Missed by the method, not the eye
In 2018, Petros Benias, David Carr-Locke and Neil Theise reported that a layer long filed under "dense connective tissue" is in fact an open, fluid-filled space threaded by a network of thick collagen bundles — draining toward the lymph nodes, and found beneath the skin and around the gut, lungs, vasculature and fascia throughout the body (Scientific Reports 8:4947).
It had been missed for a reason that is itself the point. Standard histology fixes tissue, dehydrates it, and embeds it in paraffin — which drains the fluid, so the struts holding the spaces open collapse flat. Generations of pathologists were taught to read the resulting cracks as processing artifacts. The team saw it only by imaging living tissue with confocal laser endomicroscopy (pCLE) at 60–70 µm depth, after fluorescein injection lit up a reticular pattern with no known anatomical correlate. The structure exists only when hydrated; our method of looking had been erasing it.
Struts, cables, gel, and an open interface
| Element | What it is | Role |
|---|---|---|
| Struts | Thick bundles of type I/III collagen, intersecting like scaffolding | Hold macroscopic pockets (~100–200 µm) open under load |
| Cables | Interwoven elastin fibres | Stretch and recoil — the lattice warps and springs back |
| Fill | Hyaluronic-acid / glycosaminoglycan gel + interstitial fluid | Binds water; cushions organs during movement |
| Lining | Fibroblast-like cells coating collagen strands — on one side only | Leaves an extensive interface where fluid meets bare matrix |
width
depth
vessel · fascia
Sci. Reports 8:4947
The detail that matters for this series is the last one: the collagen strands are lined by cells on only one face, leaving a large unlined interface where moving fluid is in direct contact with structural matrix. This is not a sealed vessel with an endothelial wall. It is an open, hydrated, space-filling network with an interface running through all of it.
A hydrated lattice, and an interface
Two features connect the interstitium to threads already in Principia Orthogona, and it is worth stating exactly how far the connection goes — and no further.
First, geometry. The interstitium is a space-filling network: a repeating arrangement of structural elements enclosing fluid-filled cells, quasi-ordered rather than periodic. That is the family of objects the series keeps returning to — lattices, honeycombs, the six-fold packings of the Stone Fold and the crystal tracks — now appearing as soft, wet, living tissue rather than stone or silicon.
Second, the interface. The one-sided lining leaves a boundary where fluid meets matrix across the whole network. Interfaces — boundaries across which state is exchanged under a constraint — are exactly what the series' contact-geometric operator chain (the C / Contact step) is built to describe. The interstitium is, structurally, an interface-rich medium.
The metaphor reaches for a hexagon. Asked to describe the inside, Wells offers a sheet of chicken wire embedded in a gel — collagen as the wire, hyaluronic-acid gel in the openings. Chicken wire is not a generic mesh but a hexagonal one, and this series arrives at six-fold packing from the opposite direction, by asking which network fills a plane at least cost. Noticing is not asserting: she means an open mesh in gel, and the real collagen network is quasi-ordered rather than a periodic honeycomb. What the two pictures share is the type — struts enclosing fluid-filled cells — and not the symmetry. The useful part is that the difference is measurable, which is what §5 turns it into.
An extracellular matrix has already been measured this way. Book VIII Q.2 reports, for acid-induced polylaminin, a measured Hausdorff dimension $d_H \in [1.55,\,1.70]$ against $d_H = \log b / \log \eta_3$ with a hexagonal branching factor $b \in (2.57,\,2.82)$ and $\eta_3 \approx 1.8393$ — and a stronger claim still, that the multifractal spectrum $f(\alpha)$ belongs to the contact topology rather than to the material. Polylaminin is not a remote comparison: it is self-assembled fibrous matrix in water, and so is the interstitium. One has been measured and the other has not.
And Book III says what a lattice like this is for. Chapter T argues that tubulin is computronium in the literal sense — fifteen architectures from one assembly grammar, with no computation performed on top of the structural output. The morphology is the computation. If that carries beyond tubulin, the collagen network is not a container through which fluid happens to move but what decides where fluid can go, and its vertex statistics and $f(\alpha)$ stop being descriptive and become functional. A thumb held at a point of tissue confluence is then an input to a computing medium rather than a stimulus applied to a channel. Chapter T states its theorem for tubulin dm³ systems; carrying it to a different matrix is a conjecture, not a corollary.
