Praise for Thomas Gold..."Gold is one of America's most iconoclasticscientists."

-Stephen Jay Gould

"Thomas Gold is one of the world's most originalminds."

-The Times, London

"Thomas Gold might have grown tired of tilting atwindmills long ago had he not destroyed so many."

-USA Today

"What if someone told you that [the oil crisis] wasall wrong and that the hydrocarbons that make uppetroleum are constantly refilling reservoirs.Interested? Well, you should read this book.... Goldpresents his evidence skillfully. You may not agreewith him, but you have to appreciate his fresh andcomprehensive approach to these major areas ofEarth science.... [This book] demonstrates thatscientific debate is alive and well. Science ishypothesis-led and thrives on controversy-and fewpeople are more controversial than Thomas Gold."

-Nature

.. Thomas Gold, a respected astronomer andprofessor emeritus at Cornell University in Ithaca,N.Y., has held for years that oil is actually arenewable, primordial syrup continuallymanufactured by the Earth under ultrahotconditions and pressures."

-The Wall Street Journal

"Most scientists think the oil we drill for comesfrom decomposed prehistoric plants. Goldbelieves it has been there since the Earth'sformation, that it supports its own ecosystem farunderground and that life there preceded life onthe Earth's surface.... If Gold is right, the planet's oil

reserves are far larger than policymakers expect,and earthquake prediction procedures require ashakeup; moreover, astronomers hoping forextraterrestrial contacts might want to stopseeking life on other planets and inquire about lifein them."

-Publishers Weekly

"Gold's theories are always original, alwaysimportant, and usually right. It is my belief, basedon fifty years of observation of Gold as a friendand colleague, that The Deep Hot Biosphere is allof the above: original, important, controversial, andright."

-Freeman Dyson

"Whatever the status of the upwelling gas theory,many of Gold's ideas deserve to be takenseriously.... The existence of [a deep hotbiosphere] could prove to be one of themonumental discoveries of our age. This bookserves to set the record straight."

-Physics World

"My knowledge and experience of natural gas,gained from drilling and operating many of theworld's deepest and highest pressure natural gaswells, lends more credence to your ideas than theconventional theories of thebiological/thermogenic origin of natural gas. Yourtheory explains best what we actually encounteredin deep drilling operations."

-Robert A. Hefner III, The GHK Companies,Oklahoma City, Oklahoma; From a letter to the

author

"Within the scientific community, Gold has areputation as a brilliantly clever renegade, havingput forward radical theories in fields ranging fromcosmology to physiology."

-The Sunday Telegraph, London

"In The Deep Hot Biosphere, [Gold] revealsevidence supporting a subterranean biosphereand speculates on how energy may be producedin a region void of photosynthesis. He speculateson the ramifications his concepts could have inpredicting earthquakes, deciphering Earth'sorigins, and finding extraterrestrial life."

-Science News

"Gold's theory, as explained in The Deep HotBiosphere, offers new and radical ideas to ourincomplete notions of what causes earthquakesand where we would look for life in outer space:not on planets, but in them."

-Ithaca Times

"[The Deep Hot Biosphere] now seems to besupported by a growing body of evidence."

-Journal of Petroleum Technology

"Gold knows experts are pooh-poohing his belief.It happens to Gold consistently. He has developeda reputation as someone who takes on a long-heldassumption, advances a new idea and getsrewarded when time-a decade or two-proves himright."

-The Juneau Empire

"Thomas Gold has questioned the veryfoundations of the entrenched conventionalmodels.... [The Deep Hot Biosphere] is evidentlyone of the most controversial of all bookspublished in recent history. It is bound to causemuch debate, and, if found correct, is likely torevolutionize the face of science."

-Current Science

"[Thomas Gold] is one of the few who, despite theattacks of mediocrities, is courageous enough tothink in a scientifically unconventional way.... [His]courage and original ideas are rays of hope on the

horizon of science."

-Prof. Dr. Alfred Barth, The European Academy ofSciences and Arts, Paris

Thomas Gold

With a Foreword by Freeman Dyson

Forewordby Freeman Dyson

he first time I met Tommy Gold was in1946, when I served as a guinea pig in anexperiment that he was doing on the capabilities ofthe human ear. Humans have a remarkable ability todiscriminate the pitch of musical sounds. We caneasily tell the difference when the frequency of apure tone wobbles by as little as 1 percent. How dowe do it? This was the question that Gold wasdetermined to answer. There were two possibleanswers. Either the inner ear contains a set of finelytuned resonators that vibrate in response to incidentsounds, or the ear does not resonate but merelytranslates the incident sounds directly into neuralsignals that are then analyzed into pure tones bysome unknown neural process inside our brains. In1946, experts in the anatomy and physiology of theear believed that the second answer must becorrect: that the discrimination of pitch happens inour brains, not in our ears. They rejected the firstanswer because they knew that the inner ear is asmall cavity filled with flabby flesh and water. Theycould not imagine the flabby little membranes in theear resonating like the strings of a harp or a piano.

Gold designed his experiment to prove theexperts wrong. The experiment was simple, elegant,and original. During World War II he had beenworking for the Royal Navy on radio communicationsand radar. He built his apparatus out of war surplusNavy electronics and headphones. He fed into theheadphones a signal consisting of short pulses of apure tone, separated by intervals of silence. Thesilent intervals were at least ten times longer than the

period of the pure tone. The pulses were all thesame shape, but they had phases that could bereversed independently. To reverse the phase of apulse means to reverse the movement of thespeaker in the headphone. The speaker in areversed pulse is pushing the air outward when thespeaker in an unreversed pulse is pulling the airinward. Sometimes Gold gave all the pulses thesame phase, and sometimes he alternated thephases so that the even pulses had one phase andthe odd pulses had the opposite phase. All I had todo was sit with the headphones on my ears andlisten while Gold fed in signals with either constant oralternating phases. Then I had to tell him, from thesound, whether the phase was constant oralternating.

When the silent interval between pulses was tentimes the period of the pure tone, it was easy to tellthe difference. I heard a noise like a mosquito, ahum and a buzz sounding together, and the quality ofthe hum changed noticeably when the phases werechanged from constant to alternating. We repeatedthe trials with longer silent intervals. I could still detectthe difference, even when the silent interval was aslong as thirty periods. I was not the only guinea pig.Several other friends of Gold listened to the signalsand reported similar results. The experiment showedthat the human ear can remember the phase of asignal, after the signal stops, for thirty times theperiod of the signal. To be able to remember phase,the ear must contain finely tuned resonators thatcontinue to vibrate during the intervals of silence.T h e result of the experiment proved that pitchdiscrimination is done mainly in the ear, not in thebrain.

Besides having experimental proof that the earcan resonate, Gold also had a theory to explain howa finely tuned resonator can be built out of flabby anddissipative materials. His theory was that the innerear contains an electrical feedback system. Themechanical resonators are coupled to electrically

powered sensors and drivers, so that the combinedelectromechanical system works like a finely tunedamplifier. The positive feedback provided by theelectrical components counteracts the dampingproduced by the flabbiness of the mechanicalcomponents. Gold's experience as an electricalengineer made this theory seem plausible to him,although he could not identify the anatomicalstructures in the ear that functioned as sensors anddrivers. In 1948 he published two papers, onereporting the results of the experiment and the otherdescribing the theory.

Having myself participated in the experiment andhaving listened to Gold explaining the theory, I neverhad any doubt that he was right. But the professionalauditory physiologists were equally sure that he waswrong. They found the theory implausible and theexperiment unconvincing. They regarded Gold as anignorant outsider intruding into a field where he hadno training and no credentials. For years his work onhearing was ignored, and he moved on to otherthings.

Thirty years later, a new generation of auditoryphysiologists began to explore the ear with far moresophisticated tools. They discovered that everythingGold had said in 1948 was true. The electricalsensors and drivers in the inner ear were identified.They are two different kinds of hair cells, and theyfunction in the way Gold said they should. Thecommunity of physiologists finally recognized theimportance of his work, forty years after it waspublished.

Gold's study of the mechanism of hearing istypical of the way he has worked throughout his life.About once every five years, he invades a new fieldof research and proposes an outrageous theory thatarouses intense opposition from the professionalexperts in the field. He then works very hard to provethe experts wrong. He does not always succeed.Sometimes it turns out that the experts are right and

he is wrong. He is not afraid of being wrong. He wasfamously wrong (or so it is widely believed) when hepromoted the theory of a steady-state universe inwhich matter is continuously created to keep thedensity constant as the universe expands. He mayhave been wrong when he cautioned that the moonmay present a dangerous surface, being covered bya fine, loose dust. It proved indeed to be so covered,but fortunately no hazards were encountered by theastronauts. When he is proved wrong, he concedeswith good humor. Science is no fun, he says, if youare never wrong. His wrong ideas are insignificantcompared with his far more important right ideas.Among his important right ideas was the theory thatpulsars, the regularly pulsing celestial radio-sourcesdiscovered by radio-astronomers in 1967, arerotating neutron stars. Unlike most of his right ideas,his theory of pulsars was accepted almostimmediately by the experts.

