The difficulties with corner welding... Couple things going on and it depends if it's AC or DC. DC currents are only capable of establishing electric fields whereas AC currents produce both an electric and magnetic field (with a proportion of about 80% electric, with a quadrature sinusoidal collapse pattern). The self created magnetic field from AC welding is innumerably small and has no measurable effect on the electrons path. A fixed magnet such as a clamping magnet is hundreds of millions of times stronger than the self induced mag field.
Common to both AC and DC welding are two phenomenon. 1) Current always finds the shortest path of least resistance. 2) Sometimes it takes a moment for the current to find the shortest path (due to linear momentum). Suffice it to say, when the electrode is near two edges, the electrons are trying to figure-out which way to go and they bounce around a little. Also, shortest path does not necessarily mean shortest linear distance. If there is scale etc, even a more distant path might have less resistance so, the electrons decide to go that route instead. ADDENDUM: Heat effects resistance. As heat increases, so does electrical resistance (ask me and I'll tell you why)... On an area of a weld that is really, really hot, you might find the arc jumping to somewhere else even though the electrode is closer. This is because the localized resistance is so great, the current takes a longer path but of least resistance.
Ray
EDIT: If you've ever received a little electric shock, you're probably familiar with first a sensation throughout your body then finally, a sudden sting on one localized area such as your hand etc. There are two reasons for this. 1) Capacitive discharge: Your body has many free electrons and they all start to move around when the current enters your body. This is that overall tingling sensation. 2) They figured-out the shortest path and are now traveling enmasse through one area and it's burning the hell out of some cells in your body.
EDIT (again): Someone will probably take exception to DC currents only producing electric fields. There are two distinct cases. 1) DC current traveling in a wire can indeed produce a magnetic filed. 2) We're not talking about the case of #1. We're talking about the case when an open gap has non alternating voltages at the terminals. This is a very different case and is effectively a capacitor where either air or the electron plasma itself is the dialectric. In this case, there is, can only be, an electric field. If you want to know the proof, let me know but, you will need several years knowledge of advanced algebra, calculus and matrix(vector calculus) theory. -Which are the fundamentals of Maxwell's equations.
