Notes: Chapter 3: The Constructal Law

Chapter notes for “Chapter 3: The Constructal Law”

Notes

¹ Adrian Bejan, “Constructal-theory network of conducting paths for cooling a heat generating volume,” International Journal of Heat and Mass Transfer 40 (1997): 799-816. For accessible treatment, see Adrian Bejan and J. Peder Zane, Design in Nature: How the Constructal Law Governs Evolution in Biology, Physics, Technology, and Social Organization (2012).

² Xiangyi Meng, Albert-László Barabási, et al., “Surface optimization governs the local design of physical networks,” Nature 649(8096) (2026). The research identified an exact mathematical mapping between biological network optimization and high-dimensional Feynman diagrams in string theory — not a claim of physical similarity, but a demonstration that the same mathematical structures describe vastly different phenomena.

³ The 98% synapse rate for orthogonal neural sprouts comes from the Nature study’s analysis of human brain data. Right-angle branches optimize surface area while reaching nearby targets, and those targets become connections. The geometry encodes function.

⁴ Amruthesh Thirumalaiswamy, Clary Rodríguez-Cruz, Robert A. Riggleman, and John C. Crocker, “Slow relaxation and landscape-driven dynamics in viscous ripening foams,” Proceedings of the National Academy of Sciences 122(47) (2025): e2518994122. doi:10.1073/pnas.2518994122. The Penn Engineering team found that bubbles in wet foam traverse configuration space using mathematics highly analogous to gradient descent in deep learning. Both systems benefit from exploring flat regions rather than settling into deep valleys — flexibility outperforms rigid optimization.

^4b Pahlavan, A.A. et al., “Restoring universality to the pinch-off of a bubble,” Proceedings of the National Academy of Sciences 116 (2019): 13780–13784. In unconfined fluid, bubble breakup retains memory of initial conditions — different nozzles and flow rates produce different pinch-off shapes (non-universal). In narrow tubes (< 1 mm), a first self-similar regime erases memory, after which a second regime completes the break. The overall process becomes universal regardless of tube diameter, viscosity, or formation conditions.

^5a Simonin, K.A. and Roddy, A.B., “Genome downsizing, physiological novelty, and the global dominance of flowering plants,” PLOS Biology 16(1) (2018): e2003706. The study compiled data on genome size, cell size, cell density, and photosynthetic rate for hundreds of angiosperms, ferns, and gymnosperms, demonstrating that genome downsizing triggered a cascade from smaller cells to higher stomatal and vein density to dramatically increased photosynthetic capacity — the constructal mechanism underlying the flowering plants’ ecological dominance.

^5b Gregory, T.R., “A Bird’s-Eye View of the C-Value Enigma: Genome Size, Cell Size, and Metabolic Rate in the Class Aves,” Heredity 87 (2002): 563–571. Gregory demonstrated the same genome-to-metabolism cascade in birds: smaller genomes → smaller nucleated red blood cells → more efficient oxygen transport → higher metabolic rates needed for flight. Hummingbirds have the smallest avian genomes; flightless birds the largest.

^5c Craig R. White, Dustin J. Marshall, Lesley A. Alton, Pieter A. Arnold, Julian E. Beaman, Candice L. Bywater, Catriona Condon, Taryn S. Crispin, Aidan Janber, Elia Pirtle, Hugh S. Winwood-Smith, Michael J. Angilletta Jr., and Craig E. Franklin, “Metabolic scaling is the product of life-history optimization,” Science 377(6608) (2022): 834–839. DOI: 10.1126/science.abm7649. The study presents a mathematical model of animal growth in which organisms allocate energy between growth and reproduction as they age. When lifetime reproduction is maximized, allometric (three-quarter-power) metabolic scaling emerges as the optimal strategy, without invoking physical or geometric constraints on supply networks. The result does not invalidate West’s fractal geometry account; rather, it shows that the same scaling can arise from a deeper principle (optimization under natural selection), of which fractal distribution networks are one possible implementation. The implication for constructal theory: the quarter-power pattern may be more fundamental than any single mechanism that produces it.

