Chapter 6: The Vent at the Beginning of Life

In February of 1977, a small submersible called Alvin descended two and a half kilometres to the floor of the Pacific, near the Galápagos Islands, on what was fundamentally a geology cruise.

The expedition had gone looking for hydrothermal vents. Their existence had been predicted on thermodynamic grounds — seawater must circulate through hot young crust at a spreading ridge, and it must come back out — and there were heat-flow anomalies suggesting where to look. Finding the vents would have been a satisfying confirmation of a physical model. That was the plan.

What they found, in the lights of the submersible, was an ecosystem.

Clams the size of dinner plates. Crabs. Mussels. And tube worms over a metre long, in dense white and red stands, in permanent darkness, at a depth where the standard understanding of ocean biology said there should be almost nothing at all — because the deep sea floor was known to be a place of scarcity, living on the slow rain of organic debris sinking from the sunlit water far above.

The expedition had no biologists aboard. It had not occurred to anyone that biologists would be needed. The story that has come down — and I flag it as the kind of story that improves with retelling — is that the crew, having nothing suitable for preserving specimens, used vodka from the ship's stores.

This chapter is about what that discovery started, how far the argument built on it can honestly be pushed, and where — this is the part I have been signalling since the introduction — it breaks.

Register: the vent systems and their biology are empirical. Everything about the origin of life is theoretical, and I will mark the transition explicitly.

Eating rock

The animals at the Galápagos vents were not living on debris from above. They were living, at one remove, on chemistry coming out of the ground.

Hydrothermal fluid at a volcanically driven vent is hot, acidic, and loaded with reduced compounds — hydrogen sulfide above all, along with hydrogen, methane and dissolved metals. It is chemically the opposite of the cold, oxidized seawater it discharges into. And wherever two chemically incompatible fluids meet, there is a thermodynamic gradient, which is to say a source of usable energy.

Certain bacteria and archaea make their living from exactly this. They oxidize hydrogen sulfide and use the energy released to fix carbon dioxide into organic matter — the same end product as photosynthesis, by a different route, with rock chemistry in place of light. The process is called chemosynthesis, and it sits at the base of the vent food web.

The giant tube worms turned out to be a particularly elegant case: they have no mouth and no gut as adults, and they are, functionally, containers for symbiotic chemosynthetic bacteria, which they supply with the sulfide, oxygen and carbon dioxide the bacteria need. The worm is a farm.

The headline written at the time, and repeated ever since, was that life had been found which does not depend on the Sun.

The complication in the headline

That headline is not quite right, and the correction matters a great deal for everything that follows, so I want to make it carefully.

Chemosynthesis at a black smoker runs on the reaction between hydrogen sulfide from the vent and oxygen from seawater. The sulfide is geological. The oxygen is not. Free oxygen in Earth's ocean and atmosphere is overwhelmingly a product of photosynthesis — it is biological waste, accumulated over billions of years, and it did not exist in appreciable quantities on the early Earth.

So the classic vent ecosystem is not independent of sunlight. It is independent of sunlight locally, while depending on a global oxidant that photosynthesis manufactured. Remove the Sun and these communities do not persist. They are magnificent and they are not the demonstration they were initially taken to be.

Which raises the question that reorganized this entire field. Is there a metabolism that needs no photosynthetic input at all — one that could have run on a planet with no biosphere yet in existence?

There is. Take hydrogen and carbon dioxide, and combine them to make methane and water. It releases energy. It requires no oxygen, no light, and nothing biological — only hydrogen, carbon dioxide, and a catalyst. Organisms that do this, the methanogens, are among the deepest-branching lineages known.

So the question became: where on the early Earth would you find a sustained supply of hydrogen, in contact with carbon dioxide, in conditions where the reaction could be harnessed rather than merely occurring?

And the answer to that is not a black smoker.

Lost City

On the fourth of December, 2000, an expedition working on the Mid-Atlantic Ridge identified a vent field unlike anything previously described. It sits on the Atlantis Massif, a block of upthrust mantle rock, some fifteen kilometres from the ridge axis itself, and it was given the name Lost City for its towering pale carbonate chimneys, the largest of them some sixty metres tall.

Everything about it is different from a black smoker.