Another of Gold's right ideas was rejected by theexperts even longer than his theory of hearing. Thiswas his theory of the 90-degree flip of the axis ofrotation of the earth. In 1955, he published arevolutionary paper entitled "Instability of the Earth'sAxis of Rotation." He proposed that the earth's axismight occasionally flip over through an angle of 90degrees within a time on the order of a million years,so that the old north and south poles would move tothe equator, and two points of the old equator wouldmove to the poles. The flip would be triggered bymovements of mass that would cause the old axis ofrotation to become unstable and the new axis ofrotation to become stable. For example, a largeaccumulation of ice at the old north and south polesmight cause such an exchange of stability. Gold'spaper was ignored by the experts for forty years. Theexperts at that time were focusing their attentionnarrowly on the phenomenon of continental drift andthe theory of plate tectonics. Gold's theory hadnothing to do with continental drift or plate tectonics,so it was of no interest to them. The flip predicted byGold would occur much more rapidly than continental

drift, and it would not change the positions ofcontinents relative to one another. The flip wouldchange the positions of continents only relative to theaxis of rotation.

In 1997, Joseph Kirschvink, an expert on rockmagnetism at the California Institute of Technology,published a paper presenting evidence that a 90-degree flip of the rotation axis actually occurreddur i ng a geologically short time in the earlyCambrian era. This discovery is of great importancefor the history of life, because the time of the flipappears to coincide with the time of the "CambrianExplosion," the brief period when all the majorvarieties of higher organisms suddenly appear in thefossil record. It is possible that the flip of the rotationaxis caused profound environmental changes in theoceans and triggered the rapid evolution of new lifeforms. Kirschvink gives Gold credit for suggestingthe theory that makes sense of his observations. Ifthe theory had not been ignored for forty years, theevidence that confirms it might have been collectedsooner.

Gold's most controversial idea is the non-biological origin of natural gas and oil. He maintainsthat natural gas and oil come from reservoirs deep inthe earth and are relics of the material out of whichthe earth condensed. The biological molecules foundin oil show that the oil is contaminated by livingcreatures, not that the oil was produced by livingcreatures. This theory, like his theories of hearingand of polar flip, contradicts the entrenched dogmaof the experts. Once again, Gold is regarded as anintruder ignorant of the field he is invading. In fact,Gold is an intruder, but he is not ignorant. He knowsthe details of the geology and chemistry of naturalgas and oil. His arguments supporting his theory arebased on a wealth of factual information. Perhaps itwill once again take us forty years to decide whetherthe theory is right. Whether the theory of non-biological origin is ultimately found to be right orwrong, collecting evidence to test it will add greatly

to our knowledge of the earth and its history.

Finally, the most recent of Gold's revolutionaryproposals, the theory of the deep hot biosphere, isthe subject of this book. The theory says that theentire crust of the earth, down to a depth of severalmi les , is populated with living creatures. Thecreatures that we see living on the surface are only asmall part of the biosphere. The greater and moreancient part of the biosphere is deep and hot. Thetheory is supported by a considerable mass ofevidence. I do not need to summarize this evidencehere, because it is clearly presented in the pagesthat follow. I prefer to let Gold speak for himself. Thepurpose of my remarks is only to explain how thetheory of the deep hot biosphere fits into the generalpattern of Gold's life and work.

Gold's theories are always original, alwaysimportant, usually controversial-and usually right. It ismy belief, based on fifty years of observation of Goldas a friend and colleague, that the deep hotbiosphere is all of the above: original, important,controversial-and right.

Prefai

n June 1997 I was asked by NASA to givethe annual lecture at the Goddard Space FlightCenter in Maryland. My contribution to the deep hotbiosphere theory and its implications forextraterrestrial life had won me the invitation. I wasflattered, of course, but at the same time chagrinedby the topic I was asked to address: life in extremeenvironments. I had little interest in talking about thesurface biosphere on earth, and yet, if I were to takethe topic literally, this is precisely what I was beingasked to do. The life in extreme environments is ourown surface life.

If there is one idea that I hope you will retain longafter you finish reading this book, it is this: It is wewho live in the extreme environments. And if there isone desire I hope to stimulate in you, it is a curiosityto learn more about the first and most truly terrestrialbeings-all of whom live far beneath our feet, in what Ihave come to call the deep hot biosphere.

Alas, I can only begin to satisfy this curiosity here,for at this moment in our biological and cosmicunderstanding, there are still more questions thananswers. But that is exactly what makesinvestigating the deep hot biosphere so exciting.

Thomas Gold

Ithaca, New York

December 1998

Fo revVr rd I

Freeman Dyson

Preface xi

(:i ipter t Our (I"11-dell of 1111oll

The Narrow Window for Surface Life 2

Chemical Energy for Subsurface Life 4

A Preview of This Book 7

(:11aI)t(~r., I,if't, at IIre I~) r(h rs 11

Energy Deep in the Earth 13

The Ecology of Deep-Ocean Vent Life 19

Other Borderland Ecologies 23

Deep Is Desirable 27

Beneath the Borderlands 30

r:r ,me 1)~~~I-F trIIi (,,is i,it r. 37

The Origin of Petroleum: Two Conflicting Theories38

Five Assumptions Underlying the Deep-Earth GasTheory 43

01iriter1 Evidence 1'6r I>fv(-F"irtIi (iaS 57

Petroleum Reservoirs That Refill 59

Clues in the Carbonate Record 61

The Association of Helium with Hydrocarbons 72

(;har0vr:) H;s(IV'in$2, th(' P('tr(1etnrl Par'alllx79

The Deep Hot Biosphere Solution 80

Biological Molecules in Non-Biological Petroleum 82

The Upwelling Theory of Coal Formation 86

Evidence for the Upwelling Theory 94

An Exemption for Peat 100

chapter(; Th ' SiI.jan Fxt(Tinlent 105

Drilling in Swedish Granite 107

Magnetite and Microbial Geology 114

(:banter Fxten(Iinw; theihI1mry 125

The Origin of Diamonds 127

A New Explanation for Concentrated Metal Deposits131

I~hatikr`~ IIP( hInking I~'Alrth(tuak(s 141

Mud Volcanoes 142

A Challenge to Earthquake Theory 143

Eyewitness Accounts 145

Earthquake Spots and Earth Mounds 156

Upwelling Deep Gas as the Cause of Earthquakes159

t;har)tc t) TIw Ori ;in of Life 165

The Habitability of Surface and Subsurface Realms166

The Enhanced Probability for Life's Origin 170

Darwin's Dilemma 176

(;banter III hal Next' 185

Microbial Investigations 188

Prospects for Extraterrestrial Surface Life 193

Deepening the Search for Extraterrestrial Life 201

Independent Beginnings or Panspermia? 205

Afl('rvV(r(l to IIi Iaterlklck F A111011 209

Notes 217

Ackn(wl(il ;rnents 235

I tl(I('x 237

o scientific subject holds more surprisesfor us than biology. Foremost is the surprise that lifeexists at all. How could life have started? Did oneextraordinary chance occurrence in the universeassemble the first primitive living organism, and dideverything else follow from that?

What chemical and physical circumstances wereneeded for such an unlikely event to occur? Did ourearth offer the only nurturing conditions? Or (in whathas come to be known as the "panspermia"hypothesis) did life arise somewhere else,spreading through astronomical space to take rootin any fertile spot it encountered? Or is life notunlikely after all? Perhaps life is an inevitableconsequence of physical laws and is arisingspontaneously in millions of places.

Whatever the answers to these questions, we doknow that life on the surface of the earth spans ahuge variety of forms. These forms range frommicrobes to whales, giant fungi, and enormoustrees. They include unfathomable numbers ofinsects. If we add to our reckoning the life forms thathave died out, then the diversity expands to includedinosaurs, trilobites, and vastly more.

All this living variety has much in common. Theconstruction of all known organisms involvescomplex forms of protein molecules. Those, in turn,

are built up from a set of building blocks calledamino acids, common to all known forms of life. Thechemical configuration of some of these aminoacids could occur in two forms, one of which is themirror image of the other. Yet we find that all thehuge variety of life uses only one kind of each suchpair of molecules. There thus appears to be a strongconnecting thread running through all the life formswe know.

No less important than the common constituentsof life are the common conditions under which allknown life forms can develop and survive. Theseconditions include a requirement for water in theliquid state, a limited range of temperature, andsources of energy that are delivered in (or can beconverted into) chemical form. We tend to assumethat these conditions are best-and perhaps ideally-provided on the surface of our own planet. And weconclude, sadly perhaps, that these conditions arealmost certainly not present anywhere else in thesolar system. But are these assumptions valid?

The Narrow Window for Surface Life

he universe is a harsh and severe place, arealm of extremes. Most of the universe is virtuallyempty and very cold-to be precise, 2.7 Kelvin or -270.5 Celsius, which is just 2.7C above absolutezero. This vast cold is punctuated by points ofintense heat and lightthe stars-whose surfacetemperatures reach millions of degrees.

Stars do not maintain their brilliance forever, and itis from them that the constituents of life come. Starsthat have three or more times the mass of the sunwill expire in a frenzy of violence, a supernovaexplosion that may briefly flare with the brightness ofa hundred billion stars. The explosion scatters thestellar materials into space, making the cold cloudsout of which new stars form. The different atomicnuclei created in the core of the star and during itsexplosion supply materials from which planets canform. The same stellar materials provide theelements from which we and all other living creaturesknown to us are constructed.