Vol VIII · Q.2 — Polylaminin & Contact Topology · Book III · Chapter T — Tubulin as Computronium · Vol VI · Ch 7 — Microtubule Fibonacci
This is a conceptual bridge, not a result. Nothing here derives an operator chain from the interstitium or proves a theorem about it. The claim is only that the interstitium is the kind of object — a hydrated, interface-rich, space-filling lattice — that the series' geometric methods are designed for, which makes it a candidate for later formal treatment. It is filed as an open direction, not a finding.
Established, contested, and speculative — kept apart
The interstitium attracts strong claims, and the popular retellings blur three very different tiers. Keeping them separate is the whole job.
Which is to say: a system of named sites where connective tissue converges, treated by sustained manual pressure and surface oil, is a description of mechanical loading at interface-rich points in a hydrated lattice. Whether that is why anything in the practice works is a separate question this note does not touch, and nothing here is a therapeutic claim. The observation is narrower and, for a history of structural ideas, more interesting: when a pre-modern anatomy and a 2018 imaging result are pointed at the same tissue, the older one is already using the vocabulary of junctions.
There is real experimental work here, and it is usually read backwards. Langevin and Yandow (2002) found that acupuncture points in the arm coincide with connective-tissue planes — the intermuscular fascia. Li and colleagues (2021) injected fluorescent dye at forearm points in living volunteers and watched it migrate proximally along a track corresponding to the pericardium meridian.
Both results are better evidence for the ordinary explanation than for the extraordinary one. Dye placed in interfascial tissue travels along the fascial plane, because that is the path of least resistance in a hydrated lattice; and a point map assembled over centuries by palpating living bodies would be expected to land on exactly those planes, since they are what a thumb can feel. The correlation says the meridian follows the fascia. It does not supply a channel, and it does not license chi as a circulating substance whose route has now been found.
Two things are missing from the popular account and belong here. The clinical literature's central difficulty is that needling frequently fails to separate from sham needling, which the anatomy does not address either way. And the antiquity claims — the "four thousand years" that circulates with this story — outrun the textual record: the Huangdi Neijing is roughly second century BCE, and systematised metal needling later still.
The interstitium's role as a low-resistance route for tumour-cell spread is supported and clinically important — but "fluid highway for cancer" is a mechanism under active study, not a settled quantitative model. Any formal treatment should target the transport/percolation question, where the claim can actually be made precise and tested, rather than restating the metaphor.
The contribution is what becomes possible once the math is known
The Maya chapter had a clean theorem to prove. This one has something different, and for a living tissue more useful: two measurements that can be made on micrographs already published, a prediction from elsewhere in this series that either survives them or does not, and a set of questions that stop being rhetorical the moment the numbers exist. The reason to write it down is not that the interstitium resembles a lattice. It is that knowing which lattice turns a family of arguments-by-resemblance into arithmetic.
The percolation question follows the other two rather than preceding them: treat the interstitial space as a graph of fluid-filled cells joined through the collagen network, and ask for the threshold $p_c$ above which transport spans a tissue. That is a real question with a real answer, and it is the version of "the body's hidden network" this series is equipped to pursue — but a threshold is a function of the lattice, so the lattice has to be characterised first.
There is a cheaper question that could be answered first, and the chicken-wire remark of §3 is what suggests it. A space-filling network of fluid-filled cells has an obvious null model: a wet foam. Foams at equilibrium obey Plateau's laws — four edges meeting at each vertex, at the tetrahedral angle $\arccos(-1/3) \approx 109.47^\circ$ — because they are minimising interfacial area. A flat hexagonal net, the chicken wire, has coordination three. Collagen is not a minimal surface but a fibrous solid laid down by cells, and so has no obligation to either.
Which is exactly why the vertex statistics are worth measuring. Take the published micrographs, count how many bundles meet at a junction and at what angles, and the distribution answers a real question: whether this network's geometry was set by the mechanics of the fluid it encloses, or by the fibroblasts that deposited it. That is one histogram, extractable from data that already exists, and it decides between two accounts of the same tissue — which is more than a percolation threshold can do before the network's local structure is known at all.