The fluids are not scalding but warm — roughly forty to ninety degrees. They are not acidic but strongly alkaline, with a pH above nine. They are not laden with metals and sulfide. They are rich in hydrogen and methane, and conspicuously poor in carbon dioxide. The chimneys are not sulfide but carbonate, precipitated where the alkaline fluid meets seawater. And the field is old: the deposits indicate activity sustained over a very long period, tens of thousands of years at minimum, which is unheard of for a magmatically driven vent that typically shuts off within decades as the heat source cools.

The reason for all of this is a reaction called serpentinization. When seawater reaches olivine-rich mantle rock, it reacts with it. The rock takes up water, its minerals convert to serpentine, iron is oxidized to magnetite, and hydrogen is liberated. The reaction is exothermic — it generates its own heat — and it expands the rock, cracking it open and admitting more water, so it propagates itself. Give it exposed mantle peridotite and seawater and it will run, for as long as there is unreacted rock, producing warmth, alkalinity and a steady stream of hydrogen.

That is a hydrogen source with no magma required, running for geological durations, in the deep ocean of an anoxic early Earth.

The alkaline vent hypothesis

Register: theoretical from here. What follows is a serious hypothesis with serious advocates, and it is not established.

Here is the argument that has made Lost City the most discussed square kilometre of seafloor on the planet, and it starts from an observation about you.

Every living cell on Earth — bacterium, archaeon, redwood, reader — generates the bulk of its usable energy the same way. It pumps protons across a membrane to build up a difference in concentration and charge, and then lets them flow back through a molecular turbine that manufactures the cell's energy currency. This is chemiosmosis, and it is as universal as the genetic code. Whatever the last common ancestor of all life was doing, it was doing this.

It is also, when you think about it, a very strange thing to have started doing. To use a proton gradient you need a membrane, a pump, and a turbine. To have evolved the pump and the turbine you need an energy source to build them with — but the energy source is the gradient, which is the thing the pump is for. It has the shape of a bootstrapping problem.

Unless the gradient was already there.

At an alkaline vent, hydrogen-rich fluid at pH nine or above discharges into an early ocean which — being loaded with dissolved carbon dioxide, in the absence of any biosphere to have drawn it down — was mildly acidic. Across the wall of a chimney, you therefore have a standing difference in proton concentration of several pH units. That is a proton gradient. It is geological, it is free, and it is maintained continuously for as long as the vent runs.

And the chimneys are not solid. They are porous, honeycombed with interconnected micron-scale compartments with thin inorganic walls — walls containing iron-sulfur minerals whose structures resemble the iron-sulfur clusters still sitting at the catalytic centres of ancient enzymes today.

So the proposal, developed principally by Michael Russell and later elaborated with William Martin and Nick Lane, is this: the first metabolism did not have to invent the proton gradient. It found one. A vent chimney is a natural array of tiny compartments, each with a membrane, each with a proton gradient across it, each with catalytic minerals in the walls, each supplied continuously with hydrogen and carbon dioxide — a structure that is, in the relevant respects, an inorganic approximation of a cell, present in enormous numbers, running for tens of thousands of years.

I find this a beautiful hypothesis. I want to be equally clear that it is a hypothesis, and that it faces objections from two quite different directions.

The first objection: it is contested on its own terms

The alkaline vent scenario requires that simple organic intermediates accumulate and react in those compartments. One class of proposed intermediates is the thioesters, which are attractive because they are energy-rich and because analogous chemistry sits at the heart of modern metabolism.

Work from the Earth-Life Science Institute has argued that this does not survive the thermodynamics: given the free energy of thioester hydrolysis and the corresponding equilibrium constants, these species are unlikely to have accumulated abiotically to any significant extent under Lost City conditions. If the intermediates cannot build up, a central step in the proposed pathway is in trouble.

That objection is not fatal to the whole framework, and its proponents have answers. But it is a live technical dispute inside the field, and any account which presents the alkaline vent hypothesis as settled is misrepresenting the state of play.

The second objection: it is not volcanic

And now the one this book has to face directly, because I flagged it in the introduction and because it costs me the strongest version of my own argument.

For a long time in the drafting of this book, this chapter was the keystone. The sentence I wanted was life began at a volcano. It has the great advantage of being memorable and the fatal disadvantage of being wrong.

Lost City is not driven by magma. It is driven by serpentinization — a chemical reaction between water and mantle rock that generates its own heat. There is no melt involved. There does not need to be any melt involved. The Atlantis Massif is a block of mantle peridotite exposed at the seafloor by faulting, and the vent field runs on the exposure, not on any underlying magma body. The fluids are cool, alkaline, metal-poor and CO₂-poor precisely because they are not magmatic; every property that makes Lost City interesting for the origin of life is a property that distinguishes it from a volcanic vent.