Life is thus built up from a variety of atoms forgedin nuclear furnaces deep inside giant stars. Moreprecisely, life is constructed from molecules,clumpings of atoms that are in close enough contactand cool enough for a weak attractive force to holdthem together. The interiors of stars are suitable forelement formation, but their heat is too intense forthe formation of complex molecules.

Most places in the universe do not allow thechemical action that is conducive to life. The starsare too hot, and most other places are so cold thatsubstances are in the form of a solid or a very low-density gas, whose chemical activity is exceedinglyslow. But we do see some regions in the cosmos in

which many different types of molecules have beenbuilt up. These are the large gas clouds in interstellarspaces, warmed by stars that are in or near them.Radio techniques have made it possible to identifymany different molecules there. Water is onecommon component of the gas, as arehydrocarbons-combinations of hydrogen andcarbon. It is from the materials of such clouds thatour and other star systems are believed to haveformed.

For life forms to arise and to persist, moleculesmust be awash in a liquid or a gas, so that gentlecontacts among molecules can build up othermolecules and generate a brew of the kind ofcomplexity we find in biological materials. In all of theexpressions of life known to us, this mobility isprovided by liquid water. Given the ferocious andunfriendly conditions of the universe-with points ofintense heat and vast expanses of severe cold-onewould think it rare indeed for any place to holdsurface temperatures in the range that would renderwater a liquid. Surface temperatures depend notonly on the solar irradiation intercepted by the planet,and thus on its distance from the sun and on thesun's size and surface temperature, but also on themass and composition of the planet's atmosphere.

It is the mass and composition of the atmospherethat crucially determines atmospheric pressure.Without a gas pressure, there is no such thing asliquid water. In the absence of substantialatmosphere, water is either a solid or a vapor. All inall, a planet that offers liquid water on its surface is arare occurrence. Rarer still would be the subset ofsuch places that have given rise to the intricatedesigns that we call "life."

Could there be, in this fierce universe, locationswhere perhaps a little brook runs down a hillside,with trees gently swaying in the wind, and withcreatures sitting by the side, enjoying the view? Itseems a far-fetched fantasy in this forbidding

universe. And yet we know one such place: our littleearth.

How was our planet able to bring forth theenormous abundance of surface life that we seearound us? None of the other planets and none oftheir moons have anything comparable. Indeed,because the surfaces of all other bodies in ourplanetary system offer essentially no possibility forthe existence of liquid water, it is very unlikely thatsurface life exists anywhere in our solar system otherthan on the earth. There may be only one Garden ofEden here for large life forms such as ourselves. Butliving beings small enough to populate tiny porespaces may well exist within several-and perhapsmany-other planetary bodies.

Chemical Energy for Subsurface Life

he sun provides two distinct actions. First,it is the source of heat that puts the surfacetemperature of the earth into a range suitable for thecomplex chemical reactions of molecules, and thusfor life. But ambient heat cannot be a source ofenergy, and the warmth of our surface surroundingscould not constitute an energy source for surfacelife. Only a heat flow from a hotter body to a coolerone can be converted into other forms of energy.We have such an energy flow from the hot surface ofthe sun to the cooler earth-the second action that thesun provides-and energy is taken from this flow andconverted into chemical energy in the process ofphotosynthesis.

Photosynthesis is performed today largely byplants and algae, using sunlight to dissociate watermolecules (H20) and atmospheric carbon dioxide(CO2), then reconfiguring the atoms to yieldcarbohydrates such as C6H12061 which can thanbe oxidized ("burned") as needed, back into H2Oand CO2, to yield metabolic energy. This processthen serves as the principal energy source for allsurface life. A planetary surface that does notpossess photosynthetic life would be hostile to anyof the surface life forms we know. Below the surfacethe temperature may be similar to that at the surface;but over small dimensions-like the size of livingforms there-only quite insignificant energy flowoccurs. Therefore, no energy source can existbeneath the earth's surface.

When we consider life's beginning, however, werealize that a puzzle lurks in this account of energytransformation. Photosynthesis is an exceedinglycomplex process. The microorganisms that

developed it must have already possessed intricatechemical processing systems before they acquiredthis more advanced ability. The energy source thatthese initial microorganisms drew on must havebeen chemical to begin with. The chemical energyavailable before the advent of photosynthesis couldnot have been created by solar energy or by life. Itmust have been a free gift of the cosmos.

Where exactly did such chemical energy comefrom? I propose that the original source of energy forearthly life was derived not from photosynthesis butfrom the oxidation of hydrocarbons that were alreadypresent, just as they are also present on many otherplanetary bodies and in the original materials thatformed the solar system. Spanning the range fromthe light gas methane to the heaviest petroleum,hydrocarbons are present in the earth today in largeamounts and to great depths-I believe much largerand deeper than is typically estimated. This view ofthe genesis of hydrocarbons I have called the deep-earth gas theory.'

I think we have good evidence now that a verysignificant realm of life has existed, and still exists,well below the surface biosphere that is home tohumans. This subsurface realm and its inhabitantsconstitute what I call the deep hot biosphere 2-deepbecause it may extend down to a depth of tenkilometers or more below the surface of the earth,a n d hot because, as a result of the naturaltemperature gradient of the earth, temperatures inmuch of that realm approach and even exceed100C.

The conventional notion is that hydrocarbonspresent within the earth's upper crust are derivedstrictly from plant and animal debris transformed bygeological processes-and thus that hydrocarbonscould not possibly have played a role in the origin oflife. But we shall have reason to question this, alongwith many other assumptions. And as we shall see inChapter 2, an abundance of new discoveries have

confirmed life's presence in this crustal realm andunder conditions seldom before thought tolerable toany form of life.

Chemical energy is released in chemicalreactions. The substances we call fuels in oursurface realm are really only one component of theenergy-producing reactions. The other component,oxygen, is so abundant around us that we tend toforget about it. Hydrocarbons, hydrogen, and carbonare fuels for us only because the other componentneeded for the reaction that produces energy isreadily available from the vast store of oxygenpresent in our atmosphere and dissolved inseawater as Oz. This oxygen is largely, but notentirely, created as a residue substance in theprocess of photosynthesis. It, rather than thepetroleum or the coal, represents the fossil fuel leftover from bygone vegetation.

Before photosynthesis was devised by life-andeven now at depths to which atmospheric oxygencannot penetrate-any hydrocarbonusing life musthave depended on other sources of oxygen. Oxygenis the second most abundant element (after silicon)in the crust of the earth. The rocks therefore haveplenty of oxygen in them, but most of it is too tightlybound to be useful. Clearly, sources of oxygen thatrequire more energy to free the oxygen from itsattachment in the rocks than the energy gained byoxidizing hydrocarbons with it cannot providemicrobes with an energy supply.

Subsurface life must therefore depend on sourcesof oxygen in which these vital atoms are only weaklybound with other elements. The largest sources ofweakly bound oxygen in the earth's crust are certainkinds of iron oxides and sulfates (oxidized sulfurcompounds). When oxygen is extracted from ironoxides such as ferric iron, that process leavesbehind iron in a lower oxidation state in which it ismagnetic; examples include the minerals magnetiteand greigite. When oxygen is taken from sulfates,

what is left behind may be pure sulfur or sulfidessuch as hydrogen sulfide and iron sulfide. Theexistence of such by-products of metabolic activity inthe subsurface realm helps us identify thebiochemical processes that have occurred. Theseby-products also provide a sense of the scale andreach of the deep hot biosphere.

It is crucial to the theory of subsurface life that theultimate source of up-welling hydrocarbons residesvery much deeper than the lowermost reach ofsubsurface life. The deep hot biosphere may bedeep, but it must not be excessively deep. Why isthis so? The exponential growth rates of microbes(as of all forms of life) mean that wherever liferesides, the source of energy that supports it mustarrive in a metered flow. If the earliest forms ofsubsurface life had not been checked by limits ontheir food supply, the increase in their numberswould have very rapidly consumed the entire lot in aninstant of geological time, allowing no gradualevolution to take place.

Hence energy that can be used by life must beavailable, but it must not be available all at once. Themetered energy flow for the surface biosphere isprovided by a sun that takes billions of years toconsume its own finite stores of fuels. The chemicalenergy (such as sugars) forged by photosynthesizinglife forms here on the earth is thus created throughtime in a metered way and only in areas that haveliquid water-not in the driest deserts or in theicefields of polar or high mountain regions. Thetransformation from solar to chemical energy nowtakes place at a rate sufficient to feed all the surfacelife we see. But no matter how greedy life may be,organisms simply cannot make the sun radiateenergy any faster. It is energy that supports life, butonly a metered flow of energy sustains life over along period of time.

Understanding the importance to life of a meteredsupply of energy is crucial to delimiting the

possibilities for life's origins. The oftendiscussedwarm little pond that contained nutrients forged withg rea t difficulty by surface processes is not acandidate environment, in my opinion, for thetransition from non-life to life. Such an environmentwould yield a limited amount of chemical suppliesand energy, not a long-term and continuous meteredsupply. What is needed, rather, is an environmentthat can supply chemical energy in a metered flowover tens or hundreds of millions of years, duringwhich time incomprehensibly large numbers ofmolecular experiments might take place. lies evendeeper. I will argue that photosynthesis developed inoffshoots of subterranean life that had progressedtoward the surface and then evolved a way to usephotons to supply even more chemical energy. Whensurface conditions became favorable to life (withregard to temperature, the presence of liquid water,the filtering of harsh components of solar radiation,and the termination of devastating asteroid impacts),a huge amount of surface life was able to spring up.

