Since I can't find one to purchase, I'd like to build a test speaker similar to the one shown. What output transformer in the upper left of the schematic would be used to correspond with the readings on the front of the cabinet? Would 10 watt resistors work for the field coil section?

A "universal output transformer" will work there. These were common replacement parts back in the day. One made during WW2 was the VAT-6.

You can probably get away with 10W resistors everywhere but the first one should be 20W if you want it to equal the Ferret specification (which is 15W).

Thank you Alan.
I have a couple newer universal types but they have 5 taps on the secondary side and 4 on the primary. I'd like to find one like in the Ferret.

If you are looking for a speaker load, and not necessarily for the plate transformer substitution, you might look around for a GR-1840. These are good for matching from about 1/2 Ohm to 32K Ohms. Power measurement range is from 0.001 Watt to 20 Watts.

I do have a Victory VAt-6 but at 15 watts it would be overkill for that little speaker. I have a NOS Hammond 125B which would match what would be in that test speaker. PPL pri with 6 sec ranging from 1.5 to 15 ohm to match any tubes.

Using a heftier output transformer as Alan Douglas suggested might be a good idea. I have seen a couple of ferrets with burnt output transformers. Better test speakers (e.g., Wright DeCosta) include an inductor in series with the field coil resistor substitution bank because the inductance of the field coil was used as part of the power supply filtering circuit.

Unfortunately, most (if not all) commercial test speakers of the past used custom transformers.

Here is an article showing how to construct a test speaker using standard components:

http://plymouthcolony.net/starcity/radios/pages/unispkr.html

Thank you all for the help!
Lou deGonzague- if you would like to sell that VAt-6 I'd be interested. I'm planning on using a 10" speaker I have.

That's a great article. I can use some of it in my project as I prefer his method for the field coil substitute.

The field coil substitution of the better test speakers had an additional binding post at the junction of the inductor and the resistor bank. Often a bad field coil speaker was replaced with a dynamic speaker (without a field coil) and a larger resistor (than the resistance of the field coil) was put in the power supply. The additional binding post allows a tech to experimentally determine the resistance that would work best.

Great idea. I'll add that extra binding post. Very handy