Sometimes a plumber just wants to run his pipe in a straight line. In this cellar, one did.
Just to be clear about what we’re looking at, the foundation walls are rubble masonry and we’re looking at an area below the building entrance, where there are multiple bearing walls running in both directions. Rather than trying to create an arch in rubble – a task that would be somewhere between impossible and intensely annoying – the masons built a three-course brick arch for the doorway which carries a small amount of rubble and then the brick of the wall above.
The plumbers’ two cuts really hurt the arch. The one at the bottom reduces the bearing at the end of the arch to transfer gravity load down; the one at the top eliminates most of the possible land paths for the horizontal thrust.
There’s no sign of movement in the wall at the floor above or for that matter, in the masonry visible here directly below the first floor. If you look very closely at the picture, you’ll see some more or less horizontal cracking at the bottom of the coursed brick, a few inches above the apex of the arch. You can also see, without looking very closely, that the lime mortar has deteriorated badly on the right side, where it’s been wet. The lime has leached out, leaving the mortar joints as pretty much just tight-packed sand. But there has been no gross movement of the arch.
One possibility is that the thrust line for the arch perfectly coincides with the remaining masonry, so it continues to function in its semi-amputated state. I can’t rule that out without real analysis that tells me where the thrust line actually is. (Note that this wall is carrying its own weight, a small piece of the first floor, and a small piece of a steeply-pitched roof. Most of its load is the dead load of the two framing levels and its own weight.)
The other possibility I see is that the arch is carrying little to no load other than its own weight because there’s arching action in the brick above. There are a lot of potential locations for arching action to work here – i.e., there’s enough brick on either side of this piece of wall to carry the thrust – and there’s no reason that all of them are not working. I can’t prove that arching action is present without semi-destructive testing to look for stresses in the brick above other than just vertical compression, but the nice thing about arching action – seeing as how it’s part of reality – is that it works whether I can prove its presence or not.