^5d Louis Jolyon West, Chester M. Pierce, and Warren D. Thomas, “Lysergic Acid Diethylamide: Its Effects on a Male Asiatic Elephant,” Science 138(3545) (1962): 1100–1103. The experiment took place on August 3, 1962; the dose, 297 mg delivered intramuscularly by dart, remains the largest single dose of LSD ever given to any animal. Twenty minutes after the seizures began the team injected 2,800 mg of promazine hydrochloride, and roughly an hour later an unrecorded quantity of pentobarbital sodium intravenously. Later commentators have argued that the promazine was the more probable proximate cause of death; the published record does not settle the question. That the dose itself was the error is separately supported: Ronald K. Siegel administered LSD to two elephants at far lower doses in a 1984 study of musth behavior, and both animals survived (Siegel, R.K., “LSD-induced effects in elephants: Comparisons with musth behavior,” Bulletin of the Psychonomic Society 22 (1984): 53–56). On the experiment’s provenance: West held a CIA contract at the University of Oklahoma (MKULTRA Subproject 43, “Psychophysiological Studies of Hypnosis and Suggestibility”), which is why the Tusko study is often described as an MKULTRA experiment. The published paper frames it as an investigation of musth, the periodic aggressive state of male elephants, and the CIA connection is to the investigator rather than to this protocol.

⁵ Geoffrey West, Scale: The Universal Laws of Growth, Innovation, Sustainability, and the Pace of Life in Companies, Cities, and People (2017). West’s work extends Kleiber’s 1932 metabolic scaling law (Max Kleiber, “Body size and metabolism,” Hilgardia 6 (1932): 315-353) to urban and organizational systems, demonstrating that the same power-law relationships that govern biological efficiency appear in human-built infrastructure. The approximate coordination limits discussed in the text are inferred from the scaling patterns West describes, rather than a specific finding reported in Scale; they are consistent with the constraints that three-dimensional branching optimization imposes on hierarchical organization.

⁶ A.A. Galakhova et al., “Evolution of cortical neurons supporting human cognition,” Trends in Cognitive Sciences 26 (2022): 909-922. The paper synthesizes findings on gradients in cortical structure from sensory to associative areas, documenting how neuron size, dendritic complexity, and spine density increase along the processing hierarchy.

⁷ Suzana Herculano-Houzel, “The remarkable, yet not extraordinary, human brain as a scaled-up primate brain and its associated cost,” Proceedings of the National Academy of Sciences 109 (2012): 10661-10668. Human brains contain approximately 86 billion neurons — precisely what allometric scaling predicts for a primate brain of our mass. The “special” nature of human cognition cannot be explained by neuron count alone.

⁸ H. Mohan et al., “Dendritic and axonal architecture of individual pyramidal neurons across layers of adult human neocortex,” Cerebral Cortex 25 (2015): 4839-4853. Human supragranular pyramidal neurons are roughly three times larger than their mouse equivalents and show qualitatively different dendritic structure — architecturally distinct, beyond mere scaling.

⁹ D. Beniaguev et al., “Single cortical neurons as deep artificial neural networks,” Neuron 109 (2021): 2727-2739. Detailed computational models demonstrate that the dendritic complexity of individual pyramidal neurons allows them to perform computations equivalent to multi-layer neural networks with 5-8 hidden layers.

^9a The quantitative comparison (five to eight layers, ~1,000 artificial neurons for 99% millisecond-level accuracy) is from Beniaguev et al. (2021), note 9 above. The authors tested many architectures and found the result robust: fewer layers consistently failed to capture the dendritic computations. They have shared their code to encourage attempts with fewer layers; none have succeeded. See also Allison Whitten, “How Computationally Complex Is a Single Neuron?” Quanta Magazine (September 2, 2021) for discussion by Lillicrap, Tolias, and Segev.

^9b Joshua Jacobs, Salman E. Qasim, and Lukas Kunz, “Phase precession in the human hippocampus and entorhinal cortex,” Cell 184 (2021): 3242-3255. First identification of phase precession in human spatial navigation, using data from epilepsy patients with implanted electrodes navigating a virtual environment. Phase precession was observed in 12% of monitored neurons. For non-spatial extensions: Qasim et al., “Phase precession and variable spatial scaling in a periodic attractor map model of hippocampal place cells,” eLife 10 (2021); and Liang et al. (preprint, 2021) on phase precession in serial image processing. The phenomenon was first documented in rats by O’Keefe and Recce, “Phase relationship between hippocampal place units and the EEG theta rhythm,” Hippocampus 3 (1993): 317-330.

^9c Zheng, Jieyu, and Markus Meister. “The Unbearable Slowness of Being: Why do we live at 10 bits/s?” Neuron 112 (2024). doi:10.1016/j.neuron.2024.11.008.

