The Paradox CORONA Sun’sof the Hotthe-eye.eu/public/concen.org/Scientific American...through a gap in his front teeth created in his teens when, in keeping with Himba tradition,

PLUS:A Low-PollutionEngineNorth to Mars!ControllingHair Growth

CORONA

J U N E 20 01 $ 4. 95W W W. S CI A M. COM

SIGN L ANGUAGE AND THE BRAIN ■ HOW DO FLIES FLY? (SEE P. 48)

The Paradox

of theSun’s Hot

Copyright 2001 Scientific American, Inc.Copyright 2001 Scientific American, Inc.

A S T R O P H Y S I C S

40 The Paradox of the Sun’s Hot CoronaB Y B H O L A N . D W I V E D I A N D K E N N E T H J . H . P H I L L I P SThe sun’s surface is comparatively cool, yet itsouter layers are broiling hot. Astronomers arebeginning to understand how that’s possible.

B I O M E C H A N I C S

48 Solving the Mystery of Insect FlightB Y M I C H A E L D I C K I N S O NInsects stay aloft thanks to aerodynamic effects unknown to the Wright brothers.

N E U R O S C I E N C E

58 Sign Language in the BrainB Y G R E G O R Y H I C K O K , U R S U L A B E L L U G I A N D E D W A R D S . K L I M AStudies of deaf signers illuminate how all human brains process language.

S P A C E T R A V E L

66 North to Mars!B Y R O B E R T Z U B R I NAs a first step toward building a base on Mars,scientists set up camp in the Canadian Arctic.

B I O S C I E N C E

70 Hair: Why It Grows, Why It StopsB Y R I C K I L . R U S T I N GMolecules that control hair growth may be the key to combating baldness.

A N T H R O P O L O G Y

80 The Himba and the DamB Y C A R O L E Z Z E L LA questionable act of progress may drown an African tribe’s traditional way of life.

contentsjune 2001

SCIENTIFIC AMERICAN Volume 284 Number 6 features

T E C H N O L O G Y

90 A Low-Pollution Engine SolutionB Y S T E V E N A S H L E YNew sparkless-ignition automotive engines gearup to meet the challenge of cleaner combustion.

w w w . s c i a m . c o m S C I E N T I F I C A M E R I C A N 5

48 Robotic insect

Copyright 2001 Scientific American, Inc.

columns37 Skeptic B Y M I C H A E L S H E R M E R

Lunatic conspiracies that Apollo was a fraud.

105 Puzzling Adventures B Y D E N N I S E . S H A S H ALiar, liar, liar.

106 Anti Gravity B Y S T E V E M I R S K YMeet NASA’s nose.

107 Endpoints

34 InnovationsCreating mice that make human antibodies was only half the battle for biotech firm GenPharm.

36 Staking ClaimsA court decision on patent law may give a free ride to copycats.

38 Profile: Marcia K. McNuttRoughly 95 percent of the ocean is unexplored. This geophysicist plans to change that.

96 Working KnowledgeThe evolution of golf balls.

98 ReviewsTwo books pose the question, Just how much did the discovery of dinosaur fossils change science?

102 TechnicalitiesHow much is that robo-doggy in the window?

6 S C I E N T I F I C A M E R I C A N J U N E 2 0 0 1

departments8 SA Perspectives

The Bush administration’s uncertainty about uncertainty.

12 How to Contact Us12 On the Web16 Letters18 50, 100 & 150 Years Ago20 News Scan

■ Arsenic and drinking water.■ The Milky Way’s cannibalistic past.■ Unmanned combat air vehicles leave the launchpad.■ A new superconducting compound.■ Earth: Move it or lose it.■ The high cost of monkey on the menu.■ By the Numbers: Hateful terrorism.■ Data Points: Protein projects proliferate.

24

38 Marcia K. McNutt, head of MBARI

24D

Cover ultraviolet image by NASA Goddard Space Flight Center, showing ionized helium at 60,000 kelvins, taken by SOHO on September 14, 1997; preceding page: Timothy Archibald; this page (clockwise from top left):The Boeing Company; Mark A. Garlick; Edward Caldwell

Volume 284 Number 6



SA Perspectives

Scientists are often lampooned as living in an ivorytower, but lately it seems that it is the scientists whoare grounded in reality and the U.S. political estab-lishment that is floating among the clouds. In Marchthe Bush administration gave up a campaign promiseto control emissions of carbon dioxide and withdrewU.S. support for the Kyoto Protocol. “We must bevery careful not to take actions that could harm con-

sumers,” President George W.Bush wrote in a letter to four Re-publican senators. “This is espe-cially true given the incompletestate of scientific knowledge ofthe causes of, and solutions to,global climate change.”

Yet incomplete knowledgedoesn’t seem to be a concernwhen it comes to strategic mis-sile defense. After another failedtest last summer, candidate Bushissued a statement: “While lastnight’s test is a disappointment,I remain confident that, given theright leadership, America candevelop an effective missile de-fense system....The United Statesmust press forward to develop

and deploy a missile defense system.” And press for-ward he has. The U.S. is reportedly on the verge ofwithdrawing unilaterally from the 1972 Anti-BallisticMissile Treaty.

In one case, the president invokes uncertainty; inthe other, he ignores it. In both, he has come downagainst the scientific consensus.

Presidents, needless to say, must protect the coun-try’s economic interests and shield the nation from nu-clear death. That is precisely why the administration’s

inconsistency is so worrisome. Ample research indi-cates that human activity is the main cause of globalwarming. Estimates of the economic damage by mid-century range in the hundreds of billions of dollars peryear—uncertain, to be sure, but if you’ve been smok-ing in bed, it makes sense to take out some fire insur-ance. Kyoto is far from perfect; its emissions targetsrepresent a diplomatic agreement rather than anycareful weighing of cost and benefit. But it is a start.

Regarding strategic missile defense, researchers’best guess is that a reliable system is infeasible. Theburden of proof is now on the proponents of missiledefense. Until they can provide solid evidence that asystem would work against plausible countermea-sures, any discussion of committing to building one—

let alone meeting a detailed timeline—is premature.It is one thing for a software company to hype a prod-uct and then fail to deliver; it is another when the fail-ure concerns nuclear weapons, for which “vapor-ware” takes on a whole new, literal meaning.

Perhaps the most exasperating thing about missiledefense is how the Bush administration has so quick-ly changed the terms of the debate. Journalists andworld leaders hardly ever comment anymore on thefundamental unworkability of the system or the manyways it would fail to enhance security. Now the talkis of sharing the technology so that other countries,too, could “protect” themselves.

It would be nice not to have to shell out money foremissions controls. It would be nice to have a magicshield against all nuclear threats. It would be nice tobe perfectly sure about everything, to get 365 vacationdays a year and to spend some of that time on Mars.But we can’t confuse wants with facts. As RichardFeynman said, “Science is a way of trying not to foolyourself.” The dangers of ignoring its messages aregreater than merely making politicians look foolish.

RO

GE

R R

ESS

ME

YER

Cor

bisTHE EDITORS [email protected]

Faith-Based Reasoning

PEACEKEEPER ICBM test


(WHAT YOU DIDN’T THINK YOU WANTED TO KNOW ABOUT) RECYCLED WASTEWATERThe use of effluent recycled as drinkingwater is hardly unique to such water-shortareas as Namibia [“How We Can Do It:Waste Not, Want Not,” by Diane Mar-tindale]. Every major river in the worldcarries someone’s treated effluent down-stream to another community’s source ofdrinking water. In California several indi-rect potable-water-recycling projects—

those that would put recycled water intounderground aquifers or surface waterreservoirs—have been derailed because oflocal politics and the “yuck” factor whena project is labeled as “Toilet to Tap.”

This public concern persists despitethe fact that two of the state’s majorsources of drinking water now containrecycled wastewater: the Colorado Riverreceives the treated effluent from Las Ve-gas, and the Sacramento/San JoaquinDelta is downstream of the discharge ofdozens of Central Valley communities.Several Southern California projects re-cycle more than 170,000 acre-feet ofhighly treated effluent every year into un-derground water supplies used by threemillion to four million people. Some ofthese projects have operated safely and re-liably for nearly 40 years.

ROBIN G. SAUNDERSVice President, Northern California Chapter

WateReuse AssociationSanta Clara, Calif.

UNPERSUADEDIs Robert B. Cialdini [“The Science of Per-suasion”] really trying to tell us that 17percent of our population is willing tochaperone juvenile delinquents on a daytrip to the zoo? Where I live the schoolshave a hard time getting chaperones totake a group of first-graders to the mu-seum. If your numbers are correct, thenthere would be no need for social pro-grams to help the needy, bring meals tothe terminally ill or read to the aged innursing homes. All we’d have to do is puta few people on the street asking passers-by if they would be willing to spend a fewnights alone in a cell with an inmate ondeath row. Once they rejected that, thenwe’d have 50 percent of the populationvolunteering for whatever we coulddream up for social reform.

JOHN LOMAXNovato, Calif.

CIALDINI REPLIES: We solicited volunteers withan in-person, one-on-one request: “We’re re-cruiting volunteers to chaperone a group of kidsfrom the County Juvenile Detention Center on a

PE

TER

JO

HN

SON

Cor

bis

“I HAVE X ENVY after reading ‘Why the Y Is So Weird’ [by Karin Jegalian and Bruce T. Lahn],” protestsLane Yoder of Kaneohe, Hawaii. “I learned that the human Y ‘fell into such disrepair’ that it is now in a‘severely shrunken state,’ a ‘shadow of its original self.’ The X maintained its ‘integrity’ by recombin-

ing with other Xs. A scientist across the Freudian divide might de-scribe the findings differently: Abstaining from the entanglementsof DNA swapping, the human male chromosome brought forth a fewhighly evolved genes. The estranged X’s indiscriminate couplingwith other Xs condemned it to the primitive, bloated state of its reptilian ancestors. We can’t expect complete objectivity. Still, it’s quite a stretch to say that nature consistently selected a ‘fail-ure’ that spread to thousands of new species over hundreds of mil-lions of years and now exists in its most extreme form in the mostdominant species.”

Abstaining from this particular entanglement ourselves, weinvite you to check out others, in letters about the February issue.

EDITOR IN CHIEF: John RennieMANAGING EDITOR: Michelle PressASSISTANT MANAGING EDITOR: Ricki L. RustingNEWS EDITOR: Philip M. YamSPECIAL PROJECTS EDITOR: Gary StixSENIOR WRITER: W. Wayt GibbsEDITORS: Mark Alpert, Steven Ashley, Graham P. Collins, Carol Ezzell, Steve Mirsky, George Musser, Sarah SimpsonCONTRIBUTING EDITORS: Mark Fischetti, Marguerite Holloway, Madhusree Mukerjee, Paul Wallich

EDITORIAL DIRECTOR, ONLINE: Kristin LeutwylerASSOCIATE EDITORS, ONLINE: Kate Wong, Harald FranzenWEB DESIGN MANAGER: Ryan Reid

ART DIRECTOR: Edward BellSENIOR ASSOCIATE ART DIRECTOR: Jana BrenningASSISTANT ART DIRECTORS:Johnny Johnson, Mark ClemensPHOTOGRAPHY EDITOR: Bridget GeretyPRODUCTION EDITOR: Richard Hunt

COPY DIRECTOR: Maria-Christina KellerCOPY CHIEF: Molly K. FrancesCOPY AND RESEARCH: Daniel C. Schlenoff, Rina Bander, Sherri A. Liberman, Shea Dean

EDITORIAL ADMINISTRATOR: Jacob LaskySENIOR SECRETARY: Maya Harty

ASSOCIATE PUBLISHER, PRODUCTION: William ShermanMANUFACTURING MANAGER: Janet CermakADVERTISING PRODUCTION MANAGER: Carl CherebinPREPRESS AND QUALITY MANAGER: Silvia Di PlacidoPRINT PRODUCTION MANAGER: Georgina FrancoPRODUCTION MANAGER: Christina HippeliASSISTANT PROJECT MANAGER: Norma JonesCUSTOM PUBLISHING MANAGER: Madelyn Keyes

ASSOCIATE PUBLISHER/VICE PRESIDENT, CIRCULATION:Lorraine Leib TerleckiCIRCULATION MANAGER: Katherine RoboldCIRCULATION PROMOTION MANAGER: Joanne GuralnickFULFILLMENT AND DISTRIBUTION MANAGER: Rosa Davis

PUBLISHER: Denise AndermanASSOCIATE PUBLISHER: Gail DelottSALES DEVELOPMENT MANAGER: David TirpackSALES REPRESENTATIVES: Stephen Dudley, Wanda R.Knox, Hunter Millington, Stan Schmidt, Debra Silver

ASSOCIATE PUBLISHER, STRATEGIC PLANNING:Laura SalantPROMOTION MANAGER: Diane SchubeRESEARCH MANAGER: Aida DadurianPROMOTION DESIGN MANAGER: Nancy Mongelli

GENERAL MANAGER: Michael FlorekBUSINESS MANAGER: Marie MaherMANAGER, ADVERTISING ACCOUNTING AND COORDINATION:Constance Holmes

MANAGING DIRECTOR, SCIENTIFICAMERICAN.COM:Mina C. LuxDIRECTOR, ELECTRONIC PUBLISHING: Martin O. K. Paul

DIRECTOR, ANCILLARY PRODUCTS: Diane McGarveyPERMISSIONS MANAGER: Linda HertzMANAGER OF CUSTOM PUBLISHING: Jeremy A. Abbate

CHAIRMAN EMERITUS: John J. HanleyCHAIRMAN: Rolf GrisebachPRESIDENT AND CHIEF EXECUTIVE OFFICER:Gretchen G. TeichgraeberVICE PRESIDENT AND MANAGING DIRECTOR, INTERNATIONAL:Charles McCullaghVICE PRESIDENT: Frances Newburg

Established 1845

®

LettersE D I T O R S @ S C I A M . C O M

16 S C I E N T I F I C A M E R I C A N

WASTEWATER yields drinking water in Windhoek, Namibia.


trip to the zoo. It would be voluntary, unpaid andwould require about two hours of one afternoonor evening. Would you be interested in beingconsidered for one of these positions?”

Increasingly, charity and community-basedrequests occur in an impersonal fashion. Thereis clear evidence that face-to-face, one-on-onerequests are most successful; second best arephone requests, and third best are written re-quests. Why have requesters chosen the lesseffective media? It is simply easier to use im-personal routes; moreover, it is possible toreach many, many more people that way. There-fore, even though the percentage of compliersdrops significantly, the overall number of com-pliers can actually be higher. The combinationof ease of implementation and reach has tri-umphed over impact of contact.

PYTHAGORAS, PLATO AND EVERYTHINGIf a “theory of everything” [“100 Years ofQuantum Mysteries,” by Max Tegmarkand John Archibald Wheeler] were to betotally mathematical, with “no conceptsat all,” perhaps the best interpretation ofthis would be Pythagorean. That is, todate we have assumed that the mathemat-ics describes some reality that is going on;this has led to all sorts of mental gym-nastics about what electrons and the likeare “really” doing between observations—

gymnastics that have gotten us into allkinds of trouble, not to mention many-worlds, consistent histories, “rampantlinguistic confusion” and even Zen.

All this results from assigning a de-scriptive role to mathematics. Perhaps,following Pythagoras, we should assigna prescriptive role to the math: assumethe equations are real and that matter isformless and comports itself in accor-dance with them. That is, the equationsdo not describe what matter does; rather,they tell it what to do.

ALBERT S. KIRSCHBrookline, Mass.

TEGMARK REPLIES: With such a viewpoint, whichmight also be termed Platonic, the mathemat-ical structure encapsulated by the equationswouldn’t merely describe the physical world. In-stead this mathematical structure would be

one and the same thing as the physical world,and the challenge of physics would be to predicthow this structure is perceived by self-awaresubstructures such as ourselves.

IN FORESTS, THE OLDER THE BETTERSeveral points made in “Debit or Credit?”by Sarah Simpson [News and Analysis]lose sight of the fact that forests can con-tribute legitimately to lasting reductionsin atmospheric carbon dioxide. We cancontinue to manage forests in the usualfashion globally (where deforestation isthe second-largest source of CO2) andnationally (where the forest sink has beendeclining for the past decade), or we cantake the positive steps envisioned in theKyoto Protocol, which calls for the main-tenance and enhancement of existing

forests. Significant and long-lasting gainscan be made by reducing carbon emis-sions from deforestation and by enhanc-ing carbon stocks through maintainingolder forests. In the next 50 years thesegains will be far larger than those fromnewly planted forests.

SANDRA BROWNWinrock International

Corvallis, Ore.LAURIE A. WAYBURN

President, Pacific Forest TrustSanta Rosa, Calif.

ERRATUM Re “How We Can Do It: Leaking Away,”by Diane Martindale]: Volt VIEWtech ended itsinvolvement with New York City’s ResidentialWater Survey Program in 1995. The program iscurrently overseen by Honeywell DMC Services.

w w w . s c i a m . c o m S C I E N T I F I C A M E R I C A N 17Copyright 2001 Scientific American, Inc.

JUNE 1951THE IDIOT BOX—“A survey of the programson TV was recently carried out in prepa-ration for public hearings to be held be-fore the Federal Communications Com-mission. Science and informational pro-grams amounted to about three percentof one week’s broadcasts. But most of theentertainment programs did not riseabove the rut of two-dimensional formu-la productions. A depressingly large pro-portion of the ‘entertainment’ offered onTV was uninspiring, monotonous and ul-timately derogatory of human dignity.”

NOTEWORTHY CHEMISTRY—“The HarvardUniversity chemist Robert B. Woodwardlast month announced an achievementthat was at once recognized as a mile-stone in the history of chemistry—the to-tal synthesis of a steroid. Woodward’ssteroid is strictly a synthetic product, notidentical with any natural substance. Butbecause one steroid has been convertedinto another in a number of cases, theachievement opens the way to the com-plete synthesis of natural steroids such ascortisone, testosterone and progesterone.From the synthetic steroid it maybe possible to produce cortisonein a few simple chemical steps.Cortisone is now made in 37steps from a component of bilethat is so scarce that it takes 40cattle to supply one day’s corti-sone for an arthritic patient.”

JUNE 1901KRUPP ARMAMENTS—“The Kruppmetallurgic establishments, in theRuhr River basin, now form thegreatest works in the world. Onthe first of April, 1900, the num-ber of men employed by FriedrichKrupp was 46,679. At the end of1899 the steel works of Essen hadmanufactured and sold 38,478guns. The Krupp works do notlimit their activity to the manu-

facture of guns, ammunition and acces-sories, but produce also what a pamphlet,published at Essen, calls ‘peace material’—that is to say, car wheels, rails, steel cast-ings for steamships, etc.”

YELLOW FEVER—“The Surgeon-General ofthe United States Army has approved thereport of a special medical board, whichhas reached the conclusion that the mos-quito is responsible for the transmissionof yellow fever. The medical departmentis moving energetically to put into prac-tical operation the methods of treatmentfor prevention of yellow fever. The liber-al use of coal oil to prevent the hatchingof mosquito eggs is recommended.” [Ed-itors’ note: The report was submitted byU.S. Army bacteriologist Walter Reed.]

A ONE-HORSEPOWER “IT”?—“Our illustra-tion shows a most curious invention byMitchell R. Heatherly of Mundell, Kan-sas: a single-wheel vehicle. The contriv-ance consists of a curved tongue pivotedto the harness, bearing a single wheel.Above the axle of the wheel are stirrupsfor the rider or driver.”

JUNE 1851 IRON AGE OF SHIPS—“For Lord Jocelyn’ssteam navigation committee in England,Captain Claxton gave evidence in favor ofiron steamers and of the screw, which, heavers, must ere many years elapse be ap-plied universally as the motive power ofsea-going vessels. The advantages whichhe ascribes iron-built vessels being dura-bility, inexpensiveness in repairs, greatercapacity in proportion to tonnage thanwooden vessels, healthiness, and swift sail-ing. As for durability, he described thestate of the Great Britain, lying for manymonths exposed to a series of heavy galesin Dundrum Bay, Ireland.”

TOPICAL ANESTHETIC—“The difficulty in theuse of chloroform thus far has been thedanger of suffocation, or of otherwise in-juring the body by a total stoppage ofsome of its function. A new applicationclaims the merit of escaping the danger,according to the scientific critics in Berlin.The fluid (some 10 to 20 drops) is droppedon the part affected or on a lint bandage,and then bound up in oil silk. After fromtwo to ten minutes the part becomes in-

sensible, and the pain is no longerfelt, whether it be from rheumat-ic, nervous, or other disorders.”

FALSE LIGHTS—“Three years agothere was nothing heard of in England but ‘Staite’s ElectricLight.’ It was patented, published,and puffed from one end of theworld to the other. It was to sendall the gas companies into Egyp-tian darkness in short order, andso potent was the sympathetic in-fluence of the excitement (for theshrewdest and wisest are subjectto such influences) that the stocksof gas companies were at a verylow discount. Well, a few weeksago this Electric Light became in-solvent, and it was executed by anumber of indignant creditors.”


FRO

M S

CIE

NTI

FIC

AM

ERIC

AN

50, 100 & 150 Years AgoHormones ■ Howitzers ■ Horsepower

BEFORE SCOOTERS: A single-wheel idea, 1901



SCANnews

Arsenic has long been used as a poison,most famously by the pair of elderlyaunts in the play Arsenic and Old

Lace. The murderous spinsters added a tea-spoonful to a gallon of wine, but it takes a lotless than that to prove fatal. Scientists havediscovered that arsenic may be hazardouseven in the minute quantities found in manywells and municipal water systems in the U.S.In January, just before President George W.Bush took office, the Environmental Protec-tion Agency finalized a long-awaited regula-tion reducing the amount of arsenic allowedin drinking water from 50 micrograms perliter—the U.S. standard since 1942—to 10micrograms per liter, which is the standardused by the European Union and the WorldHealth Organization. But in March the EPA—

under the new leadership of Bush’s appoin-tee, Christie Whitman—withdrew the pend-ing rule. And in April the agency asked theNational Academy of Sciences (NAS) to re-assess the research on arsenic, delaying a finaldecision until February 2002.

The scientists who have studied arsenic’shealth effects immediately assailed Whitman’sdecision. A growing number of epidemiolog-ical studies indicate that drinking arsenic-tainted water can cause skin, lung, liver andbladder cancers. A 1999 report by the NAS

estimated that daily ingestion of water con-taining 50 micrograms of arsenic per liter

would add about 1 percent to a person’s life-time risk of dying from cancer. That’s aboutthe same as the additional risk faced by a per-son who’s living with a cigarette smoker. “Theevidence against arsenic is very strong,” saysepidemiologist Allan H. Smith of the Univer-sity of California at Berkeley. “But the EPA hascreated a false appearance of uncertainty.”

Perhaps the best evidence comes from along-term study of 40,000 villagers in south-western Taiwan whose wells had high arseniclevels. (Because arsenic seeps into aquifersthrough the weathering of rocks and soils, it’s

PU

BLI

C H

EAL

TH

A Touch of PoisonTHE EPA MAY WEAKEN A REGULATION LIMITING ARSENIC IN WATER BY MARK ALPERT

Arsenic is in food as well as water,but researchers say a typical daily

diet contains only 10 to 15micrograms of inorganic arsenic—

the compounds that are hazardous(food contains much more organic

arsenic, but that form passesharmlessly through the body).

Although toxicologists aren’t surehow arsenic attacks the body’s

cells, a new study by scientists atDartmouth Medical School indicates

that the substance disrupts theactivity of hormones called

glucocorticoids, which help toregulate blood sugar and suppress

tumors. Arsenic interferes withthese processes by binding to the

glucocorticoid receptors in cells and changing their structure. The

study suggests that arsenic,instead of causing cancer by itself,promotes the growth of tumorstriggered by other carcinogens.

