IN tudying tory of Yelloone National Park, BobCiansen of ted States Geological Survey became puzzled about somet,oddly,  troubled anyone before:  find t ime t Yelloone ure—t’s ed for all itsgeysers and oteamy features—and t volcanoes is t tty conspicuous. But Ciansen couldn’t find tone volcanoanyructure known as a caldera.

    Most of us, es in a symmetrical mound.

    t Parícutin in Mexico, a farmer artled tosee smoke rising from a patc  opped out at almost fourteen  andogeten trusivelyvisible volcanoes on Eart a fe. But ted type of volcano t doesn’t involve mountain building. t t open in a single migure, leaving be subsided pit,tin one obviously ype,but Ciansen couldn’t find the caldera anywhere.

    By coincidence just at time NASA decided to test some neude cameras bytaking pograpone, copies of o ties on tion t t make a nice bloors’ centers. As soon as Ciansen saos o spottually ta crater more ty miles across—mucoo o be perceived from any some time in t Yelloone must o humans.

    Yelloone, it turns out, is a supervolcano. It sits on top of an enormous  spot, areservoir of molten rock t rises from at least 125 miles do fromt spot is one’s vents, geysers,  springs, and popping mudpots. Beneat is about forty-five miles across—roug eig its t point. Imagine apile of tNt about t miles into to about t of t cirrus clouds, and you  visitors to Yellooneare sop of. t sucs on tabove ed Yelloone and about territory about1,700 feet  bleaclysm is pretty o Professor Bill McGuire of University College London, “you be able to get ers of it” ing. t followed would be even worse.

    Superplumes of type on s are ratini glasses—t spreading out as to create vast boable magma.

    Some of to 1,200 miles across. According to tal explosively but sometimes burst fort, continuous outpouring—aflood—of molten rock, sucraps in India sixty-five million years ago.

    (trap in text comes from a Sributed to tainly didn’t gassings. Superplumesmay also be responsible for ts t cause continents to break up.

    Suc all t rare. t ty active ones on t t, and t-knole Pitcairn int apart from Yelloone test idea one’s ended up beneatinental plate. Only tain: t t at Yelloone is t t is . But  spot or  is t is tter of ed (as it e. tinentalnature of t makes a o its eruptions. end to bubble aeadily and in a comparatively benign fasone blo doesn’t en, but  to stand well back.

    Since its first knoion 16.5 million years ago, it  a imes, but t recent tions are t get ten about. t eruptionimes greater t of Mount St.  imesbigger, and t nobody knoly   least ty-five imes greater t.  per timesmore monstrous.

    e ely noto compare it to. t blast in recent times  ofKrakatau in Indonesia in August 1883, ed around ter slos if you imagineted material from Krakatau as being about t of tone blasts  about  St. han a pea.

    tone eruption of t out enougo bury Neate to a depty-seven feet or California to a depty. tmade Mike Voorern Nebraska. t blast occurred in  a rate of about one inc raveled over it, so t today it is directly under nort yoming. (t spot itselfstays in one place, like an acetylene torc a ceiling.) In its  leaves t ofric are ideal for groatoes, as Idas like to joke, Yelloone eppingaround geysers.

    t Yelloone eruption covered all or parts of nineteen ernstates (plus parts of Canada and Mexico)—nearly ted States  of t of America, an area t produces roug is  like a big sno  in ted to grooput all t took t monto clear 1.8 billion tons of debrisfrom teen acres of trade Center site in Ne ake to clear Kansas.

    And t’s not even to consider tic consequences. t supervolcano eruptionon Eart toba, in nortra, seventy-four te   it obablast  least six years of “volcanic er” and goodness kno. t, it is t, may  to tinction, reducing tion to no more t means t all modern ion base,y. At all events, tosuggest t for t ty total number of people on Eart any time. t is, needless to say, a long time to recover from asingle volcanic blast.

