That number AO-21055-(2 or Z) looks very familiar from my research and from what I've found (and you likely will also) is that it has a 1.75" voice coil and will accept a cone kit for most of the G12 series of Celestions.
An ohm meter on the 2 different coils would be a good starting point. If neither coil is "open" or shorted, it would seem that one of the next steps would be to figure out how to best fire up that amp, possibly a slow Variac fire up to gently/hopefully "reform" the caps a little and hopefully be able to test voltages without having to do a pile of repairs on the amp!


Also identifying the amp, possibly finding a schematic for it would be helpful. It would be interesting, out of curiosity, to see if the field coil power is taken from it's own B+ node or if it shares a node with other parts of the amp.
I'm a fairly rabid search monster and if you can get some numbers off the amp, I can likely track down some info. Remember that while firing up the amp, you need to have a suitable impedance load on the output transformer.
Once we know the voltage we're dealing with, we can either build or acquire a similar voltage power supply in a variable form. I would suggest that for initial testings, we use a really low, 5 or so, watt amp to avoid smokin' the voice coil!
Sorry for the long post, but I'm pretty stoked about this!

Thank You Sir,
Gene