¹⁰ G. Testa-Silva et al., “High bandwidth synaptic communication and frequency tracking in human neocortex,” PLoS Biology 12 (2014): e1002007. Human synapses recover from synaptic depression approximately three times faster than rodent synapses, enabling ninefold higher information transfer rates during sustained activity.

^11a Eleni Katifori, Gergely J. Szöllősi, and Marcelo O. Magnasco, “Damage and fluctuations induce loops in optimal transport networks,” Physical Review Letters 104 (2010): 048704. See also Katifori and Magnasco, “Quantifying loopy network architectures,” PLoS ONE 7 (2012): e37994. The authors modeled flow networks under two pressures, steady-state efficiency and resilience to random link failure, and showed that the optimal architecture shifts from pure tree to hierarchically nested loops as the damage probability increases. The Tokyo slime mold experiment: Atsushi Tero et al., “Rules for biologically inspired adaptive network design,” Science 327 (2010): 439–442.

^11b Pablo Blinder, Philbert S. Tsai, John P. Kaufhold, et al., “The cortical angiome: an interconnected vascular network with noncolumnar patterns of blood flow,” Nature Neuroscience 16 (2013): 889–897. Kleinfeld’s team mapped the complete surface vasculature of the rodent somatosensory cortex, confirming the random loop lattice and quantifying its redundancy. For the vulnerability of penetrating arterioles: Nishimura, N. et al., “Penetrating arterioles are a bottleneck in the perfusion of neocortex,” PNAS 104 (2007): 365–370.

¹² J.J. Hutsler et al., “Comparative analysis of cortical layering and supragranular layer enlargement in rodent carnivore and primate species,” Brain Research 1052 (2005): 71-81. The disproportionate expansion of layers 2 and 3 across primate evolution reflects increasing allocation of cortical resources to cortico-cortical integration rather than direct sensorimotor processing.

¹⁴ M.A. Topinka et al., “Coherent branched flow in a two-dimensional electron gas,” Nature 410 (2001): 183-186. The discovery that electrons in semiconductors spontaneously form branching filaments through smooth random potential variations, rather than spreading diffusely, established branched flow as a distinct transport phenomenon.

¹⁵ A. Patsyk et al., “Observation of branched flow of light,” Nature 583 (2020): 60-65. The first demonstration of branched flow in optics, using soap films as the randomly varying medium. The phenomenon had been theoretically predicted but observing it required only a laser pointer and careful attention — technology available for decades before anyone thought to look.

¹⁶ E.J. Heller et al., “Branched flow,” Physics Today 74 (2021): 44-51. This review traces branched flow across scales from quantum electrons to ocean waves, documenting the tsunami connection and establishing the phenomenon’s universality. The same mathematical framework describes focusing in semiconductor nanostructures and in thousand-kilometer oceanic wave propagation.

¹⁷ Jane Jacobs, The Nature of Economies (2000). Jacobs argues that economic development follows the same principles as biological development — not by imitating nature but by being a form of natural development. The quotes are from Chapter 1 (“Evading the Issue”) and Chapter 4 (“Development”). Spencer Beebe, founder of Ecotrust, met with Jacobs in the mid-1990s when she was writing the book; she served on Ecotrust’s board for several years. Beebe describes Jacobs’ framework as foundational to his work on “conservation economy” approaches to forestry and fisheries. See Spencer Beebe, “Biomimicry in the Rainforests of Home,” Long Now Foundation seminar (2005).

²⁰ Franziska L. Sendker, Yat Kei Lo, Thomas Heimerl, et al., “Emergence of fractal geometries in the evolution of a metabolic enzyme,” Nature 628(8009) (2024): 894–900. The catalytic form of S. elongatus citrate synthase is a triangular hexamer; these hexamers self-assemble into Sierpiński triangles of 18 or 54 subunits (first- and second-order fractals). The key to the fractal is that protein chains make slightly different interactions at different positions — asymmetric tiling that creates large internal voids rather than the regular lattices typical of protein oligomers. Cryo-electron microscopy and ancestral sequence reconstruction suggest the fractal-forming capacity arose from a small number of mutations, and genetic modification eliminating the fractal assembly had no measurable effect on cell growth. Research conducted at Max Planck Institute for Terrestrial Microbiology (Marburg), with structural data at the ESRF synchrotron and cryo-EM at Philipps University Marburg.