Alkaline vents and black smokers are often described together because they are both hydrothermal and both near spreading centres. They are better understood by their differences. One is a bonfire; the other is a chemical hand-warmer that happens to be nearby.

So what, if anything, does the volcanic system contribute?

The honest reformulation is this. What Lost City requires is mantle rock exposed to seawater. That exposure is produced by extensional faulting at a slow-spreading ridge — by the tectonic system that volcanism accompanies, drives, and is part of, but which is not itself volcanism. On a planet with no interior heat engine there would be no spreading ridges, no exposed peridotite, no serpentinization and no Lost City. The dependency is real. It is just one link further back in the chain than I wanted it to be, and the link it runs through is tectonic rather than magmatic.

That is a weaker claim. I think it is also a more interesting one, and it is the version this book will defend: not that volcanism made life, but that the same planetary heat engine which produces volcanism also produces the conditions in which the best-supported origin scenario becomes possible.

The other hypothesis, which is volcanic

Having given away the marine argument, I should tell you that there is a competing origin-of-life hypothesis which is straightforwardly volcanic, which is less famous, and which has some strikingly good arguments behind it. It deserves better than the footnote it usually gets — not least because it is the one that would rescue this chapter, and I want to be careful that I am not being seduced by that.

The setting is not the deep ocean but subaerial hydrothermal fields — hot springs, geysers, mud pools and fumaroles on volcanic land, of the kind you can walk around today in Iceland, Yellowstone or Kamchatka. Genuinely magmatic, in the sense that the heat comes from a shallow magmatic system.

Two arguments are made for it.

The first is chemical, and it is the kind of argument I admire: it reasons from the composition of living cells back to the geology of their birthplace. Every cell, without exception, maintains a high concentration of potassium and a low concentration of sodium in its interior, and works hard to keep it that way. But the earliest protocells would have had leaky membranes and no sophisticated ion pumps — so their interiors should have resembled their surroundings. Seawater is the opposite: sodium-rich, potassium-poor. If life began in the sea, the universal ionic signature of the cell is difficult to explain.

Vapour-dominated geothermal fields, however, produce condensates that are potassium-rich relative to sodium, with an ionic profile that resembles cytoplasm considerably better than seawater does. The argument is that the composition of the inside of your cells is a fossil of the water in which life started, and that the water was on volcanic land.

The second argument is about wet–dry cycling. Making polymers — chains of nucleotides or amino acids — requires removing water, which is thermodynamically awkward in the middle of an ocean. On land, a hot spring pool that repeatedly fills and evaporates provides exactly the alternation needed: dehydration to drive polymerization and concentrate reactants, rehydration to redistribute the products. Laboratory work has shown lipid vesicles forming and encapsulating polymers under simulated cycles of this kind.

The objections are equally real. A pool on land is exposed to intense ultraviolet radiation, since there was no ozone. It is subject to dilution by rainfall and to complete drying. Above all, it does not offer the sustained, continuously renewed disequilibrium that a vent provides for tens of thousands of years — it offers episodes, not a permanent supply.

The honest position, in my reading, is that the marine alkaline vent scenario has the better energetics and the subaerial hot spring scenario has the better chemistry, that neither is established, and that they are not necessarily exclusive. If you want a volcanic origin of life, this is where the argument lives — and it is much less secure than the one I gave up.

What is not in doubt

Set the origin question aside, and there remains something in this chapter that is not speculative at all.

Beneath the seafloor and beneath the continents, in the pore spaces and fractures of rock, there is life. A great deal of it. Microbial communities living kilometres down, at temperatures and pressures that would have been dismissed as impossible within living memory, growing extraordinarily slowly — with generation times that may be measured in centuries — and making their living from the chemistry of water in contact with rock.

A substantial proportion of that deep biosphere is hosted in basalt: the oceanic crust, which is volcanic rock, through which seawater circulates continuously and reacts. The hydrogen, methane, reduced iron and sulfur species produced by those reactions are what the organisms eat.

The total mass is genuinely uncertain and estimates have moved a lot, but on any current reckoning this is one of the largest habitats on the planet, and quite possibly comparable in carbon terms to everything living on the surface.