A Preview of This Book

n the remaining chapters, I shall set forth thetheory that a fully functioning and robust biosphere,feeding on hydrocarbons, exists at depth within theearth and that a primordial source of hydrocarbons

In retrospect, it is not hard to understand why thescientific community has typically sought only surfacelife in the heavens. Scientists have been hindered bya sort of "surface chauvinism." And because earthscientists did not recognize the presence ofchemical energy beneath their feet, astronomers andplanetary scientists could not build a subsurfacecomponent into their quests for extraterrestrial life.Unfortunately, this misunderstanding lingers. Theidea that hydrocarbons on earth are the chemicalremains of surface life that has long been buried andpressurecooked into petroleum and natural gas hasbeen exceedingly difficult to unseat. I have beentrying to do so since 1977, and I discovered alongthe way that some pioneering Russian scientistswere my forebears.3 The reason for this continuingconfusion in understanding how hydrocarbons cameinto being is a story in itself; I shall take it up inChapter 3.

As long as Western scientists continue to assumea biological origin for all terrestrial hydrocarbons, themajor sources of the earth's chemical energy will notbe recognized. And as long as this substantial foodsupply goes unrecognized, the prospect that a largesubterranean biosphere may indeed exist, and existdown to great depth, will likewise fail to attractscientific attention. Thus the particular importance ofChapter 3, in which I will examine the considerationsthat favor the deep-earth gas theory.

Surface evidence for that theory follows in Chapter

4. Most important, I introduce a set of observationsthat cannot be explained at all by a sedimentaryorigin of hydrocarbons-the strong association ofhydrocarbons with a gas that can have no chemicalinteractions either with plant materials or withhydrocarbons: the inert element helium. How canpetroleum have gathered up clearly biologicalmolecules but also an inert gas that is normallysparsely distributed in the rocks? I call thisassociation the "petroleum paradox." Its resolution(in Chapter 5) suggests that multitudes of microbiallife must exist in the pore spaces of the rocks. In myview, hydrocarbons are not biology reworked bygeology (as the traditional view would hold) butrather geology reworked by biology. In other words,hydrocarbons are primordial, but as they upwell intoearth's outer crust, microbial life invades.

Chapter 6 presents the striking results of a large-scale drilling project that I initiated in Sweden to testthe deep-earth gas theory and also to look for deepmicrobial life. In Chapters 7 and 8, I undertake toshow how the deep-earth gas theory can account forconcentrated deposits of certain metal ores in thecrust and also for important features of earthquakes.

In Chapters 9 and 10 I use the deep-earth gas anddeep hot biosphere theories to offer newspeculations on what are perhaps the two mostprofound mysteries of the biological sciences: theorigin of earth life and the prospects forextraterrestrial life. As background, I begin with acomparison of the two biospheres. In what majorways might the surface biosphere and the deepbiosphere differ, beyond the simple fact that onedraws on chemical energy and the other on solar? Ithen revisit the question of life's origin, explainingwhy I believe that surface life is the descendant of anoriginal form of life that began at depth, rather thanthe other way around.

If this sequence from depth to surface bestexplains the origin and expansion of terrestrial life,

then subsurface life on many other planetary bodieswould seem very probable. There are many bodiesin the solar system whose internal conditions arethought to be similar to those of our earth but whosesurfaces do not offer the extraordinary advantagesfor life that ours has. It would be unlikely indeed forsubsurface life to develop just in the one unique bodythat could support surface life as well. This reasoningled me in 1992 to make the tentative prediction thatour own solar system harbors not one but ten deephot biospheres.4

We surface creatures may well be alone in thesolar system, but the denizens of the terrestrial deepseem likely to have many-possibly independentlyevolved-peers. Only when we recognize theexistence of a thriving subterranean biosphere withinour own planet will we learn the right techniques tobegin the search for extraterrestrial life in otherplanetary bodies. Some such techniques and furthersuggestions for future research will be presented inChapter 10.

Our journey will begin in the next chapter with alook at the borderland regions between the twobiospheres. Along hydrothermal vents and petroleumseeps of the ocean, and in hot springs and methane-rich caves on land, we encounter some extraordinaryambassadors from the deep hot biosphere. Here wecan also begin to comprehend why deep may, infact, be desirable for life.

n February and March of 1977, the smalldeep-sea-diving submarine Alvin descended to adepth of 2.6 kilometers along the East Pacific Rise.This region, northeast of the Galapagos Islands,was known to be a center of sea floor spreading. Aresearch ship had drawn a camera over the areathe previous year, confirming the existence of aseries of cracks in the ocean floor that appeared tobe volcanically active. But the occupants of Alvinsaw much more.

Far below the deepest possibility forphotosynthetic life, Alvin's searchlight revealed apatch of ocean bottom teeming with life, in sharpcontrast with the surrounding barrens. This patchwas covered with dense communities of seaanimals-some exceptionally large for their kind.Anchored to the rocks, these creatures thrived in therich borderland where hot fluids from the earth metthe marine cold. New to science were species oflemon-yellow mussels and white-shelled clams thatapproached a third of a meter in length. Most strikingof all were the tube worms, which lurk inside verticalwhite stalks of their own making, bright red gillsprotruding from the top. Like the tube worms ofshallow waters, these denizens of the deep liveclustered together in communities, with tubesoriented outward resembling bristles on a brush. Butunlike their more familiar kin, the tube worms of the

deep are giants, reaching lengths in excess of twometers.

Further investigations soon revealed that thisstrange and isolated community of life was by nomeans unique. Populations of the same organismswere discovered at other points along that ocean rift,a t hydrothermally active vents elsewhere in thePacific, and in the Atlantic and Indian Oceans too.This was clearly a global phenomenon. Theseunsuspected oases represent an entirely new habitatfor life. Where did these creatures come from? Whatsources of energy and nutrients could support suchastonishing fecundity and in such a patchworkdistribution?

Through the windows of Alvin, the 1977 discoverycrew witnessed not only strange life forms but alsostreams of milky fluids and black "smoke" emergingfrom vents in the sea floor. These streams ofhydrothermal fluids, heated and enriched in gasesand minerals, are now known to be the sources ofchemical energy at the base of the vent community'sfood chain. Two decades later, however, we haveonly begun to understand how it all works.

Because we are surface creatures, we readilyadopt the outlook that surface life is the only possiblekind. We marvel at the exotic life along the deep-ocean vents. We assume, of course, that the ventswere originally colonized by emigrants from asurface ecosystem-pioneers in evolving theadaptations necessary to subsist on energy drawnf r o m chemical sources rather than bundled inphotons, the units of energy in which light isdelivered. This top-down scenario is reasonable forthe large animals. Tube worms and clams surely didmigrate down from shallow waters. But no animal ofany kind can serve as the base of a food chain. Allanimals depend on chemical energy stored in thebodies of organisms they consume. Something,therefore, must have already been growing aroundthe ocean vents when the worms and clams arrived.

In my view, the base of the food chain in the deepocean vents is more likely to have emerged frombelow than to have descended from above. Themicrobes (bacteria and archaea) that today supportt h e whole complex enterprise are offspring ofmicrobial communities that lived and still live withinthe earth's crust. Whereas the large life forms canexist only where there is considerable space forthem, the micro bial life that feeds them occurs inunits small enough to inhabit minute cracks in therocks of the sea floor and elsewhere throughout theearth's upper crust. The total volume of rock that isaccessible to such microbes is enormous; as weshall see in Chapter 5, the microbial content of theearth's upper crust may well exceed in mass andvolume all surface life. Indeed, microbes from therealm that I call the deep hot biosphere probablyinvaded this borderland between the twoworldsbetween the deep biosphere and the surfacebiosphere-long before photosynthesis evolved onthe surface. In fact, the chemical differencesbetween the two worlds may have been slight priorto the advent of photosynthesis, because it wasphotosynthesis that transformed the earth's surfaceinto a zone pervaded by free oxygen-molecules ofOZ.

Energy Deep in the Earth

hotosynthesis is an exceedinglycomplex process for turning the energy of light intochemical energy. But why does the route that energytakes have to include chemical forms? Why cannotthe sunlight be made to drive directly all theprocesses that the organism requires? There aresome compelling reasons. First, the energy requiredto run cellular metabolisms must be available inincrements no more than a tenth as powerful as thatsupplied by even a single solar photon. Expecting acell to use a photon directly to synthesize a sugarwould be more ludicrous than expecting a baseballplayer to field bullets from a machine gun. Rather, lifehas devised an extremely sophisticated apparatusto perform the initial task of catching the bullets.

Second, a photon has no patience. Make use of itnow or lose it forever. Sunlight cannot be captured ina jar and stored on a shelf. But its energy can beused to set up molecules such as sugars, that willdeli ver energy on combining with atmosphericoxygen. Our breathing demonstrates this: we take insuch "reduced" (unoxidized) carbon compounds inour food and we inhale oxygen and exhale carbondioxide. This describes the overall metabolic activity,but in fact there are various stages in between, alldependent on the energy provided by the oxidationof the reduced carbon compounds we eat, eventuallyto C02- Sugars or other intermediate molecules canbe stored on the cellular shelf, and the rate of"combustion" can be controlled. Chemical energythus carries the advantage of availability, offering anadjusted amount where and when it is needed.