And there is a second measurement on the same micrographs, which this series is in the unusual position of having already committed to. Substrate-blindness is a prediction: if $f(\alpha)$ belongs to the contact topology and not to the material, then an extracellular matrix realising that topology should show the polylaminin spectrum whatever it is made of. The interstitium is collagen and hyaluronic acid rather than laminin, self-assembled in water on a different scale, in a different tissue. It is exactly the substrate the claim has not been tried on.
Q.2 is explicit that substrate-blindness is a theoretical claim and not independently tested outside its source, and that the $d_H$ values it matches were taken from the source deposit rather than re-derived. That is the condition in which a prediction is worth making and worthless to assume.
So the test has teeth in both directions. A box-counting dimension and an $f(\alpha)$ curve extracted from the published interstitial micrographs either land on the polylaminin spectrum or they do not. If they do not, substrate-blindness is narrower than Q.2 states it, and this chapter will have been the thing that narrowed it.
What follows from having the numbers is the actual contribution, and it is worth setting out plainly, because each item below is currently phrased as a metaphor somewhere in the literature and each becomes a calculation instead.
None of this requires new imaging. It requires counting what is in micrographs published since 2018 and comparing two curves. The object was identified eight years ago; the geometry is named here; what remains is arithmetic nobody has had a reason to do — until there was a framework in which a wrong answer is as informative as a right one.
A second instance, running now
Everything above is one instance of the grammar this series follows, worked out in tissue: a structure compressed out of the picture by the method that founded modern anatomy, hidden for the 475 years between Vesalius and 2018, then disclosed by a change of instrument. A second instance is running in mathematics as this is written — the formal-verification turn, in which a referee's assent is replaced by a compiler's verdict against a fixed axiom set — and it can be watched mid-unfold rather than reconstructed afterwards.
It is treated separately, because it is a paper about method and not about tissue: WP-115 · The Fold in Formal Verification names the five moves and checks each against the primary sources, including what a deposit claiming a machine-checked proof does not disclose. That the same apparatus was derived independently in Sanskrit before the common era is WP-116 · Akṣapāda Had the System.
- Benias, P. C., Wells, R. G., Carr-Locke, D. L., Theise, N. D., et al. (2018). Structure and Distribution of an Unrecognized Interstitium in Human Tissues. Scientific Reports 8:4947. nature.com · DOI 10.1038/s41598-018-23062-6
- Meet Your Interstitium, a Newfound "Organ." Scientific American, 2018. scientificamerican.com
- Cooper, A. Z. (11 May 2026). Inside the Interstitium, the Human Body’s Hidden Pathways. The New York Times Magazine; illustrations by Jérôme Berthier. nytimes.com (cited as the popular framing this chapter’s §4 responds to — in particular the linking of the interstitium to acupuncture — and not as evidence for any anatomical or clinical claim. Every such claim above is sourced to the primary literature below.)
- Langevin, H. M., & Yandow, J. A. (2002). Relationship of acupuncture points and meridians to connective tissue planes. The Anatomical Record 269(6), 257–265. doi:10.1002/ar.10185
- Li, T., et al. (2021). In Vivo Visualization of the Pericardium Meridian with Fluorescent Dyes. Evid. Based Complement. Alternat. Med. doi:10.1155/2021/5581227 · PMID 33854554
- Langevin, H. M., Churchill, D. L., & Cipolla, M. J. (2001). Mechanical signaling through connective tissue: a mechanism for the therapeutic effect of acupuncture. The FASEB Journal 15(12). doi:10.1096/fj.01-0015hyp
- Langevin, H. M., Bouffard, N. A., Churchill, D. L., & Badger, G. J. (2007). Connective tissue fibroblast response to acupuncture: dose-dependent effect of bidirectional needle rotation. J. Altern. Complement. Med. PMID 17480137
- Suśruta Saṃhitā, Śārīrasthāna 6 (marmavibhāga śārīra) — the classical enumeration of marma as sites of tissue confluence. Classical source; consult a scholarly edition and translation rather than a practitioner summary.
- Chen, E., Cummins, C., Grubisic, D., Haller, L., Ono, K., Zhang, J., et al. (2026). Fel's Conjecture on Syzygies of Numerical Semigroups. arXiv:2602.03716 · Lean sources: AxiomMath/fel-polynomial
- Tao, T. (18 August 2026). Palomar — a registry of Lean verified mathematics. terrytao.wordpress.com