In the real world, crystals grow, and they start off by a single nucleus, may be a single atom, then two, then 3, then 4 then 5 etc. For such small clusters, the six-fold symmetry, the space-filling pattern, may not be the stable configuration, it may not even be defined. Now, it turns out that growth of facetted crystals requires nucleation events on each facet in order to grow one new facet. This is to say that, like onion skins, each new layer is difficult to grow, because you need a new nucleation event on each facet. ...Actually, this is true only of regular space-filling crystals. A kinetic study shows that a pentagonal nucleus requires only a single nucleation event to grow a new layer. This is a subtle geometrical fact, related to how the atoms are positioned along corners of the crystals. In a hexagonal or a cubic lattice, a new nucleation event is required along each facet, because the wave of depositing crystals on each facet cannot propagate from one facet to an other, when it has reached the edge (it cannot turn the corner). While on pentagonal germs, the crystallization wave along one facet can propagate "around the corners". Therefore, pentagonal macles are kinetically favored.

For me, animals are some sorts of soft macles. This is to say that exactly the same physical rules, may select a complex 3D form, composed, in detail, of the association of parts in which all the positions of the entities are not equivalent, although they follow a single physical rule. An example of consequence, is that the tetrapod body may be viewed as a 4-fold macle, whose pattern is inherited from the germ of 4 cells which propagates "upwards" in the form of the tetrapod body.