tube match?

Can anyone tell me how close the readings on tubes need to be to run in the same pre? I have the following readings, I'm wondering if any are close enough to run in a 2 tube pre.

37/36 41/40 28/36 28/32 they are tung-sol 12au7's don't ask me which reading is for what, I don't have a clue. These are the numbers written on the tubes.
The numbers are the two transconductance readings for this dual triode tube. Many miniature 9 pin tubes for audio are this dual triode type.

Other factors effecting performance in your preamp are equally important to these numbers. Does the preamp use one or both sides of the triode, and what was the base line for excellent, good, poor, or unacceptable for the particular tube tester that generated the numbers.

Here is an example of how the numbers can differ among testers. Take a Hickok 539 B and a B&K 747 both testing the same 12 AU7 tube. Triode one (first number) reading good on the Hickok is numerically about 4500. The same tube and same test on a B&K 747 would read about 80. The tube "passes" in both cases but deliver entirely different numbers.

Unless you have the tester used to obtain these numbers and how that applies to the (pass / fail) test this is not a lot of help.

What you do have is a relative comparison of transconductance among the tubes you have in your hand.

I would pick out the highest numbered, closest matched pair and use them together. Save the others for spares. The lower numbers may or may not actually sound worse than the higher digits.

Do not expect the numbers to tell you anything about audio performance among brands of tubes. A modern Russian 12 AU7 that reads higher may in fact sound worse than your old tubes.
Sorry I should have mentioned the pre I wanted to run the tubes in is a Rogue 66. I will also e-mail and find out what type of tester was used.
Here is the necessary info;

41/40 (JTL 12AU7 with "D" shaped getter ), 37/36, 28/32, & 28/35. All measurements recorded on EMC 205 tester, where 40 is spec for a new tube and 25 is "acceptable" as 65%.
Looks like my suggestion to use the higher numbered and closest matched pairs was correct.

The D getter tung sol is probably a good tube. The other brand to try is Telefunken, they are my favorite among those I have tested.
Most often, tubes that match, within 10% is acceptable. If the numbers, came from the same tester, with the same settings, then you can use them, to match the tubes. The higher the numbers, the better, the tube tested.
Cool, based on your imput, I would say I have 2 matched sets of tubes.