Smoke........ I thought that you know everything.
He does, but is too modest to admit it. I am Mr. Knowitall (in theory).
I have said it many times before that swelling will occur if the auxin flow is interrupted. The only auxin that matters is the auxin transported from one cambium cell to another, so it comes down to interrupting the cambium bucket brigade. This is exactly what happens when girdles a stem for a layer OR applies a tourniquet to it OR tightly wraps wire around it OR threads the stem though another (i.e., a thread graft).
The magical part of layering is that excess auxin in the cambium can make stem cells morph into root tissue! It is not guaranteed this magic will happen as it seems to go away with maturity of the tree and distance from the ground level in many species, but if it happens, it happens. If the beginnings of roots emerge into a moist environment, they can continue growing into roots. If not, they get air pruned. Hence, the long-standing advice to keep grafts dry. Conversely, pack sphagnum around a branch with which wiring has 'bitten in' and you might get roots. Hence the flare that occurs at the top of a girdle and/or above a tourniquet. Keep it dry and there will not be roots; damp, likely.
So, when roots are growing from the swelling, auxin is consumed and the tendency to swell is lost because there is no longer any interruption of the cambium and hence no accumulated excess of auxin in it. Which brings us to an idea: if you want flare, apply a tourniquet (or tightly wrapped spiral of wire) and keep it dry until you've gotten what you want (or it is clear that is all you're going to get). Then pack it in a moist environment of sphagnum, pearlite, your favorite substrate, etc., to generate roots.
In many respects, it would seem to make sense that screwing a layer or flat bottomed trunk to a board might help generate flare. My results have been mixed. So, I am skeptical that it does much, in fact. On the other hand, I've had acer palmatum layers produce just one or two sizable roots just on one side (no roots anywhere else around the stem). Nevertheless, a big fat 'skirt' formed around the entire periphery of what was the top of the girdle. It takes several seasons, but roots will and/or can eventually form around the entire periphery.
I haven't tried any of this specifically with elms, but the basics are the same. I've got a multiple trunks threaded through a pot bottom similar to
@Smoke's experiment, but I'm using horse chestnuts. I had some flare and a few roots above the pot bottom from last year, this spring, so I don't think I'm going to get a whole lot of flare - should have left it above ground and dry last year, IMHO (hmmm, I could go out there and cut those off; just a question of whether the old roots are nearly starved to death or not).