²¹ Alfred Russell Wallace, “On the zoological geography of the Malay Archipelago,” Journal of the Proceedings of the Linnean Society of London 4 (1860): 172–184. The Wallace Line was refined by subsequent biogeographers, notably Max Carl Wilhelm Weber’s 1902 faunal balance line and Richard Lydekker’s 1896 Australasian boundary. During Pleistocene glacial maxima, sea levels dropped ~120 meters, exposing the Sunda and Sahul continental shelves — but the deep-water Makassar and Lombok Straits between them never closed. Fifty million years of continuous deep-water separation produced one of the sharpest biogeographic discontinuities on Earth.

²² Javier Montenegro, Jessica Kolbusz, Yakufu Niyazi, Alan J. Jamieson, Joan J. Soto-Angel, Aino Hosia, Allen G. Collins, and Dhugal J. Lindsay, “An unexpected journey—the arctic deep-sea halicreatid trachymedusa Botrynema brucei ellinorae off Florida: a reassessment under an integrative taxonomic approach,” Deep Sea Research Part I: Oceanographic Research Papers 223 (2025), doi:10.1016/j.dsr.2025.104551. Montenegro’s team at the University of Western Australia’s Minderoo-UWA Deep-Sea Research Center used integrative taxonomy (morphological and genetic analysis) to show that knobbed and smooth forms of B. brucei ellinorae belong to the same genetic lineage despite their morphological divergence. The smooth form has never been recorded south of ~47°N, corresponding to the North Atlantic Drift transition zone. The knobbed form’s global range may be facilitated by the Deep Western Boundary Current, which connects Arctic deep-water communities to lower-latitude basins.

²³ Garret Sutherland, “Quantum-Physical Softmax via Rydberg Blockade: Experimental Validation of T³ Semantic Geometry on QuEra Aquila,” MirrorEthic LLC (2026, unpublished manuscript). Sutherland’s T³ cellular automata experiments demonstrated that cooperative coupling alone (1/r², “survival coupling”) produces autocatalysis but not differentiation. Adding competitive coupling (C₆/r⁶, equivalent to Rydberg blockade interaction) produced spontaneous differentiation and specialization. The dual-coupling requirement was independently confirmed on QuEra Aquila quantum hardware, where the van der Waals blockade mechanism implements winner-take-all dynamics identical to softmax competition in transformer attention layers. See also the Becoming Minds chapter for the substrate-independence implications.

²⁴ Farhad Yusef-Zadeh et al., analysis of the Snake galactic center filament (G359.1-0.2) using Chandra, XMM-Newton, NuSTAR, and radio telescopes (2025). The study identified pulsar G359.13 near the Snake’s major kink through X-ray and radio emissions, proposing that the filament’s deformation was caused by the pulsar passing through at 500–1,000 km/s. Galactic center filaments were first discovered by Yusef-Zadeh, Morris, and Chance in 1984; over a thousand have since been cataloged. The vertical filaments are perpendicular to the galactic plane, up to 150 light-years long, emit synchrotron radiation from relativistic electrons, and often appear in regularly spaced pairs. The pulsar wind nebula hypothesis for filament formation, that neutron stars interacting with interstellar magnetic fields generate the filamentary structures, remains preliminary, based on this single confirmed pulsar-filament association. See also Chapter 14 for the φm implications.

²⁵ Evan M. Gora et al., “How some tropical trees benefit from being struck by lightning: evidence for Dipteryx oleifera and other large-statured trees,” New Phytologist (2025). The team tracked 93 lightning-struck trees in the Barro Colorado Nature Monument, central Panama, using antenna triangulation to locate strikes, field sensors, and cameras to assess damage. All nine Dipteryx oleifera individuals survived direct strikes; other species suffered 5.7 times more canopy loss, with 64% dying within two years. Each strike on a Dipteryx killed an average of 9.2 neighboring trees and 78% of lianas, destroying 2.1 Mg of competitor biomass. The mechanism appears to be physical: high sap conductivity allows current to flow through the tree without explosive resistance, while the root architecture disperses energy laterally to surrounding vegetation. Over 40-year monitoring periods, proximity to Dipteryx was a significant hazard for all neighboring plant life — a pattern consistent with repeated lightning-mediated competitive advantage.

