... or the later dual-circuit master (image from Steve Shoyer).

... but definitely was with the early North American dual-circuit (image from Dave's Cars).

AAU7152 with the two concentric rings indicated on the cylinder body on the mounting flange side. The other arrow is indicating an alternative but unspecified identification mark (image from Parts5)

GMC 1007 for a 4-cylinder (image from Sussex Classic Cars)

GMC 1011 for the factory V8 and also some Midgets (image from eBay)

GMC1011 installed on Hans Duinhoven's MGB. Enough room to access the cap even though it is an LHD so the clutch master is more tucked into a corner of the engine compartment that it would be on an RHD. Hans's car is a 4-cylinder, so strictly speaking should have the GMC1007. Originally the V8 master was 1.2mm smaller bore than the 4-cylinder, which with a 4-cylinder slave would give a slightly lighter pedal but less throw, i.e. the biting point would be a bit nearer the floor if the remainder of the mechanical linkage at the master remains the same.

Is this the one that causes problems with the unboosted dual ciruit master? Whilst the main casting and the diameter of the reservoir seem to be much the same as for GMC1007, the cap is obviously much bigger. (Image from Victoria British)

Showing how fluid fills the space between the two seals when the pedal is operated. If the secondary seal (nearest the pedal fulcrum) leaks you will get fluid running down the pedal, but the clutch otherwise operates normally (until all the fluid is lost!).

Showing how the primary seal must clear the bypass hole to allow fluid expansion and contraction to pass in and out of the reservoir. If this doesn't happen and the fluid expands, pressure will be applied to the release bearing which could cause the clutch to slip and will definitely accelerate wear, as well as rendering it impossible to bleed.