So the present-tense claim is safe even where the origin claim is not: volcanic rock, and the chemical disequilibrium that water-rock reaction maintains within it, supports an enormous fraction of Earth's biology right now. No inference about four billion years ago is required.

What this chapter licenses

Let me state the conclusion precisely, because this is a chapter where imprecision would be self-serving.

It does not license: that life began at a volcano. The best-developed origin hypothesis is set at a vent that is not volcanic, and the volcanic alternative, while real, is less well supported.

It does license, and I think securely: that every serious origin scenario except panspermia requires a wet rocky planet maintaining chemical disequilibrium at its surface and in its crust — sustained gradients of hydrogen, carbon and oxidation state that do not run down. A geologically dead planet does not provide that. It equilibrates, and then nothing further happens to it.

Whether the specific setting was an alkaline chimney on the flank of a spreading ridge, or a hot spring on a volcanic island, or somewhere nobody has yet proposed, all of the candidates are features of a planet with a working interior heat engine. The engine is not the author of life. It is the condition under which the question of life's origin can even be posed.

That is a more modest claim than the one I set out to make. It is also, unlike the other one, defensible — and it is the version that the rest of Part III will build on.

Chapter 7 makes the case that the same engine supplied the air and the ocean, manufactured the continents, and has been running the thermostat ever since. That argument is on much firmer ground than this one, and I will not have to give any of it back.

Draft notes — verification status

Standing convention. Lost City's core facts and the ELSI objection were resolved during the thesis-paper verification pass. The origin-of-life literature is the most contested material in the book and every attribution needs checking.

Already verified — do not re-check:

Lost City identified 4 December 2000, on the Atlantis Massif, via DSV Alvin and ROV Argo II.
Serpentinization-driven and not created by volcanic processes; fluids roughly 40–90 °C, pH above 9, hydrogen- and methane-rich, poor in CO₂, H₂S and metals; carbonate chimneys.
The Earth-Life Science Institute objection regarding thioester hydrolysis free energy and abiotic accumulation.

Pending verification:

The February 1977 Alvin Galápagos Rift dive: depth (~2,500 m), the absence of biologists aboard, and the vodka preservation anecdote. Flag the anecdote in the text as uncertain or cut it — it is exactly the sort of story that improves in retelling, and the chapter says so, but it should still be checked.
The 1979 Science publication of the Galápagos vent findings (Corliss et al.) and the discovery's framing at the time.
Riftia pachyptila physiology — absence of mouth and gut in adults, endosymbiotic chemoautotrophs — and Colleen Cavanaugh's identification of the symbiosis.
The claim that vent chemosynthesis depends on photosynthetically derived oxygen. Directionally correct and important to the chapter's argument; confirm the standard framing and check whether anaerobic pathways at black smokers complicate it.
Methanogenesis (H₂ + CO₂ → CH₄ + H₂O) as energy-yielding and as a deep-branching metabolism; the phylogenetic claim needs care, since the "deepest-branching" question is itself disputed.
Lost City chimney heights (draft: largest ~60 m) and field longevity (draft: "tens of thousands of years at minimum"; sources give >120,000 years for the oldest deposits — reconcile).
Distance of the field from the ridge axis (draft: ~15 km).
Typical lifetimes of magmatically driven vent fields ("decades") for the contrast being drawn.
The serpentinization reaction as stated: olivine + water → serpentine + magnetite + H₂; exothermic; volume expansion driving self-propagation.
Chemiosmosis as universal; Peter Mitchell's chemiosmotic hypothesis and its acceptance.
Attribution of the alkaline vent hypothesis: Michael Russell (and Allan Hall) originally, with William Martin and Nick Lane as principal later developers. Check the attribution order before naming anyone in print.
The pH differential across a chimney wall between alkaline fluid and an acidic Hadean ocean, and the claim that the early ocean was mildly acidic due to high dissolved CO₂.
Iron-sulfur minerals in chimney walls and their structural resemblance to enzyme cofactors — this parallel is frequently overstated; verify how strong the resemblance actually is.
The subaerial hot spring hypothesis: Mulkidjanian et al. on the K⁺/Na⁺ argument from cytoplasmic composition, and David Deamer's wet–dry cycling work on vesicle formation and polymerization. Confirm both attributions and the strength of the experimental results.
Deep biosphere: current mass estimates and the proportion hosted in oceanic basalt; generation times ("centuries"). Estimates have been revised substantially — use the most recent synthesis and state its uncertainty.