²⁶ David M. Romps, Jacob T. Seeley, David Vollaro, and John Molinari, “Projected increase in lightning strikes in the United States due to global warming,” Science 346:6211 (2014): 851–854. The study modeled lightning flash rate as a function of precipitation rate and convective available potential energy (CAPE), projecting approximately 12% more strikes per degree Celsius of warming. While focused on the continental United States, the underlying physics (warmer air carries more moisture, releases more latent heat during convection, drives stronger updrafts) applies globally, and the tropics, which already host roughly 70% of global lightning, are expected to see proportional or greater increases. The specific feedback mechanisms proposed for Dipteryx—reduced fire ignition, liana suppression, altered carbon dynamics — are inferences from established tropical forest ecology applied to the Gora et al. findings; no study has yet measured these feedbacks directly for Dipteryx-dominated forest patches.

²⁷ Eduardo Mercado III and Jessica Zhuo, “Do rodents smell with sound?” The hypothesis that rodent ultrasonic vocalizations serve vibro-acoustic particle manipulation rather than (or in addition to) social communication. Ultrasonic vocalization in rats was first reported by John Anderson in 1954 (“Production of ultrasonic sounds by laboratory rats and other mammals,” Science 119(3101) (1954): 808–809). For seven decades, USVs were interpreted primarily as social signals: distress calls, courtship vocalizations, emotional expressions. Mercado, whose background is in humpback whale song analysis, noticed that rats consistently initiated sniffing behavior immediately after ultrasonic calls, suggesting the vocalizations may function as active sensory enhancement rather than communication. The proposed mechanism, vibro-acoustic clustering of airborne particles at ultrasonic frequencies, is physically well-established in materials science but has not previously been identified as a biological strategy. If confirmed, this would represent the first known active olfactory sensor in any organism.

²⁸ Xin Wen, Qian Ma, Alessandro Mannino, Marivi Fernandez-Serra, Shengping Shen, and Gustau Catalan, “Flexoelectricity and surface ferroelectricity of water ice,” Nature Physics (2025), doi:10.1038/s41567-025-02995-6. The team at ICN2 (Barcelona), Xi’an Jiaotong University, and Stony Brook University prepared ice capacitors (layers of ultrapure ice frozen between gold-coated aluminum electrodes) and applied oscillating three-point bending to measure the resulting charge. The bulk flexoelectric coefficient (the voltage generated per unit of bending) is 1.14 ± 0.13 nC/m, comparable to strontium titanate and titanium dioxide. Below ~160 K, the surface (though not the bulk) undergoes a ferroelectric phase transition, with the coefficient peaking at 7.6 nC/m (fivefold amplification). The effect depends on electrode material: platinum electrodes produced a larger ferroelectric peak than gold, while aluminum showed almost none, implying electron transfer driven by the surface ferroelectricity. Above −25°C (248 K), the flexoelectric coefficient reverses sign, aligning with decades of experimental evidence on temperature-driven polarity reversal in thunderstorm charging. The researchers calculated that flexoelectric charge densities from ice-graupel collisions are comparable to observed charge transfer in thunderclouds, providing a candidate mechanism for the long-standing puzzle of thunderstorm electrification. For background on flexoelectricity as a universal property of dielectrics: Mieczysław Kaczmarek (Mesco Kadesh), review articles on flexoelectric coupling in crystalline and amorphous materials.

²⁹ Gajigan, A.P., Schvarcz, C.R., Laughlin, A.B., Weatherby, T.M., Culley, A.I., Edwards, K.F., and Steward, G.F., “A dinoflagellate-infecting giant virus with a micron-length tail,” bioRxiv preprint (2025). doi: 10.1101/2025.07.19.665647. PelV-1 (Pelagodinium lytic virus 1) was isolated from 25 m depth in the North Pacific Subtropical Gyre near Hawaii, infecting the dinoflagellate Pelagodinium. Capsid diameter approximately 200 nm; tail extends up to 2,300 nm — the longest viral appendage on record, surpassing the “Rapunzel” phage P74-26 (875 nm) and the Tupanvirus (~1,850 nm). The tail-to-capsid ratio of approximately 11–12x appears consistent across several unrelated tailed viruses, but has not yet been formally tested as a scaling law. PelV-1 exhibits at least five distinct morphological configurations, suggesting the tail may develop extracellularly after host cell lysis. A second co-occurring virus (co-PelV) was discovered during genome assembly. Research conducted at the University of Hawai’i at Mānoa.