    All tically interesting until 1973, ous: er in Yelloone Lake, in t of to run overt t te end of ter mysteriously flos did a y survey and discovered t a largearea of ting up one end of ter to run out at ted one side of a cral region of t   formallysurveyed. tral part of t incnoo be swelling again.

    ts realized t only one tless magma chamber.

    Yelloone  te of an ancient supervolcano; it e of an active one. It  about time t to  t tone’seruptions averaged one massive blo one, interestingly enougone, it appears, is due.

    “It may not feel like it, but you’re standing on t active volcano in tone National Park geologist, told me soon after climbing off an enormousorcycle and s at ters atMammot Springs early on a lovely morning in June. A native of Indiana, Doss is anamiable, soft-spoken, extremely tful man ail. A smallsappud graces one ear. A sligrains against his crisp Park Service uniform.

    employee. In fact,   t place into do it,”  off in a bouncy, battered four-o let me accompany  doing  is a park geologist does. t assignment today is to give anintroductory talk to a neour guides.

    Yelloone, I  out, is sensationally beautiful, atelymountains, bison-specked meadoumbling streams, a sky-blue lake, ing. “It really doesn’t get any better t,” Doss says. “You’vegot rocks up at Beartoot are nearly ters of to Eart mineral springs s at t springs from s title—“—or prettier.”

    “So you like it?” I say.

    “O,” y. “I mean I really love it ers are toug too , but ’s just—”

    errupted o point out a distant gap in a range of mountains to t, o vieains, old me, ins.

    “t gap is sixty or maybe seventy miles across. For a long time nobody could understandiansen realized t it o be because tains  bloy miles of mountains just obliterated,you knoty potent. It took Ciansen six years to figureit all out.”

    I asked  caused Yelloone to blow w did.

    “Don’t knorange t understandt all. Vesuvius, in Italy, ive for til an eruption in 1944and t just stopped. It’s been silent ever since. Some volcanologists t it isrectle . But nobody knows.”

    “And  if Yelloone o go?”

    time it bleand possibly some cterns of beeam vents, butnobody really knows.”

    “So it could just blo warning?”

    fully. trouble,  nearly all t itute  in some measure at Yelloone. “Eartions, but ts of eart year. Most of too small to be felt, but theless.”

    A ctern of geyser eruptions migaken as a clue,  too vary unpredictably. Once t famous geyser in tused to erupt regularly and spectacularly to s of t, but in 1888 it juststopped. t erupted again, to a  of eig. SteamboatGeyser is t geyser in t bloer four  intot tervals bets eruptions tle as four days to almostfifty years. “If it bleoday and again next   tell us anyt all about miger or ty years from now,” Doss says.

    “tile t it’s essentially impossible to draanyt happens.”

    Evacuating Yelloone s some tors ayear, mostly in tively fe intentionally narroly to sloraffic, partly to preserve an air ofpicturesqueness, and partly because of topograpraints. At t of summer, itcan easily take o cross to get anyop, bison jams. e get wolf jams.”

    In tumn of 2000, representatives from tional ParkService, along  and formed sometoneVolcanic Observatory. Four sucence already—in on—but oddly none in t volcanic zone in t actually a t more an idea—an agreement to coordinate efforts at studying andanalyzing t tasks, Doss told me, o draion in t of a crisis.

    “t one already?” I said.

    “No. Afraid not. But there will be soon.”

    “Isn’t t just a little tardy?”

    ’s just say t it’s not any too soon.”

    Once it is in place, t tiansen in Menlo Park, California,Professor Robert B. Smit ty of Utaential cataclysm and advise tendent. tendent ake to evacuate tone o blo tes.

    Of course it may be tens of t day comes. Doss t come at all. “Just because ttern in t doesn’t mean t it stillrue,” o suggest t ttern may be a series ofcatastrop. e may be in t no most of tallizing. It is releasing itsvolatiles; you need to trap volatiles for an explosive eruption.”

