I ran into something similar.

I was trying to build a period (1920s or 30s) Hartley oscillator transmitter, using a type 210.

The tube I used was special... an original DeForest brand with a white ceramic base. It is

beautiful, and aesthetically PERFECT for the rig I was building.

The oscillator started and ran all right, but that big dome globe was filled with a yellow / purple

glow completely surrounding the plate.

Startled, I backed off of plate voltage on my bench supply; the glow died down somewhat,

and I let it run at reduced plate voltage while I investigated.

As I probed with the Simpson (I suspected a bias voltage problem for some reason), I noticed

that the plate glow was gradually reducing. Cautiously, I inched up the plate voltage again to

get more of the glow... same thing; in a few minutes the glow was reduced as the oscillator ran.

Repeat the plate voltage increase, gradually working to the intended Ep... and within 15 - 20 minutes,

the oscillator was running nicely at 450 VDC with no signs of glow.

I came to the conclusion that the tube was indeed gassy, but running it with a substantial plate

current caused enough heat to reactivate the tube's getter, clearing the gas contamination.

That was about 10 years ago. I still have the same breadboard transmitter, with the same DeForest

210.  Occasionally I fire it up on 80 meters, and the tube still works nicely.

You might give the rig a try with a variable plate voltage supply. but the easiest method in your case

is to just swap out tubes; 1625s are cheap and common.

Mr. T., W9LBB