The River, the Leaf and the Lightning: The Shape Family in Branching Forms
When a flow must spread across a surface or drain from it, the medium branches, and the pattern is set by the geometry of the problem rather than the substance doing the flowing.
From the airplane window
Look down from thirty thousand feet over a dry country and the rivers draw trees on the ground: a trunk, limbs, twigs, finer and finer toward the hills. Look at a leaf held to the sun and the veins draw the same tree. A lightning strike photographed at night draws it again, and so does the human lung if you could see it whole. None of these has anything in common in what it is made of. They have in common a problem.
The problem is distribution. Something must be carried across a surface or collected from it — water to the sea, sap to the leaf's edge, charge to the ground, air to the blood. Wherever that problem is posed, the medium branches, and the branching has a recognizable family of shape.
The acrylic witness
The cleanest witness is the one with no life in it. Fire a beam of electrons into a block of clear acrylic, let the charge build, then tap the block. The charge escapes in an instant and burns its path into the plastic: a Lichtenberg figure, a frozen tree of light branching out from the point of release. Acrylic has no biology, no hydrology and no growth. It branches because the geometry of the discharge admits branching and refuses a flat, even spread.
That is why the figure belongs in this family. It removes every excuse. There is no gene, no river bed, no history of evolution in the block. There is a medium, a flow it must carry, and the shape the carrying permits.
Set by the constraint
The branching pattern is set by the geometry of the constraint, not by the substance. A river network settles toward routes that drain its basin with least loss; a vascular tree tends toward branchings that deliver with least work; a discharge follows the field toward the paths that open first and feed themselves. The same instability seen in the growing crystal of Chapter 86 runs here: a branch that reaches a little further gathers a little more, and grows.
The honest limit is that least-cost principles explain much branching but not all. A tree's limbs carry its developmental history, and some biological networks are constrained by the order in which they were built. The weaker claim is enough, and it is the one the chapter makes: the family recurs across media sharing nothing but the geometry of spreading.
The catalog closes
With branching, the gathering that began with light and water is complete for this volume: the galaxy above, the pan and the crystal on the bench, the solid and the growing sheet, and now the tree drawn by every flow that has to reach. Across crystal growth, soap films, elastic sheets, chemistry and discharge, the disciplines do not share a department, and they already share the admission rule. Geometry sets which shapes can amplify.
The next chapter turns from the catalog to the way of seeing it — how a theorist recognizes a shape that persists across media before the proof or the language has caught up, and how the gathered family turns toward evolution as the process running on the rule.
Equations borrowed
- Lichtenberg figures in electron-irradiated acrylic and dielectric breakdown models of branching discharge.
- Optimal channel network theory for river basins (Rodríguez-Iturbe and Rinaldo).
- Murray's law and minimum-work accounts of vascular branching; West, Brown and Enquist on space-filling networks.
- Companion chapters: 85 ‘The Arms Are Not Made of Stars’, 86 ‘The Shape Without a Template’, 87 ‘The Sheet That Grows Its Own Fold’.
Validity band
Branching flows across surfaces and volumes where distribution or drainage dominates. Developmental history and material limits shape many biological networks beyond what least-cost accounts predict. The family claim concerns the shared constraint, not a shared mechanism.
Falsifier
Branching leaves the family if networks with the same distribution problem in unrelated media were shown to have no shared statistics of shape — if river, discharge and vein branching differed as much as unrelated patterns do. Measured branching statistics run the other way, within stated limits.
Where this chapter is weakest
Branching is so familiar that it invites the loosest kind of analogy; the chapter guards the line by naming the shared problem rather than claiming a shared cause. References should be matched to specific papers before publication.
The volume-wide audit of these weak points is collected in Where This Volume Is Weak.