Thanks to both who replied! I now have those PDF documents and I've been able to get the meter working. The manual for the model 248 revealed the polarity of the battery pack -- in fact, the 248 looks like a re-implementation of the 245 using LEDs instead of Panaplex displays. The 245 design dates from 1974, but the 248 is nearer 1978. They look like they share the same case design and battery pack (6 cell AA NiCad).
So, with a Farnell bench supply rigged up to the battery terminals, I switched on and... it worked! It's drawing about 100mA from the PSU at 7V or so (simulating a somewhat run-down NiCad pack). The meter flashes the +/- polarity indicator when the battery voltage drops below about 6V, which makes sense as a low-battery warning. The auto-zero feature on the 10V range appears to work.
Testing with a 1k Ohm, 0.02% resistor gave a reading of 1.000 on the 10k range and 0.9997 on the 1k range. Not bad! And the DC voltage and current ranges seem OK, too.
Photo shows the meter under test -- note that the red crocodile clip is on the battery terminal, but it isn't shorting with the range switch, it just looks like it is!
Thanks again for assistance rendered, and I'll update with any further progress!