Because photosynthesis is such a complexprocess, and because the energy derived fromphotons must be converted into chemical energy

before the cell can make use of it, researchers whoprobe the possible origins of earthly life havebecome convinced that the first living cells tappednot sunlight but chemical energy present in theenvironment. Where this chemical energy came fromand what it consisted of remain hotly debatedissues, but the widespread assumption is that eitherthis primordial energy source has long since beenused up or the conditions that produced it billions ofyears ago no longer prevail. I shall return to thisquestion in later chapters. For now it is important toremember only that it would be far more difficult todesign a living cell that could construct chemicalenergy from photons than it would be to design aliving cell that scavenged chemical energy from itssurroundings.

The cells that perform this complex function ofphotosynthesis must have access to liquid water, asalready noted, and they must have access to carbonand nitrogen for the fabrication of proteins, theprincipal building blocks for their chemicalmachinery. The solar energy is used to "reduce"(unoxidize) compounds that will serve to provideenergy as they are later oxidized again. Oxygenmust therefore also be available, as must catalysts(enzymes) that initiate and control the reaction ratesand thereby the power output.

Life as we know it depends fundamentally on thepresence of carbon; earth life is sometimes referredto as "carbon-based life," to distinguish it from thetheoretically possible (but unknown) "silicon-basedlife." Carbon atoms constitute the skeletal structureof all proteins and of all genetic materials of all thelife forms we know. In the surface biosphere, carbonis provided by carbon dioxide, which is present insmall proportion in the atmosphere. Each of theseveral varieties of photosynthesis that life hasevolved begins with carbon dioxide, from which thecomplex molecules of life are then forged. In themost common form of photosynthesis, energeticphotons from the sun are employed to dissociate

water and thus to gain access to atoms of hydrogen.The hydrogen is next used to "reduce" (take oxygenaway from) the molecule of carbon dioxide. Thismakes available unoxidized carbon, which can thenbe used for construction materials and for a varietyof functional materials such as proteins. Unoxidizedcarbon can also be used to construct the varioussugar-like substances (saccharides andpolysaccharides) that provide storable sources ofchemical energy.

When the photosynthetic organism dies, and whenthe other organisms that have benefitted from itsproducts die, microbial decay will return to theatmosphere all the materials that have been takenout. Depending on the type of microbe undertakingthe decomposition, the carbon will be returned to theatmosphere either as carbon dioxide or as methane(CH4). Because the atmosphere is rich in oxygen,a ny methane released into it will spontaneouslytransform into carbon dioxide and water on a timescale of about ten years. So far as the energybalance is concerned, no chemical energy derivedfrom the earth has been used up. Carbon dioxidereturns as carbon dioxide, and water returns aswater.

It may thus seem that carbon cycles through thesurface biosphere in a complete and closed manner.If the atmosphere and the exposed rocks initiallypossess the volumes of raw materials required bylife, the process should go on for as long as the sunshines and temperatures allow water to remain in aliquid state. But as we will see in Chapter 4, the paththat carbon follows through the cycle ofphotosynthesis and oxidation is far from a closedloop. Several times as much carbon as is taken upby living materials is constantly extracted from theatmosphere and taken out of circulation, as long-lived or permanent carbonate rock. The surfacebiosphere must therefore have been kept alive by anongoing and large supply of carbon in the form ofeither methane or CO2 (or, as some observations

would indicate, by a mix of the two). CO2 will be thefinal addition to the atmosphere in either case.

In the surface biosphere, all the energy drivingbiochemical transformations ultimately comes fromsunlight. Life in the deep hot bio sphere does nothave access to sunlight, so the source of energycould not work in the same way. But even there,carbon is the basic building block of life. What is thesource of this carbon in the subsurface realm?

The notion, derived from surface biology, thatCO2 is the standard carbon supply for all life hasbeen applied by some investigators to the deep lifealso. While the ocean water contains plenty of C021it does not have any energy source to reduce this.The reduced carbon that trickles down from thesurface layers would be quite inadequate. No energycan be derived from a process that both starts andends with oxidized carbon. If unoxidized carbon wereavailable at the outset, in the form of hydrocarbonmolecules migrating upward, then these moleculeswould be the logical candidate for a carbon supplythat would also yield an energy-producing sequence,ending up with CO2.

The hot ocean vents are not themselves provincesof the deep hot biosphere; they are borderlandsbetween two worlds, between surface andsubsurface. Nevertheless, their food chains aredriven by processes so different from that of thesurface realm that they are a good place to beginour explorations of deep hot biosphere energy. Theamounts of carbon that sink down from oceansurface life are quite inadequate to supply theexceptionally fertile ocean vent biology. The volcanicrocks of the sea floor contain only a very smallfraction of carbon-about 200 parts per million (ppm).To extract carbon from this source would be difficultand very energy-consuming. There is, however, amuch larger carbon source in all these communities:hydrocarbons. Methane (CH4) is generally the mostabundant, but the heavier members of the series,

such as ethane (CZH6) and all the way up to oilsconstituted of twenty to thirty carbon atoms, are alsofound along the same fault lines, though in regionswhere less volcanic heat is in evidence. As the nexttwo chapters will show, these hydrocarbon fluidsshow many features that suggest they have come upfrom much deeper regions.

The chemical energy supply, we might thensuspect, is driven by the oxidation of thesehydrocarbons. Starting out with hydrocarbons avoidsthe first and energetically most demanding step inthe surface energy cycle. The chemical energy that ismade available at the ocean vents is very similar tothat made available by burning natural gas (which islargely methane) and turning it into water and carbondioxide. There is one snag, however. When methaneis burned in a furnace, there is an unlimited amountof oxygen from the atmosphere available all the time.In the ocean vents, a borderland between the surfacea nd the deep biospheres, there may be someatmospheric oxygen available that was carried downin solution in the cold ocean water. If this weresufficient for converting all the methane suppliedfrom the vents into carbon dioxide and water, thenthis borderland province would be dependent onsurface biological processes, and it would not be anoutpost of what I suggest is an independent realm oflife stretching down into the rocks below. It seemsdoubtful that the prolific life at these concentratedlocations on the ocean floor could receive enoughwaterborne atmospheric oxygen, but a firm answeris not yet known. However, this issue is not of centralimportance. We now know of many cases where wecan probe so far down into the deep biosphere thatatmospheric oxygen has absolutely no access, andwe observe generally similar metabolic processestaking place there. Where does the necessaryoxygen come from?

There is plenty of oxygen bound in the rocks, asnoted earlier, but most of it is so strongly bound thatmore energy would be required to remove it than

could be derived by using it subsequently to oxidizehydrocarbons. There are just two commonsubstances in which oxygen atoms are boundloosely enough that more energy would be obtainedfrom using oxygen so acquired than is spent inacquiring it. These two common substances arehighly oxidized iron (Fe2O3 and associatedcompounds) and oxidized sulfur (such as SO2 andHZSO4 in compounds that are called sulfates). Ifmicrobes at or beneath the ocean vents secure theiroxygen needs from ferric iron oxides, what willremain is a less oxidized form of iron-magnetite orgreigite. Microbial action leaves a clear fingerprintbehind: The crystals of these products are muchsmaller than those of the same substances that havefrozen out in the cooling of rocks from liquid to solidform.

The water of the oceans includes the secondsource of lightly bound oxygen, sulfate, in greatquantities. Sulfate (SO4) is the second mostabundant ion of negative charge in seawater. Theamount of oxygen that could be derived from marinesulfate ions may well exceed the convectedatmospheric oxygen available at the ocean vents. Ifoxygen is, in fact, primarily available near the ventsin the form of sulfate, then the microbes that makeuse of the hydrocarbons will be in an ideal situation:The chemical transformations for extracting thechemical energy from upwelling hydrocarbons willnot run by themselves, because an initial energysupply is required for the first step of freeing oxygenatoms from sulfate. The microbes will be amplycompensated for this energy-demanding step,however, when the second step is taken.

The task of brokering such transactions is left tothe world of microbes. Here, it is important toremember that a chemical fuel is useless to life if itcombusts spontaneously. Dinner would do you nogood if the food burst into flames on your plate. For asubstance to qualify as "food," it must becomeoxidized only with the help of a catalyst created and

deployed by life. This is a fundamental requirementboth for the organisms at the base of the food chainsof the surface and deep biospheres and also for allorganisms that stand later in line.

The removal of oxygen from sulfates at the oceanvents would produce either pure (elemental) sulfur orsulfides, which are unoxidized sulfur compounds.The large quantities of metal sulfides that are foundheaped up at the edges of ocean vents suggeststhat such biologically facilitated transformations areindeed taking place.

A further requirement for the construction oforganisms-be they inhabitants of the surfacebiosphere or the subsurface biosphere-is a supply ofvarious metals required in the protein moleculesknown as enzymes that catalyze chemical reactions.Also required for biological construction or chemicalprocessing are some reactive molecules thatcontain elements such as sulfur, phosphorus, andchlorine. The required quantities of these are smallenough that the upper crust of the earth can usuallysupply them. The deep biosphere and the landportions of the surface biosphere are thusadequately nourished. But the surface waters of theopen oceans may be impoverished, particularly withrespect to phosphorus and iron.