Arsenic-induced effects appeared at concentrations as low as

two micrograms per liter.

THE MYSTERIOUSCARCINOGEN

ARSENIC-TAINTED water can cause various cancers.

KE

VIN

HO

RAN

Sto

ne



newsSCAN

generally more concentrated in groundwaterthan in lakes or streams.) In villages with themost severely contaminated wells, the deathrates from bladder cancer were dozens oftimes above normal. Similar studies in Ar-gentina and Chile later corroborated thosefindings. In a region of northern Chile, for ex-ample, researchers determined that 7 percentof all deaths among people over the age of 30could be attributed to arsenic.

In the Taiwan study, the lowest medianlevel of arsenic was 170 micrograms per liter.To determine the risk at the 50- and 10-mi-crogram levels, epidemiologists extrapolatedthe health effects in a linear way (that is, halfthe exposure leads to half the cancer risk).Some toxicologists have criticized this ap-proach, saying that arsenic concentrationsmay have to exceed a threshold level to causecancer. But new research suggests that if thisthreshold exists, it is most likely well below10 micrograms per liter.

In the U.S., most public water systemswith high arsenic concentrations are in thewestern states [see table at right]. The EPA

originally proposed lowering the arsenic stan-dard to five micrograms per liter, but theagency doubled the allowable level after rep-resentatives of the water systems complainedabout the expense of removing the carcino-gen. In the regulation issued in January, theagency estimated that 4,100 systems servingsome 13 million people would have to pay atotal of $180 million annually to implementthe 10-microgram standard. The EPA claimedthat the rule would prevent 21 to 30 deathsfrom lung and bladder cancer each year, but

some epidemiologists say the standard couldsave 10 times as many lives.

So what prompted the EPA to suddenlycall for a reassessment of the standard? Someenvironmentalists speculate that industrygroups such as the National Mining Associ-ation, which filed a court petition in Marchto overturn the arsenic rule, put pressure onthe Bush administration. The tailings frommines are often laced with arsenic. Becausethe EPA’s cleanup regulations are based ondrinking-water standards, tightening the re-strictions on arsenic could vastly increase thecost of decontaminating abandoned mines,many of which are Superfund sites.

Whitman has asked the NAS to review theEPA’s risk analysis of arsenic. Many research-ers fear that she will use the new report to jus-tify a limit of 20 micrograms per liter, a stan-dard that would cost about $110 million lessthan the stricter regulation but save only halfas many lives. “The weaker standard wouldnot be sufficient to protect public health,”says Chuck Fox, who headed the EPA’s Officeof Water until the change of administrations.“The standard for arsenic should be as closeto zero as feasible.”

F or years, archaeologists have beenspeaking the language of astronomers.Remote-sensing techniques have found

lost cities; celestial alignments have shed lighton temples and pyramids. But lately the flowof ideas has reversed. Astronomers have re-alized that our galaxy is an intricately layeredplace—a Tel Galaxia that encodes a rich his-

tory like buried strata of an ancient city. Ce-lestial excavations are starting to provide amuch needed reality check on theories notjust of the galaxy but also of the broader cos-mos. “It’s not all that easy to find experi-mental verification of these theories,” saysHeather L. Morrison of Case Western Re-serve University. “Studies of the Milky Way

Galactic ArchaeologyDIGGING INTO THE MILKY WAY’S PAST EXPOSES ITS LIFE AS A CANNIBAL BY GEORGE MUSSER

ASTR

ON

OM

Y

Large municipal water systems withaverage arsenic levels above theproposed 10-microgram standard:

CITY ARSENIC LEVEL (micrograms per liter)

Norman, Okla. 36.3

Chino Hills, Calif. 30.2

Lakewood, Calif. 15.1

Lancaster, Calif. 14.5

Albuquerque, N.M. 14.2

Moore, Okla. 12.6

Rio Rancho, N.M. 12.4

Victoria, Tex. 11.6

Midland, Tex. 11.1

Scottsdale, Ariz. 11.1

SOURCE: Natural Resources Defense Council

NEED TO KNOW:DANGER ZONES

ABANDONED MINE in Butte, Mont., is laced with arsenic.

WAL

TER

HIN

ICK

Mon

tan

a S

tan

da

rd/A

P P

hot

o



newsSCAN

NIG

EL

C.

HAM

BLY

Un

iver

sity

of

Edin

bu

rgh

;SI

MO

N T

. H

OD

GK

IN U

niv

ersi

ty o

f C

am

bri

dg

e

can give us some pretty solid constraints.”She and other galactic archaeologists tell

layers apart by playing a game of which-stars-are-not-like-the-others. The sun and mostother stars swirl around the galactic centerwithin a thin circular disk. Nearly a centuryago, however, astronomers noticed that somestars orbit not within a disk but within asphere—a “halo” that envelops the disk. Halostars are older than disk stars, and their irreg-ular orbits suggest they formed before mater-ial orbiting every which way had a chance tolose energy, flatten out and fall into lockstep.

The disk, the halo—astronomers thoughtthat was all. Over the years, though, and es-pecially in the past decade, they have foundstrange patterns in the halo: anomalouslyyoung stars, stars separated by vast distancesyet flying in formation, even entire galaxiesembedded within. As for the disk, astron-omers have given up talking about “the” disk.There is a thin disk, at least one thick disk andmaybe a so-called protodisk—stacked like lay-ers of an Oreo cookie. This mess of a galaxymust have taken shape over time rather thanin one fell swoop, as once thought.

This past January one of the ongoing digs,the 2dF Old Stellar Populations Survey,delved into the origins of the thick disk. Rose-mary F. G. Wyse of Johns Hopkins Univer-sity and her colleagues mapped 1,500 sunlikestars located outside the thin disk. Two thirdslooked like the usual halo or thick-disk stars.The rest, however, had half the expectedamount of orbital angular momentum—infact, a value characteristic of the Milky Way’ssmall satellite galaxies.

The results argue for an 1980s-era theorythat the thick disk arose when the Milky Waydevoured one of its satellites. In the process,whatever stars were around at the time gotstirred up, puffing the thin disk into the thickdisk. Interstellar gas presumably got stirredup, too, but gas (unlike stars) can easily dis-sipate energy and settle back into a thin disk.Subsequent generations of stars, forged fromthis gas, constitute the thin disk we see today.A corollary is that no sizable mergers haveoccurred since the thick disk took shape anestimated 12 billion years ago, although more

modest mergers continue to the present day.Other surveys have focused on the solar

neighborhood. Halo and thick-disk stars oc-casionally pass through, moving at conspicu-ously high velocities relative to the sun. Awhole new breed of interloper has recentlyemerged: very cool, very dim, very old whitedwarf stars. In March, Ben R. Oppenheimerof the University of California at Berkeley andhis colleagues reported 38 white dwarfs with-in 480 light-years of the sun. “This populationmay trace the oldest building blocks of the gal-axy,” says Rodrigo A. Ibata of Strasbourg Ob-servatory, who has conducted similar surveys.

Unfortunately, astroarchaeology has atragic flaw: it does not pin down the fullthree-dimensional distribution of objects. Anintense debate has erupted over whether theskulking dwarfs are part of the halo, thindisk, thick disk or putative protodisk. Simi-larly, astronomers dispute whether shardsfrom galactic mergers account for the wholehalo or just a small part of it. Depending onhow these issues shake out, the newly dis-covered populations could explain the resultsof dark-matter surveys over the past decade—

which hinted at undetected bodies but couldnot identify them—and thereby complete theinventory of ordinary matter in the galaxy.

That still leaves the extraordinary matter,the cold dark matter, which seems to makeup its own, far vaster halo. Galaxies ruled byit should grow the same way that planets do:from the agglomeration of smaller units. Thelayering of the Milky Way bears that out. Onthe other hand, cold-dark-matter theorieshave trouble explaining the inferred numberof satellite mergers, the shape of stellarstreams and the rate of disk formation. What-ever the fate of this or that theory, astrono-mers’ perspective on our home galaxy has fun-damentally changed. They have come to seeit not as a thing, sculpted long ago and left forus to admire, so much as a place, an arenawhere empires of stars rise and fall over thecourse of cosmic time.

Cosmological models suggest theMilky Way originally had dozens of

small satellite galaxies. Now there are 11. The closest is the

Sagittarius dwarf galaxy,discovered seven years ago on the

opposite side of the Milky Way fromthe sun. Despite its distance from

us, it spans a quarter or more of theway across our sky—a sure sign

of its being stretched, shreddedand assimilated by our galaxy.

The unexpected extent of theSagittarius dwarf galaxy emerged

recently from several surveys: the“Spaghetti” Survey (so named

because the stellar streams pulledoff incoming galaxies look like

spaghetti), the APM carbon starsurvey and the Sloan Digital Sky

Survey. Such studies haveunearthed shards of at least

five hapless galaxies.

PREY OFTHE MILKY WAY

WHITE DWARF, seen drifting across the sky over a period of 43 years, may represent a hithertounrecognized population of stars.


24B S C I E N T I F I C A M E R I C A N J U N E 2 0 0 1

T. H

. JO

HAN

SEN

ET

AL.

Un

iver

sity

of

Osl

o

newsSCAN

You can buy magnesium boride ready-made from chemical suppliers as ablack powder. The compound has been

known since the 1950s and has typically beenused as a reagent in chemical reactions. Butuntil this year no one knew that at 39 degreesabove absolute zero it conducts electric currentperfectly—it is a superconductor. Although itssuperconducting temperature is far below thatof the copper oxide high-temperature super-conductors, the compound has set off a flur-ry of excited activity among researchers. Mag-nesium boride overturned theorists’ expecta-tions and promises technological applications.

Jun Akimitsu of Aoyama-Gakuin Univer-sity in Tokyo announced the surprising dis-covery at a conference in Japan on January10, after he and his co-workers stumbled onmagnesium boride’s properties while trying tomake more complicated materials involvingmagnesium and boron.Word of the discovery spedaround the world by e-mail, and in three weeksthe first research papers byother groups were postedon the Internet. In earlyMarch a special session onmagnesium boride washastily put together in Seat-tle at the American Physi-cal Society’s largest annualconference: from 8 P.M.until long after midnight,nearly 80 researchers pre-sented ultrabrief summa-ries of their results.

Until January standard wisdom ruled outthe possibility of a conventional supercon-ductor operating above about 30 kelvins.Conventional superconductors are under-stood by the so-called BCS theory, formulat-ed in 1957. The magnesium boride resultseemed to imply that either a new supercon-ducting mechanism had been discovered orthat the BCS theory needed to be revised.

Almost all the experimental evidence so farsupports the idea that magnesium boride is astandard BCS superconductor, unlike the cop-per oxides. For example, when researchers usethe isotope boron 10 in place of boron 11, the

material’s critical temperature rises slightly, asexpected, because the lighter isotope alters vi-brations of the material’s lattice of atoms, akey component of BCS theory. How, then, hasthe magic 30 kelvins been exceeded? “Thosepredictions were premature,” says RobertCava of Princeton University, with 20/20 hind-sight. Magnesium boride has a combinationof low-mass atoms and favorable electronstates that was overlooked as a possibility.

Physicists are trying to push the BCS limiteven further to produce higher critical tem-peratures by doping the material with care-fully selected impurities. Groups have addedaluminum or carbon (neighbors of boron inthe periodic table), but these both decrease thecritical temperature. Calcium is expected towork better, but no one has succeeded in pro-ducing calcium-doped magnesium boride.“It’s like Murphy’s Law,” Cava gripes.

Even undoped, magne-sium boride has several at-tractive features for appli-cations. First, the high-er operating temperaturewould allow cooling of thesuperconductor by refrig-eration instead of by ex-pensive liquid helium, as isneeded for the most wide-ly used superconductors.The high-temperature cop-per oxide superconductorsbeat magnesium boridehands-down on that count,but they have proved dif-ficult to manufacture into

convenient wires. Also, the supercurrent doesnot flow well across the boundaries of mi-croscopic grains in copper oxides.

Magnesium boride, in contrast, has al-ready been fashioned into wires using simpletechniques, and the supercurrent flows effort-lessly between grains. One drawback, how-ever, is that magnesium boride loses its su-perconductivity in relatively weak magneticfields, fields that are inescapable in applica-tions. But with the progress seen already in ascant few months, and with many tricks stillup their sleeves, researchers are confident theycan overcome such problems.

New Trick from Old DogA MAGNESIUM COMPOUND IS A STARTLING SUPERCONDUCTOR BY GRAHAM P. COLLINS

PH

YSIC

S

MAGNETIC FLUX at low levels, seen herepenetrating a film of magnesium boride,destroys the material’s superconductivity.

The high-temperature superconductor mercury-

barium-calcium copper oxidesuperconducts below 164kelvins—but only when it is

subjected to tremendous pressure,greater than 10,000 atmospheres.

At ordinary atmospheric pressurethe record temperature is held by the same substance, but at

138 kelvins. Hints of room-temperature superconductivityhave often emerged over the years,but none of the claimed results has

ever been successfully reproduced.Recently some press reports have

hyped such claims by a group at theUniversity of Zagreb in Croatia, but superconductivity experts

have concluded that those results have no merit.

Among researchers, such sightingsof irreproducible, anomalously highsuperconducting temperatures are

known as USOs, for unidentifiedsuperconducting objects.

NEED TO KNOW:GETTING HIGH



THE

BO

EIN

G C

OM

PAN

Y

newsSCAN

L ater this summer, or perhaps in early au-tumn, a small pilotless plane will riseinto the clear air over southern Califor-

nia’s desert salt flats on its maiden flight.From all appearances, the new aircraft willlook similar to the many other unmanned ve-hicles that have soared into the sky on solospy missions and scientific surveys in recentyears. This robotic airplane, however, willdiffer significantly from its predecessors.Rather than toting surveillance cameras andradars, it will carry “smart” bombs and mis-siles, should the system eventually be de-ployed in the field. Moreover, this new un-manned combat air vehicle (UCAV) is de-signed to fly for the most part autonomously,in squadrons that will sweep over heavily de-fended battle zones in waves making coordi-nated ground strikes.

The first of a new generation of pilotlessattack aircraft, the X-45A is one of a pairbuilt by Boeing Phantom Works in St. Louisas part of a $131-million program sponsoredby the U.S. Air Force and the Defense Ad-vanced Research Projects Agency (DARPA).Though strictly a technology demonstrator,the Boeing aircraft is designed to meet real re-quirements for hazardous combat missions inwhich airplanes fly directly into the teeth ofsurface-to-air missile batteries. If the conceptproves itself in flight tests planned for the nexttwo years, production UCAVs could be in theair by around 2010.

With a shovel-shaped nose, boomerang-like swept wings, a fuselage resembling amanta ray and no tail, the stealthy prototypecan haul up to a ton and a half of weapons topoints as far as 1,000 miles away. Video cam-eras, a Global Positioning System (GPS) andradar carry out precision-targeting tasks.

The X-45A is one of several UCAVs beingdeveloped. The U.S. Navy and DARPA are de-signing an unmanned bomber that will oper-ate from naval vessels. The Pentagon is re-portedly developing another, still classifiedUCAV design. Meanwhile Sweden’s SaabAerospace and France’s Dassault have intro-duced their own robotic combat aircraft.

Although keeping pilots out of harm’sway is one benefit, it’s not the main purposeof unmanned aircraft. First, UCAVs should

have greater chances of survival than theirmanned counterparts, explains Rich All-dredge, Boeing’s UCAV Advanced Technolo-gy Demonstration (ATD) program manager.Being smaller, they would be harder to detectby radar. In addition, the lack of a cockpitmeans that the engine air intake can be buriedin the upper fuselage, which is the most favor-able position for maintaining low observabil-ity. Second, “the pilot in the cockpit doesn’talways have the best idea of what’s happen-ing out on the battlefield,” Alldredge contin-ues. “With UCAVs, we can put the operatorin the combat air operations center, rightwhere all the intelligence is collected.”

Cockpit-less UCAVs should also be cheap-er to build and operate than conventionalstrike aircraft, says Col. Michael Leahy,DARPA UCAV ATD program manager. “Pi-lots need hundreds of hours each year in theair to maintain combat readiness,” he notes.“UCAV operators can train in simulators.”The team believes that the UCAV can be de-ployed for a third the acquisition price of thenew Joint Strike Fighter. Operational and sup-port costs are expected to total three quartersof that needed for a manned tactical squadron.

A UCAV is more than just an unmannedaerial vehicle with weapons. The aircraft willbe able to execute predetermined “scripts” atcertain points in its mission. All decisions re-garding lethal force will be left in the handsof the operator, however. “You always haveto have a person confirm a target and then, atthe last possible moment, make the decisionto deploy the weapons or not,” Leahy says.

Robotic BombersUNMANNED STRIKE AIRCRAFT BEGIN TO TAKE OFF BY STEVEN ASHLEY

WE

APO

NS

TEC

H

Why UCAVs may be moresuitable than missiles:

“Every time you fire a cruise missileyou lose all your high-cost targeting

sensors. With UCAVs you keep thesensors on the vehicle and release

the cheapest ordnance you can.Also, cruise missiles are fine if you

know exactly where the target is, but they can’t hunt down mobile,

relocatable targets.”—Col. Michael Leahy, DARPA UCAV

ATD program manager

Why they may not be:

Precision standoff missiles, whichare launched from afar by

manned aircraft, could accomplishmany of the same tasks as UCAVs.

In fact, some Pentagon planners areunconvinced that UCAVs are

worth developing.

NO PILOT: Unmanned combat air vehicles (UCAVs) willkeep humans in control but out of danger.

BETTER THANTODAY’S MISSILES?


24D S C I E N T I F I C A M E R I C A N J U N E 2 0 0 1

MAR

K A

. G

ARLI

CK

newsSCAN

Sending a giant rock toward Earthevery 6,000 years has its dangers:

• Collision The asteroid could hitEarth, rather than flying by it.

• Orbital destabilization Thechange in Earth’s orbit could disturb

the motions of the other planets.

• Loss of the moon Most likely, the moon would be stripped away

from Earth unless some additionalenergy-expensive shepherding were arranged. The moon helps

to stabilize Earth’s axial tilt, and its absence could radically upset

our planet’s climate.

NEED TO KNOW:DRAWBACKS

B IOKO ISLAND, EQUATORIAL GUINEA—“How would you like it if I cooked por-cupine tonight?” our cook asks hope-

fully. After four weeks in Central Africa, I hadbecome accustomed to eyebrow-raising ques-tions. “How about fish?” I suggest. Fish isreadily available on Bioko, an island 32 kilo-meters off the coast of Cameroon that forms

part of the tiny African nation of EquatorialGuinea. On the mainland, however, seafoodisn’t always an option. Across tropical Africa,where timeless village ways are meeting thecash economy, the bushmeat trade—huntingwildlife for food—is fast becoming big busi-ness. In a surprise even to battle-worn con-servationists, the trade is eradicating mam-

Unfair GameTHE BUSHMEAT TRADE IS WIPING OUT LARGE AFRICAN MAMMALS BY JOSEPHINE HEARN

CO

NSE

RVA

TIO

N

One billion years—that’s about all thetime we have until the increasing lumi-nosity of the aging sun cooks our plan-

et to near death. But it does not have to bethis way. Researchers argue that graduallymoving Earth farther from the sun is possible.

Since the sun formed 4.6 billion years ago,it has steadily grown and gotten brighter. Al-ready it shines about 30 to 40 percent brighterthan it did when it first entered the main se-quence, its current long-lived period of sta-bility. In about one billion years the sun willbe 10 percent more luminous than it is now—

more than adequate to make land-based lifedifficult or even impossible.

A team led by Donald G. Korycansky ofthe University of California at Santa Cruz hasdeveloped an ambitious yet feasible plan thatcould add another six billion years to ourplanet’s sell-by date. The process is an un-usual application of the well-known gravita-tional slingshot. As a spacecraft closes in ona planet, gravity accelerates the probe, and itshoots away with added energy. That extraenergy does not come free, though: the plan-et suffers equal and opposite changes in en-ergy and momentum.

In the same way, the team’s paper, pub-lished in the March Astrophysics and SpaceScience, shows how Earth’s orbit can be in-creased very slightly if a suitable asteroid (orany object about 100 kilometers across and

weighing about 1016 metric tons) can bemade to fly in front of Earth as it moves in itsorbit. In doing so, the asteroid imparts someof its orbital energy to Earth, shifting it to aslightly larger orbit. The orbit of the asteroidis engineered such that, after its flyby of Earth,it heads toward Jupiter or Saturn, where inthe reverse process it picks up the orbital en-ergy it lost to Earth. Then, when the asteroidreaches its farthest distance from the sun, aslight course correction is applied—by, say,firing engines on the asteroid using fuel man-ufactured from materials mined there—send-ing it once more toward Earth.

Korycansky and his collaborators calculatethat for Earth to enjoy the same intensity ofsunlight it does now, our planet would have tobe nudged outward about once every 6,000years, on the average, for the entire remainingmain-sequence lifetime of the sun. In 6.2 bil-lion years Earth would be just beyond the cur-rent orbit of Mars. The scenario sounds likescience fiction, but it actually uses technologythat is mere decades away from being reality.

Ambitious though the scheme is, it is no so-lution when the sun encounters its fate—as acool, dim white dwarf. At the very end, escap-ing to another star system is ultimately the onlyoption.

Mark A. Garlick, a former astronomer, is awriter and artist based in Brighton, England.

Save the EarthDELAYING OUR PLANET’S ULTIMATE DEMISE—BY SHIFTING ITS ORBIT BY MARK A. GARLICK

ASTR

O-

EN

GIN

EE

RIN

G

COMETS or asteroids could rescue Earth.



PAU

L T.

TE

LFE

R

newsSCAN

mals from the rain forests with remarkableferocity. According to the Convention on In-ternational Trade in Endangered Species(CITES), the practice affects 30 endangeredspecies—among them gorillas, chimpanzees,elephants, duikers (forest antelopes) and mon-keys. At stake is not only biodiversity but alsovital information about the origins of virus-es such as Ebola and HIV, which have beenlinked to the eating and handling of bushmeat.

For decades, conservationists worriedmost about habitat destruction, but these days“bushmeat is recognized as the most signifi-cant threat to wildlife populations,” saysHeather Eves, director of the Bushmeat CrisisTask Force based at the American Zoo andAquarium Association in Silver Spring, Md.Many forests in West and Central Africa havebeen hunted so heavily that no species largerthan a squirrel survives. David Wilkie, a bi-ologist at Boston College and with theWildlife Conservation Society, estimates thatin the Congo Basin alone, bushmeat con-sumption measures one million metric tonsannually, equivalent to seven million head ofcattle. But that figure, he notes, is far belowthe actual volume of meat taken from the for-est. Many animals rot in wire snares beforehunters collect them. “In some places, thatwastage rate is up to 80 percent,” Wilkie says.

Many conservationists have been workingto unravel the myriad factors that influencebushmeat hunting and consumption. Hu-mans have been stalking prey in tropicalforests for thousands of years, but only re-cently has the bushmeat crisis become appar-ent. Perhaps the single biggest factor is log-ging. In Cameroon, Gabon and the Republicof Congo, multinational loggers have domin-ion over more than 60 percent of the land. Al-though their techniques tend to be ecological-ly friendly—felling only a few select trees perhectare—the roads they construct throughonce impenetrable forests offer hunters easyaccess. Sometimes the logging camps them-selves have hundreds, or even thousands, ofhungry workers, creating instant demand.One logging camp in Congo harvested 8,251animals in a single year.

