![]() |
![]() ![]() ![]() The above figures are two of the most interesting artifacts
from my career in mathematics. Mainly they're just nice to look at. But they share a
special property as well. Don't know what it is? Well, here's a hint: The center
square in the right-hand figure should really be a sixth color, instead of simply red.
Okay, here's another hint. Think about the discriminantal arrangement of
hyperplanes determined by the generating vectors (including multiplicities) of the
zonotopes, and the natural connection between chambers of the
discriminantal arrangement and the set of tilings of those zonotopes. It's easy to
see it now, right?
Okay, so by now you're probably sorry you asked. The answer is that each of these geometric
doohickeys is an example of an incoherent tiling of a 2-zonotope. In other words,
there is no convex (ball-like) three-dimensional object which looks like either of these
things when you take its picture. An "easy" way to see this for the figure on the right
is to imagine that each set of borders shaped like >- is an arrow. Then going around
the outside, there's a collection of arrows going counterclockwise around the figure,
with the head of one arrow pointing at the tail of the next. If there really were a
ball-like object which looked like this when you took it's picture, it would have to
have infinite depth. In short, it's an optical illusion, like M.C. Escher's print
Ascending and Descending. The two-dimensional representation makes it look like
the object can exist in three dimensions, but it really can't.
![]() A similar thing is true for the figure on the left, only now the arrows look like >->-. If you're a real hard-core mathematician (or if you're having trouble sleeping), you can actually read my entire thesis, which Vic is nice enough to maintain under his web page. It's in postscript format. Back to Dave's Room |