Tim

Firstly, you are not overheating. Forget that possibility. You would know it, you would smell it, your rad fan would be on continually and your rad would be overflowing.

Your water temp gauge is effected by a limited number of things.

1. Your voltage stabilizer. This items alters and maintains your 12 Volt to a 10V supply to your fuel gauge and water temp gauges. If the screw holding it at the tab looses, the
stabilizer's earth wire, held by the same screw, can lose a proper contact and the current to the gauges will jump to your alternator's output (14 +or-.5V). That will make your fuel level and water temp  jump
considerably. First, verify that the stabilizer is earth by testing the earth to the body of the stabilizer. 90% of problems are resolved here.

2. Your voltage stabilizer has failed. I have had 4 fail in the space of 2 weeks. I switched to another source from the Caerbont stabilizer (who sells to Holden, Moss, Europe, Morgan..ect etc) and resolved the problem.

You can test this item. You cannot detect the proper output from a stabilizer, but you can substitute something you know is putting out 10V. If you passed #1, get another battery and let it run down to 10V. Earth it on the car and have it feed your fuel of temp gauge directly (no stabilizer in between). If your
temps now behave, the stabilizer is blown.

BTW, I purchased the other 10V stabilizer (1.5 amp) from http://www.digikey.com #UA7810CKC They are 40 pence each rather than 8 pounds. Read this article . Try your neighbourhood Radio Shack or computer repair place.

3. The wire from the temp sender on the engine block is earthed somewhere. To test, you could simply put in another wire from the sender temporarily and loop it around to the gauge.

4. If you pass..#1-3 the sender is gone. Not unheard of. It must be replaced.

5. I have not seen this yet, but I imagine the gauge itself can go.

The key is that if it is the voltage stabilizer is abnormal in some way, both the fuel level and the water temp will be effected.

TESTING YOUR COOLANT FOR DAMAGING ELECTROLYSIS
from Tim at Ron Davis Racing

A voltmeter capable of reading both AC and DC currents is required to test cooling systems. The meter needs to read zero to the maximum voltage of the system being tested in tenths of a volt. The meter leads must be long enough to reach between the coolant and the groundside of the battery. An ohm function of a voltmeter is very helpful to pinpoint areas of resistance in as electrical system that will cause an electrical current to ground through the coolant rather than the engineered electrical circuit.

PROCEDURE

1.  Attach the proper meter lead to the groundside of the battery, negative-to-negative or positive-to-positive.
 
2.  Install the second lead in the coolant touching the coolant only.
 
3.  Read the DC and AC voltage with all systems off. If a block heater is present, also take a reading with the heater turned on. If an automatic battery charger is present, as a standby system, also take a reading with this system running.
 
4.  Read the DC and AC voltage with the electrical starter engaged.

5.  Read the DC and the AC voltage with the engine running and all systems turned on: lights, coolers, fans, heaters, air conditioning, cell phone, two-way radio, including the phone and radio on both standby and transmit.
 
6. The above procedure will test a complete system except for an electrical current, which can be generated by the rear end transmission. This is particularly true with air bag suspensions, rubber pad suspensions and rubber-mounted transmissions. Any current generated will travel up to the drive shaft to ground through the engine coolant. Grounding rear ends and transmissions is strongly recommended.
 
7.  Voltage of zero to .3 is normal in a coolant of cast iron engine. Such an engine will be destroyed with time by .5 volts, and engine manufactures are reporting .15 volts will destroy an aluminum engine.
 
8. The current will be AC if the problem is due to static electricity.
 
 9. If the coolant shows an electrical problem with all the equipment turned on; turn off one system at a time until you finally turn off the system that stops the electrical current. When the current stops, this will indicate the electrical system causing the problem.
 
10.  Be partially careful of starters. They can cause as much damage to a cooling system as much as a direct connection to an arc welder! This is due to the amperage present.

11.  Always change the coolant if a current is detected. The electrical current will destroy the protecting chemicals in a properly inhibited coolant.