velocities. But the sonic booms pro-duced by the 3,500-pound tails ofthese behemoths would probablyhave sounded more like cannon firethan the crack of a bullwhip.

Dr. Myhrvold and Dr. Currie de-scribed the research in interviewsand in a report in this month's issueof the journal Paleobiology. Theyconducted a variety of computersimulations, testing different as-sumptions about the biomechanicalcapabilities of these giant dinosaurs .They compared tails with whips intheir computer analysis to see howsimilarly they behaved.

"In all cases, it was easy to findsimulations that produced superson-

. . .

Did Dinosaurs Break the Sound Barrier?By JOHN NOBLE WILFORD

MOVE over, Chuck 'Yeager,and give way to supersonicdinosaurs.

At least 150 million years beforeMr. Yeager in 1947became the firsthuman to break the sound barrier ina rocket plane, the largest dinosaurs,a group known as sauropods, couldhave mustered the right stuff to sendsonic booms resounding over theMesozoic landscape. No, the 100-toncreatures never got off the ground.All they would have had to do wasflick their long tails like a bullwhip.

The idea had occurred to somepaleontologists examining fossils ofthe enormous sauropod tails, which

tapered to thin tips. Could they havebeen used like whips to defend them-selves or to produce a loud "crack"to intimidate predators or communi-cate with fellow sauropods, includingpotential mates? As physicists haveknown since 1958,the crack of a whipis actually the shock wave, or sonicboom, caused by the thin tip of a whipexceeding the speed of sound for onemoment..

No one had put the idea to a test. until a master of computer simula-tions, Dr. Nathan P. Myhrvold of the

, Microsoft Corporation, struck up anextended electronic-mail conversa-tion with a leading dinosaur expert,Dr. Philip J. Currie of the RoyalTyrrell Museum of Paleontology inDrumheller, Alberta. The result

•

A tail might havebeen as fast as awhip, but louder.

could be the beginning of a new re-search specialty that Dr. Myhrvoldcalls cyberpaleontology.

In analyzing fossils and developingcomputer models of sauropod tails,the two researchers said they hadfound evidence that dinosaurs likeApatosaurus (also known as Bronto-saurus) and Diplodocus could indeedhave flicked their tails to supersonic Continued on Page 7

A Computer Spawns a Noisy DinosaurScientists have long speculated that dinosaurs like this Apatosaurus used their tails to attack prey or defend themselves. Now researchersusing computer models have proposed that Apatosaurus used its tapered tail like a bullwhip, making noise to attract a mate or establish dominance.Energy waves would have picked up speed as they rippled along the tail from the wide base to the narrow tip. The very end of the tail couldhave moved faster than sound, creating a loud sonic "crack."

The New York Times; Illustration by Gregory S. Paul

Science Times TUESDA Y, DECEMBER 2, 1997

NE Cl

. .mbtNt\tJ !l!ork mimesDid Dinosaurs UseWhiplike Tails r-T-h--'e=--·~i-no-s-au-r-Sw-in-g--"-------'--------------,

For Breaking the Sound Barrier?COlltimH'd From First SCi('t1('C,' PaJ~(~

ic motion," the scientists wrote."The geometric scaling of vertebraldimensions found in the various. dl-plodoctds strongly suggests that anyof them. or mm-diplndocld sauropodswith ,',&,hiplash' tails, would sharethis capability."

With one side-to-side flick, the re-searchers determined, a wave of en-ergy could accelerate through thelength of one of the tapering, seg-mented tails, gaining momentum topropel the tip of the tail to velocitiesof more than 750 miles an hour, fast-er than the speed of sound.

"We must confess that it is pleas-ing to think that the first residents ofEarth to exceed the sound barrierwere not humans, but rather the di-plodocid sauropods," Dr. Myhrvoldand Dr. Currie concluded.

Other dinosuar experts are sharp-ly divided over. theresearch. Greg-ory S. Paul, .~nindependent specialistin dinosaur anatomy who Is based inBaltlmnre.. said- he thought the con-cept of sauropods with supersonictails was physically plausible. But ata recent paleontology conference, hesaid, he heard "other people who just

The possibleexplanations fortail cracking rangefrom love to war.

hate the idea.",One critic is Dr. Kenneth Carpen-

ter, a paleontologist at the DenverMuseum of Natural History. "To beblunt," he said in an interview, "thecomputer simulations are anothercase of garbage in, garbage out."

Dr. Carpenter questioned whetherthe bony segments of the dinosaurtails could have produced a super-son ice boom. Even if that was possi-ble, he said, using the taillike a whipmight have been both painful anddamaging to dinosaurs: The last fewsegments might even snap off.