³⁰ Smil, Vaclav, Energy and Civilization: A History (2017). MIT Press. Smil provides comprehensive data on per-capita energy consumption across historical eras, documenting the roughly hundredfold gap between biological metabolic rate (~90 watts) and the social metabolic rate of industrial civilization (~11,000 watts).

³¹ Alan Turing, “The Chemical Basis of Morphogenesis,” Philosophical Transactions of the Royal Society of London B 237 (1952): 37-72. Turing’s final major scientific contribution, published two years before his death. The activator-inhibitor mechanism he proposed has been confirmed across biological systems from embryonic digit formation to seashell patterns and bacterial colony spacing.

³² Du, K., et al., “Hybridization reveals divergence in reproductive compatibility between gar species,” Proceedings of the Royal Society B 289 (2022): 20221267. Documents the conservation of gar morphology over 240 million years and the high fertility of interspecific hybrids despite 105 million years of divergence, suggesting exceptionally efficient DNA repair mechanisms.

³³ Raichle, M.E. and Gusnard, D.A., “Appraising the brain’s energy budget,” PNAS 99 (2002): 10237-10239. The brain’s disproportionate energy consumption (20% of metabolic energy for 2% of body mass) reflects the thermodynamic cost of maintaining neural computation at the edge of chaos.

³⁴ Yuki Fuseya, Hiroyasu Katsuno, Kamran Behnia, and Aharon Kapitulnik, “Nanoscale Turing patterns in a bismuth monolayer,” Nature Physics 17 (2021): 1031-1036. The Turing mechanism at atomic scale, with vertical and in-plane displacements replacing chemical morphogens. The patterns self-heal — a property Fuseya connects to the robustness of the activator-inhibitor mechanism. For the broad range of Turing patterns across scales, see Irving Epstein and John Pojman, An Introduction to Nonlinear Chemical Dynamics (Oxford, 1998).

³⁵ Villaescusa-Navarro, F., et al., “Cosmology with one galaxy?” The Astrophysical Journal 929 (2022): 132. The CAMELS (Cosmology and Astrophysics with Machine Learning Simulations) project trained neural networks on ~1 million galaxies across 2,000 simulated universes with matter densities ranging from 10% to 50%. A network given only ~17 properties of a single galaxy predicted the parent universe’s matter density (Ωm) to within ~10%. The result held across galaxy morphologies and was replicated using two independent simulation codes (IllustrisTNG and SIMBA), though cross-prediction between the two recipes remains imperfect — indicating the network finds simulation-specific patterns alongside any universal signal.

^36d Gábor Domokos, Douglas J. Jerolmack, Ferenc Kun, and János Török, “Plato’s Cube and the Natural Geometry of Fragmentation,” Proceedings of the National Academy of Sciences 117(31) (2020): 18178–18185. The paper develops a mathematical framework classifying all possible mosaic patterns from fragmentation, showing that Platonic cubes are the generic three-dimensional attractor for random fracture. In two dimensions, the attractor is rectangles. The framework also describes exceptional cases (hexagonal Voronoi patterns arising from tensile stress or cyclic cracking) as distinct geological signatures within the same geometric vocabulary.

^37d The tectonic plate vertex count (5.77) matches the theoretical prediction for Voronoi tessellation on a sphere. A hexagonal mosaic on a flat plane averages six vertices per cell; spherical topology requires the inclusion of pentagonal cells (as on a soccer ball), reducing the average. That the actual plates match the spherical Voronoi prediction suggests the lithosphere cracked under tensile stress — consistent with models where Earth’s crust formed as a separate system that expanded, grew brittle, and fragmented, rather than being passively shaped by mantle convection alone. See Joshua Sokol, “Scientists Uncover the Universal Geometry of Geology,” Quanta Magazine (November 19, 2020), for an accessible account of the full research program.

³⁶ Dabiri, J.O., Colin, S.P., Costello, J.H., and Gharib, M., “Flow patterns generated by oblate medusan jellyfish: field measurements and laboratory analyzes,” Journal of Experimental Biology 208 (2005): 1257-1265. Foundational measurement of vortex ring formation during jellyfish propulsion, using digital particle image velocimetry in the field (Palau) and laboratory. Demonstrated that the wake structure of Aurelia aurita consists of a series of vortex rings whose geometry correlates with swimming efficiency. See also Dabiri, J.O., “On the estimation of swimming and flying forces from wake measurements,” Journal of Experimental Biology 208 (2005): 3519-3532, which formalizes the relationship between vortex ring properties and propulsive force across aquatic and aerial locomotion.