    In time ty of otone, as atingly evident on t of August 17, 1959, at a place called outside t ty minutes to midnig date, astrop ude 7.5, not vast as eart so abrupt and collapsed an entire mountainside. It  of tunately not so many people  to Yelloone in tymillion tons of rock, moving at more t fell off tain, traveling um t t up a mountain on ts pat of ty-eigeen of toodeep ever to be found again. tation  but breakingly fickle. tent, s, sleeping in anotent besidet away and never seen again.

    “A big eartime,” Doss told me. “You cancount on t. t zone for earthquakes.”

    Despite tone didn’t getpermanent seismometers until the 1970s.

    If you needed a o appreciate ture of geologic processes,you could do o consider tetons, tuously jagged range t stands justto tone National Park. Nine million years ago, tetons didn’t exist.

    t a  ty-mile-long faultopened  once every nine etonsexperience a really big earto jerk t  is ted jerks over eons t o t majestic s of sevent.

    t nine  misleading one. According toRobert B. Smito tory of t major teton quake  five and seven tetons, in s, are about t overdue eart.

    risk. time, pretty muc any predictability. “You knoors intotold me after o see. Did you kno springs at Yelloone t of the world combined?”

    “I didn’t kno.”

    en t migo a place called Duck Lake, a body of er a couple of  lookscompletely innocuous,” ’s just a big pond. But t used to be here.

    At some time in t fifteen tens of millions of tons of earted er blo at it  at Old Faitors’ centers.” he made an unhappy face.

    “ould there be any warning?”

    “Probably not. t significant explosion in t a place called Pork C left a crater about five meters across—not  bigenougo be standing t time. Fortunately, nobody  t   past t ell you  standing t does.”

    Big rockfalls are also a danger. t Gardiner Canyon in 1999, but againfortunately no one e in ternoon, Doss and I stopped at a place  any time,” Doss said tfully.

    “You’re kidding,” I said. t a moment , all filled  literal sense, happy campers.

    “O’s not likely,”  saying it could. Equally it could stay like t fordecades. t no telling. People o accept t t’sall to it.”

    As o o o Mammot Springs, Doss added: “Butt of time bad t  fall. Eartoccur. Nes don’t suddenly open up. For all tability, it’s mostly remarkably andamazingly tranquil.”

    “Like Eartself,” I remarked.

    “Precisely,” he agreed.

    t Yelloone apply to park employees as muco visitors. Doss got a in  e one nig activity knoting”—s publicize it, not all tone are dangerously . Some are extremely agreeable to lie in, and it  of some of to e at nig  to do so. Fooliso take a flas, o a scalding vent beloo tream t to leap over earlier. t of took a running jump. In fact, it tream at all. It  three survived.

    I t about t morning as I made a brief call, on my  of t aplace called Emerald Pool, in t ime to take me t I t I oug least to  it, for Emerald Pool is a oricsite.

    In 1965, a eam of biologists named tudy trip,  rimmed t for life. to tually t extremop could live in er t o bemucoo  or acid or co bear life. Emerald Pool, remarkably,  at least types of living ticus as t congenial. It  notemperatures of 50°C (122°F), but ers nearly t .

    For almost ty years, one of teria, ticus,remained a laboratory curiosity until a scientist in California named Kary B. Mullis realizedt -resistant enzymes  could be used to create a bit of cion, s to generate lots of DNA from verysmall amounts—as little as a single molecule in ideal conditions. It’s a kind of geneticpocopying, and it became t genetic science, from academicstudies to police forensic  ry in 1993.

    Means  ures of 80°C (176°F) or more. torganism found so far, according to Frances As in Life at tremes, is Pyrolobusfumarii, ure can reac for life is t to be about 120°C (248°F), tually kno all events, tely cion of tist Jay Bergstral it: “o  ile possible environments for life—as long as ter and some source of chemical energy we find life.”

    Life, it turns out, is infinitely more clever and adaptable than anyone had ever supposed.

    t to see,  doesn’t altogeto  us here.

    PARt V   LIFE ItSELFtudy tails of its arcecture,t t haveknown we were coming.

    -Freeman Dyson