In their report, Dr. Myhrvold and,Dr. Currie emphasized that only thelast two or three inches of the dino-saur tail would have exceeded thespeed of sound. The possibility ofpain or damage might be minimized

A lop view shows that the Apatosaurus lairs base. segmenlst !tHough 17. was slill and muscular and mi~lhl havelunclioned like a bullwhip handle.In soqrnents 18 throuqh 25. R transition mea. vertebraeare slighlly tonqcr than those at tho tail's base or and.Thouo vl':.!rluh"H~ nhow sj~Jns of mechanical stress Ih:11c()tilc.l Il:JVe I){)orl, C(ltl~';()(JI ly ;1 wl1irlpin9 action.The linal lail seqments were Ilexible, like awhiplash, in movements parallel 10 the ground.The lip could cover n moter-lonq arc.

Segment 25

Segment 18 <,

I

ing "could have been used as a nonle-thaI form of male-versus-male domi-nance contests."

What if the loud crack of a tail wasa male's way of calling or attractingfemales? A test of this idea, Dr.Myhrvcld said, would be to studymany more fossils of dinosaur tailsto see if the stress scarring wasconfined to males. That would meantail cracking had been a H malething." But paleontologists are notsure that they can distinguish a malesauropod fossil from a femaie fossil.

Even if the issue of dinosaur tailcracking is never resolved, Dr. Cur-rie said, the computer simulations-cyberpaleontology - promise to be"rile simple first steps for more com-plex models or dinosaur move-rncnts.'

Prances Roberts for The New York Times

Dr. Nathan P. Myhrvold of the Microsoft Corporation, who createscomputer models of dinosaurs, caUs his work cyberpaleontology,

or eliminated, they pointed out, if themost extreme part of the tail extend-ed past the last vertebra as a piece, ofskin. tendon or keratin, the proteinthat can take the form of scales,claws or feathers. "If whips madefrom the skins. of cows and kanga-roos are able to withstand supersonicmotion," they said, "'why not dino-saur skin and tendons?"

But the two researchers agreedwith the patecnmtogtsts who now reojeer the idea that the sauropods regu-larly used their tails defensively.The animals would probably havesustained as much Injury to theirtails as they inflicted on attackers.

As chief technology offlcer nt M i-croson, Dr. Myhrvold presumablyhas mere pressing research mattersthan dinosaur tails. But dinosaurshave fascinated him since childhood,and he has probably' never' met aresearch problem he did not try touse a computer to solve. He got in thehabit of stretching the imaginationwhen he studied cosmology underDr. Stephen W. Hawking at Cam-bridge University in England.

"I don't claim it's relevant to Mi-crosoft," he said of the dinosaur sim-ulations. "It's just an int.erestingproblem to me."

'In. particular. Dr. Myhrvold wagintrigued.by the analogy of the bull-whip to explain the sauropod tails, assuggested a few years ago by Dr.R. McNeill Alexander of the Univer-sity of Leeds in England. The pro-gressive rate of tapering from, the

base of the tail to the tip is compara-hie to that of a bullwhip from the gripto the tip. Each successive vertebrain a sauropod tail is about 6 percentsmaller than its predecessor.

An analysis of fossils, especiallythose of an apatosaur at the CarnegieMuseum of Natural History in Pitts-burgh, showed that in {l-fnot tails of80 vertebrae, the lengths of eachconnecting segment reached 11 maxi-mum of some 4 inches in the sccuonbetween the IHIIl and 2~lh vent-brae.

._-----------,-,._--,

That happens to correspond to allarea of ossification, fusing pairs ofvertebra and otherwise indicating in-juries from repeated stress." As Dr. Myhrvold and Dr. Currie

noted, the injury is consistent withoverextension of the tail joints fromwhiplike motions in a plane parallelto the ground. Fossils suggest thatthe sauropods could probably movetheir tails about 30 degrees side toside; for purposes of the computersimulations, the reseuchers limitedthe molion to 9 degrees side to side.The animals were presumably ex-tremely limited in vertical move-ments of their tails.

Next, Dr. Myhrvold recalled trying10 learn everything posstble aboutwhips. As a computer expert, he nat-urally COIl"tIlICd the Internet, but

that directed him mainly to discus-sions of sado-masochistic practices,Finally, he learned that the man whomade whips for Hollywood movieslived nearby in Seattle. Purchasingone, after pledging to use it only fordinosaur studies, he began to under-stand the Newtonian physics of whipdynamics, and he applied -this inadapting commercial software lorconducting the computer simulationsGf dinosaurian whiplike tails.

The two researchers offer severalexplanations for why sauropodswould have engaged in tail cracking.It might have been a way to enforcediscipline within a group, as herdersof cattle and horses sometimes dowith whips. It might have been a wayto resolve disputes without resortingto comhut. Or, 1I11'Ywrote, tail crack-