Maybe none. Color is often indicative of the temperature seen by the tube and comes from the formation of a thin oxide passive film - the thickness of the surface film forms the color you see and is generally a function of highest temperature the tube saw.
XL83NL is correct. I held back because there has to be enough time at highest temperature for the film to grow and passivate, so the tube could have seen a higher temperature than the color indicates but only for a brief time. I never really thought about how the discoloration would differ in different mediums (I just assumed it would be the same for any oxygen-bearing medium). Do you have a reference on the subject?
The oxygen-bearing medium is what XL83NL is referring to, according to me.
If it happened in an inert atmosphere, there would be no discoloration, no matter how high the temperature went.