fsxnp, the four orange lights shine into the curved mirror to form the 'meatball'. The greenie/blue datum lights are either side of the mirror all shining up the glideslope. The concave mirror shape allowed the meatball and datums to be seen easily before halfway around base turn. At beginning of A4G ops we did low circuits at 300 feet daytime to take advantage of that having a very short sharp straightaway (minimising errors).
{EDIT}: Forgot to add that this base turn was LEVEL, thus making one less distraction, no need to try to figure out any descent rate during base turn. A level turn meant we picked up the ball at 300 feet and we did not descend further unless meatball allowed. Not sure if after my time the circuit height was raised onboard. We did FCLP at 500 feet at night at NAS Nowra because it was too dangerous, with uneven rising ground to west & south-west of airfield to go lower. We could descend a bit during base turn because on RW 26 there was a big gully to fall into, so large that it was possible to go below airfield elevation before crashing.
And to be clear - a night deck landing could only be done via a CCA (Carrier Controlled Approach) similar to a GCA (Ground) with a precision radar and a controller guiding pilot to 1& 1/4 NM at around 1000 feet or less, to then have the pilot look ahead to land visually. The CCA controller would keep quiet then to allow pilot and LSO to communicate.
Some mirror pics attached now with explanations. This first pic shows 5 wires with No.4 being the target wire. If less than five set then No.3 would become target. No.4 allowed best 'hook to ramp' clearance which was only 6 feet.
A 13Mb .WMV video here explains how the mirror was invented and how it works:http://www.filefront.com/16719379/MirrorExplainedFAAM.wmv(13Mb) (Left mouse click on URL to go to download page first)
The source (sauce?) lights that form the orange 'meatball' can be varied in intensity along with the blue/green datum lights. The mirror could be moved to change glideslope and it was stabilised only in pitch not for roll (which could be a problem). However the transition of the meatball in the mirror as it moved was smooth unlike IFLOLS which jumps between light cells. Fourth last pic shows the MCO Mirror Control Officer who is looking down the glideslope which he checks with a sight to see that it is setup for A4G or S2E. A 'helper' checks that the mirror is uncovered and clean etc. Third last pic shows HMAS Melbourne mirror now in the musesum FAAM Fleet Air Arm Museum at NAS Nowra, (south of Sydney on the coast more ore less) NSW, Australia.