In summary, there are important differences andimportant similarities between the two biospheres.The surface biosphere runs on solar energyconverted into chemical energy; the deep biospherebegins with chemical energy freely supplied from thedepths of the earth. Both biospheres rely onunoxidized carbon as the building block of life, butsurface life extracts it initially, with the help ofsunlight, from carbon dioxide in the atmosphere,whereas deep life extracts it from the samesubstances used as the energy source:hydrocarbons. Oxygen is a requirement in bothrealms, since chemical energy is provided only in theprocess of oxidation. For surface creatures, oxygen

is available largely in the form of pure, molecularoxygen. Inhabitants of the subsurface must workharder to gain their supply, extracting oxygen atomsthat are loosely bound in iron oxides and sulfates.

The Ecology of Deep-Ocean Vent Life

ecause we are surface creatures, we areinclined to regard an ecosystem based on chemicalenergy rather than photosynthetic energy as astrange, if wonderful, adaptation of life. We marvela t the ecology of the deep ocean vents as a deftadjustment of surface life to an inhospitable realm.The evidence argues otherwise. Microbes and evenanimals are thriving at these vents; growth rates arethought to exceed those in even the most productivesurface realms. If the theory of the deep hotbiosphere is correct, we would infer that themicrobial pioneers invaded from below. Manyviewpoints would have to be changed as aconsequence.

The communities of life at the deep ocean ventsdiffer from other marine ecosystems not so much intheir garish macrofauna but in their unseenmicrobes-the bacteria and archaea at the base ofthe food web. Two decades of studies haverevealed that these microbes feed on moleculesgushing from the vents: hydrogen (H2), hydrogensulfide (1-12S), and methane (CH4), each of whichcan supply energy only if oxygen is available.' Noknown animal can feed on any of these chemicalsdirectly, but animals can feed on microbes that do.What is particularly remarkable about the deep-ocean vent communities is that many of themacrofauna seem to be dependent on symbioticpartnerships with the microbes.

Clams and mussels have entered into symbioticpartnerships with microbes bound in their gilltissues. The giant tube worm species, however, hastaken partnership to a new dimension. Its interiorguests are so skilled in producing food forthemselves and their host that coevolution has

themselves and their host that coevolution hasatrophied the worm's digestive system and deprivedit of a mouth. Utterly dependent now on the excessproduction of its symbionts, the tube worm hasevolved a large and specialized organ deep insideits body for the microbes to inhabit. The wormsupplies its microbes with the materials they need byemploying feathery red gills to filter useful moleculesout of seawater. Then it volunteers its own circulatorysystem to deliver what the gills have gathered.

The greatest challenge to organisms alonghydrothermal vents is posed by the risk of beingswept out of range of the vent and thereby losing thechemical supplies and the temperature range theyrequire. The bivalves and tube worms solve theproblem by anchoring themselves in place. Crabsand shrimps and snails that live among the fixedorganisms can, of course, creep and clutch asneeded. The microbes that constitute the primarystep in the food chain have found ways to hold theirplace, too. The most heat-adapted varieties can livevery close to (and even inside of) the vent. Whereverit is too hot for animal grazers such as snails tointrude, microbes cling to the rocks in communalmats of slime. Those that take to the water columnabove the venting fluids possess a whip-likeflagellum by which to locomote, sensing temperatureor chemical stimuli to guide their directionalmovements and thus staying within or next to the ventstream. The most audacious bacterial entrepreneursare those that have made themselves welcomeguests within the very tissues of the bivalves andtube worms. There they are protected from prowlinggrazers as well as errant currents.

The hydrogen, hydrogen sulfide, and methanefuels consumed by both free-living and symbioticmicrobes in the vent communities are exploited bymicrobes that access oxygen atoms loosely boundin ferric iron oxide carried up from the depths in ventfluids, oxygen derived from sulfate that pervadesseawater, and perhaps also free oxygen in theseawater.

All animals, however, depend on molecularoxygen for their metabolic needs. No animals areknown to extract oxygen directly from oxidizedmaterials in their surroundings. Many investigatorshave therefore assumed that the macrofauna at thevents depend on the molecular oxygen carried downin seawater. Thus these species would still bedependent-indirectly-on surface photosynthesis.They would still be members of our food club.

The great abundance of molecular oxygen in theatmosphere is mainly due to its production as awaste product of photosynthesizers- by plants onland and algae and cyanobacteria near the surfaceof the sea. Molecular oxygen diffuses into surfacewaters, especially at high latitudes, because thesolubility of oxygen in seawater is greatly increasedat low temperatures. Oxygen-rich waters from theArctic and Antarctic plunge to the deeps and thenslowly snake along the ocean floor, following valleys,toward the equator. A global-scale system ofatmospheric and oceanic circulation thus bringsmolecular oxygen to some deep areas of the oceanfloor.

Most of the ocean vents that have beendiscovered are situated at volcanic ridges and highspots of ocean floor, and such areas do not receivethe cold, oxygen-rich flows of polar waters. Whetherthe oxygen that had diffused to these locations andis made available only by slowly moving water wouldbe sufficient to foster the extremely rapid growthobserved at the vents is doubtful. Although themacrofauna cannot extract oxygen from othersources, microbial life can. If the supply of oxygen isthe limiting factor for the vent community, then wehave to suspect that symbiotic exchange may haveadvanced to such a state that the symbiont microbeswithin the animals are stripping oxygen atoms fromseawater sulfate not only for themselves but also fortheir animal hosts. No doubt further researches willdetermine whence the macrofauna derive theiroxygen. But, as we shall discuss in the next section,

many microbial communities have been identifiedthat certainly have no access to atmospheric oxygen.

Life thrives at the ocean vents because these aresites at the borders between two worlds. Anabundance of chemical energy can be extractedfrom the chemicals that meet there and that had noopportu nity to reach equilibrium with one another.Upwelling fluids from the world below are rich in"reduced" molecules, such as hydrogen andmethane. Hydrogen sulfide is also present, but wedo not yet know whether this is a primary fluid fromthe depths of the earth or a product of microbes asthey utilize a combination of hydrogen and sulfate forenergy needs.

Of the three major sources that provide energywhen reacted with oxygen (hydrogen, hydrogensulfide, and methane), hydrogen sulfide hasattracted the most research interest, because itseems to be the fuel on which the microbialsymbionts of the giant tube worms and clamsdepend. But the carbon atoms that form the core ofall organic molecules must be obtained elsewhere.The presence of methane in the output of oceanvents thus assumes particular importance; it can bethe source of the required carbon as well as thesource of chemical energy.

Hydrocarbons bear a structural resemblance tofoods we eat that are derived fromphotosynthesizers. For example, the only materialdifference between a molecule of hexane (a six-carbon form of petroleum) and a molecule of glucose(a six-carbon sugar, common in foods at the surface)is that hydrogen atoms surround the chain of carboni n hexane, whereas water molecules surround thechain of carbon in the sugar. The hexane C6H14 is ahydrocarbon, whereas the sugar C6H1206 is acarbohydrate. The terminological difference is subtlebut important. For us animals the carbohydrate isfood, the hydrocarbon poison. Nevertheless, thebiological idiosyncrasies of our own tribe of complex

life should not be allowed to constrain our judgmentas to the possibilities-indeed preferences-amongthe multi-talented microbes. They might well have ametabolism that requires an input of petroleum.

Microbes that utilize methane as a source ofenergy in the presence of oxygen, and also as asource of carbon, are known to be present in thehydrothermal vent communities. Suchmethanotrophs ("methane eaters") have beenidentified as symbionts within the macrofauna- thusfar, only in mussels-but they are presumably free-living as well.2 They can consume heavierhydrocarbons, too.

Are the methanotrophs of the deep-ocean ventsambassadors from this other, deeper, and perhapsindependent world? We know that clams and wormsdo not venture any deeper than the thin skin ofsurface rock and sediments. But what about thebacteria and archaea? If microbial slimes on therocks near and within the vents thrive on methaneand sulfide gases that rise up from below, might theynot also find suitable habitat within cracks and porespaces deep below the crustal surface?

Other Borderland Ecologies

ithin the past three decades, manya n d various borderland ecosystems have beendiscovered and their secrets probed. First tocapture scientific attention was a type that had longbeen enjoyed by crowds of tourists: the microbialcommunities that colorfully coat the rocks within hotpools of Yellowstone National Park. Serious study ofthe metabolisms of Yellowstone's thermophilic(heat-loving) microbes began in the mid-1960s.3 Itwas here that scientists first came to appreciate theextraordinary talents of the earth's seeminglysimplest forms of life. For example, one bacteriumdiscovered in Yellowstone's hot pools, Therm usaquaticus, provided the enzyme that launched themolecular biology industry by making DNAreplication fast and easy. Today, Yellowstone's hotsprings offer rich prospecting for scientists seekingto add new names to the list of microbes classifiedin the taxonomic domain of Archaea.

