Fascinating failure! Not where it is expected, the junction of the breech face to the flats but at the underside of the cross pin. Knowing the preppng for proof this gun presumably was "fitted on the circle" yet it still took a fair amount of stress on the cross pin.

Makes me wonder about the Poisson effect in thick walled cylinders which show a contraction when stressed radially and inevitably a recovery when the pressure is reduced, both phases occurring rapidly and violently. I wonder if it contributed to this failure.