Timothy Murphy wrote:
Bill wrote:

I've worked out that there must be a break in the connection
between 2 of the lights on this circuit.

Did you work this out in theory or have you actually been into the
fittings and tested?

Thanks for your (and all the other) responses.
This was in fact a purely theoretical deduction ...

If you haven't been into the fittings then it maybe as simple as a
loose screw allowing the feed to the next light to become
disconnected. Much quicker to fix than your proposed temporary
solution. Also rather a lot safer :-)

I've looked at all the ceiling roses that are affected,
particularly the last one working and the first one affected.
The connections on them all look fine to me.

If course the problem could be at a working light fitting!

Unfortunately for my experiment,
the ceiling roses are not as in the online diagram at
http://wiki.diyfaq.org.uk/index.php?...ers#Lightin g
and elsewhere online.
Each ceiling rose has two terminals on the ceiling side,
one of which has 3 incoming wires attached to it,
and the other has 1 incoming wire attached to it.
The two wires to the light are attached to the two terminals
underneath the rose.

Is the ceiling rose similar to?

http://www.tlc-direct.co.uk/Products/ASPCR2000.html

The only way I can make sense of this is that the single wire
goes to the switch and thence to the live supply,
while the 3 wires are negative and go to the next light,
the last light and possibly the switch?

Does every fitting have 3 neutrals?

I should say the wiring is fairly old - 25 years -
but was installed by an electrician who I am sure
would have followed whatever rules were in place at the time.
(He worked for the electricity supply company.)
I checked in an electrical store here,
and the ceiling roses sold now are as shown in the online texts.

Any advice on this setup gratefully received.

You need a voltstick
http://www.tlc-direct.co.uk/Products/TMPL107.html
to help trace the problem

--
Adam