Logging, as well as farming and ranching,has fragmented many forests. According toJohn Oates, a primatologist at Hunter Collegeof the City University of New York, “becausethe areas are getting small, [hunters can] havea devastating impact on what little wildlife isleft.” Last fall Oates reported the probable ex-tinction of Miss Waldron’s red colobus (Pro-

colobus badius waldroni), a monkey that hadsurvived only in isolated chunks of forest inGhana and the Ivory Coast. Miss Wal-dron’s—named after the traveling companionof the collector who discovered the species in1933—is the first primate lost in centuries.

To stem the bushmeat trade, conservation-ists are taking several approaches. They aredeveloping partnerships with loggers to limithunting on logging concessions and to offeralternative sources of protein to workers. Andthey are encouraging loggers to adopt codesof good conduct set by the Forest StewardshipCouncil and other groups. On Bioko Island,the Bioko Biodiversity Protection Programworks with the local university to inform peo-ple about the educational and scientific valueof wildlife. The program also employs vil-lagers to guard one of the island’s protectedareas. Similar strategies have succeeded in ahandful of parks across Africa. “No single an-swer is going to solve the bushmeat problem,”Eves remarks. “It’s a mosaic of solutions.”

Deep within Bioko’s southern protectedarea, one solution may be falling in place. Aftera morning hike through a section abundant inwildlife, Claudio Posa Bohome leans over tome and says conspiratorially, “You know, Iused to be a hunter.” Bohome, now an agron-omist at the National University of Equatori-al Guinea, explains how he pursued game notfar from here. “But now,” he states, motion-ing to the tape that marks the conservationtrails, “I think this is the right thing to do.”

Josephine Hearn lives in Washington, D.C.

Social forces play a significant rolein the bushmeat business. On Bioko,

logging and habitat fragmentationare not threats; still, the island is

not the wildlife paradise one mightexpect. Its seven species of

monkeys—five of them endangeredthroughout their ranges—are

heavily hunted for bushmeat andthen marketed at prices only the

upper classes can afford. Foursubspecies are found nowhere else

in the world. As the big city on theisland grows and becomes affluent,more people demand the foods and

smells of their ancestral villages. Ina 39-month study, more than

26,000 animals passed throughthe main bushmeat market, far

above sustainable levels.

WHEN BUSHMEATIS A DELICACY

WORTH MORE DEAD than alive, monkeys such as thismandrill are prime hunting targets throughout Africa.



RO

DG

ER

DO

YLE

newsSCAN The American Terrorist

A PINCH OF POLITICS, A POUND OF HATE BY RODGER DOYLE

BY

THE

NU

MB

ER

S

Say “terrorism,” and most people thinkOsama bin Laden and Timothy Mc-Veigh, but they are just a small, if scary,

part of a much larger American problem.The accumulation of solid data on U.S. ter-rorism is only now beginning, most notablywith the FBI’s tabulation of hate crimesstarting in the early 1990s. These and otherreports suggest that the number of terror-ist acts against Americans worldwide overthe past 20 years is 250,000 to 300,000.During this time, at least 1,500 Americanshave died in terrorist incidents that, in theirtiming, were utterly unpredictable. Mostdied as a result of bombings.

Fewer than 3,000 of the terrorist acts werecommitted abroad, most prominently byMuslim groups, who have killed about 600Americans since 1982 (the majority of themin the bombings of the U.S. Marine barracksin Lebanon in 1983 and Pan Am Flight 103 in1988). The biggest domestic terrorist act in re-cent years was, of course, the 1995 bombingof the Federal Building in Oklahoma City.

Political beliefs have little to do with do-mestic terrorism. The McVeigh group andother organized, overtly ideological extrem-ists—whether right-wing, left-wing, anti-Mus-lim, pro-Muslim, anti-Castro, Puerto Ricannationalist, eco-terrorist, animal liberationistor cyber-terrorist—probably accounted for asmall number of incidents since 1982, whenthe FBI began keeping systematic records.Rather the largest categories of terrorist of-fenses are racial/ethnic crimes (mostly againstblacks), followed by religious (mostly anti-Semitic) and anti-gay crimes. The occurrenceof racial/ethnic offenses declined during the1990s, while religious and anti-gay offensesheld steady. Many, perhaps most, of these in-cidents were spur-of-the-moment acts by in-dividuals or ad hoc groups.

Another important category, one that isnot adequately covered by official statistics,is attacks against and harassment of abor-tion-services providers. The National Abor-tion Federation, in its incomplete tabulationof violence and threats, estimates such inci-

dents at 12,000 from 1984 to2000, with a substantial de-cline since 1988–89, the peakyears of clinic protests.

Unfortunately, there areno reliable statistics on otherkinds of terrorism, such asstudent attacks on other stu-dents, exemplified by the infa-mous Columbine shootings in1999, and police violenceagainst civilians, as in the no-torious Rodney King episodeof 1992. It would be useful toinclude such acts, as well asanti-abortion terrorism, in thenational reporting system.That way, Americans willhave a comprehensive and re-liable picture of all types ofterrorist acts.

Rodger Doyle’s e-mail [email protected]

S O U R C E S : F B I ; U . S . D e p a r t m e n t o f S t a t e ; A n t i - D e f a m a t i o n L e a g u e ; N a t i o n a l A b o r t i o n F e d e r a t i o n

10,000

1,000

100

10

1

1980 1985 1990 1995 2000

TERRORISM AT HOME AND ABROAD

Racial/EthnicAntiabortion

Anti-Semitic Anti-Gay

Deaths from Anti-American Acts

Worldwide

Marine Barracksin Lebanon:241 Killed

Pan Am Flight 103:278 Killed

Oklahoma City:168 Killed

U.S. Embassiesin Kenya and

Tanzania:301 Killed

Year

Num

ber o

f Ter

roris

t Act

s

There is no agreement on whatconstitutes terrorism. The U.S. State Department says it is

“premeditated, politically motivatedviolence by subnational groups or

clandestine agents, usually intendedto influence an audience.”

The FBI says it is “the unlawful use,or threatened use, of force or

violence against persons or propertyin furtherance of political or social

objectives.” In this article “terrorism”is defined as the use or threat of

violence to make a statementabout ideological or cultural

beliefs. The conscious aim may ormay not be to coerce a government

or a group of people into granting the terrorists’ demands.

DEFININGTERRORISM


newsSCAN

UN

IVE

RSI

TY O

F C

ALIF

OR

NIA

, B

ER

KE

LEY,

MIC

RO

-RO

TAR

Y C

OM

BU

STIO

N L

AB (

top

); M

ATT

CO

LLIN

S (m

idd

le l

eft)

; N

ASA/

JPL/

MAL

IN S

PAC

E S

CIE

NC

E S

YSTE

MS

(mid

dle

rig

ht)

; FR

ED

SP

OO

R,

© N

ATIO

NAL

MU

SEU

MS

OF

KE

NYA

(b

otto

m)

A S T R ONOM Y

Not So WateryIn the April 1 Geophysical Research Letters,researchers propose that the gullies seen onMars last year may have been carved by car-bon dioxide rather than by water. They the-orize that the arrival of spring warms liquidCO2 trapped in the pores of the rocky surface;after expanding and bursting through the sur-face, the liquid quickly vaporizes, and some ofit condenses into CO2 snow. Along withrocky debris, the snow becomes suspended inthe remaining CO2 gas; this suspension, the re-searchers say, flowed and carved the gullies.The model would explain why the gullies arelocated where the planet is coldest and whereunderground liquid CO2 is most likely to bestable. Another strike against Martian waterappears in the April 5 Nature. Ridge featureson the planet’s northern hemisphere werethought to be remnants of an ancient shore-line. New data gathered by the Mars GlobalSurveyor, however, suggest that tectonic stresscreated the ridges. —Philip Yam

ENGINE MITE

CARVED bycarbon dioxide?

As a sequel to the Human GenomeProject, scientists in early April

discussed plans for an analogoussearch for proteins, called the

human proteome project.Meanwhile a joint venture of MyriadGenetics, Hitachi and Oracle called

Myriad Proteomics promises to identify all human proteins

for $500 million.

Number of human chromosomes:46

Length of unraveled DNA: 0.7 to 3.3 inches

Estimated number of genes: 30,000

Estimated number of proteins:300,000 to a few million

Time that the Human GenomeProject took to finish first draft:

10 years

Time that Myriad Proteomics says itwill finish its project:

3 years

S O U R C E S : C e l e r a G e n o m i c s ; H u m a nG e n o m e P r o j e c t ; M y r i a d P r o t e o m i c s

DATA POINTS:GET YOUR PROTEINS

E NGIN E E R ING

The Little EngineThat MightCarlos Fernandez-Pello and his colleagues atthe University of California at Berkeley havebeen boosting the output of their penny-sizerotary internal-combustion engine—the small-est engine ever to deliver continuous power.This mini engine, which runs on a high-en-ergy liquid hydrocarbon such as butane orpropane, produced four watts of electricity asof April, up from 0.7 watt in February. A re-fined, more powerful version might replacebatteries in laptop computers and other port-

able devices. Further-more, a version fash-ioned out of siliconcould someday shrinkthe engine down tothe size of a pinhead.

—Alison McCook

A N T H R OP OLO G Y

Lucy, Meet KenWhen the American Association of Physical Anthropologists gathered in Kansas City, Mo., inMarch, Kenyanthropus platyops stole the show. Meave Leakey of the National Museumsof Kenya talked about the 3.2-million to 3.5-million-year-old fossil remains from northernKenya’s Turkana Basin. Previously, the only hominid thought to have existed during thattime was Australopithecus afarensis, the species to which the famed Lucy fossil belongs and

from which all later hominids—including ourselves—ap-peared to be descended. But the new fossil leaves Lucy’sancestral status uncertain. This early hominid diversi-ty, Leakey says, may have resulted from adaptations tonew ecological niches opened up by the spread of so-called C4 plants, which created bushy grasslands andgrassy woodlands—a shift that has been used to explaindiversification among other mammals from that period.Not everyone agrees that the new fossil warrants a newgenus, however. “Time will tell whether we were rightor wrong,” Leakey remarked. “At least this makes peo-ple ask more questions.” —Kate Wong

NEW ENTRY intothe hominid ranks




newsSCAN

■ Cognitive behavioral therapyseems to help insomniacs,offering an alternative to long-termdrug use. /041101/2.html

■ Scientists created a compositematerial that has a negativeindex of refraction. It may lead to unusual lenses andelectromagnetic devices./040901/3.html

■ Researchers have discovered justhow the mutant protein inHuntington’s disease does itsneuron-destroying job—and havereversed the impending cell deathin the lab dish. /032301/4.html

■ Insulin-like hormones dictate theaging process across severalspecies—a possible explanationfor why low-calorie diets, whichreduce insulin levels, extend life./040601/1.html

WWW.SCIAM.COM/NEWSBRIEF BITS

A ID S

Locating the Latent EnemyFrustrating treatment for HIV-positive patients is the virus’s abilityto hide in T cells. These immune system cells must be turned on by aforeign particle (antigen) but can later turn off and hibernate in theblood for many years. Scientists have found hidden copies of the virusin retired T cells and, more recently, in newborn T cells, which haveyet to be activated. In the April Nature Medicine, researchers suggest that some of these HIV-infected “naive” T cells originate from an HIV-infected thymus, the organ that makes T cellsand releases them into the blood. To test this theory, they added substances that mimicked theaction of a T cell antigen to a culture of HIV-infected thymus tissue that was extracted from amouse. Within 24 hours the amount of viral genes in the culture jumped 30-fold. These resultsmay explain why most patients experience a resurgence in viral levels years after becoming in-fected and may help in developing new therapies against latent HIV. —Alison McCook

BIOLO G Y

Boning UpA purring cat is not necessarily a happy one;many species—including cheetahs and somelions—also purr when wounded or anxious.Some researchers speculate that this lovelyrumble may serve a function: to heal fracturesand strengthen bones. In an as yet unpublishedstudy from the Fauna Communications Re-search Institute in Hillsborough, N.C., inves-

tigators determinedthat the frequency atwhich many catspurr, between 27and 44 hertz forhouse cats, matchesthe frequency thatseems to help hu-man bones strength-en and grow. If cor-rect, the theory mayexplain why catsheal so quickly afterinjury.

—Alison McCook

C OM P U T E R S

Copy UnprotectedIt’s strike one for proponents of hardware-embedded copyright protection. In April thecommittee that designates technology stan-dards voted against a proposal to install aprogram called content protection for re-cordable media (CPRM) directly onto a com-

puter’s hard drive. Opponents have longfeared that CPRM, which would block usersfrom downloading copyright-protected ma-terial, could compromise open-source soft-ware and copying for personal use [see “ToProtect and Self-Serve,” Cyber View, byWendy M. Grossman, March]. But becausecompliance with these technology standardsis voluntary, the group that produced CPRMcan still sell it. —Alison McCook

HIV virus via computer modeling

PURRING as bone builder

S O C IOLO G Y

Aborted CrimeWave, Part 2Two years ago Steven D. Levitt of the Univer-sity of Chicago and John J. Donohue III ofStanford University achieved notoriety byproposing that up to 50 percent of the dropin crime in the 1990s was attributable to thelegalization of abortion: fewer unwanted chil-dren meant less crime. Now another econo-mist has analyzed the same crime data, aswell as other indicators, and has reached adifferent conclusion. “There is nothing tosuggest anything related to legalized abor-tion,” says Theodore J. Joyce of Baruch Col-lege. Based on his analysis, Joyce believes in-stead that the most plausible explanations arethe waning of the crack epidemic and a com-bination of police action, incarceration andeconomic growth. Donohue and Levitt’s re-port, now finally peer-reviewed, appears inthe May Quarterly Journal of Economics.Joyce plans to submit his for publication in afew months. —Marguerite Holloway

MIC

HAE

L FR

EE

MAN

Cor

bis


Monoclonal antibodies are biotechnology’s biggestcomeback story. Until the late 1990s, monoclonals,which had been dubbed magic bullets, appeared to beshot from a gun that couldn’t shoot straight.

A monoclonal is an exact copy of a single antibodythat binds to a specific antigen—a molecule on, say, abacterium, virus or cancer cell. It then triggers a cascadeof events in the immune system that destroys or neu-tralizes the interloper. Although they had the potentialof being highly targeted drugs, monoclonals did not

fulfill their promise. The antibodies, manufactured inmice, provoked an immune response in humans thatmade them unusable as pharmaceuticals.

The race to rectify this early defect generated fero-cious competition among start-ups and provides a com-pelling example of how biotechnology companies weath-er the legal struggles that may prove more critical tosurvival than technical and scientific prowess. A num-ber of researchers responded to the early debacles withmouse antibodies by creating transgenic mice that pro-duce antibodies that are mostly human but still partlyrodent. Ideally, a transgenic mouse bearing genes for anentire human antibody would produce a fully humanmonoclonal. The antibody-making cell could then beisolated to generate an unlimited supply of antibodies.

In 1989 Nils Lonberg, a postdoctoral student atMemorial Sloan-Kettering Cancer Center in New YorkCity, was hired by GenPharm International, then locat-ed in South San Francisco, Calif., to create just such amouse. During the early 1990s he and his group laboredon several big technical challenges: they had to inactivatethe key genes the mouse uses to produce its own anti-bodies and then insert human genes. They could onlyhope that the transplanted human antibody genes, whichdiffer from those of the mouse, would succeed in initi-ating the maturation cycle that leads to antibodies ableto bind tightly enough to antigens to prove effective.

Their plan worked remarkably well. “We werelucky we didn’t encounter problems,” Lonberg says.“There was no way to guarantee that the differencesbetween mouse and human genes wouldn’t have severeconsequences. We just had to try it.” A culminationcame at an industry conference in late 1993, when Lon-berg gave a presentation on a mouse that produced ful-ly human antibodies with high affinities for a target, afeat that Cell Genesys in Foster City, Calif., GenPharm’schief rival, had yet to achieve.

For GenPharm, the announcement became a publicdeclaration that the company had arrived—and it served


OLI

VIE

R L

AUD

E

Innovations

The Mice That WarredNatural selection picks the best antibodies to fight invading microbes—and it also determines who survives to sell these molecules as drugs By GARY STIX

NILS LONBERG developed mice that produce human antibodies.



as a claim of leadership in this race to produce a mousecapable of making human antibodies. Shortly there-after, GenPharm planned its first public stock offering,which would give it the financial wherewithal to launchclinical trials of monoclonal antibody drugs and to setup the manufacturing facilities needed to supply anti-bodies to pharmaceutical company partners.

On February 1, 1994, a few days before Gen-Pharm’s filing for an IPO, Cell Genesys, which had al-ready received a cash infusion by going public, sued thecompany, charging it with having stolen a trade secretfor inactivating a mouse gene. “They were clearly be-hind in terms of technology but clearly ahead in termsof money,” Lonberg says. “We were on our last legs interms of money. We had 110 people and a significantburn rate. We couldn’t figure out what we were beingsued for. I didn’t take it that seriously. The people onthe ground at the technical level thought it was ridicu-lous. But it derailed our public offering. It was harderand harder to get money from venture capitalists.”

The company fired everyone but a skeleton staffand began to try to find a buyer, but the lawsuit stoodin the way. GenPharm, which at one point had a mere$15,000 in cash and owed large sums to lawyers andbanks, had to survive on barter; it tended mouse cagesfor another biotechnology concern in exchange for asliver of laboratory space. It held a fire sale to get rid offurniture, laboratory equipment and patents. A pro-fessor of developmental biology from Stanford Uni-versity came down and inspected a surgical microscopeas a possible toy for his kid.

GenPharm countered with an antitrust lawsuit andtwo patent suits against Cell Genesys. Two weeks beforethe trial in early 1997, Cell Genesys dropped its suit—purportedly because GenPharm had gained a patent thatgave the company a superior intellectual-property posi-tion. Within months, the two companies hammered outa cross-licensing agreement that provided access to eachother’s technologies. Meanwhile GenPharm scrappedits remaining litigation. As part of the accord, CellGenesys and a partner agreed to pay GenPharm nearly$40 million. The former legal foes were now poised toshare a lock on this potential bonanza technology.

GenPharm’s technology—if not the original com-pany—survived to become a significant player in whathas become perhaps the hottest area of biotechnology.After the debacle, the company needed cash fast to meetthe obligations venture-capitalist firms had to their in-vestors, so it began to seek a buyer. “We were on the roadimmediately shopping the company,” Lonberg says.

In the fall of 1997 Medarex, an antibody company

based in Annandale, N.J., bought GenPharm, a deal thatprovided manufacturing facilities and other resources ithad been unable to acquire during the years of the law-suit. Lonberg, who now holds the title of scientific di-rector at Medarex, is the only remaining employee whohas worked on the program since its inception. His bit-terness remains. “The final story is not that we prevailedbut that [Cell Genesys] actually succeeded in its strate-gy. It was able to use litigation to capture a technology.”

Cell Genesys spun off its mouse technology into aseparate company: Abgenix in Fremont, Calif. Abgenixdisputes the contention that it used lawsuits to catch up,saying that scientific papers show that its mouse was de-

finitively better than the Medarexrodent. It doesn’t really matter any-more who is right. Things havebeen good for both Medarex andAbgenix, which have become the

Coke and Pepsi of the antibody world. Monoclonalshave boomed. More than 90 monoclonal antibodiesare now in clinical trials, most using the older tech-nologies that retain some of the properties of themouse. About 10 have made it to market, includingdrugs for breast cancer and non-Hodgkin’s lymphoma,and constitute an estimated $2.1 billion in revenue for2001. The market may grow to more than $5 billionby 2004.

The advantages of all-human antibodies in im-munogenicity and in speed and cost of developmenthave revived Lonberg’s work from its near-death en-counter. Beginning in 1998, Medarex began to strikepartnerships with large drug companies and biotechconcerns to provide monoclonal-producing mice. By2000 it was entering into an agreement with anothercompany nearly every month—it now has 31 partner-ships in addition to launching its own clinical trials ofa few drugs. And last year the mouse technology pro-pelled a $400-million Medarex stock offering. Human-antibody mice mark a step toward fulfilling the dreamfor these drugs. But Lonberg’s experience also confirmsthe musings of immunologist and Nobel Prize winnerPaul Ehrlich, who, around the start of the 20th centu-ry, conceived of the notion of a “magic bullet” againstdisease. He said, “Magic substances like the antibodies,which affect exclusively the harmful agent, will not beso easily found.” That may be true, though perhaps notfor the technical reasons Ehrlich contemplated.

Lonberg’s work on monoclonals has beenrevived from its near-death encounter.



BR

IAN

STA

UFF

ER

Staking Claims

Should someone be able to patent an invention that blatantly duplicates a previously patented creation ex-cept for some minor alterations—changing a rivet to a bolt, for instance? The Court of Appeals for the Fed-eral Circuit, the judiciary that handles appeals in patentcases, has effectively said yes. Its recent ruling dra-matically weakens a body of common law that letspatent holders expand coverage of their patents to fend

off imitators. Whether itcasts a chill on innova-tion or enlivens it still re-mains to be seen.

Some legal analystshave termed the Novem-ber 29, 2000, decision—

Festo v. SMC—a fatalstrike against the so-calleddoctrine of equivalents,which protects an inven-tor against a copycat whocreates a different butfunctionally equivalentproduct. To prevent abuseof that principle, a pastrestriction has prohibitedpatent applicants fromnarrowing a claim to per-suade examiners that aninvention is original and

then, after the patent is issued, using the doctrine ofequivalents to broaden the scope of the claims. UnderFesto, the reach of this rule grows: virtually any nar-rowing of a claim, even one that clarifies the patent lan-guage, precludes later use of the equivalence argument.

If the ruling stands, the fallout may be huge. Ap-plicants routinely make changes in filings in the back-and-forth negotiations with patent examiners. Manypatents—some critics say virtually all—would be af-fected by the decision. A copycat can now examinewhich claim provisions in a patent have been amended

and then design an invention with only minimal alter-ations to those components.

The case may not be closed, though. Festo, a man-ufacturer that brought an infringement suit against ri-val SMC over a part used in a robotic arm, wants totake the case to the Supreme Court. If the SupremeCourt lets Festo stand, the effect of this case, which ap-plies retroactively to patents issued as far back as themid-1980s, could cheapen the value of existing patentportfolios. An exclusive license issued from a patentholder would be worth less if someone else could read-ily manufacture and market virtually the same tech-nology without infringement. “It has the potential todramatically decrease the value of patents and, as a re-sult, dramatically decrease the incentives for innova-tion,” says Jay Alexander of the law firm Kirkland andEllis in Washington, D.C. The patent applicationprocess could become longer and more expensive ascompanies spend more time drafting claims that wouldnot need amendment later, a burden in particular forsmall companies and individual inventors.

A number of large corporations, including IBM,Ford and Kodak, all of which filed friend-of-court briefsin the case, welcomed the Festo decision. Before Fes-to, they contend, it was impossible to tell if a new wid-get would infringe on a competitor’s patent because thedoctrine of equivalents might be invoked. Big compa-nies worry about getting broadsided by a lawsuit froman individual or small company that is trying to rakea large firm. Arthur Neustadt, the attorney for SMC,argues that the decision will encourage more innova-tion because patent claims will be clearer.

Whether Festo introduces more certainty intopatent law is still unknown. But both sides will con-tinue to promote themselves as champions of innova-tion and technological advancement in their bids togain the upper hand in this epic battle.

