GB Battery +/- polarity confusion

[Nowhere-Man]

I'm just making a replica Drydex H.1001 Grid Bias battery and have got myself into a state of confusion  

It would seem there is still a bit of juice in the original upon which the replica is based, but the polarity is the opposite of what I expected. As you can see from the attached picture I have the positive DMM probe in the + terminal socket and the negative probe in the 1.5 negative socket, the voltage reads (pos) +1.15. Surely, it should read (neg) -1.15?
Wouldn't this be putting a positive voltage on the valve grid?

Can someone help this dim-wit please?

Nick.

From what I gather from my very old sets, the Grid Bias battery would be connected + to chassis and - to grid. Its purpose, of course, is to stop the grid becoming too positive. It can be a bit confusing at first as both HT and LT are shown as + - connections with - to chassis but the GB battery is shown differently.

One thing that had me thinking was also there was a need to use a grid leak as well. And now it starts to get a bit more involved.

Starting off with the bias on the grid, most modern mains sets will do this a large resistor in series with the cathode and leading to chassis. This prepares to set the bias. The end of the resistor close to cathode will be more positive and the chassis end will be negative and can be tapped into. Add an RF bypass capacitor and that eliminates any R.F. there So, now you can add another resistor - the Grid Leak. In the very early sets this would go between the grid and the + filament. It seems any correct positive connection will do the job at the correct voltage to allow electrons to escape from the grid if it becomes too negative. So, in a way, it's a bit like another bias on an established bias - all to make sure the grid doesn't go out of wack

Then it starts to tie