³⁷ Gharib, M., Rambod, E., Kheradvar, A., Sahn, D.J., and Dabiri, J.O., “Optimal vortex formation as an index of cardiac health,” PNAS 103 (2006): 6305-6308. Demonstrated that left ventricular vortex ring formation during diastolic filling follows a universal optimal formation number, and that deviations from this optimum correlate with cardiac pathology. Subsequent clinical studies have confirmed that vortex ring signatures in echocardiographic imaging can detect diastolic dysfunction before conventional structural markers become abnormal.

³⁸ Dabiri, J.O., “Potential order-of-magnitude enhancement of wind farm power density via counter-rotating vertical-axis wind turbine arrays,” Journal of Renewable and Sustainable Energy 3 (2011): 043104. Theoretical analysis followed by field validation at a test site in the Antelope Valley, northern Los Angeles County, using up to 24 vertical-axis wind turbines in fish-school-inspired configurations. Measured power densities approximately ten times those of conventional horizontal-axis wind farms. The mathematical correspondence between fish-school wake dynamics and vertical-axis turbine aerodynamics is developed in Whittlesey, R.W., Liska, S., and Dabiri, J.O., “Fish schooling as a basis for vertical axis wind turbine farm design,” Bioinspiration & Biomimetics 5 (2010): 035005.

³⁹ Tlili, S., Gsell, S., Merkel, M., and Lenne, P.-F., “Marangoni-like tissue flows drive symmetry breaking during mouse gastruloid elongation,” bioRxiv (March 2025). The team at Aix Marseille University observed cell flows in mouse gastruloids that match the Marangoni effect: genes produce an asymmetric protein distribution that lowers surface tension locally, driving tissue circulation and elongating the gastruloid into a head-tail axis. For the original description of the Marangoni effect: Thomson, J., “On certain curious Motions observable at the Surfaces of Wine and other Alcoholic Liquors,” The London, Edinburgh, and Dublin Philosophical Magazine 10:67 (1855): 330–333. See Demming, A., “Genes Have Harnessed Physics to Help Grow Living Things,” Quanta Magazine (October 2025), for an accessible account of the broader revival of mechanical explanations in developmental biology.

⁴⁰ Shyer, A.H., Rodrigues, A.R., et al., “Emergent cellular self-organization and mechanosensation initiate follicle pattern in the avian skin,” Science 357(6353) (2017): 811–815; updated in Shyer et al., Science (2023), establishing that morphogens influence tissue-level material properties rather than instructing individual cells, with mechanical forces producing the follicle spacing pattern. For D’Arcy Thompson’s original thesis: Thompson, D’Arcy Wentworth, On Growth and Form (Cambridge University Press, 1917; revised 1942). Thompson argued that physical forces (surface tension, mechanical stress, diffusion) shape organisms alongside natural selection, a position increasingly vindicated by modern biophysics. For the role of actin production in cellular elasticity: Doubrovinski, K., et al., “Measurement of cortical elasticity in Drosophila melanogaster reveals roles for cell shape changes and actin polymerization in tissue mechanics,” Physical Review Letters (June 2025).

⁴¹ Devauchelle, O., Petroff, A.P., Seybold, H.F. and Rothman, D.H., “Ramification of stream networks,” Proceedings of the National Academy of Sciences 109 (2012): 20832–20836. Demonstrated that groundwater-driven channel growth in humid landscapes converges on a characteristic bifurcation angle of 72 degrees (one-fifth of a circle), confirmed across 4,966 junctions in the Florida Panhandle.

^41a Seybold, H.J., Kite, E. and Kirchner, J.W., “Branching geometry of valley networks on Mars and Earth and its implications for early Martian climate,” Science Advances 4 (2018): eaar6692. doi: 10.1126/sciadv.aar6692. Junction angles sort by climate: networks in humid and permafrost regions peak close to 72 degrees, while arid, runoff-dominated networks branch far more narrowly (the mode for the Upper Colorado-Dirty Devil basin is about 41 degrees). Martian valley networks match the arid, narrow-angle distribution, which the authors read as evidence that overland flow, rather than groundwater seepage, did most of the carving, and therefore that early Mars had an active hydrologic cycle.