Please let us know about interesting or unusualpatents. Send suggestions to: [email protected]

A License for Copycats?A court decision may clarify what is patentable while giving a free ride to knockoffs By GARY STIX


The price of liberty is, in addition to eternal vigilance,eternal patience with the vacuous blather occasionallyexpressed from behind the shield of free speech. It is acost worth bearing, but it does become exasperating,as when the Fox Broadcasting Company aired its high-ly advertised special “Conspiracy Theory: Did WeLand on the Moon?” NASA, viewers were told, fakedthe Apollo missions on a movie set.

Such flummery should not warrant a response, butin a free society, skeptics are the watchdogs against irra-tionalism—the consumer advocates of ideas. Debunkingis not simply the divestment of bunk; its utility is in of-fering a better alternative, along with a lesson on howthinking goes wrong. The Fox show is a case study, start-ing with its disclaimer: “The following program dealswith a controversial subject. The theories expressed arenot the only possible explanation. Viewers are invited tomake a judgment based on all available information.”That information, of course, was not provided, so let’srefute Fox’s argument point by point in case the statis-tic at the top of the show—that 20 percent of Americansbelieve we never went to the moon—is accurate.

Claim: Shadows in the photographs taken on themoon reveal two sources of light. Given that the sun isthe only source of light in the sky, the extra “fill” lightmust come from studio spotlights. Answer: Setting asidethe inane assumption that NASA and its co-conspiratorswere too incogitant to have thought of this, there are ac-tually three sources of light: the sun, the earth (reflect-ing the sun) and the moon itself, which acts as a pow-erful reflector, particularly when you are standing on it.

Claim: The American flag was observed “waving”in the airless environment of the moon. Answer: The flagwaved only while the astronaut fiddled with it.

Claim: No blast crater is evident underneath theLunar Excursion Module (LEM). Answer: The moonis covered by only a couple of inches of dust, beneathwhich is a solid surface that would not be affected bythe blast of the engine.

Claim: When the top half of the LEM took off fromthe moon, there was no visible rocket exhaust. TheLEM instead leaped off its base as though yanked upby cables. Answer: First, the footage clearly shows thatthere was quite a blast, as dust and other particles goflying. Second, without an oxygen-rich atmosphere,there is no fuel to generate a rocket-nozzle flame tail.

Claim: The LEM simulator used by astronauts forpractice was obviously unstable—Neil Armstrongbarely escaped with his life when hissimulator crashed. The real LEM wasmuch larger and heavier and thus im-possible to land. Answer: Practicemakes perfect, and these guys practiced.A bicycle is inherently unstable, too, un-til you learn to ride it. Also, the moon’sgravity is only one sixth that of theearth’s, so the LEM’s weight was less destabilizing.

Claim: No stars show in the sky in the photographsand films from the moon. Answer: Stars don’t routine-ly appear in photography shot on the earth, either.They are simply too faint. To shoot stars in the nightsky, even on the moon, you need to use long exposures.

The no-moonie mongers go on and on in this vein,weaving narratives that include the “murder” of astro-nauts and pilots in accidents, including Gus Grissom inthe Apollo 1 fire before he was about to go public withthe hoax. Like most people with conspiracy theories,the landing naysayers have no positive supporting evi-dence, only allegations of cover-ups. I once asked G.Gordon Liddy (who should know) about conspiracies.He quoted Poor Richard’s Almanack: “Three peoplecan keep a secret if two of them are dead.” To think thatthousands of NASA scientists would keep their mouthsshut for years is risible rubbish.


BR

AD H

INE

S

Skeptic

Fox’s FlapdoodleTabloid television offers a lesson in uncritical thinking By MICHAEL SHERMER

Michael Shermer is the founding publisher of Skepticmagazine (www.skeptic.com) and the author of How We Believe and The Borderlands of Science.

NASA, viewers were told, fakedthe Apollomissions on amovie set.


MOSS LANDING, CALIF.—“Come on, just be a little bitcareful because the tide’s low.” We hop on board theday boat Point Lobos, and Marcia Kemper McNutt—a marvel of efficiency on land—noticeably relaxes. Thisold hulk is the size of a tugboat, converted from ser-vicing offshore oil rigs to plying the canyons of Mon-terey Bay for science. “Hey, Knute, how was your daytoday?” she calls out to the pilot of a remotely operat-ed vehicle. She seems to know not just the first name

but the welfare of every one of the 200-some engineers,scientists and operations crews who work for her.

McNutt raised more than a few eyebrows when sheleft an endowed chair at the Massachusetts Institute ofTechnology four years ago. Not only was she odds-onfavorite to become department head in a year, but shealso held a key post associated with the Woods HoleOceanographic Institution, an estimable leviathan ofocean research. Instead she headed west to direct theMonterey Bay Aquarium Research Institute (MBARI,pronounced “em-BAHR-ee”), a relatively backwater in-stitute substantially overshadowed by its namesake sis-ter 20 miles south, the actual tourist-beloved aquarium.

By McNutt’s logic, improving on the past depart-ment chair’s job at M.I.T. would be impossible. MBARIseemed “poised to make a huge impact,” she says. Allit needed was some tweaking. Now, as president ofMBARI, this 49-year-old Minneapolis native finds her-self one of the world’s most influential ocean scientists.

Offbeat choices are nothing new for McNutt. Shechose Colorado College even though her perfect-800SAT scores could have gained her entry nearly any-where. Her adviser there discouraged her from takingphysics, deeming it unsuitable for women. McNutt’sresponse: to switch advisers. She graduated summacum laude with a physics degree in three years.

Then, in the early 1970s, she read John F. Dewey’sarticle on plate tectonics in Scientific American. “Thisis so beautiful, so simple,” she recalls thinking of the thenrelatively new theory. “It’s got to be right.” She went onto obtain a Ph.D. in earth sciences at the Scripps Insti-tution of Oceanography in La Jolla, Calif. She beganto travel to sea, sometimes to study the midocean ridgesystem, where plates meet and new oceanic crust forms.

Science ships are tight for space, so students need-ed a skill to justify their presence. “I went out to be theshooter,” she remarks. A summer with the U.S. NavySEALs taught McNutt how to handle explosives, wrapthem in detonation cords and time the charges precisely P

HO

TOG

RAP

HS

BY

ED

WAR

D C

ALD

WE

LL

Profile

Piloting through Uncharted SeasThe privately funded Monterey Bay Aquarium Research Institute enables scientists and engineers to engage in radical pursuits. As long as Marcia K. McNutt likes their ideas BY JOHN ADAM

■ Husband, Ian Young; daughters Meredith and twins Ashley and Dana■ Best-known fact: president of 38,000-member American Geophysical Union■ Least-known fact: Navy SEAL-certified demolitions expert■ On research: “In principle, we still retain the concept of international waters.

But the fact is that she who owns the technology owns the oceans.”

MARCIA K. MCNUTT: GOING DEEP



so that a clean blast would acousticallymap ocean geology. “I love going to sea,”she says. “There’s a camaraderie. Every-one is focused on the same mission.” Sheeventually met her husband, a captain,when she was chief scientist at sea. Mc-Nutt manages to continue her geophysicsresearch—in May she joined an instituteresearch vessel off Hawaii to examine hotspots. But she spends most of her timekeeping MBARI shipshape.

The institute is the brainchild of DavidPackard. The billionaire engineer and co-founder of Hewlett-Packard took a keeninterest in oceans during the last years ofhis life, thanks in part to his daughters,who studied marine science. Packardfounded MBARI in 1987 as a private-sector complement to gov-ernment-dominated ocean research. Engineers working besidetop scientists could, he believed, open up deep-ocean research.

The David and Lucile Packard Foundation pours about $40million a year into MBARI. Researchers there, unencumberedby teaching or the federal grant application process, can movenimbly, assuming McNutt likes their ideas. They can also car-ry out risky long-term technology-intensive projects that mightotherwise be quashed under peer review. One example: 12years of monitoring for global temperature change from Mon-terey Bay ultimately paid off, McNutt explains, when “wecould see a trend in the data” showing that the bay’s relative-ly small increase in temperature resulted in disproportionatelylarge decreases in its algal biomass productivity.

Unlike other institutes, MBARI schedules its growing fleet ofvessels on its own. (Other institutes, beholden to federal funds,cooperatively schedule their fleets for the nation’s marine scien-tists.) Such autonomy endows McNutt’s post with great influ-ence. It is “much more powerful than a typical institution’s di-rector” position, says G. Ross Heath of the University of Wash-ington, who was McNutt’s predecessor at MBARI. Shortlyafter Packard died in 1996, McNutt became director. AlthoughPackard is still revered at MBARI, his presence made it toughfor others to move without fear of being second-guessed.

McNutt stepped in just as MBARI’s new buildings, its tworesearch ships, its two remotely operated vehicles (ROVs) andits systems for acquiring and cataloguing information were op-erational, or nearly so. “We underestimated the time it wouldtake to build a new institute,” confesses Julie Packard, who as-sumed the chair of MBARI’s board from her father. Only now,she says, with all the equipment working well, are the scientif-ic benefits being reaped, as shown by a rise in MBARI-affiliat-ed authorship in prestigious journals.

Credit McNutt for smoothing relations and squeezing the

most out of her diverse crew. One of herfirst actions was to shift some engineersinto coveted ocean-view offices that hadbeen an exclusive province of scientists.(Her own office is efficiently austere, witha view of the twin smokestacks across theharbor.) MBARI engineers bring scien-tists vicariously to ever greater depths.The institute’s first ROV, modified froman oil-industry machine, dove to depthsof nearly two kilometers; its second hadmore homemade innovations and reachedfour kilometers. McNutt proudly showsme MBARI’s yellow autonomous under-water vehicle, still in the shop; the sub-mersible will soon be swimming unteth-ered as deep as 4.5 kilometers and will

launch from MBARI’s third research vessel, the Zephyr.For the most part, the ocean’s major events remain unob-

served. “Plankton bloom. Volcanoes erupt. Plates slip in earth-quakes. Fish spawn,” McNutt says. “The chance of being in theright time and in the right place to catch such events in actionis very small.” Eventually schools of swimming robots couldremedy that. In short, it’s a race to see whether these tools canbe made to address such pressing oceanic problems as globalwarming, energy production and sustainable fisheries.

With all the various interests seeking to exploit or conservethe ocean, McNutt keeps an open mind and has learned to mod-erate controversy with lessons learned from home. “So manytimes my twins get into an argument. Both are absolutely firmin their convictions. And if you say to either one of them they’rewrong, then they start tuning you out.” So McNutt tries to makethe 15-year-olds aware of the other’s position and values.

That trick apparently worked when McNutt recentlychaired the President’s Panel on Ocean Exploration. First shemade sure the committee had no deadweight. “We didn’t wantpeople sitting around the table, taking up space and wastingour time when they aren’t in a position to give first-class scienceinput,” she recalls. The group of eminent scientists and educa-tors reached a consensus calling for a 10-year, $750-million ef-fort to inventory and explore the Exclusive Economic Zone (anarea that extends 200 nautical miles from all U.S. coasts), con-tinental margins, the Arctic and other regions.

With roughly 95 percent of the ocean unknown and unex-plored, it would seem that McNutt has much work ahead. Sheintends to add another 60 permanent positions and bring in awider assortment of visiting scientists and student interns tokeep MBARI connected with the broader research communi-ty—and, she hopes, a step or two ahead.

John Adam is a technology writer based in Washington, D.C.


PRIDE OF MBARI: McNutt with a submersible andon a data-gathering mooring (opposite page).


Like a boiling teakettle atop a COLD stove,

the sun’s HOT outer layers sit on the relatively cool surface.

And now astronomers are FIGURING OUT WHY


SUSPENDED IN MIDAIR, a prominence (wispy streamon right side) has erupted off the sun’s surface into its

atmosphere—the corona. The coronal plasma is invisiblein this image, which shows ultraviolet light from coolergas in the prominence and underlying chromosphere.

White areas are high density; red are low density.

coronaparadoxof the Sun’s hot

the

by Bhola N. Dwivedi and Kenneth J. H. Phillips


On August 11, 1999, tens of millions of people acrossEurope and Asia were witness to one of the most beau-tiful spectacles in all of nature: a total eclipse of the sun.The two of us were among them. One of us (Phillips)watched from Bulgaria as the glaring disk of the sunwas blotted out by the cool black moon, bringing forththe full glory of the gleaming corona. The other (Dwive-di) watched from India as the glaring disk of the sunwas blotted out by a dull haze of clouds at just thewrong time. But all was not lost, for the spectacle in theheavens was replaced by one on the ground. Across theholy river Ganges, chants reverberated as vast crowdswaded in and prayed for the sun god to reappear.

Millions more will have their view this month as themoon’s shadow sweeps across southern Africa. As-tronomers will get another of their rare opportunitiesto make detailed studies of the enigmatic corona fromEarth’s surface—another chance to make sense of oneof the most enduring conundrums in astronomy.

The sun might look like a uniform ball of gas, theessence of simplicity. In actuality it has well-defined lay-ers akin to a planet’s solid part and atmosphere. Solarradiation, on which all life on Earth ultimately depends,derives from nuclear reactions deep in the core. The en-ergy gradually leaks out until it reaches the visible sur-

face, known as the photosphere, and escapes into space.Above that surface is a tenuous atmosphere. The low-er part of the atmosphere, the chromosphere, is visi-ble as a bright red crescent during total eclipses. Beyondit is the pearly white corona, extending out millions ofkilometers. And from the corona’s outer reaches em-anates the solar wind, the stream of charged particlesthat blows through the solar system.

As you might expect, the sun’s temperature dropssteadily from its core, 15 million kelvins, to the pho-tosphere, a mere 6,000 kelvins. But then an unexpect-ed thing happens: the temperature gradient reverses.The chromosphere’s temperature steadily rises to10,000 kelvins, and going into the corona, the tem-perature jumps to one million kelvins. Parts of the coro-na associated with sunspots get even hotter. Consid-ering that the energy must originate beneath thephotosphere, how can this be? It is as though you gotwarmer the farther away you walked from a fireplace.

The first hints of this mystery emerged in the 19thcentury when eclipse observers detected spectral emis-sion lines that no known element could account for. Inthe 1940s physicists associated two of these lines withiron atoms that had lost up to half their normal retinueof 26 electrons—a situation that requires extremely high

The loops, ARCHES and holes appear to trace out the sun’s MAGNETIC FIELDS.

NAS

A G

OD

DAR

D S

PAC

E F

LIG

HT

CE

NTE

R (

pre

ced

ing

pa

ge)

; TR

ACE

/NAS

A(b

elow

)

CORONAL LOOP, seen in ultraviolet light by the TRACE spacecraft, extends 120,000 kilometers off the sun’s surface.


temperatures. Later, instruments on rockets and satel-lites discovered that the sun emits copious x-rays andextreme ultraviolet radiation—as can be the case only ifthe coronal temperature is measured in megakelvins.Nor is this mystery confined to the sun: most sunlikestars appear to have x-ray-emitting atmospheres.

At long last, however, a solution seems to be with-in our grasp. Astronomers have implicated magneticfields in the coronal heating; where those fields arestrongest, the corona is hottest. Such fields can trans-port energy in a form other than heat, thereby side-stepping the usual thermodynamic restrictions. The en-ergy must still be converted to heat, and researchers aretesting two possible theories: small-scale magnetic fieldreconnections—the same process involved in solarflares—and magnetic waves. Important clues have come

from complementary observations: spacecraft can ob-serve at wavelengths inaccessible from the ground,while ground-based telescopes can gather reams of da-ta unrestricted by the bandwidth of orbit-to-Earth ra-dio links. The findings may be crucial to understand-ing how events on the sun affect the atmosphere ofEarth [see “The Fury of Space Storms,” by James L.Burch; Scientific American, April].

The first high-resolution images of the corona camefrom the ultraviolet and x-ray telescopes on board Sky-lab, the American space station inhabited in 1973 and1974. Pictures of active regions of the corona, located


X-RAY IMAGE from the Yohkoh spacecraft shows structures bothbright (associated with sunspots) and dark (the polar coronal hole).

INST

ITU

TE O

F SP

ACE

AN

D A

STR

ON

AUTI

CAL

SC

IEN

CE

, JA

PAN

; LO

CK

HE

ED

-MAR

TIN

SO

LAR

AN

D A

STR

OP

HYS

ICS

LAB

OR

ATO

RY;

NAT

ION

AL

ASTR

ON

OM

ICAL

OB

SER

VATO

RY

OF

JAP

AN;

UN

IVE

RSI

TY O

F TO

KYO

; N

ASA


above sunspot groups, revealed complexes of loopsthat came and went in a matter of days. Diffuse x-rayarches stretched over millions of kilometers. Awayfrom active regions, in the “quiet” parts of the sun, ul-traviolet emission had a honeycomb pattern related tothe granulation of the photosphere. Near the solarpoles were areas of low x-ray emission—the so-calledcoronal holes.

Connection to the Starry DynamoEACH MAJOR SOLAR SPACECRAFT since Skylabhas offered a distinct improvement in resolution. Since1991 the x-ray telescope on the Japanese Yohkohspacecraft has routinely imaged the sun’s corona,tracking the evolution of loops and other featuresthrough one complete 11-year cycle of solar activity.The Solar and Heliospheric Observatory (SOHO), ajoint European-American satellite launched in 1995,orbits a point 1.5 million kilometers from Earth on itssunward side, giving the spacecraft the advantage of anuninterrupted view of the sun [see “SOHO Reveals theSecrets of the Sun,” by Kenneth R. Lang; ScientificAmerican, March 1997]. One of its instruments, theLarge Angle and Spectroscopic Coronagraph (LASCO),observes in visible light using an opaque disk to maskout the main part of the sun. It has tracked large-scale

coronal structures as they rotate with the rest of thesun (a period of about 27 days as seen from Earth).The images show huge bubbles of plasma known ascoronal mass ejections, which move at up to 2,000kilometers per second, erupting from the corona andoccasionally colliding with Earth and other planets.Other SOHO instruments, such as the Extreme Ultra-violet Imaging Telescope, have greatly improved onSkylab’s pictures.

The Transition Region and Coronal Explorer(TRACE) satellite, operated by the Stanford-LockheedInstitute for Space Research, went into a polar orbitaround Earth in 1998. With unprecedented resolution,its ultraviolet telescope has revealed a vast wealth ofdetail. The active-region loops are now known to bethreadlike features no more than a few hundred kilo-meters wide. Their incessant flickering and jouncinghint at the origin of the corona’s high temperature.

The loops, arches and coronal holes appear to traceout the sun’s magnetic fields. The fields are thought tooriginate in the upper third of the solar interior, whereenergy is transported not by radiation but by convec-tion. The circulation acts as a natural dynamo, con-verting about 0.01 percent of the outgoing radiationinto magnetic energy. Differential rotation—wherebylow latitudes rotate slightly faster than higher lati-


DO

N D

IXO

N

CORONA

CONVECTIVE ZONE

PHOTOSPHERE

CORE

RADIATIVE ZONE

CHROMOSPHERE

SOLAR WIND

TRANSITIONZONE


tudes—distorts the lines of magnetic force into char-acteristic patterns. At sites marked by sunspot groups,ropelike bundles of field lines pierce the photosphereand extend outward into the corona.

For a century, astronomers have measured the mag-netism of the photosphere using magnetographs, whichobserve the Zeeman effect: in the presence of a mag-netic field, a spectral line can split into two or morelines with slightly different wavelengths and polariza-tions. But Zeeman observations for the corona have yetto be done; for the spectral lines that the corona emits,the splitting is too small to be detected with present in-struments. So astronomers have had to resort to math-ematical extrapolations from the photospheric field.These extrapolations predict that the magnetic field ofthe corona generally has a strength of about 10 gauss,20 times Earth’s magnetic field strength at its poles. Inactive regions, the field may reach 100 gauss.

Space HeatersTHESE F IELDS ARE WEAK compared with thosethat can be produced with laboratory magnets, butthey have a decisive influence in the solar corona. Thisis because the corona’s temperature is so high that it isalmost fully ionized: it is a plasma, made up not of neu-tral atoms but of protons, electrons and atomic nuclei(mostly helium). Plasmas undergo a wide range of phe-nomena that neutral gases do not. The magnetic fieldsof the corona are strong enough to bind the chargedparticles to the field lines. Particles move in tight heli-cal paths up and down these field lines like very smallbeads on very long strings. The limits on their motionexplain the sharp boundaries of features such as coro-nal holes. Within the tenuous plasma, the magneticpressure (proportional to the strength squared) exceedsthe thermal pressure by a factor of at least 100.

One of the main reasons astronomers are confidentthat magnetic fields energize the corona is the clear rela-tion between field strength and temperature. The brightloops of active regions have a temperature of about fourmillion kelvins, whereas the giant arches of the generalcorona have a temperature of about one million kelvins.

Until recently, however, ascribing coronal heatingto magnetic fields ran into a serious problem. To con-vert field energy to heat energy, the fields must be ableto diffuse through the plasma, which requires that thecorona have a certain amount of electrical resistivity—

in other words, that it not be a perfect conductor. A

perfect conductor cannot sustain an electric field, be-cause charged particles instantaneously repositionthemselves to neutralize it. And if a plasma cannot sus-tain an electric field, it cannot move relative to the mag-netic field (or vice versa), because to do so would in-duce an electric field. This is why astronomers talkabout magnetic fields being “frozen” into plasmas.

This principle can be quantified by considering thetime it takes a magnetic field to diffuse a certain distancethrough a plasma. The diffusion rate is inversely pro-portional to resistivity. Classical plasma physics as-sumes that electrical resistance arises from so-calledCoulomb collisions: electrostatic forces from chargedparticles deflect the flow of electrons. If so, it should takeabout 10 million years to traverse a distance of 10,000kilometers, a typical length of active-region loops.

Events in the corona—for example, flares, whichmay last for only a few minutes—far outpace that rate.Either the resistivity is unusually high or the diffusiondistance is extremely small, or both. A distance as shortas a few meters could occur in certain structures, ac-companied by a steep magnetic gradient. But researchershave come to realize that the resistivity could be higherthan they traditionally thought. Over the past few years,physicists have observed instabilities in laboratory plas-mas such as those in fusion devices. Those instabilitiescan stir up small-scale turbulence and fluctuations in thebulk electric charge, providing a source of resistancemore potent than random particle encounters.

Raising the MercuryASTRONOMERS HAVE TWO basic ideas for coronalheating. For years, they concentrated on heating bywaves. Sound waves were a prime suspect, but in the late1970s researchers established that sound waves emerg-ing from the photosphere would dissipate in the chro-mosphere, leaving no energy for the corona itself. Sus-


BHOLA N. DWIVEDI and KENNETH J. H. PHILLIPS began collaborating on solarphysics a decade ago. Dwivedi teaches physics at Banaras Hindu University inVaranasi, India. He has been working with SUMER, an ultraviolet telescope onthe SOHO spacecraft, for more than 10 years; the Max Planck Institute for Aeron-omy near Hannover, Germany, recently awarded him one of its highest hon-ors, the Gold Pin. As a boy, Dwivedi studied by the light of a homemade burnerand became the first person in his village ever to attend college. Phillips is theleader of the solar research group at the Rutherford Appleton Laboratory in Did-cot, England. He has worked with x-ray and ultraviolet instruments on numer-ous spacecraft—including OSO-4, SolarMax, IUE, Yohkoh and SOHO—and has ob-served solar eclipses using CCD cameras.

THE

AU

THO

RS

The sun might look like a UNIFORM BALL, the essence of simplicity. In actuality, it has well-defined LAYERS.


picion turned to magnetic waves. Such waves might bepurely magnetohydrodynamic (MHD)—so-called Alf-vén waves—in which the field lines oscillate but thepressure does not. More likely, however, they sharecharacteristics of both sound and Alfvén waves.

