Mystery RCA Transistor Part Number

I've searched but haven't found it in any RCA documentation.

It's not for a repair. I have 3 of them, and wanted to know the intended application and specification.

I thought perhaps the part number was a code that RCA used to obfuscate the actual part number, and that over time, its equivalent became known.
 
The 5 digit part RCA part numbers were their own. Some of them were industry standard (ie 2N) types tested to higher specification, and some were RCA’s own with no prototype 2N type. Numbers starting in 3 were ancient. Doesn’t surprise me that people selling old stock have no data at all, and that they’re priced through the roof. They figure if somebody needs one they will pay. And the sale of just one pays for their entire lot.
 
Do you happen to know how ancient? Could it be an alloy junction transistor?

I measured all 3 on a curve tracer. Two seem fairly standard, and have 50 and 80 hFE (Ib 100uA, Vce 5V) . The third is odd. The hFE increases dramatically at Vce 12V with low base current (4uA) to 400ish. Very nonlinear increase. All must have a low Vce rating as they experience a forward avalanche breakdown at 15V.

Thankfully, the ones I have were free along with many others discarded by a surplus guy I know.
 
The 3’s were back when germanium was still in widespread use. The only 3’s I have are Ge PNP TO-3’s. I don’t have any data on those either, but I tried a pair in a classic totem pole circuit on 26V and they worked fine. Silicon was around, but it’s not what it is today. Alloy junction would be consistent with the low vceo. They probably didn’t even know how to grow epitaxy yet. Or at least not very well.