In 1977 the exciting exotica we have alreadydiscussed were discovered beneath the sea-theelaborate assemblages of microbes and animals atthe edges of hot springs on the ocean floor. In 1984came the discovery of more assemblages ofsymbiotic microbes, tube worms, and bivalves-not,this time, in the abyssal depths but on the muchshallower continental shelves.4 Similar in form, buttaxonomically different at the species or even genuslevel, tube worms and bivalves on the continentalshelves were making their living in "cold seep"regions, where crude oil and hydrocarbon volatilesseep up through the sediments. No hot springs orother hydrothermal action is associated with theseseeps. Unlike the hydrothermal vents, which arepoint sources restricted in size, cold petroleum

seeps offer marine life chemical energy over vastexpanses of the continental shelves that are toodeep to support photosynthesis. (In even the clearestocean waters, photosynthesis is impossible anydeeper than about 200 meters beneath the oceansurface, and continental shelves often sink to a depthof a kilometer or more.) Growth rates in the regionsof hydrocarbon seeps are not, however, as high asthey are at the actively venting rift zones of the deepocean.

On land, too, an ecosystem border realm hascaptured scientific and public attention. In 1986 acave in Romania-until then, isolated from theatmosphere-was discovered and found to contain athriving ecosystem based on the chemical energy ofreduced gases emanating from below. Ten yearslater, when its biological inventory was published,this cave habitat was touted by the media as the firstinstance of a terrestrial ecology that was not basedon photosynthesis and yet was able to support notjust microbes but land animals as well.5 Feeding onthe bacterial base of the food web are more thanforty species of cavedwelling invertebrate animals,including spiders, millipedes, centipedes, pillbugs,springtails, scorpions, and leeches. Thirty-three arenew to science. As with the deep-ocean vent habitat,hydrogen sulfide was identified as the reduced gassupporting the base of the food chain in this cave,though I suspect that methane also plays a role.Indeed, methane consumers may well be generatinghydrogen sulfide as a waste product when sulfate isused to oxidize methane, in which case the sulfideconsumers would be a notch up from the base of thefood chain. Hydrogen sulfide, converted by waterinto sulfuric acid, probably carved out the limestonecave.

In 1997 another cave ecosystem based entirely onchemical energy was explored in southern Mexico.That cave, too, appears to have been carved out oflimestone by a flow of sulfuric acid. The acid fumesin this cave are so intense that scientists were able

to venture a mile into its tunnels only with theassistance of breathing masks. Microbial life is soprolific throughout that the walls are shrouded inslime.' Feasting on the microbes is a community ofinvertebrates, but this ecosystem also supportsvertebrates: tiny fishes in the waist-deep water thatoccupies the tunnel system.

Also very recently, Russian scientists have beenpreparing to explore a vast lake-as large as LakeOntario-that was discovered in central Antarcticabeneath four kilometers of ice.' Lake Vostok owesi ts existence to the entrapment of heat upwellingeverywhere within the earth. The thick glacial ice,strangely enough, acts as a thermal insulator,segregating the heat from the intense cold ofAntarctic air. Remote sensors indicate a water depthof perhaps 600 meters in some places, underlain bysediments 100 meters thick. Drilling was halted 250meters above the water line, pendingimplementation of procedures that could ensuresterile contact. If life is present down there, it willunquestionably be based on chemical energy wellingup from below. To test that possibility, it isimperative to prevent contamination of the pristinelake by surface microbes. NASA has expressedinterest in fostering technologies for sterileexploration of Lake Vostok, which would probablyhappen no sooner than 2001. One reason forNASA's interest is that a subglacial lake offers anextraordinary analog for the subsurface environmentof Europa, a moon of Jupiter that is covered with athick layer of ice and may have liquid waterunderneath that.'

An important discovery of very large amounts ofmethane was made in the last two decades.Methane hydrates, crystals of water ice that entrapmethane molecules within their lattices, exist in greatquantities on many areas of the ocean floor. Thepresence of methane raises the freezing point ofwater by an amount depending on the ambientpressure, and therefore this ice can form in regions

where water is supplied in liquid form and thenfreezes where methane is added.

For methane hydrates to form, temperatures mustbe no greater than about 7C and pressures no lessthan about 50 atmospheres. This means that muchof the sea floor that is outside of volcanic zones andcovered by water to a depth of 500 meters or morecould support methane hydrates.`' Within the pasttwo decades we have learned, both by remotesensing and by direct sampling, that methanehydrates do indeed exist in great quantities in manyareas of the ocean floor. They produce a clear andunique signature on sonar and are remotely sensedas a distinct layer in ocean muds, sometimes lyingdirectly on the bedrock of the ocean floor. A largearea of the continental shelf has been surveyed inthis fashion. Results indicate that methane hydratesmay, in fact, be present in all areas where thepressure and temperature allow them to form.10 Ithas been estimated that methane hydrates (thosewithin the Arctic permafrost layer as well as thoseunder the sea) contain more unoxidized carbon thanall other deposits of unoxidized carbon known in thecrust, such as crude oil, natural gas, and coal."

Often there is more carbon in the methane atomstrapped in a deposit of hydrate than in all of thesediments associated with that deposit. In suchinstances the conventional explanation of its source(biological materials buried with the sediments)cannot account for the production of so muchmethane. The methane embedded in the ice latticesmust have risen from below, through innumerablecracks in the bedrock. Once a thin, capping layer ofthe solid forms, the genesis of more such hydrateunderneath becomes an inevitability, providedmethane continues to upwell.

This conclusion-that the source of methane liesbeneath, not within, the crustal sediments-isstrengthened by evidence of pockets of freemethane gas beneath some regions of hydrate

ice12 and also beneath permafrost layers of Arctictundra.13 In these regions, downward migration ofmethane gas from overlying sediments does notseem conceivable. Gases, after all, do not migratedownward in a liquid of greater density. If there isany flow, it is in the reverse direction.

Lake Vostok, which we have just discussed, willbe an ideal place to check on the quantity ofhydrocarbons that have come up from below sincethe ice cover formed. The quantities of methanehydrates contained there may be very large, theymay even represent the major component of thelake.

In the domain of high methane hydrates there isalso macro-life, just as at the ocean vents. Littleworms are found there that plough through themethane hydrates and the overlying water.14 Theirexistence indicates that such methane hydrates havebeen there long enough to allow life to adapt to thestrange circumstances. Most probably symbioticmicrobes inside the worms use energy derived fromthe oxidation of methane. The carbohydrates andother biological molecules the microbes produce arethen shared with their animal hosts.

Hydrates made up with CO2 rather than methanecan exist also, though over a smaller stability rangeof temperature and pressure than methane hydrates.Nevertheless, there are substantial areas of oceanfloor that could support CO2 hydrates, but few-if any-such samples have been found. The conclusion mustbe that the "gentle" but widespread addition ofcarbon to the atmosphere is a global phenomenono f diffusion from the ground of methane and otherhydrocarbons, no doubt at different rates at differentlocations and at different times. The dominance ofCO2 over methane from volcanoes is the exceptionand not the rule. This conclusion then agrees with thefinding that methane is far more abundant than CO2in wellbores (to the good fortune of the petroleumindustry), and also with the evidence from meteorites

t h a t hydrocarbons and not CO2-producingcompounds will have been the principal input ofcarbon in the forming earth. (Chapters 3 and 4 willexplore these points in detail).

Deep Is Desirable

n light of the discoveries of thrivingchemicalbased ecosystems associated withmethane hydrates, hot ocean vents, and coldpetroleum seeps on the ocean floor, along with thoseassociated with hot springs and gas-rich caves onland, we can conclude that methane, hydrogensulfide, and other energy-rich gases (those thatcould provide large amounts of energy if combinedwith supplies of oxygen) are attractive to life formsthat span a wide range of temperature. Very close tothe hot ocean vents, however, and wherever hotsprings on land are more than merely warm-above,say, 45C-these habitats do not support animals. Butheat-loving (thermophilic) microbes are abundant inthese places.

As temperatures rise even more, thermophilesdrop out, but hyperthermophiles-microbes that growbest at 80C or higher15-go about their businessunperturbed. The waxy cell membranescharacteristic of hyperthermophiles facilitatematerial exchange at temperatures at which fattymembranes like our own would simply melt.16Hyperthermophiles can grow and reproduce only atsuch high temperatures. At lower temperatures theirmembranes stiffen to the point where materials canno longer pass through as needed. Molecules calledheat-shock proteins enshroud the DNA and regularproteins of hyperthermophiles, guarding theintricately folded structures against the unravelingthat such high heat would otherwise bring about.

What are the highest temperatures thathyperthermophiles can tolerate? We are stilluncertain. But we do know that temperature alone isno more determinative of an environment's livabilitythan it is determinative of a fluid's boiling point. One

than it is determinative of a fluid's boiling point. Onemore factor must be considered: pressure.

Although the boiling point of water is 100C at sealevel, it rises to a full 300C at a depth of just 876meters. At that depth, the water column exerts apressure of 87 atmospheres, which means 87 timesmore than the pressure exerted by the atmosphereat the surface of the sea. This pressure is sufficientto prevent water molecules at even 299C fromexpanding into a vapor phase. Deeper still, at adepth of 2.25 kilometers, the "critical point" isreached. Here the pressure is so great that nomatter what the temperature, there is no longer anydistinction between vapor and liquid. Rather, it ismore appropriate to refer to water beyond the criticalpoint as existing as a fluid-specifically, a "super-critical" fluid.

Now consider that the first community ofhydrothermal vent organisms ever witnessed in theabyssal realm of the sea was found at a depth of 2.6kilometers. Here water is a supercritical fluid. Waterat temperatures of about 300C has been detectedissuing from the vents, but it is cooled quickly as aresult of mixing with surrounding water. Boiling is notan issue for organisms at that depth, because watercannot boil there. Melting of membranes andunraveling of proteins, rather, may become thelimiting factors for life at high temperatures."