MHD theory combines two theories that are chal-lenging in their own right—ordinary hydrodynamicsand electromagnetism—although the broad outlinesare clear. Plasma physicists recognize two kinds ofMHD pressure waves, fast and slow mode, dependingon the phase velocity relative to an Alfvén wave—

around 2,000 kilometers per second in the corona. Totraverse a typical active-region loop requires about fiveseconds for an Alfvén wave, less for a fast MHD wave,but at least half a minute for a slow wave. MHD wavesare set into motion by convective perturbations in thephotosphere and transported out into the corona viamagnetic fields. They can then deposit their energy in-to the plasma if it has sufficient resistivity or viscosity.

A breakthrough occurred in 1998 when the TRACEspacecraft observed a powerful flare that triggeredwaves in nearby fine loops. The loops oscillated backand forth several times before settling down. The damp-ing rate was millions of times faster than classical theo-ry predicts. This landmark observation of “coronal seis-

mology” by Valery M. Nakariakov, then at the Uni-versity of St. Andrews in Scotland, and his colleagueshas shown that MHD waves could indeed deposit theirenergy into the corona.

Despite the plausibility of energy transport bywaves, a second idea has been ascendant: that coronalheating is caused by very small, flarelike events. A flareis a sudden release of up to 1025 joules of energy in anactive region of the sun. It is thought to be caused byreconnection of magnetic field lines, whereby oppo-sitely directed lines cancel each other out, convertingmagnetic energy into heat. The process requires thatthe field lines be able to diffuse through the plasma.

A flare sends out a blast of x-rays and ultraviolet ra-diation. At the peak of the solar cycle (now occurring),several flares per hour may burst out all over the sun.Spacecraft such as Yohkoh and SOHO have shownthat much smaller but more frequent events take placenot only in active regions but also in regions otherwisedeemed quiet. These tiny events have about a millionthof the energy of a full-blown flare and so are called mi-croflares. Hard x-ray emission from them was first de-tected in 1980 by Robert P. Lin of the University ofCalifornia at Berkeley and his colleagues with a bal-loon-borne detector. During the solar minimum in1996, Yohkoh also recognized events as small as 1017

joules.Flares are not the only type of transient phenome-

na. X-ray and ultraviolet jets, representing columns ofcoronal material, are often seen spurting up from thelower corona at several hundred kilometers per second.But tiny x-ray flares are of special interest because theyreach the megakelvin temperatures required to heat thecorona. As we and Pawel T. Pres of Wroclow Univer-sity in Poland have found, following up work by re-nowned solar physicist Eugene N. Parker of the Uni-versity of Chicago, the observed flare rates can beextrapolated to even tinier events, or nanoflares. Thetotal energy could then account for the radiative out-put of the corona, about 3 × 1018 watts.

Which mechanism—waves or nanoflares—domi-nates? It depends on the photospheric motions that per-turb the magnetic field. If these motions operate ontimescales of half a minute or longer, they cannot trig-ger MHD waves. Instead they create narrow currentsheets in which reconnections can occur. Very high res-olution optical observations of bright filigree structuresby the Swedish Vacuum Tower Telescope on La Palma

Spacecraft imaging is TOO SLUGGISH to capture the periodic FLUCTUATIONS caused by magnetic waves.


LAU

RIE

GR

ACEX-RAY SPECTRA of the twin stars Capella (between flares) and the

sun (during a flare) both indicate a temperature of six millionkelvins—typical for Capella but anomalously high for the sun.

CAPELLA

SOLAR FLARE

Rela

tive

Inte

nsity

Rela

tive

Inte

nsity

1.0

0.75

0.5

0.25

1.0

0.75

0.5

0.25

012 14 16 18 20

Wavelength (angstroms)

Wavelength (angstroms)

0

108

12 14 16 18 20108


in the Canary Islands—as well as SOHO and TRACEobservations of a general, ever changing “magnetic car-pet” on the surface of the sun—demonstrate that mo-tions occur on a variety of timescales. Although the ev-idence now favors nanoflares for the bulk of coronalheating, waves may also play a role.

FieldworkIT IS UNLIKELY, for example, that nanoflares havemuch effect in coronal holes. In these regions, the fieldlines open out into space rather than loop back to thesun, so a reconnection would accelerate plasma out intointerplanetary space rather than heat it. Yet the coro-na in holes is still hot. Astronomers have scanned forsignatures of wave motions, which may include peri-odic fluctuations in brightness or Doppler shift. Thedifficulty is that the MHD waves involved in heatingprobably have very short periods, perhaps just a fewseconds. At present, spacecraft imaging is too sluggishto capture them.

For this reason, ground-based instruments remainimportant. A pioneer in this work has been Jay M.Pasachoff of Williams College. Since the 1980s he andhis students have used high-speed detectors to look formodulations in the coronal light during eclipses. Analy-ses of his best results indicate oscillations with periodsof one to two seconds. Serge Koutchmy of the Instituteof Astrophysics in Paris, using a coronagraph, has foundevidence of periods equal to 43, 80 and 300 seconds.

The search for those oscillations is what led Phillipsand his team to Shabla, a small town on the Black Seacoast of Bulgaria, for the August 1999 eclipse. Our in-strument consists of a pair of fast-frame CCD camerasthat observe both white light and the green spectral lineproduced by highly ionized iron. A tracking mirror, orheliostat, directs sunlight into a horizontal beam that

passes into the instrument. During the two minutes and23 seconds of totality, the instrument took 44 imagesper second. Analyses by Pawel Rudawy of Wroclawand David A. Williams of the Queen’s University ofBelfast have revealed localized oscillations, generallyalong loop structures. The periods are between two and10 seconds. Elsewhere, however, our instrument de-tected no oscillations. Therefore, MHD waves are like-ly to be present but not pervasive or strong enough todominate coronal heating. We will take our equipmentto Zambia for the June 21 eclipse and later adapt it fora coronagraph. (Although the opaque disk inside a coro-nagraph allows year-round observing, it cannot maskout the sun as effectively as the moon during an eclipse.)

Insight into coronal heating has also come from ob-servations of other stars. Current instruments cannotsee surface features of these stars directly, but spectros-copy can deduce the presence of starspots, and ultra-violet and x-ray observations can reveal coronae andflares, which are often much more powerful than theirsolar counterparts. High-resolution spectra from theExtreme Ultraviolet Explorer and the latest x-ray satel-lites, Chandra and XMM-Newton, can probe temper-ature and density. For example, Capella—a stellar sys-tem consisting of two giant stars—has photospherictemperatures like the sun’s but coronal temperaturesthat are six times higher. The intensities of individualspectral lines indicate a plasma density of about 100times that of the solar corona. This high density impliesthat Capella’s coronae are much smaller than the sun’s,stretching out a tenth or less of a stellar diameter. Ap-parently, the distribution of the magnetic field differsfrom star to star. For some stars, tightly orbiting plan-ets might even play a role.

The mystery of why the solar corona should be sohot has intrigued astronomers for more than half a cen-tury, but the reason is now within our grasp, given thelatest findings from spacecraft and fast imaging of thecorona during eclipses. But even as one mystery beginsto yield to our concerted efforts, others appear. The sunand other stars, with their complex layering, magneticfields and effervescent dynamism, still manage to defyour understanding. In an age of such exotica as blackholes and dark matter, even something that seems mun-dane can retain its allure.


JEAN

MO

UE

TTE

AN

D S

ER

GE

KO

UTC

HM

Y, ©

IN

STIT

UT

D’A

STR

OP

HYS

IQU

E D

E P

ARIS

M O R E T O E X P L O R EGuide to the Sun. Kenneth J. H. Phillips. Cambridge University Press, 1992.

The Solar Corona above Polar Coronal Holes as Seen by SUMER on SOHO. Klaus Wilhelm et al.in Astrophysical Journal, Vol. 500, No. 2, pages 1023–1038; June 20, 1998.

Today’s Science of the Sun, Parts 1 and 2. Carolus J. Schrijver and Alan M. Title in Sky &Telescope, Vol. 101, No. 2, pages 34–39; February 2001; and No. 3, pages 34–40; March 2001.

Glorious Eclipses: Their Past, Present and Future. Serge Brunier and Jean-Pierre Luminet.Cambridge University Press, 2001.

Nearest Star: The Exciting Science of Our Sun. Leon Golub and Jay M. Pasachoff. HarvardUniversity Press, 2001.

ORDINARY LIGHT, EXTRAORDINARY SIGHT: the corona photographed in visible light on August 11,1999, from Chadegan in central Iran.


FLYING ROBOTIC INSECT prototype is under development at the Vanderbilt School of Engineering’s Center for IntelligentMechatronics. Such devices will rely on aerodynamics moreakin to that of insects than conventional aircraft.


n a two-ton tank of mineral oil, a pair of mechanical wings flap con-tinuously back and forth, taking a leisurely five seconds to completeeach cycle. Driven by six computer-controlled motors, they set the flu-id swirling, a motion that is revealed by millions of air bubbles im-mersed in the liquid (the tank has a strong resemblance to a giant glassof beer, albeit one with a 60-centimeter-wingspan mechanical flythrashing around in it). Flashing sheets of green laser light illuminatethe scene, and specialized video cameras record the paths of the glis-

tening, churning bubbles. Sensors in the wings record the forces of the fluid acting onthem at each moment.

My research group constructed this odd assortment of specialized equipment tohelp explain the physics of one of the commonest of occurrences—the hovering of atiny fruit fly. The fly knows nothing of the aerodynamics of vortex production, de-layed stall, rotational circulation and wake capture; it merely employs their practi-cal consequences 200 times each second as its wings flap back and forth. The fly’s me-chanical simulacra, dubbed Robofly, imitates the insect’s flapping motion, but at athousandth the speed and on a 100-fold larger scale. Awed by the rapidity and thesmall size of the real thing, my colleagues and I pin our hopes on Robofly for under-standing the intricate aerodynamics that allows insects to do what they do so rou-tinely—that is, how they are able to fly.


Solving the Mystery of

INSECT FLIGHT

INSECTS USE A COMBINATION OFAERODYNAMIC EFFECTS TO REMAIN ALOFT

BY MICHAEL DICKINSONPhotographs by Timothy Archibald

ICopyright 2001 Scientific American, Inc.

As measured by sheer number of spe-cies, ecological impact or total biomass,insects are the dominant animals on ourplanet. Although numerous factors con-tribute to their extraordinary success, theability to fly ranks high on the list. Flightenables insects to disperse from theirbirthplace, search for food over large dis-tances and migrate to warmer climeswith the changing seasons. But flight isnot simply a means of transport—manyinsects use aerial acrobatics to captureprey, defend territories or acquire mates.Selection for ever more elaborate and ef-ficient flight behavior has pushed the de-

sign of these organisms to the limit.Within insects we find the most sensitivenoses, the fastest visual systems and themost powerful muscles—all specializa-tions that are linked one way or anotherto flight behavior. Until recently, howev-er, an embarrassing gap has marred ourunderstanding of insect flight: scientistshave had a difficult time explaining theaerodynamics of how insects generate theforces needed to stay aloft.

That difficulty has even made its wayinto an urban legend of science, typical-ly recounted as “a scientist ‘proved’ thata bumblebee can’t fly” and often cited asan inspiring example for persevering inthe face of overbearing dogma. The bum-blebee story can be traced back to a 1934book by entomologist Antoine Magnan,who refers to a calculation by his assis-tant André Sainte-Laguë, who was an en-gineer. The conclusion was presumablybased on the fact that the maximum pos-sible lift produced by aircraft wings assmall as a bumblebee’s wings and trav-

eling as slowly as a bee in flight would bemuch less than the weight of a bee.

In the decades since 1934, engineersand mathematicians have amassed abody of aerodynamic theory sufficient todesign Boeing 747s and stealth fighters.As sophisticated as these aircraft may be,their design and function are based onsteady-state principles: the flow of airaround the wings and the resulting forcesgenerated by that flow are stable overtime. The reason insects represent such achallenge is that they flap and rotate theirwings from 20 to 600 times a second.The resulting pattern of airflow createsaerodynamic forces that change continu-ally and confound both mathematicaland experimental analyses.

In addition to resolving an old scien-tific puzzle, understanding how insects flymay have practical applications. Recent-ly engineers have begun to explore thepossibility of developing thumb-size fly-ing robots for applications such as searchand rescue, environmental monitoring,surveillance, mine detection and plane-tary exploration. Although humans havesucceeded in constructing model aircraftas small as a bird, no one has built a fly-size airplane that can fly. The viscosity ofair has greater importance at such tinysizes, damping out the kind of airflowsthat keep larger aircraft aloft. Insects flaptheir wings not simply because animalshave never evolved wheels, gears and tur-bines, but because their Lilliputian di-mensions require the use of different aero-dynamic mechanisms. Robotic insects ofthe future may owe their aerodynamicagility to their natural-world analogues.

A BLUR OF WINGSIT IS APPARENT to the casual observ-er that a hovering insect, its wings a blur,does not fly like an aircraft. Much lessobvious is the complexity of the flappingmotion. Insect wings do not merely os-cillate up and down like paddles on sim-ple hinges. Instead the tip of each wingtraces a narrow oval tilted at a steep an-gle. In addition, the wings change orien-tation during each flap: the topside facesup during the downstroke, but then thewing rotates on its axis so that the un-derside faces up during the upstroke.

MICHAEL DICKINSON began his career as aneurobiologist, focusing on the cellular ba-sis of behavior. His interest in flight devel-oped from an investigation of tiny sensorystructures that sense the bending of a wingas it flaps. He now attempts to study behav-ior in a more integrated fashion, by synthe-sizing the tools and analyses of biology,physics and engineering. He is a professorin the department of integrative biology atthe University of California, Berkeley.

THE

AU

THO

R

ROBOFLY FLAPPING SLOWLY in viscous mineral oil simulates the aerodynamics of fruit-fly wings flapping rapidly in air. Laser beams illuminate air bubbles in the oil to reveal the intricate flows produced, and sensors in the wings record the forces generated.



The earliest analyses of insect flighttried to apply conventional steady-stateaerodynamics, the approach that worksfor aircraft wings, to these complex mo-tions. Such attempts are not as naive asthe infamous bumblebee computation,because they take into account the chang-ing velocity of the wings as they flapthrough the air. Imagine freezing the in-sect’s wing at one position in the strokecycle and then testing it in a wind tunnelwith the wind velocity and the wing ori-entation set to mimic the precise move-ment of the wing through the air at thatinstant. In this way, one could measurethe aerodynamic force acting on the wingat each moment.

If this steady-state theory were suffi-cient, the average force, computed byadding up the forces for all the differentwing positions throughout the stroke,should point upward and equal the in-sect’s weight. Even in the late 1970s ex-perts disagreed about whether suchanalysis could explain how insects stayaloft. In the early 1980s Charles Elling-ton of the University of Cambridge care-fully reviewed all available evidence andconcluded that the steady-state approachcould not account for the forces required.The search for dynamic, “unsteady flow”mechanisms that could explain the en-hanced performance of flapping wingstook off with renewed vigor.

The distribution of velocities andpressures within a fluid is governed bythe Navier-Stokes equations, which wereformulated in the early 1800s. (For thepurpose of analyzing aerodynamics, airis simply a very low density fluid.) If wecould solve these equations for a flappinginsect wing, we could fully characterizethe aerodynamics of the insect’s flight.Unfortunately, the complex motion ofthe wing renders this problem excruciat-ingly hard to simulate with even the mostpowerful computers.

If we can’t solve the problem by pure


TETHERED FLY is seen against the backdrop of avirtual-reality arena (top). A computer controls the thousands of green diodes to produce theillusion (for the fly) of objects moving according to the fly’s aerodynamic maneuvers. A similararena is mounted on a gimble (bottom) to simulatethe turns, rolls and yaws of free flight.


theory and computation, can we insteaddirectly measure the forces generated bya flapping insect wing? Several groupshave made informative and valiant ef-forts and are developing imaginative newapproaches, but the delicate size and highspeed of insect wings make force mea-surements difficult.

To circumvent these limitations, biol-ogists studying animal locomotion fre-quently employ scale models—the sametrick used by engineers to design planes,boats and automobiles. Engineers scaletheir vehicles down in size, whereas in-sect-flight researchers enlarge and slowthe wings to a more manageable size andspeed. Such models produce meaningfulaerodynamic results provided they meeta key condition regarding the two forcesthat an object encounters within a fluid: apressure force produced by fluid inertiaand a shear force caused by fluid viscosi-ty. The inertial force is essentially thatneeded to push along a mass of fluid andis larger for denser fluids. Viscosity is morelike friction; produced when adjacent re-gions of fluid move at different velocities,it is what makes molasses hard to stir. Theunderlying physics of the real and themodel animals is identical as long as bothhave the same ratio of inertial to viscousforces, called the Reynolds number.

The Reynolds number increases inproportion to an object’s length and ve-locity and the density of the fluid; it de-creases in proportion to the fluid’s vis-cosity. Being large and fast, aircraft

operate at Reynolds numbers of about amillion to 100 million. Being small andslow, insects operate at Reynolds num-bers of around 100 to 1,000 and under100 for the tiniest insects, such as thrips,which are a common garden pest.

DELAYED STALLTO GAIN SOME INSIGHT into how aflapping fruit-fly wing generates aerody-namic force, in 1992 Karl Götz and I,both then at the Max Planck Institute forBiological Cybernetics in Tübingen, Ger-many, built a model wing consisting of afive-by-20-centimeter paddle connectedto a series of motors that moved it with-in a large tank of thick sugar syrup. Thatcombination of increased size and viscos-ity, and the slower flapping rate, resultedin the same Reynolds number, and thusthe same physics, as a fruit-fly wing thatis flapping in air.

We equipped the wing with a forcesensor to measure the lift and drag gen-erated as it moved through the sticky flu-id. We put baffles at the ends of the wingto inhibit flow along the length andaround the edge of the wing. Simple aero-dynamic models often use this technique:it effectively reduces the flow from threedimensions to two, which makes theanalysis easier but at the risk of missingimportant effects.

Our experiments with this modelwing and work in other laboratorieshelped to uncover one possible solutionto the conundrum of insect flight: de-


ILLU

STR

ATIO

NS

BY

BR

YAN

CH

RIS

TIE

FRUIT FLY USES three differentaerodynamic mechanisms to support its

weight in the air. During much of the wing stroke (1), a leading-edge vortex

forms and increases lift, a process called delayed stall because the vortex

does not have time to detach, which is what happens when an aircraft stalls. At the end of a stroke (2, 3, 4), the wing

rotates, which produces rotational lift analogous to a tennis ball hit with

backspin. At the start of the upstroke (5),the wing passes back through the wake

of the downstroke. The wing is oriented sothat this increased airflow adds further

lift, a process called wake capture.

AIRCRAFT WING generates lift bysteady-flow aerodynamics (top).

Smooth flow over the top of the wing isfaster than that under the wing,

producing a region of low pressure andan upward force. If the angle of attack is too great (bottom), the wing stalls.

When a stall begins, a leading-edgevortex forms with a high flow velocity

that momentarily increases lift. Thevortex quickly detaches from the wing,

however, greatly reducing lift.

STARTING VORTEX

WING STROKE

LIFTHIGH VELOCITY

LOW VELOCITY

LEADING-EDGEVORTEX

LIFT

WING ANGLE

12 3

45

DELAYED STALL ROTATIONAL LIFT

LIFT

21LEADING-EDGE VORTEX

WING ROTATION


layed stall. In an aircraft, stall occurs ifthe angle that the wing cuts through theair—the angle of attack—is too steep. Atshallow angles of attack, the air splits atthe front of the wing and flows smooth-ly in two streams along the upper andlower surfaces. The upper flow travelsfaster, resulting in a lower pressure abovethe wing, which sucks the wing upward,producing lift. When the angle of attackis too steep, however, the upper flow can-not follow the contour of the upper sur-face and separates from the wing, result-ing in a catastrophic loss of lift.

How can stall, which is disastrous foran airplane, help to lift an insect? The an-swer lies in the rate at which the wingsflap. Wings do not stall instantly; it takessome time for the lift-generating flow tobreak down after the angle of attack in-creases. The initial stage of stall actuallybriefly increases the lift because of a short-lived flow structure called a leading-edge

vortex. A vortex is a rotating flow of flu-id, as occurs in tornadoes or the littlewhirlpool in a draining bathtub.

The leading-edge vortex forms justabove and behind the wing’s leadingedge, like a long cylindrical whirlpoolturned on its side. The airflow in the vor-tex is very fast, and the resulting very lowpressure adds substantial lift. This effectwas first recognized by aeronautics engi-neers in England in the early 1930s, but itis too brief to be of use to most aircraft.Very quickly, the vortex detaches fromthe wing and is shed into the aircraft’swake, and lift drops precipitously, as doesthe plane. The wing strokes of insects,however, are so brief that the wing flipsover and reverses direction, producing anew vortex in the opposite direction im-mediately after the previous one is shed.

These results, obtained with simplifiedtwo-dimensional models, were extendedto three dimensions in the mid-1990s byEllington and his co-workers at Cam-bridge. His group studied the large hawk-moth, Manduca sexta, flying tethered ina wind tunnel, as well as a fully three-di-mensional robotic moth. Lines of smoke(called a smoke rake) revealed that a vor-tex is indeed attached to the wings’ lead-ing edges during the downstroke. Elling-ton’s team suggested that an axial flow ofair from base to tip of the wings en-hanced the effect by reducing the strengthof the vortex but increasing its stabilityand allowing it to remain attached to thewings throughout the stroke. Such axial

Per second, flying expends about 10 times more energy thanlocomotion on the ground. On the other hand, per kilometertraveled, flying is four times moreenergy-efficient than groundlocomotion. Thus, flying is veryhard to achieve but has greatvalue for organisms that can do it.

The first animals to evolve active flight were insects.

Most insects have two pairs ofwings. The hind wings of flies,however, have evolved into tinysensory organs that function asgyroscopes, monitoring theorientation of the fly’s body.


WAKE FROM PREVIOUS

STROKE

3 4WAKE CAPTURE5

AIRFLOW AROUND a tennis ball that is hit withbackspin generates rotational lift. Insects use thesame phenomenon by rotating their wings at the end of each stroke.


flow might be especially important forlarge insects such as hawkmoths anddragonflies that flap their wings over agreat distance during each stroke.

Although identifying this effect solveda major piece of the puzzle, various linesof evidence suggested that insects har-nessed other mechanisms in addition todelayed stall. First, the extra force pro-duced by delayed stall is enough to ex-plain how an insect remains airborne butinsufficient to explain how many insectscan lift almost twice their body weight.Second, several investigators have at-tempted to measure the forces an insectcreates by tethering it to a sensitive forcetransducer. Such experiments must beviewed cautiously, because tethered ani-mals may not behave identically to freelyflying animals, but the precise timing ofthe forces is not easily explained by de-layed stall. For example, when Götz useda laser diffraction technique to measurethe forces generated by a fruit fly, hefound that the greatest forces occurredduring the upstroke—at a time when theforces resulting from delayed stall are ex-pected to be weak.

To search for additional unsteady

mechanisms, in 1998 Fritz-Olaf Lehmann,Sanjay P. Sane and I constructed a largemodel of a flapping fruit fly, Drosophilamelanogaster—the Robofly describedearlier. The viscous mineral oil within thetank makes the 25-centimeter robot wingsflapping once every five seconds dynam-ically similar to 2.5-millimeter fruit-flywings flapping 200 times a second in air.We measured two critical properties—

the aerodynamic forces on the wings andthe fluid flow around them—that arenearly impossible to determine on real flywings. Although Robofly is designed tomimic a fruit fly, by programming the sixmotors that drive the two wings, we canre-create the wing motion of numerousinsect species. In addition, we can makeRobofly flap its wings in any way re-quired to test specific hypotheses—a lux-ury not afforded by real animals, whichtend to get temperamental under labora-tory conditions.

