Kinda guessing, but I'd assume from temperature - i.e. chaotic (a.k.a. brownian) motions of the particles resulting in them randomly getting close enough to the

Kinda guessing, but I'd assume from temperature - i.e. chaotic (a.k.a. brownian) motions of the particles resulting in them randomly getting close enough to the enclosure's walls that the surface tension (which from what I see on Wikipedia seems to be basically electric attraction of particles) "glues" them to the wall, and/or to particles that got already "glued" to the wall just moments before. I would suppose "in real life" the resulting cooling is probably basically irrelevant and gets immediately "replenished" from the surroundings, and (again also from what Wikipedia seems to say) the equilibrium seems to be reached vs. gravity, presumably mostly of the "non-glued to walls" part of the water column. I.e. I guess when the freeriding bunch of suckers in the middle of the column pull you down more than the sweet, sexy wall pulls you up... errr, I mean, the sum of the gravity force acting on the water molecules in the middle of the water column equals the sum of the wall-sticking force of all the wall-sticking water molecules (distributed locally through the water-to-water-sticking force of the water molecules).