Because of the effect of pressure, if one mustcope with temperatures approaching or exceeding100C, then deep is certainly desirable. Howwidespread are zones of such high temperature?Hot springswhether on the sea floor or on land-arefar from the norm. They occur where heat generateddeep within the planet finds a rapid escape route tothe surface, by way of fluids buoyed up from below.These are active volcanic zones. Far more commonare non-volcanic regions, such as those over whichyou and I are probably sitting right now.

The earth generates its own heat from

compression, gravitational sorting, and radioactivedecay deep within its core and mantle. In a non-volcanic region the temperature of the rock,beginning at the surface, increases steadily withdepth and at a rate fairly uniform over the entireglobe. This phenomenon is referred to as earth'sthermal gradient. The temperature of the crust nearits contact with the atmosphere is approximately20C over most of the area. The temperatureincreases at a rate of between 15C and 30C perkilometer of depth in nonvolcanic regions.

Hyperthermophiles are known that can grow attemperatures of 110C. This means that, on averageand provided that the necessary chemical resourcesare present, life as we know it could survive down toa depth of six kilometers in regions of crust thatexhibit the low temperature gradient (15C perkilometer or less) and three kilometers where thetemperature gradient is high (30C per kilometer). Itis not yet clear whether hyperthermophiles exist thatcan tolerate higher temperatures still. Somemicrobiologists consider that the temperature limitfor microbial life may be as high as 150C.18 In thatcase, life might extend to deeper levels, in somecool areas possibly to a depth of ten kilometers.

It is crucial to remember that because of thesteady rise in pressure with depth, nowhere withinthe earth's crust (with the exception of volcaniczones) does the combination of temperature andpressure ever permit water to boil. What aboutmethane, the lightest and hence quickest to boil of allhydrocarbons? Moving downward along any thermalgradient, methane becomes denser at the greaterpressures of increasing depths, even as it remains avapor. What does this increase in density mean forsubterranean life forms that feed on methane?

For one thing, the greater density means thatmethane is actually easier for life to access at depth.At a depth of six kilometers, for example, methanewould be 400 times as dense as it would be on the

surface at atmospheric pressure. Also, highertemperatures that coincide with greater depthsescalate the rate at which methane molecules collidewith the cell membranes of microbes. Both factorsenhance the rate at which methane would beexpected to diffuse across cell membranes. Deep isthus desirable not only to ease some of thebiological problems created by high temperaturesbut also to assist methane consumers in accessingtheir food.

Up here in the surface biosphere, where methaneexists only as a diffuse gas, methane consumers area curious group. But methanotrophs may be far fromtangential members of the food web in the deepbiosphere. Indeed, they may be the foundation ofthat system.

Beneath the Borderlands

o study the deep hot biosphere andsample its inhabitants, we must probe far beneaththe borderland regions of hot springs, hydrothermalvents, oil seeps, methane hydrates, and gas-richcaves. We must peer into the bottom of deep wellsdrilled into the earth's crust.

When I began developing the deep hot biosphereidea in the early 1980s, and when my "Deep, HotBiosphere" paper was published in 1992,19 apersistent criticism was that microbes brought up insamples from the depths of oil and gas wells werenot native inhabitants but opportunists introducedfrom the surface in biologically contaminated drillingfluids.20 This contamination argument was at firstdifficult to refute. But in 1995 a key paper publishedin one of the top scientific journals demonstrated thatmicrobes discovered at a depth of 1.6 kilometers inFrance were truly "members of a deep indigenousthermophilic community."21 The following yearanother report of indigenous microbes, this timefrom an oil well in Alaska, established active biologyat a depth of 4.2 kilometers and a temperature of110C.22 In 1997 the discovery of microbial fossilsembedded in granitic rock at a depth of 200 metersconfirmed the indigenous interpretation; fossilscannot be introduced by drilling fluids into solidgranite.23

Thus far, the deepest indication of active biologywas detected in 1991, at a depth of 5.2 kilometers inSweden, as we will see in Chapter 6.24Significantly, the well in which these microbes weredetected had been drilled into solid granitic bedrock,not the sedimentary strata that generally attractpetroleum prospectors. A sample that had been

taken and sealed at depth and then drawn up wascultured in the laboratory. It yielded previouslyunknown strains of anaerobic microbes thatreproduced only in the temperature range from whichthey had been sampled, 60C to 700C.

The term I coined, deep hot biosphere, issometimes mentioned in scientific papers or mediacoverage interpreting such findings of microbial lifediscovered at depth.25 Many of these reports,however, misunderstand my argument and, I believe,misinterpret the facts in ways that are far from trivial.These errors are of two types.

First, microbes drawn from deep oil wells arerightly interpreted as feeding on hydrocarbons. Butthere is an implicit assumption that the hydrocarbonsare the reworked remains of life that once belongedto the photosynthetic food club-algae and the like.26This is the standard Western view of a putativebiogenic origin of petroleum, which I will challenge inthe next chapter. As long as petroleum is regardedas biogenic, then no matter how far down life may befound in oil wells, it will always be regarded as anovelty-a thrilling extension of the surface biospheredownward as it mines its own earlier remains.

A second error in reports heralding the discoveryof a deep biosphere, or even a deep hot biosphere,is the characterization of indigenous microbes as"rock-eating." This second error requires a bit moreexplanation than the first. To begin with, "rock-eating" is the usual interpretation of microbialmetabolism when microbes are discovered in wellsdrilled in igneous rock. Because igneous rocksformed from a melt, the only hydrocarbons they couldpossibly contain must have migrated fromsomewhere else after the magma cooled into rock.T he standard way of thinking would have thosehydrocarbons seep into the cracks and pores ofigneous rock from a sedimentary "source" rock(such as black shale) nearby. When there is nonearby source rock, this explanation is of no use.

Reports of microbial life within igneous rocks areconsiderably less widespread than reports ofmicrobes detected in sedimentary rocks. Thereason for their scarcity is simple: If we believe thatoil and gas are the reworked remains of surface lifelong buried in sediments that consolidated into rock,then why would anyone drill in igneous territory? Thenumber of boreholes drilled in sedimentary rock isso much larger than the number drilled in igneousrock that this disparity alone can readily account forthe difference in the number of reports of microbiallife from the two domains.

Deep drilling into non-sedimentary rock hasnevertheless been undertaken, either forexplorations of a general kind or for an altogetherdifferent purpose: to assess the radioactivecontamination of ground water. During the Cold War,radioactive wastes generated by plutoniumproduction were not always disposed of carefully.This was the case at the Hanford nuclear processingfacility in central Washington state, which was builton Columbia River basalt. A test well drilled 400meters into the igneous rock to sample radioactivecontamination of aquifers had the side effect ofrevealing bacteria.27 What were they living on?Because everyone believed that such an extensivebasalt could not possibly contain hydrocarbons, theplentiful supply of methane detected there28 wasinterpreted as a metabolic by-product of a laterstage in the food chain (with what source ofcarbon?)-rather than, as I would have it, the fuelsource for the primary producers.

In igneous rocks, methane is by far the mostcommon fluid, second only to water. Methane is themost likely fuel source, carbon source, and hydrogensource at the base of the food chain. To my way ofthinking, carbon dioxide is largely a product ofmicrobial oxidation of hydrocarbons, not the sourceof carbon for the base of the food chain. This view-that hydrocarbons provide the carbon source as wellas the fuel for biosynthesis at depth-has been greatly

strengthened by a paper published in 1994.29 PetraRueter and colleagues cultured a moderatelythermophilic microbe in conditions that confirmedthat this metabolic strategy was in use, with sulfateproviding the oxidant.

For many reasons, therefore, I do not agree withthe ecological interpretations of the researchersworking on the Columbia basalt aquifer.Nevertheless, I can well understand howmisinterpretations could have been made. It isdifficult to sample, culture, and identify the presenceof indigenous life at depth. It is even more difficult todetermine the foundation of the food web and thefuel and material sources on which the primarymetabolism is based. Until the primary metabolismis identified, however, one cannot be sure whether aparticular chemical constituent is original resource orbiological product.

It should now be clear that the best way to learnabout the deep hot biosphere-and even to testwhether this hypothesized realm of life does indeedexist-is to drill into rocks and examine what is downthere. Few if any holes have yet been drilled with theexpress aim of searching for deep life. Wells aredrilled to search for commercial quantities ofhydrocarbons, to test for contamination of groundwaters, or to provide data for understandinggeological processes. Any microbial lifeencountered during such ventures is almost alwaysdismissed as biological contamination from thesurface, introduced in the drilling fluids.

The borderland habitats are exciting, but theycannot demonstrate with certainty whether and whatbiological processes may be active at depth. Thusfar, we have had only glimpses of what may prove tobe a vast expression of earth life awaiting ourexploration. There has, fortunately, been a recentsurge in demand to study microbes hauled up fromdepth. Interest in a deep hot biosphere (though notnecessari ly my stringent interpretation of an

independent, hydrocarbon-based deep biosphere)has blossomed. Part of this interest has beenstimulated by the numbers of deep wells that havetested positive for biological inhabitants. Life is notsupposed to be down there, so our curiosity ispiqued. Another substantial part of the interest isattributable to the success of the University of Illinoisevolutionist Carl Woese in convincing biologi