ROBOFLY’S RESULTSWHEN ROBOFLY FLAPPED like a fruitfly, we measured a curious pattern offorces. The wings generated momentarystrong forces at the beginning and end ofeach stroke that could not be easily ex-plained by delayed stall. These force peaksoccurred during stroke reversal, when thewing slows down and rapidly rotates,suggesting that the rotation itself might beresponsible.

Rotating objects moving through theair produce flows similar to those that lifta conventional wing. A tennis ball hitwith backspin pulls air faster over thetop, causing the ball to rise. Conversely,topspin pulls air faster underneath, push-ing the ball down. A flat wing is differentfrom a spherical ball, but rotation of awing should produce some lift by thesame general mechanism.

We tested our hypothesis by modify-ing the precise moment in the stroke cy-cle when the wing flips. If a wing rotatesat the end of one stroke, as in a normal flystroke, the wing’s leading edge rotatesbackward relative to the direction in


BLOWFLY IS WIRED for studies that relate theelectrical activity in steering muscles to changes inwing motion that occur during steering maneuvers.

Air is more kinematicallyviscous than water: its ratio ofviscosity to density is higher.That ratio is what matters for fluid dynamics.

Flight muscle of insectsexhibits the highest-knownmetabolic rate of any tissue.

Insects possess the mostdiverse wing structure andkinematics of all flying animals.


which the wing is moving and the wingshould develop some upward force—

analogous to a tennis ball hit with back-spin. If the wing rotates late, at the begin-ning of the next stroke, the leading edgemoves forward relative to the direction ofmotion, and the wing will develop a down-ward force analogous to topspin. Robo-fly’s data were in complete agreementwith these expectations, indicating thatflapping wings develop significant lift byrotational circulation.

There remained, however, anothersignificant force peak in Robofly’s data,occurring at the start of each downstrokeand upstroke, that rotational circulationcould not easily explain. Several sets ofexperiments indicated that this peak wascaused by a phenomenon called wakecapture—the collision of the wing withthe swirling wake of the previous stroke.

Each stroke of the wing leaves behinda complicated wake consisting of the vor-ticity it produced by traveling and rotat-ing through the fluid. When the wing re-verses direction, it passes back throughthis churning air. A wake contains energylost from the insect to the fluid, so wakecapture provides a way for the insect torecover some of that energy—to recycle it,one might say. We tested the wake-cap-ture hypothesis by bringing Robofly’swings to a complete stop after flappingback and forth. The stationary wings con-tinued to generate force because the fluidaround them was still moving.

Although wake capture must alwaysoccur at the start of each stroke, as withrotational circulation the fly can manip-ulate the size and direction of the forceproduced by changing the timing of wingrotation. If the wing rotates early, it al-ready has a favorable angle of attackwhen it collides with the wake, producinga strong upward force. If the wing rotateslate, collision with the wake generates adownward force.

Together wake capture and rotationalcirculation also help to explain the aero-

dynamics of flight control—how flies steer.Flies are observed to adjust the timing ofwing rotation when they turn. In somemaneuvers, the wing on the outside of aturn rotates early, producing more lift,and the wing on the inside of a turn rotateslate, generating less lift; the net force tiltsand turns the fly in the desired direction.The fly has at its disposal an array of so-phisticated sensors, including eyes, tinyhind wings that are used as gyroscopes,and a battery of mechanosensory struc-tures on the wings that it can use to pre-cisely tune rotational timing, stroke am-plitude and other aspects of wing motion.

BRIDE OF ROBOFLYTHE WORK of numerous researchers isbeginning to coalesce into a coherent the-ory of insect flight, but many questions re-main. Insects have a vast array of bodyforms, sizes and behaviors, ranging fromtiny thrips to large hawkmoths; fromtwo-winged flies such as fruit flies tolacewings that flap two pairs of wingsslightly out of sync and tiger beetles thathave two large stationary wings (their ely-tra, which form their carapace when onthe ground) in addition to the two wings

that flap. To what extent do the results forfruit flies apply to these myriad cases?

Also, the studies so far have focusedon hovering flight, which is the hardestcase to explain because the insect can gainno benefit from onrushing air. But do in-sects use other significant mechanisms toproduce lift when they are moving? Manyresearchers are preparing to study thesechallenging questions. My group, for ex-ample, is building “bride of Robofly,”which will live in a tank large enough forit to fly forward and make turns, to test,for instance, our hypotheses about howflies make their characteristic remarkablysharp turns by adjusting the timing oftheir wing strokes. After uncovering thebasic set of tricks that insects use to stayin the air, the real fun now begins.


M O R E T O E X P L O R EThe Biomechanics of Insect Flight: Form,Function, Evolution. Robert Dudley. PrincetonUniversity Press, 2000.

The Web site of the author’s research group isavailable at http://socrates.berkeley.edu/~flymanmd/

An account of the origins of the bumblebee mythis online at www.math.niu.edu/~rusin/known-math/98/bees

PROTOTYPE MICROMECHANICAL flying insect isbeing developed by the Robotics and IntelligentMachines Laboratory at the University of Californiaat Berkeley. The design parameters are based on the blowfly Calliphora.



NE OF THE GREAT MYSTERIES of thehuman brain is how it understands andproduces language. Until recently, most ofthe research on this subject had been basedon the study of spoken languages: English,French, German and the like. Starting inthe mid-19th century, scientists made largestrides in identifying the regions of thebrain involved in speech. For example, in1861 French neurologist Paul Broca dis-

covered that patients who could understand spoken languagebut had difficulty speaking tended to have damage to a part ofthe brain’s left hemisphere that became known as Broca’s area.And in 1874 German physician Carl Wernicke found that pa-tients with fluent speech but severe comprehension problemstypically had damage to another part of the left hemisphere,which was dubbed Wernicke’s area.

Similar damage to the brain’s right hemisphere only veryrarely results in such language disruptions, which are calledaphasias. Instead right hemisphere damage is more often asso-ciated with severe visual-spatial problems, such as the inabili-

ty to copy a simple line drawing. For these reasons, the lefthemisphere is often branded the verbal hemisphere and theright hemisphere the spatial hemisphere. Although this di-chotomy is an oversimplification, it does capture some of themain clinical differences between individuals with damage tothe left side of the brain and those with damage to the right.

But many puzzles remain. One that has been particularlyhard to crack is why language sets up shop where it does. Thelocations of Wernicke’s and Broca’s areas seem to make sense:Wernicke’s area, involved in speech comprehension, is locatednear the auditory cortex, the part of the brain that receives sig-nals from the ears. Broca’s area, involved in speech production,is located next to the part of the motor cortex that controls themuscles of the mouth and lips [see illustration on page 60]. Butis the brain’s organization for language truly based on the func-tions of hearing and speaking?

One way to explore this question is to study a language thatuses different sensory and motor channels. Reading and writ-ing, of course, employ vision for comprehension and handmovements for expression, but for most people these activitiesdepend, at least in part, on brain systems involved in the use of

How does the human brainprocess language? New studies of deaf signershint at an answer

OCopyright 2001 Scientific American, Inc.

BRAINin the

SIGN

Illustrations by Peter Stemler

TRANSLATION of the phrase “Sign Language in the Brain” into American Sign Languageis shown in these artist’s renderings, which are based on photographs of a deaf signer.

languageby Gregory Hickok, Ursula Bellugiand Edward S. Klima


a spoken language. The sign languages of the deaf, however, pre-cisely fit the bill. Over the past two decades, we have examinedgroups of deaf signers who have suffered damage to either theright or the left hemisphere of their brains, mostly as a result ofstrokes. By evaluating their proficiency at understanding andproducing signs, we set out to determine whether the brain re-gions that interpret and generate sign language are the same onesinvolved in spoken language. The surprising results have illu-minated the workings of the human brain and may help neu-rologists treat the ills of their deaf patients.

The Signs of Language

MANY PEOPLE MISTAKENLY BELIEVE that sign language isjust a loose collection of pantomime-like gestures throwntogether willy-nilly to allow rudimentary communica-

tion. But in truth, sign languages are highly structured linguis-tic systems with all the grammatical complexity of spoken lan-guages. Just as English and Italian have elaborate rules forforming words and sentences, sign languages have rules for in-dividual signs and signed sentences. Contrary to another com-mon misconception, there is no universal sign language. Deafpeople in different countries use very different sign languages.In fact, a deaf signer who acquires a second sign language as anadult will actually sign with a foreign accent! Moreover, signlanguages are not simply manual versions of the spoken lan-guages used in their surrounding communities. American SignLanguage (ASL) and British Sign Language, for example, aremutually incomprehensible.

Sign and spoken languages share the abstract properties oflanguage but differ radically in their outward form. Spoken lan-

guages are encoded in acoustic-temporal changes—variations insound over time. Sign languages, however, rely on visual-spatialchanges to signal linguistic contrasts [see box on opposite page].How does this difference in form affect the neural organizationof language? One might hypothesize that sign language wouldbe supported by systems in the brain’s right hemisphere becausesigns are visual-spatial signals. Accordingly, one could contendthat the sign-language analogue of Wernicke’s area in deaf sign-ers would be near the brain regions associated with visual pro-cessing and that the analogue of Broca’s area would be near themotor cortex controlling hand and arm movements.

When we began to test this hypothesis in the 1980s, twofundamental questions needed to be answered: Did deaf sign-ers with brain damage have sign-language deficits? And if so,did the deficits resemble either Wernicke’s aphasia (compre-hension problems and error-prone speech) or Broca’s aphasia(good comprehension but difficulty in producing fluent speech)?The answer to both questions was a resounding yes. One of thefirst patients studied by our group signed fluently, using all theproper grammatical markers of ASL, but the message conveyedby his signing was often incoherent. An English gloss of one ofhis utterances reads:

And there’s one (way down at the end) [unintelligible]. The manwalked over to see the (disconnected), an extension of the (earth)room. It’s there for the man (can live) a roof and light with shadesto (keep pulling down).

The patient’s disorganized signing and apparent lack of com-prehension of others’ signs were very similar to the symptomsof hearing patients with Wernicke’s aphasia. Another deaf pa-tient we studied early in the research program had extreme dif-ficulty producing signs. She had to struggle to shape and orienther hands to perform the proper movement for virtually everysign she attempted. Most of her utterances were limited to iso-lated signs. This was not merely a motor control problem: whenasked to copy line drawings of objects such as an elephant or aflower, she did so accurately. Also, in contrast to her severesign-language production problems, her comprehension of signlanguage was excellent. This profile of language abilities par-allels the symptoms of Broca’s aphasia.

But where was the brain damage that caused these signaphasias? The answer was surprising. Both patients had lesionsin their left hemispheres. And the lesions were located justabout where you’d expect to find them in hearing patients with

GREGORY HICKOK, URSULA BELLUGI and EDWARD S. KLIMA haveworked together on sign language aphasia for a decade. Hickok is as-sociate professor in the department of cognitive sciences at the Uni-versity of California, Irvine, and director of the Laboratory for Cogni-tive Brain Research, where he studies the functional anatomy oflanguage. Bellugi is director of the Salk Institute’s Laboratory for Cog-nitive Neuroscience in La Jolla, Calif., where she conducts researchon language and its biological foundations. Much of her research is incollaboration with Klima, who is professor emeritus at the Universityof California, San Diego, and a research scientist at Salk.

THE

AU

THO

RS

T WO OF THE REGIONS of the brain’s left hemisphere that playimportant roles in language processing are Broca’s area and Wer-nicke’s area (there are several others). Broca’s area is activatedin hearing individuals when they are speaking and in deaf peoplewhen they are signing. Wernicke’s area is involved in the compre-hension of both speech and signs.

Where Language Lives

WERNICKE’S AREAAUDITORY CORTEX

BROCA’S AREA

MOTOR CORTEX



S ign languages, like spoken languages, haveseveral kinds of linguistic structure, in-cluding phonological, morphological and

syntactic levels. At the phonological level, signsare made up of a small set of components, just asspoken words are composed of a small set of con-sonants and vowels. The components of signs in-clude hand shapes, the locations around the bodywhere signs are made, the movements of thehands and arms, and the orientation of the hands(for example, palm up versus palm down). InAmerican Sign Language (ASL) the signs for “sum-mer,” “ugly” and “dry” have the same hand shape,movement and orientation but differ in location[see illustrations at left]. Likewise, signs such as“train,” “tape” and “chair” share hand shape, ori-entation and location but differ in movement.

At the morphological level, ASL has grammat-ical markers that systematically change themeaning of signs. Morphological markers in En-glish include fragments like “-ed,” which can beadded to most verbs to indicate past tense (“walk”becomes “walked”). Whereas in English the mark-ers are added to the beginning or end of a word, inASL the signs are modified using distinctive spa-tial patterns. For example, adding a rolling move-ment to the sign “give” (and to most ASL verbsigns) changes the sign’s meaning to “give con-tinuously.” Signers can use different patterns tomodify the verb to mean “give to all,” “give to each,”“give to each other” and many other variations.

At the syntactic level, ASL specifies the gram-matical relations among signs (that is, who is do-ing what to whom) in ways that do not occur inspoken languages. In English the order of thewords provides the primary cue for the syntacticorganization of a sentence such as “Mary criti-cized John.” Reverse the order of the nouns, andyou reverse the meaning of the sentence. Signersof ASL can use word-order cues as well, but theyneed not. Instead signers can point to a distinctposition in space while signing a noun, thus link-ing the word with that position. Then the signercan move the verb sign from Mary’s position toJohn’s to mean “Mary criticized John” and in theother direction to mean the reverse.

“SUMMER”1 2

2

The Building Blocks of Sign Language

“DRY”1

“UGLY”

2

1

LOCATION of a sign is a critical element inconveying meaning. In American Sign Language,“summer” is articulated near the forehead, “ugly” near the nose and “dry” near the chin.



similar problems. The deaf signer withcomprehension difficulties had damagethat included Wernicke’s area, whereasthe patient who had trouble making signshad damage that involved Broca’s area.

These observations showed that theleft hemisphere plays a crucial role in sup-porting sign language. But what about theright hemisphere? One would think thatdamage to the right hemisphere, whichappears to be critically involved in manyvisual-spatial functions, would have adevastating effect on sign-language abili-ty as well. But this assumption is appar-ently wrong. Signers with damage to theright hemisphere were fluent and accuratein their production of signs, used normalgrammar and comprehended signs withease. This held true even in patients whosenonlinguistic visual-spatial abilities hadbeen severely compromised by their braindamage. One signer with damage to theright hemisphere , for example, could notcreate or copy recognizable drawings andfailed to notice objects in the left part ofhis visual field (a condition known ashemispatial neglect). Yet he could com-municate very efficiently in sign language.

Subsequent research using largergroups of deaf signers confirmed the ear-ly case studies. A study published by ourteam in 1996 compared the sign-languageabilities of 13 left hemisphere–damaged(LHD) signers with those of 10 righthemisphere–damaged (RHD) signers. Asa group, the LHD signers performedpoorly across a wide range of sign-lan-guage measures: They had trouble com-prehending isolated signs and signed sen-tences and were likely to have problemswith fluency as well. They also had diffi-culty with picture-naming tasks and fre-quently made paraphasic errors—slips ofthe hand—in which they inadvertentlysubstituted one sign for another or onecomponent of a sign, such as hand shape,for another. In contrast, the RHD signersperformed well on all these tasks. Thestudy also showed that difficulties withsign-language fluency were not caused bymore general problems in controllingvoluntary hand or arm movements: pa-tients who had trouble making signswere often capable of producing non-meaningful hand and arm gestures.

We obtained similar results in anoth-

er study, this one focusing on sign-lan-guage comprehension in 19 lifelong sign-ers with brain lesions, 11 with damage tothe left hemisphere and eight with dam-age to the right. The LHD group per-formed significantly worse than the RHDgroup on three tests that evaluated theirunderstanding of single signs, simple sen-tences and complex sentences. The mostimpaired signers were those with damageto the brain’s left temporal lobe, whereWernicke’s area is located.

Taken together, these findings suggestthat the brain’s left hemisphere is domi-nant for sign language, just as it is forspeech. The organization of the brain forlanguage does not appear to be particu-larly affected by the way in which lan-guage is perceived and produced.

The Story Gets Complicated

AS WE NOTED at the beginning ofthis article, the assumed left-rightdichotomy of the brain—with ver-

bal abilities concentrated in the left hemi-sphere and visual-spatial abilities clusteredin the right—is an oversimplification. Re-search over the past few decades has

COMMON PROBLEM experienced by left hemisphere–damaged (LHD) deaf signers is the production of para-phasias— slips of the hand— analogous to the slips of thetongue experienced by LHD hearing patients. The illustra-tion at the right shows the correct form of the sign for “fine,”whereas the drawing on the opposite page shows an erroroften made by LHD signers. In the latter figure, the signer ar-ticulated the location and movement of the sign correctlybut used the wrong hand shape, resulting in something thathas no meaning in ASL— a nonsense sign, equivalent to“bline” or “gine” in English.

Although the hand shape in this paraphasia is incorrectfor “fine,” it is used in many other ASL signs, such as “play”and “California.” Similar paraphasias include errors in pro-ducing the proper location, movement and hand orientationof a sign, as well as mistakes in rendering the morphologicaland syntactic structure of the language.

The brain’s left hemisphere is dominant for sign language, just as it is for speech.

CORRECT SIGNFOR “FINE”


S C I E N T I F I C A M E R I C A N 63

shown that most cognitive abilities can bedivided into multiple processing steps. Atsome levels, brain activity may be lateral-ized (taking place primarily in one hemi-sphere), whereas at others the activity maybe bilateral (occurring in both).

Language ability, for instance, hasmany components. A hearing person mustbe able to perceive and produce individualspeech sounds and the words they makeup; otherwise, one could not distinguish“cup” from “pup.” In addition, one mustbe able to recognize morphological addi-tions (“walking” vs. “walked”), syntacticconstructions (“the dog chased the cat” vs.“the dog was chased by the cat”), andmelodic intonations (“the White House”vs. “the white house”). Finally, to conductan extended discourse one must be able toestablish and maintain a coherent con-nection between characters and eventsover the course of many sentences.

Of all these aspects of linguistic abili-ty, the production of language is the onemost sharply restricted to the brain’s lefthemisphere. Damage to the left hemi-sphere often interferes with the ability toselect and assemble appropriate soundsand words when speaking. Right hemi-

sphere damage rarely does. One excep-tion to the left hemisphere’s monopolyon language production is the creation ofa coherent discourse. Patients with righthemisphere damage may be able to con-struct words and sentences quite well,but they frequently ramble from one sub-ject to the next with only a loose threadof a connection between topics.

The perception and comprehensionof language appear to be less confined tothe left hemisphere than language pro-duction is. Both hemispheres are capableof distinguishing individual speech sounds,and the right hemisphere seems to have arole in the comprehension of extendeddiscourse. But deciphering the meaningof words and sentences seems to takeplace primarily in the left hemisphere.This may explain why language wasoriginally considered to be the exclusiveprovince of the left hemisphere: the mostcommon tests for aphasia evaluated thecomprehension and production of wordsand sentences, not longer discourses.

Nonlinguistic spatial abilities can alsobe broken down into components withdiffering patterns of lateralization. Al-though the most severe impairments of

spatial abilities occur more commonlyfollowing damage to the right hemisphere(both in deaf and hearing populations),researchers have observed some visual-spatial deficits in LHD hearing people.The symptoms typically involve difficul-ties in perceiving or reproducing the lo-cal-level features of a visual stimulus—

such as the details in a drawing—eventhough the LHD patients can correctlyidentify or reproduce the drawing’s over-all configuration. RHD hearing peopletend to show the opposite pattern. Thus,it has been suggested that the left hemi-sphere is important for local-level spatialperception and manipulation, whereasthe right hemisphere is important forglobal-level processes.

This more sophisticated picture of thebrain raises an interesting question: Is thedivision of visual-spatial abilities betweenthe two hemispheres—local level in theleft, global level in the right—related tothe division of sign-language abilities? In-dividual signs and signed sentences can bethought of as pieces of the language,whereas an extended discourse can rep-resent how those pieces are put together.Perhaps the left hemisphere is dominantfor producing and comprehending signsand signed sentences because those pro-cesses are dependent on local-level spatialabilities. And perhaps the right hemi-sphere is dominant for establishing andmaintaining a coherent discourse in signlanguage because those processes are de-pendent on global-level spatial abilities.

We set out to test this hypothesis. Ourresearch confirmed that many RHD sign-ers have trouble with extended discourse:their narratives are full of tangential ut-terances and even confabulations—justthe kind of difficulties that hearing RHDpatients often have. But some RHD sign-ers face another type of problem. Dis-course in sign language has a unique spa-tial organization: when telling a storywith many characters, the signer identi-fies each one using a different location.The space in front of the signer becomesa sort of virtual stage on which eachcharacter has his or her own spot. Ourstudies found that some RHD signerswere able to stay with a topic in their dis-course but failed to maintain a consistent

INCORRECT SIGN FOR “FINE” TYPICALLY PRODUCEDBY A SIGNER WITH LEFT HEMISPHERE DAMAGE


spatial framework for the characters intheir narratives.

Is either of these types of discourseproblems in RHD deaf signers causallyconnected to deficits in their nonlinguis-tic spatial abilities? It would appear not.We studied one RHD signer whose spa-tial abilities were severely impaired yetwho had no trouble signing a coherentstory. Another RHD patient had onlymild visual-spatial problems yet couldnot sustain a proper spatial frameworkfor the characters in the narrative. Clear-ly, the cognitive systems in the righthemisphere that support nonlinguisticspatial abilities are different from theones that support extended discourse.

What about deaf signers with damageto the left hemisphere? Are their sign-lan-guage aphasias caused by impairments inlocal-level spatial abilities? To address thisissue, we asked a group of deaf signers toreproduce line drawings and hierarchicalfigures, which have recognizable local andglobal features. (An example would bethe letter “D” fashioned out of a constel-lation of small “y”s.) Just like hearing pa-

tients with left hemisphere damage, theLHD deaf subjects tended to reproducethe global configuration of the drawingscorrectly but often left out some of the de-tails. (The RHD deaf subjects exhibitedthe reverse pattern, drawing pictures withlots of detail but a disorganized whole.)We found no correlation between theseverity of the local-level spatial deficits inthe LHD subjects and the severity of theirsign-language aphasias. Contrary to allexpectations, the sign-language abilitiesof lifelong deaf signers appear to be inde-pendent of their nonlinguistic spatial skills.

It is possible that we have missed somefine distinctions in the organization of thebrain for language in hearing patientsand signers. Studies of patients with brainlesions are limited in their precision: toascertain exactly which parts of the brainare involved in sign language, researcherswould need to examine dozens of deafsigners with lesions in just the right plac-es, and it would take decades to find themall. But the introduction of noninvasivebrain imaging techniques—functionalmagnetic resonance imaging (fMRI) and

positron-emission tomography (PET)—has given scientists new tools for probingthe neural roots of language.

Researchers have employed these tech-niques to investigate the role of Broca’sarea in speech and sign production. Imag-ing results have shown that Broca’s areais indeed activated in hearing patientswhen they are speaking and in deaf pa-tients when they are signing. Brain imag-ing has also confirmed that the regionsthat play a role in sign-language compre-hension are much the same as those in-volved in the understanding of spokenlanguage. In a 1998 study researchers usedfMRI methods to observe the brain ac-tivity of lifelong deaf signers who werewatching videotapes of sentences in ASL.The investigators found regions of activ-ity in several parts of the left temporallobe, including parts of Wernicke’s area,and in several regions of the left frontallobe, including Broca’s area.

