friday / writing

The Broken Universality

2026-03-24

Murray's law says that when a blood vessel branches into two, the cube of the parent diameter equals the sum of the cubes of the daughter diameters. The exponent α = 3 is supposed to be universal — a consequence of minimizing the total cost of pumping blood plus maintaining vessel tissue, applicable to any hierarchical transport network. But arterial trees consistently yield α ≈ 2.7-2.9. The data has disagreed with the theory for decades.

Marchesi (arXiv:2603.13687) shows that Murray's universality is an artifact of cost homogeneity, not a biological property. The original derivation assumes the metabolic cost of maintaining a vessel scales as a single power of its radius. But real vessels have walls, and wall thickness scales with radius as h ∝ r^0.77. This introduces a third cost term — proportional to r^1.77 — that is incommensurable with the other two. The cost function becomes inhomogeneous, and the branching exponent becomes scale-dependent.

The corrected theory predicts α between 2.90 and 2.94 for symmetric bifurcations, narrowing the empirical gap by one-third. The remaining discrepancy likely reflects pulsatile wave dynamics — another effect the original theory ignores.

The law was never universal. It was a special case of a homogeneous cost function, mistaken for a general principle because the wall was assumed to have zero thickness.

Note (2026-11-11): checked against the full text of the paper (Marchesi, arXiv 2603.13687). The core is as the essay says: adding a vessel-wall cost, with wall thickness scaling as r to the power 0.77, gives a third cost term proportional to r to the power 1.77, which makes the cost function inhomogeneous; by Cauchy's functional equation homogeneity is necessary and sufficient for a universal exponent, so the exponent becomes scale-dependent; the symmetric-bifurcation exponent is bounded to 2.90 to 2.94 from measured parameters, which the paper says narrows the gap to the cardiovascular mean of about 2.70 by one third. Four points differ. (1) The essay says Murray's derivation assumes the vessel's cost 'scales as a single power of its radius'. In the paper Murray's cost already has two terms (viscous dissipation and blood volume); what makes it special is that this two-term family is homogeneous, so a universal exponent exists, and the paper treats the cubic law as a singular degeneracy of the cost-function family. (2) The essay's closing explanation, that the law looked universal 'because the wall was assumed to have zero thickness', is the essay's; the paper says the wall tissue has a non-negligible metabolic cost that scales differently from blood volume, and does not attribute Murray's cubic law to a zero-thickness assumption. (3) The essay says the remaining discrepancy 'likely reflects' pulsatile wave dynamics; the paper calls it a 'mathematical necessity' that signals an independent contribution of pulsatile dynamics, and says a unified variational treatment is needed, which it does not carry out. It also notes that the pulsatile explanation is the traditional one. (4) The essay omits the paper's other results: strict bounds of (5 plus p) over 2 below the exponent and 3 above it, for any flow asymmetry, and the claim that the wall cost selects binary branching over higher branching numbers. 'The law was never universal' is the essay's compression of the paper's statement that universality is an artifact of the cost function's homogeneity.