The study also found regions of ac-tivity in the right temporal lobe and rightfrontal lobe. This result has led some re-searchers to suggest that sign-language

The brain is a highly modular organ, with each module organized around

a particular computational task.


SEQUENCE OF DRAWINGS below shows the correct maintenance of aspatial framework for an extended discourse in American Sign Language.The signer is describing a series of pictures that show two children paint-ing each other’s faces as they sit side by side at a table. At the start of thediscourse, the signer linked each child to a particular location in space:

Alice on the signer’s right and Bob on the signer’s left ( not shown). Sub-tle shifts in the signer’s body position and the direction of the movementof the sign for “paint” (from Alice’s location on her right to Bob’s locationon her left) indicate that Alice is painting Bob ( a, b). The reverse move-ments ( c, d) indicate that Bob is painting Alice.

a b c d


comprehension may be more bilaterallyorganized than spoken-language com-prehension. But bilateral activity has al-so been detected in studies of hearingsubjects listening to speech. More re-search is needed to clarify the role of theright hemisphere in sign-language pro-cessing. In any case, the studies of brainlesions make it clear that if differences ex-ist between spoken and sign language,they are likely to be subtle.

Lessons from Sign Language

S IGN LANGUAGE involves bothlinguistic and visual-spatial pro-cessing—two abilities that are sup-

ported by largely distinct neural systemsin hearing individuals. But contrary to allexpectations, the neural organization ofsign language has more in common withthat of spoken language than it does withthe brain organization for visual-spatialprocessing. Why should this be the case?

The answer suggested by our line ofresearch, as well as the work of others, isthat the brain is a highly modular organ,with each module organized around aparticular computational task. Accord-ing to this view, the processing of visual-spatial information is not confined to asingle region of the brain. Instead differ-ent neural modules process visual inputsin different ways. For example, visual in-puts that carry linguistic informationwould be translated into a format opti-mized for linguistic processing, allowingthe brain to access the meanings of signs,extract grammatical relations, and so on.But visual stimuli that carry a different

kind of information—such as the featuresand contours of a drawing—would betranslated into a format that is optimizedfor, say, carrying out motor commandsto reproduce that drawing. The compu-tational demands of these two kinds ofprocessing tasks are very different, andthus different neural systems are involved.

Viewed in this way, it is not so sur-prising that comprehending and produc-ing sign language appear to be indepen-dent of visual-spatial abilities such ascopying a drawing. Although they bothinvolve visual inputs and manual out-puts, the tasks are fundamentally differ-ent. Consequently, we would expect themto share brain systems to some extent atthe peripheral levels of processing—forinstance, at the primary visual cortex thatreceives signals from the optic nerve—butto diverge in more central, higher-levelbrain systems.

The situation with spoken and signlanguages is just the opposite. These twosystems differ radically in their inputs andoutputs but appear to involve very similarlinguistic computations. We therefore ex-pect that spoken and sign languages willshare a great deal of neural territory at themore central, higher-level brain systemsbut diverge at the more peripheral levels of

processing. At the sensory end, for exam-ple, the peripheral processing of speech oc-curs in the auditory cortices in both hemi-spheres, whereas the initial processing ofsigns takes place in the visual cortex. Butafter the first stages of processing, the sig-nals appear to be routed to central lin-guistic systems that have a common neur-al organization in speakers and signers.

These findings may prove useful toneurologists treating deaf signers whohave suffered strokes. The prognosis forthe recovery of the signers’ language abil-ities will most likely be similar to that ofhearing patients with the same braindamage. Furthermore, when neurosur-geons remove brain tumors from deafsigners, they must take the same precau-tions to avoid damaging the languagecenters as they do with hearing patients.

A major challenge for future researchwill be to determine where the peripher-al processing stages leave off and the cen-tral stages begin (or even if there is sucha sharp boundary between the two). Morestudy is also needed to understand the na-ture of the computations carried out at thevarious levels of linguistic processing. Thesimilarities and differences between spo-ken and sign languages are ideally suitedto answering these questions.

MORE TO EXPLOREThe Signs of Language. Edward S. Klima and Ursula Bellugi. Harvard University Press, 1979. Reprinted 1988.

What the Hands Reveal about the Brain. H. Poizner, Edward S. Klima and Ursula Bellugi. MIT Press, 1987.Reprinted by Bradford Books, 1990.

The Neural Organization of Language: Evidence from Sign Language Aphasia. G. Hickok, U. Bellugi and E. S.Klima in Trends in Cognitive Sciences, Vol. 2, No. 4, pages 129–136; April 1998.

The Signs of Aphasia. G. Hickok and U. Bellugi in Handbook of Neuropsychology, Vol. 3. Second edition. Edited byR. S. Berndt. Elsevier, 2001.


e f g h

MANY SIGNERS with right hemisphere damage make mistakes in theirspatial organization of a discourse. They can correctly link the charactersin the narrative to positions in space, but they often fail to reference thesepositions consistently. In the drawings above, the signer does not link

the sign for “paint” to the positions of Alice and Bob. An English equiva-lent of this lack of specificity might be: “Alice and Bob were sitting at atable, painting. Suddenly someone painted on someone’s face (e, f ), andthen someone painted on someone’s face (g, h).”


MARS!ALIEN LANDSCAPEof Canada’s Devon Island is the setting for a simulated Mars base.

NORTH TO


EVON ISL AND IS A BARRENbut weirdly beautiful place in theCanadian Arctic, only 1,500 kilo-meters from the North Pole.About 23 million years ago a me-teorite struck there with the force

of 100,000 hydrogen bombs, gouging out the20-kilometer-wide Haughton Crater. The im-pact killed every living thing on the island anddestroyed its soil, leaving a bizarre landscape ofcondensed powder ridges and shock-fracturedstones. Because most of the island is devoid oftrees, bushes and grasses, it looks and feels likean alien world. In fact, for the past four years ateam of scientists from the National Aeronau-tics and Space Administration has been study-ing the island’s geology to learn about Mars[see box on next page].

In 1998 I founded the Mars Society, an or-ganization dedicated to promoting the explo-ration of the Red Planet. Pascal Lee, the geolo-gist who headed NASA’s Devon Island team,proposed that the society build a simulatedMars base at Haughton Crater. Researchers atthis station could explore Devon under manyof the same conditions and constraints thatwould be involved in a human mission toMars. The things we could learn from such aprogram would be invaluable. For example,the crew at the Devon Island station could tryout equipment intended for a Mars mission,such as water-recycling systems and spacesuits,putting the gear through months of roughtreatment in the field instead of merely testingit in the laboratory. While studying Devon Is-land’s geology and microbiology, the researchers


DTO PAVE THE WAY FOR A MISSION TOMARS, A BAND OF SCIENTISTS DECIDED TO GO TO THE CANADIAN ARCTICby Robert Zubrin


could find out how much water an active Mars explorationteam would really need. They could also investigate key issuesof crew psychology, such as the effects of isolation and closeconfinement, team dynamics and how the layout of the stationwould affect the astronauts’ performance.

In addition, the Devon Island crew could determine whichoperational procedures would work well on the Red Planetand which would simply hamstring future astronauts. Andthey could develop the art of combined operations—that is,coordinating the mission’s human, robotic and vehicular ele-ments to maximize their effectiveness. In short, the Devon Is-land crew could learn how to live and work on Mars.

The Mars Society raised $600,000 in private funds for theproject and in January 2000 contracted the manufacturing ofthe station to Infrastructure Composites International (Infra-comp). The company had invented a honeycomb-like fiber-glass that was originally intended for building floating citieson the sea. This ultrastrong and relatively lightweight mater-ial proved perfect for our Mars habitat. The shape of the struc-ture—a domed cylinder with a diameter of about eight meters,containing two decks with about three meters of headroomeach—mirrored the design proposed by NASA for a six-personMars habitation module.

Look Out Below!T H E P L A N W A S T O D E L I V E R the large fiberglass componentsto Devon Island by parachuting them from U.S. MarineCorps C-130 transport planes. On July 5 the first five airdrops,

which contained the walls, legs and some of the dome sectionsof the habitat, delivered the payloads safely to the ground, al-though gusty winds caused most of them to fall several hun-dred meters from the construction site. The sixth drop, on Ju-ly 8, carrying the remaining dome sections and other equip-ment, also went well. The final drop that day, however, was adisaster. The payload separated from the parachute at an alti-tude of 300 meters and plunged to the ground. The habitat’sfiberglass floors were completely destroyed, along with thetrailer needed to move the 360-kilogram wall panels and thecrane needed to lift them.

With the loss of the trailer, the floors and the crane, theconstruction crew that the Mars Society had hired to assem-ble the station declared the task impossible and fled the island.At this point, the project seemed doomed. Indeed, one jour-nalist asked me if I saw a parallel between the failure of the air-drop and that of the Mars Polar Lander, the unmanned space-craft that apparently crashed on the Martian surface in De-cember 1999. I replied: “There’s a parallel in that we both hita rock. But the difference is that we have a human crew here,and we are going to find a way out of this.”

Lee and I assembled a new, makeshift construction teamconsisting of Mars Society scientists and volunteers, Inuityouth hired from the town of Resolute Bay (about 170 kilo-meters away) and a few journalists. Frank Schubert, a gener-


ARCTIC STATIONis based on the

design of a Mars habitation

module.

A list of similarities and differences, compiled by Pascal Leeof the SETI Institute and the NASA Ames Research Center:

SIMILARITIES ENVIRONMENT: Devon Island and Mars are po-lar deserts: cold, dry, windy, rocky and dusty. GEOLOGY: HaughtonCrater’s thick deposits of impact breccia—rock fragments createdby the meteorite strike—may be similar to the rubble on the sur-face of Mars. Also, the valley systems created by erosion of De-von’s terrain look like many of the valleys and gullies observed onthe Red Planet. BIOLOGY: Some types of microbes on Devon Islandlive inside impact-damaged rocks. Researchers have also foundstromatolites— fossils of colonies of primitive microorganisms.Similar life-forms or their fossils might exist on Mars.

DIFFERENCES GRAVITY: Martian gravity is only about one thirdas strong as Earth’s. TEMPERATURE: On Devon Island the averagetemperature is about –15 degrees Celsius, compared with –55 de-grees C on Mars. ATMOSPHERE: Whereas Earth’s atmosphere ismostly nitrogen and oxygen, the Martian atmosphere is 95 per-cent carbon dioxide, and the surface pressure is less than 1 per-cent of Earth’s. RADIATION: Ultraviolet radiation on Mars is about800 times more intense than it is on Devon Island in the summer.

AND

Y AN

DE

RSO

N (

pre

ced

ing

an

d o

pp

osit

e p

ag

es);

RO

BE

RT

ZUB

RIN

(a

bov

e)

Devon Island vs. Mars


al contractor from Denver and a founding member of the MarsSociety, flew in to direct the construction, with the assistanceof his foreman, Matt Smola, and Infracomp vice president JohnKunz. A new trailer, “the Kunzmobile,” was constructed outof wood and parts of a wrecked baggage cart from the ResoluteBay airport. Over three days of freezing rain we used this trail-er to haul all the scattered habitat components to the con-struction site.

Next we designed wooden floors to replace the ruined fiber-glass decks, using timber transported by small aircraft fromResolute Bay. To erect the 20-ton habitat without a crane, wehad to resort to construction techniques that have been in usesince the time of the ancient Romans. A 12-person crew wouldhave to lift each six-by-two-meter wall panel into place usinga winch, a scaffold, bracing timbers and guy ropes. We knew itwould be too risky to try this method if Devon’s frequent 40-knot winds came up. But on the evening of July 19 the weath-er cleared, leaving us with sunny skies, negligible winds andtemperatures in the low 40s. On Devon Island, it doesn’t getany better than that. That night we held a council of war in themess tent at the base camp. We decided to go for it.

Arctic OvertimeB E C A U S E W E D I D N O T K N O W how long the good weatherwould last, we raced against the clock. We took advantage ofthe Arctic’s perpetual summer daylight to institute 14-hourworkdays. In three incredible days, we got the walls up. Theweather was still clear. We partly installed the wooden floors,then used block-and-tackle gear to haul the first two 160-kilo-gram dome sections onto the upper deck. Then we construct-ed a scaffold and erected the two sections over a central core tocreate an arch.

Now the task before us was to lift the other dome sectionsinto place. By this time, however, the long workdays had start-ed to take their toll. Near exhaustion, the crew made mistakes.We had two close calls when improperly guyed scaffolds camecrashing down. But we pressed on. On the evening of July 26,with the wind finally beginning to rise and rain threatening, weheaved the last dome section into place. The only remainingtask was to build the bunks, airlock and toilet area inside the habitat. The freezing rain was back, but it didn’t matter. We could do our work inside the best house on Devon Island.

On July 28 we held a ceremony to mark the completion ofthe habitat. Because of the construction delays, we had time foronly a four-day mission simulation. Commanded by CarolStoker of the NASA Ames Research Center, the six crew mem-bers lived and worked in the habitat, supporting a series of ex-ploration traverses and the field-testing of a Mars spacesuit pro-

totype. The team maintained radio contact with a simulatedmission control based in Denver; 20-minute delays were in-serted between the transmissions to duplicate the time lags thatwould ordinarily occur in radio communications betweenEarth and Mars.

On August 4 the habitat was sealed for the winter. TheMars Society is now preparing for the summer of 2001, whenthe station will support eight weeks of research under strict mis-sion-simulation protocols. Everyone who leaves the habitat willhave to wear a simulated spacesuit and spend 20 minutes in theairlock exiting and returning. Under these and similar con-straints, we will attempt to conduct a sustained program of ge-ology and microbiology field exploration. We are certain to runinto operational problems—but those problems are exactlywhat we are trying to uncover.

Many things did not go right on Devon Island. In all likeli-hood, many things will also go awry when astronauts maketheir first journey to Mars. When venturing into the unknown,one can expect the unexpected. But as we learned from our Arc-tic experience, a resourceful crew can overcome nearly any ob-stacle. On the first human mission to Mars, the crew will be thestrongest link in the chain.


MARS SPACESUIT PROTOTYPE designedby aerospace company Hamilton

Sundstrand is tested on Devon Islandby geologist Pascal Lee.

THE

AU

THO

R ROBERT ZUBRIN, author of The Case for Mars (Simon & Schuster,1996) and Entering Space (Tarcher Putnam, 1999), is presidentand founder of the Mars Society (www.marssociety.org). He is alsothe founder of Pioneer Astronautics, which is involved in the re-search and development of space exploration technology. Readerswho wish to express their opinions on the Mars Society’s Devon Is-land project can send their comments to [email protected]




HAIRWHY IT GROWSWHY IT STOPS

Scientists are rapidly discovering the molecules that control hair production.

In so doing, they could be unearthing the key to combating both baldness

and excessive hair growth

By Ricki L. Rusting • Photographs by Jeff Mermelstein



BY AGE 50,about half of all menand women are grappling with hair loss, typically a recedinghairline and a balding crown in men and diffuse thinning inwomen. Countless people also fret over having too much hairwhere they don’t want it.

Remedies exist but could certainly stand improvement. Onelogical way to correct hair disorders would be to deliver drugsdesigned specifically to influence the molecules that normallyorchestrate hair production. To develop such targeted drugs,however, pharmaceutical makers would need to know the iden-tities of those hair-regulating molecules.

Five years ago biologists were still very much in the dark.Now several research groups are beginning to uncover the mol-ecular controls on hair development. Once the picture is com-plete, they should be in good position to determine which onesgo awry in specific hair conditions and how to bring the defec-tive regulatory system back into line.

Surprising as it may seem, people suffering from serious dis-orders unrelated to hair may also benefit from the recent re-search on hair production. Earlier this year investigators pin-pointed the hiding place of the unspecialized stem cells that re-place lost hair-producing cells and constantly rejuvenate theepidermis at the skin surface. If scientists can turn these mal-leable cells into other kinds of tissue, particularly nerves andmuscles, they will have a readily accessible source of stem cells

for potentially treating Alzheimer’s, Parkinson’s and other dis-eases, while avoiding the complex ethical issues raised by har-vesting stem cells from embryos.

Fashioning a FollicleTO TRACE THE MOLECULAR CONTROLS over any giv-en process, scientists first need to know the basic outlines of theprocess itself. By 1995 microscopists and others had developeda good sketch of the incredible steps that lead to the formationof hair follicles (the tiny bulbs that produce the hair shaft) in thedeveloping embryo. They had also described the hair cycle—theperiodic phases during which follicles produce or stop produc-ing hair; follicles undergo this cycle repeatedly in a lifetime.

Hair follicles develop during gestation in response to crosstalk between the ectoderm, or top layer of cells, composing ayoung embryo and the mesoderm underneath. First, a patchof the mesoderm signals the overlying ectoderm to make an ap-pendage. In response, the ectoderm cells organize, proliferateand invade the mesoderm, becoming an elongated structurecalled a hair germ.

Next, the hair germ directs the underlying mesoderm cellsto cluster together to form the dermal papilla of the hair folli-cle. This structure becomes a kind of command central: it in-structs the germ cells to multiply further and develop into a full-fledged hair follicle. In the end, the upper, permanent section ofthe follicle contains an oil-producing sebaceous gland and abulge—a swelling now known to be the home of most or all ofthe stem cells that replenish the hair, the sebaceous gland andthe epidermis throughout life. The lower segment of the folli-cle, below the bulge, becomes the hair-producing region andis the part that cycles through different stages after embryonicdevelopment is complete.

During development, this bottom section arises as cells fromthe hair germ spread downward, like a growing root, deep intothe dermis of the skin, where they form a bulb-shaped matrix ofcells surrounding the dermal papilla. The dermal papilla prodsthe matrix cells into dividing. As matrix cells get pushed upwardand lose their contact with the dermal papilla, they stop divid-ing and mature, a process known as terminal differentiation.

The matrix cells sitting directly over the apex of the dermalpapilla mature into hair cells—that is, they produce the fibrouskeratin proteins characteristic of hair. More peripheral matrix

■ Biologists now understand many of the steps leading to thedevelopment of hair follicles in embryos and to hairproduction throughout life.

■ Signaling proteins in a family known as Wnt play a role indirecting many of those steps. Researchers are uncoveringadditional regulatory molecules as well.

■ Baldness often arises not because follicles die but becausethey shrink and malfunction. Drugs that manipulate Wnts or other regulatory proteins might one day protectthreatened follicles and prod shrunken ones into producinghair normally again.

■ Knowledge of the controls on hair production could also leadto new ways of eliminating unwanted hair.

Overview /Hair Growth



JOE

ZE

FF

LEF1/Tcf

DSH

GSK3

Wnt

ß-CAT

ß-CAT

ß-CAT

Frizzled(Wntreceptor) Wnt directs GSK3 to

stop marking β-CAT for destruction

Extra β-CATsurvives and goes to thenucleus

NUCLEUS

PROTEINS IN JUNCTIONSWITH OTHER CELLS

CELL INVOLVED INFORMING A HAIR

FOLLICLE OR HAIR

GENE

Bone morphogenic proteinFibroblast growth factorNogginSonic hedgehogSoxTransforming growth factor-betaWinged-helix nude

Hair shaft

Muscle

Bulge

Sebaceousgland

Basement membraneOuter root sheathInner root sheath

Matrix

Dermal papilla

THE HAIR FOLLICLE ...

EPIDERMIS

DERMIS

PERMANENTSEGMENT

CYCLINGSEGMENT

BulgeUnmoored

hair

Basementmembrane

Dermalpapilla

Bulge

Dermalpapilla

Dermalpapilla

New inner

root sheath

New outer

root sheath

New hair

Dying cells

Old hairready

to fall out

FORMERPOSITION

Recent studies suggest that Wnt proteins play a major role in inducing embryonic skin cells to make hair follicles and, later, in prodding follicle cells to make hair throughout life. Wnts instruct cells to stop breaking down another protein, beta-catenin (ß-CAT), which joins with LEF1 or a related protein to help activate specific genes. Those genes, in turn, give rise to proteins that cause the cells to specialize and contribute to follicle or hair formation. Molecules in, or interacting with, the Wnt signaling pathway could one day serve as targets for drugs aimed at enhancing or diminishing hair production.

MOLECULARCONTROLS ONHAIR PRODUCTION

PROTEINS THAT MAY INTERACT WITH THE Wnt SIGNALING PATHWAY IN THE SKIN

Proteins are produced, such as cyclin D (in embryonic skin cells) and hair keratins (in follicle cells that become hair cells)

THE MATURE HAIR FOLLICLE consists of a bulbous, vaselike structure that produces the hair shaft. Throughout life, follicles cycle through hair-growing and nongrowing phases. Many molecules involved in controlling cycling have now been identified. These discoveries suggest new strategies for treating hair disorders.

... AND ITS THREE-PART LIFE CYCLE

1 CATAGEN Cells in the cycling

segment die, and the shaft loses its mooring. Meanwhile the basement membrane pulls the dermal papilla up to the bulge

2 TELOGEN The follicle

rests. Hair may fall out now or in the next phase

3 ANAGEN On instructions from

the dermal papilla, the lower part of the follicle rebuilds and begins producing hair again

THE HAIRY DETAILS


J U N E 2 0 0 1

cells differentiate to form the inner root sheath, which serves asa vase for the hair. An outer root sheath arises, too, and encas-es the inner sheath. As new hair cells push the older ones up-ward, causing them to break through the skin surface, the old-er cells die, forming a shaft of dead, keratin-rich cells. Thesedead cells are virtually indestructible, persisting to decorate ourheads and blanket our bodies with hair.

At the end of gestation, a newborn human enters the worldwith five million to six million hair follicles distributed in a ge-netically determined pattern over the body. No new ones formthereafter. The palms of the hands and the soles of the feet arepretty much the only places that truly lack follicles; other ar-eas that seem to be hair-free actually produce short, thin hairs.

Born to CycleFOLLICLES BEGIN CYCLING within two or three years af-ter birth. The cycle has been shown to have three main stages.In catagen, the epithelial cells below the bulge essentially com-mit suicide, leaving behind only the dermal papilla and a mem-brane (the basement membrane) that formerly encased the nowdying region. As the cells die, this membrane contracts anddraws the dermal papilla up to the bulge. (In the scalp, this triptakes about two weeks.) Meanwhile the hair shaft loses its an-chor deep in the dermis. It therefore becomes prone to sheddingin the next two stages.

When the dermal papilla reaches the bulge, follicles enterstelogen, the resting phase. Telogen can last about three monthsin the human scalp, but its duration can be modulated by arange of factors. Plucking hair or wounding the follicles canshorten it, for instance.

Anagen follows telogen. Early on some of the stem cells fromthe bulge divide and travel down along the basement membraneto become matrix or outer root sheath cells. Once formed, thematrix cells proliferate and ultimately give rise to the hair cellsand the inner root sheath, repeating the steps that occur duringembryonic development. This repetition implies that the eventsof anagen are probably controlled by a number of the same sig-naling molecules that operate during development.

The reconstituted anagen follicles produce an inch of hairevery two months or so in the scalp and usually keep this up forsix to eight years. The length of anagen determines how long asingle hair can grow. On any given day in the life of a typical20-year-old, about 90 percent of the scalp follicles are in theproductive, hair-growing phase, and about 10 percent are de-caying or inactive; approximately 50 to 100 hairs are shed.

Hair thinning generally happens not because follicles dis-appear but because the ratio of follicles in the growing and non-growing phases shifts unfavorably. Also, many follicles in bald-ing people shrink progressively, ultimately producing onlysmall, colorless hairs.

As is true during follicle development in the embryo, dur-ing anagen signals from the dermal papilla instruct the matrixcells to divide and subsequently differentiate into hair cells. Forthis reason, scientists have become very interested in uncover-ing the nature of the signals issued by the dermal papilla dur-

JULIUS CAESAR reportedly combed his “scantylocks” forward to cover his baldness; he also had

“superfluous” hairs plucked out

GOOSE BUMPS FORM, and hair stands on end, whentiny muscles that are linked to hair follicles

contract in unison

COMMON BALDNESS in men can be inherited fromeither side of the family; many genes

appear to be involved

BODILY STRESS, such as severe infection or surgery,can cause hair to fall out, but the loss may

be delayed for up to six months

SMOKING can hasten hair loss

HAIR-REPLACEMENT SURGERY moves follicles fromthe sides and back of a man’s head to the bald areas

SOME WOMEN LOSE HAIR after giving birth or discontinuing birth-control pills

THE AVERAGE SCALP FOLLICLE can generate morethan 30 feet of hair in a lifetime

FOLLICLES with round cross sections yield straighthair; flatter follicles yield curlier locks


ing development and cycling. They don’t have the answer yet,but in the past few years Elaine Fuchs and her colleagues at theUniversity of Chicago have discovered that the dermal papil-la’s signals probably convey their directives largely by activat-ing still other signaling molecules—members of the Wnt fami-ly of proteins. Wnt proteins have long been recognized as keyregulators of varied developmental processes in mammals andother organisms.

The Hand on the HelmFUCHS STUMBLED ACROSS the first clues to the importanceof Wnts to hair about six years ago. At the time, for reasons un-related to treating human hair disorders, she wanted to identi-fy the signaling molecules that instruct certain matrix cells tobegin producing hair keratins.

Often a cell will initiate a behavior, such as making newproteins, after a molecule from the outside binds to a receptoron the cell surface and triggers a cascade of molecular interac-tions on the inside. These signaling cascades frequently lead tothe activation of specific genes in the nucleus, culminating inthe production of the proteins the genes encode. Knowing this,Fuchs began her search for the molecules that dictate the con-version of matrix cells to hair cells by trying to identify the mol-ecules in the nucleus that switch on the hair keratin genes.

In 1995 her group discovered that a regulatory proteincalled lymphocyte enhancer factor 1 (LEF1) participated in ac-

tivating the hair keratin genes. It was also present during hairfollicle formation in the embryo, where it appeared in the ear-liest clusters of ectoderm cells as well as in the cells destined toform the dermal papilla. Presumably on orders from some out-side signal, LEF1 became active and helped to turn on genesneeded for follicle formation or hair growth. Consistent withthis conclusion, Rudy Grosschedl and his co-workers, then atthe University of California at San Francisco, discovered thatwithout LEF1, mice fail to make a furry coat. And when Fuchs’steam engineered mice that produced excess LEF1 in the skin,the animals produced more hair follicles than normal.

At about the same time, other groups demonstrated thatLEF1 cannot activate genes on its own; rather it must first cou-ple with a second protein, beta-catenin. The only mechanismknown to trigger this coupling was the activation of the sig-naling cascade that begins with the binding of a Wnt moleculeto the cell surface. Beta-catenin normally helps to form junc-tions with neighboring cells. In the absence of Wnt signaling,an enzyme inside the cells marks any unused beta-catenin fordestruction. Wnts instruct cells to handcuff that enzyme. Withthe enzyme out of commission, beta-catenin becomes free to ac-cumulate and to pair with LEF1 or one of its relatives.

Combined with Fuchs’s discoveries, these results suggest-ed that Wnts and rescued beta-catenin molecules might be cen-tral in both follicle formation and hair production. Subsequentstudies in mice added support to that notion. For instance,


SOME DRUGS that can cause hair loss includeblood thinners, antidepressants, high blood

pressure pills and anabolic steroids

CUTTING HAIR does not make it grow faster or get thicker

THE AVERAGE ADULT SCALP sports about 100,000 follicles

CANCER DRUGS cause balding because they killrapidly dividing cells, including ones that multiply

transiently during hair production. Hair growsback eventually because the drugs usually do not

harm the stem cells that replenish lost hair cells

BRISTLES are the only hairs on whales, elephants and rhinoceroses

HAIR CAN BE STRETCHED by about a third itslength in water without being harmed


Fuchs’s group devised a way to flag cells that activated LEF1binding genes in response to a Wnt signal in a developing em-bryo. Those experiments implied that Wnt is the mesoderm-issued signal that instructs the overlying ectoderm to beginforming an appendage and is likewise the ectodermal signal thattells the underlying mesoderm to form the dermal papilla. Whatis more, much later in development, after follicles have formed,

Wnt appears to be the message that directs matrix cells abovethe dermal papilla to differentiate into hair cells.

Even more dramatic evidence for the central importance ofWnts came when Fuchs’s group created mice that, after birth,could not degrade beta-catenin in their epidermal cells, a fea-ture that made the cells behave as if they were endlessly receiv-ing a Wnt signal. As adults, these rodents acquired an unusu-


FEW PEOPLE ACCEPT hair loss with equa-nimity. “People often say to me they can dealwith the loss of a kidney . . . but not with hairloss,” says Vera H. Price, director of the Hair Re-search Center at the University of California atSan Francisco. So how can science help?

The good biological news is that in the mostcommon types of thinning, hair follicles don’tdie. In classic male- and female-pattern hair loss(androgenetic alopecia), for instance, folliclesbecome miniaturized and their growing phaseabbreviated; they then produce extremely short,fine hairs. “Even guys who are bald still have lit-tle hairs on the top of their head,” explains BruceA. Morgan of Harvard’s Cutaneous Biology Re-

search Center. In a rarer condition, alopecia area-ta (affecting nearly 2 percent of people), the fol-licles’ growth phase ends prematurely under au-toimmune attack, causing hair to fall out inpatches or, in extreme cases, all over the body.But, again, the follicles survive.

Treatment for alopecia areata typically fo-cuses on quelling the wayward immune sys-tem, but treatment for male- and female-pat-tern hair loss must increase the size of Lilliputianfollicles as well as hair length. Minoxidil—intro-duced as Rogaine in 1988—was the first drugapproved by the U.S. Food and Drug Administra-tion for this purpose and is the only one licensedfor use in both sexes. Scientists still debate how

minoxidil, which is applied topically, producesthicker, longer hairs: perhaps it increases bloodsupply, better nourishing the follicles, or per-haps it alters cellular concentrations of sub-stances that regulate hair growth.

The mechanism of the second approveddrug, finasteride, is clearer. This compound,marketed as Propecia for male hair loss (and asProscar in a higher-dose formulation for pros-tate enlargement) is taken orally. In the body, itinhibits an enzyme that converts testosteroneto a hormone called dihydrotestosterone (DHT)and thus reduces DHT production. DHT is criticalto the development of male fetuses, but later itcan be a troublemaker. It stimulates some folli-

SAVE THE HAIRS! by Mia Schmiedeskamp

Drug companies are busy searching for the next generation of hair-promoting compounds


ally lush coat by forming new hair follicles between the onesthat were laid down during embryonic development.

The production of new follicles is certainly exciting, but acouple of other findings from the mouse studies might makebalding readers think twice before tracking down Fuchs andbegging her for vials of Wnt to pour on their heads. As the fur-ry rodents aged, they acquired benign lumps that resembled a

common human scalp tumor called pilomatricoma. Fuchs’slaboratory subsequently demonstrated that in humans these tu-mors arise when a mutation in the beta-catenin gene preventsthe protein’s breakdown. Wnts and extra beta-catenin have al-so been implicated in cancers of the colon, liver, breast and re-productive tract.

To Fuchs, all these results, including the unfortunate mouse


cles to produce thick, long hair (on the cheekand chin, for example), and it induces scalp hairthinning in susceptible people, sometimes asearly as the preteen years. Some researcherssuspect that DHT disrupts hair follicles by act-ing on a region called the dermal papilla, alter-ing its production of substances that influencehair growth.

In a study published in 1998 of more than1,200 men between the ages of 18 and 41 withmild to moderate hair loss, about 83 percentmaintained the hair they had on the top of theirhead after two years of finasteride use. Morethan half had at least mild regrowth. But JerryShapiro, director of the University of British Co-lumbia Hair Research and Treatment Center,cautions, “I think the big job is to keep patients’expectation levels appropriate, so that theyknow it’s not going to be luxuriant hair. The em-phasis should be on prevention. Regrowth is al-so a possibility but shouldn’t be stressed, es-pecially in men with more advanced hair loss.”

Men can expect similar results with 5 per-cent minoxidil, advises Marty E. Sawaya of ARATEC, which conducts clinical trials for vari-ous companies. She estimates that 25 to 30percent of men gain moderate-to-dense re-growth with either product. Some men hedgetheir bets by using both over-the-counter mi-noxidil and prescription finasteride. In stump-tailed macaques, at least, the two drugs to-gether worked better than either one alone.

Shapiro and Price advise their patients to al-low about a year to see whether hair-growthdrugs work—and to commit to using them cor-rectly. Minoxidil solution is applied directly tothe scalp twice daily; Propecia is a once-dailypill. To maintain results, either drug must beused consistently and indefinitely. “Nothing re-verses thinning completely,” Price says. “But dothese drugs work? Yes, they work.”

For female-pattern hair loss, only 2 percentminoxidil has FDA approval. About 60 percent ofwomen achieve maintenance of hair and some

regrowth with this option. Those with the oppo-site woe of unwanted facial hair now have anFDA-approved option, too: topical eflornithinecream, marketed as Vaniqa, inhibits an enzymenecessary for cell proliferation, thus retardinghair growth. But experts once more caution pa-tients to keep their expectations reasonable.About 58 percent of women see slight improve-ment or better, usually after a couple of months.The cream slows hair growth but does not stopit, so treated women do continue to tweeze orremove hair by other means.

Looking AheadMUCH RESEARCH done by companies is keptsecret until drug candidates reach clinical trials.Probably, though, the next hair-loss treatmentsto come down the road will work on familiar prin-

ciples. Several companies have developed mol-ecules that inhibit the same enzyme as finas-teride. Hoechst AG has done some laboratorytesting of a drug that would be applied to thescalp to block DHT from binding to hair folliclecells. And Bristol-Myers Squibb has a drug in ear-ly clinical studies that is thought to functionsimilarly to minoxidil.

Perhaps the most promising compound toenter human trials is called dutasteride, fromGlaxoSmithKline. Like finasteride, it inhibits theenzyme that produces DHT, but it blocks twoforms of the enzyme instead of just one.Sawaya says the preliminary results suggestthat dutasteride is more effective at increasinghair count than finasteride, even at a lower dose.

“ We don’t have a product yet that’s going tobe ‘wow!’ for over 50 percent of people,” shenotes. “I do think dutasteride will be that prod-

uct if the company goes forward.” When or if FDA

approval might be sought for using dutasterideagainst hair loss remains uncertain, however.Gaining approvals is time-consuming and cost-ly, and GlaxoSmithKline may choose to pursueit first as a prostate drug, as happened with finasteride.

Scientists are on the prowl for new drugs allthe time. Morgan’s Harvard colleague MichaelDetmar discovered earlier this year that abun-dant amounts of a growth factor that increas-es the blood supply make mice grow hair fasterand thicker. Now, Morgan says, the hunt is onfor small molecules that will either mimic or ac-tivate the factor. But drugs like that or ones in-tended to interact with molecules that directlyregulate hair growth—such as Wnt or beta-catenin—are a long way off. Much more re-

search needs to be done before the right targetscan be manipulated without risking such con-sequences as cancer.

A fundamental understanding of hair biolo-gy may someday let physicians replace a de-fective gene in hair follicles through gene ther-apy or grow hairs in a petri dish for use in graftsurgery. “The complexity of the question is likeunderstanding how a limb forms. It’s ambitious.But we are discovering a lot and discovering alot quickly,” muses Kurt S. Stenn, chief scientif-ic officer of Juvenir Biosciences, a company re-cently spun off from Johnson & Johnson to fo-cus predominantly on hair research. “This is awonderful time to be working in hair biology. Somany breakthroughs are coming.”

Mia Schmiedeskamp is a science writer based in Seattle.

“This is a wonderful time to be working in hairbiology. So many breakthroughs are coming.”

—Kurt S. Stenn, Juvenir Biosciences


tumors, provide useful information for scientists interested intreating hair disorders. They teach that Wnts are major regu-lators of follicle development and cycling but that simply de-livering Wnts by constant application would not be feasible asa human therapy, because of the tumor risk. The trick to cor-recting hair maladies, Fuchs contends, may be to deliver Wntsin a pattern that mimics nature better or to manipulate othersteps in the Wnt signaling cascade.

To do that, scientists need still more information about theWnt signaling pathway and about other factors in the skin thatinfluence it. Which Wnts, and which of their numerous recep-tors, are involved at different steps of the hair cycle and in fol-licle development, and what molecules control their produc-tion? And which molecules inside Wnt target cells determinehow those cells respond to Wnts, such as whether they becomehair cells or other parts of a follicle?


JOE

ZE

FF;

SOU

RC

E:

O’T

AR T

. N

OR

WO

OD

IN THE PAST 10 years several studies havehinted that baldness is more than an em-barrassment: it can be a visible warning ofincreased risk for heart disease. Last yearthe largest study yet conducted confirmedthat notion.

The investigation, headed by JoAnn E.Manson of Harvard Medical School’s Brighamand Women’s Hospital, looked at partici-pants in the Physicians’ Health Study, a long-term project that examined the risks forheart disease in some 22,000 male physi-cians. Eleven years into the project, the doc-tors, who were then between the ages of 51and 95, indicated which of five pictures mostclosely approximated their hair patternwhen they were 45. Most men who go baldlose their hair in a standard sequence, al-though at different rates. First, the hairlinerecedes near the temples. Next, the hair atthe crown, or vertex, begins to go. Then thehairline rises farther and the bald spot at thecrown grows, until the two areas meet.

Manson and her colleagues correlated

the hair patterns with heart problems thathad arisen in 19,112 subjects who had nocardiovascular problems at the start of thePhysicians’ Health Study. Physicians whodied during the 11-year period were not eval-uated, so the link between baldness and fa-tal heart attacks could not be assessed. Butthe researchers could look at the connectionbetween hair loss and other “coronaryevents,” namely nonfatal heart attacks,angina or treatment for heart disease (by-pass surgery or angioplasty).

When potential confounding influenceswere eliminated, the results showed that, re-gardless of age, men with frontal baldnessalone were only slightly more likely (9 per-cent) to face heart problems than were menwho retained all their hair. But those withmild thinning at the crown had a 23 percenthigher risk of heart disease, and those withmoderate or severe balding at the crown hadmore than a 30 percent higher risk.

Worst off were severely bald men withhigh cholesterol levels or high blood pres-

sure. Those with elevated cholesterol werealmost three times more likely to have heartdisease than were men with high cholesteroland hair on their pates. Bald subjects withraised blood pressure faced almost twice therisk encountered by their counterparts withlusher hair.

Researchers can only speculate aboutwhy bald men would be more susceptible toheart disease. Genetic inheritance could beat fault, or high levels of male hormones (an-drogens) or increased sensitivity to themcould be the common denominator. Andro-gens, Manson notes, play a part in male-pat-tern baldness and appear to contribute toatherosclerosis and increased blood clot-ting, both of which promote heart disease.

Obviously, baldness does not causeheart attacks. It can, however, serve as a dai-ly reminder to take preventive measures.The drill should be familiar by now: avoidsmoking, get regular exercise, eat right andkeep blood pressure and cholesterol levelsin the normal range.

Will taking drugs aimed at preventinghair loss also protect against heart disease?Manson received a flood of e-mails last yearasking that very question. Sorry, she says.No evidence suggests it will. —R.L.R.

STANDARD CLASSIFICATION SCHEME for the progression of male-pattern baldness, the Norwood-Hamilton scale, includes the heads depicted here. In a less typical pattern (not shown), the hairlinerises progressively to the back of the head but a discrete bald spot never forms. The heart studydescribed above presented a simplified version of the scale.

TAKING HAIR LOSS TO HEART

Shiny domes may signify an elevated susceptibility to heart attacks


Research into signaling pathways that interact with the Wntpathway will surely offer some clues to the answers. Scientistswho study many different tissues and organisms have identifieda host of proteins in such pathways, including sonic hedgehog,transforming growth factor-beta, bone morphogenic protein,noggin and fibroblast growth factor, to name just a few.

Sonic hedgehog could be particularly crucial for hair growth.Like Wnt proteins, it carries a signal from one cell to anotherand is known to participate in the proper development of em-bryos. Further, sonic-hedgehog and Wnt-signaling pathwaysoften influence each other. Investigations reveal that althoughsonic hedgehog is not needed for formation of the hair germ,it is needed for subsequent conversion of the germ to a full-fledged follicle. And last year Ronald G. Crystal of Weill Med-ical College of Cornell University found that when hair folliclesin adult mice are induced to make the protein during the rest-ing, telogen stage, the follicles shift prematurely into the hair-producing, anagen stage. Thus, sonic hedgehog can stimulatedormant follicles to begin producing hair.

Although treatment with sonic hedgehog might seem an at-tractive idea for inducing hair growth, too much signaling bythis molecule results in basal-cell skin cancers in humans. Todevelop therapies that involved sonic hedgehog, Wnts or oth-er proteins able to induce cell division, pharmaceutical manu-facturers would first have to make sure that those moleculeswere properly controlled.

The effects of bone morphogenic proteins and differentforms of transforming growth factor-beta on Wnt signaling areproving difficult to sort out. But some scientists suspect thatwhen that task is done, those proteins, too, could prove usefulfor stopping or starting hair growth.

Styling TherapyIDENTIFYING THE MULTITUDE of molecules that coor-dinate the development and cycling of hair follicles is clearly adaunting job. Yet thanks to the rapid pace of technological ad-vancement, researchers should soon be able to discern all thegenes that are activated in purified populations of cells at dif-ferent stages of follicle development and cycling. With this in-formation at the ready, they could assess how these complex

patterns of gene activity are altered in people who have hair dis-orders. The technology should also enable skin biologists to un-cover new proteins important to hair production as well as tospecify which ones contribute to different disorders.

As researchers become more sophisticated in their knowl-edge of the molecular interactions underlying hair growth, theycan begin animal testing of compounds that might restore or-der to deranged regulatory pathways and revive dormant fol-licles. If those tests go well, human scalp skin can be trans-planted onto mice incapable of rejecting it to determine whetherhuman and mouse follicles respond comparably to the agents.And if those results are good, investigators may attempt humantrials of the most promising drug candidates.

No one can predict how soon dermatologists and pharma-ceutical companies will be able to produce new therapies builton the discoveries emerging from basic research into hair folli-cle development and cycling. But that research is progressingremarkably fast. If the pace continues, Fuchs predicts, muchof the information that is needed to understand the complexcontrols on hair manufacture will probably be in hand withinthe next five years.

Ricki L. Rusting is a staff editor and writer.


M O R E T O E X P L O R EThe Secret Life of the Hair Follicle. Margaret H. Hardy in Trends in Genetics,Vol. 8, No. 2, pages 55–61; February 1992.

The Biology of Hair Follicles. Ralf Paus and George Cotsarelis in New EnglandJournal of Medicine, Vol. 341, No. 7, pages 491–497; August 12, 1999.

Multiple Roles for Activated LEF/TCF Transcription Complexes During HairFollicle Development and Differentiation. Ramanuj Das Gupta and ElaineFuchs in Development, Vol. 126, No. 20, pages 4557–4568; October 1, 1999.

Stem Cells: A New Lease on Life. Elaine Fuchs and Julia A. Segre in Cell,Vol. 100, No. 1, pages 143–155; January 7, 2000.

Involvement of Follicular Stem Cells in Forming Not Only the Follicle but Alsothe Epidermis. G. Taylor, M. S. Lehrer, P. J. Jensen, T. T. Sun and R. M. Lavkerin Cell, Vol. 102, No. 4, pages 451–461; August 18, 2000.

Morphogenesis and Renewal of Hair Follicles from Adult Multipotent StemCells. H. Oshima, A. Rochat, C. Kedzia, K. Kobayashi and Y. Barrandon in Cell,Vol. 104, No. 2, pages 233–245; January 26, 2001.

General information about hair can be found at www.keratin.com

HAIRS DETECT MECHANICAL STIMULIand relay the information to

the nervous system

AN EYEBROW HAIR grows for about two months before falling out

CONTINUALLY WEARING tight bands orbraids or buns may cause permanent

patches of baldness


HimbaDam

A questionable act of progress may drown this African tribe’s way of life. Similar dramas are playing out around the world

In our world a dam is a small thing that gives cattle water. What you are talking about is something else and will finish the Himba.

—Chief Katjira Muniombara

By Carol EzzellPhotography by Karin Retief

AND THE

THE


Jakatunga Tjiuma, counselor to a Himba headman inNamibia, washes his hands in the Kunene River, thesite of a proposed hydroelectric dam.


Turning, I point to hills in the east. “And the water would backup behind the dam to make a lake that would stretch to there.”I can see the shock and incredulity in his eyes as he begins to un-derstand how high the water would rise up the faraway hill-sides, flooding more than 140 square miles of Himba settle-ments, grazing land and grave sites. He clutches a blanketaround his shoulders and crouches on a rock, speechless.

Tjiuma is a counselor to one of the headmen for the Him-ba tribe, an essentially self-sufficient band of 16,000 peoplewho eke out an existence from the barren, rocky terrain ofnorthwest Namibia, living off the milk and meat of their cattleand goats, along with the occasional pumpkin or melon. TheHimba are sometimes called the Red People because they tra-ditionally cover their bodies, hair and the animal skins theywear with a mixture of butterfat and a powder ground from theiron ore ocher. They say they use the ocher-butter mixture be-cause they like the way it looks, although it undoubtedly alsoprotects their skin against the arid climate.

For decades, the Himba have lived in relative isolation. Noother tribes wanted their hardscrabble land, and the Germanswho colonized the area in the late 19th century rarely interact-ed with them. More recently, the Himba’s main contact withoutsiders has been with soldiers during the fight for Namibia’sindependence from South Africa (which was won in 1990),with marauding combatants spilling over from Angola’s on-going civil war, and with the occasional caravan of hippieAmericans or Europeans. But if the Namibian government hasits way, by 2008 more than 1,000 foreign workers will havesettled in a temporary village just downstream from EpupaFalls, the site the government favors for the dam. With themwill come a cash economy, alcohol, prostitution and AIDS—

as well as improved roads, better access to medical care, schoolsand perhaps even electricity.

The situation surrounding the proposed dam on the KuneneRiver can be viewed as a microcosm of dam projects around theworld that are affecting indigenous peoples. A survey by theWorld Commission on Dams, which issued its controversial fi-nal report last November, found that 68 of the 123 dams world-wide they studied would displace people, many of them in tribesthat had little prior contact with the technological world. Thelargest dam project, the massive Three Gorges Dam on the Yang-tze River, will require the resettlement of up to two million Chi-nese. Nearly all the dams will change local peoples’ livelihoodsand cultures—for good or ill, or some combination of the two.

How should global society weigh the right of such peoplesto be left alone against, in some cases, the very real necessity fordeveloping countries to take advantage of their resources?Should such countries have the autonomy to decide what is inthe best interests of all their citizens, even if some of them don’twant to change? Perhaps most important, how can traditionalpeoples decide such issues for themselves when they have onlya shaky idea of how more developed societies live and whatthey might be getting themselves into?

Into the Desertkaokoland, the corner of namibia where the Himbalive, is truly the back of beyond. We arrive at Epupa Falls, themodest waterfall on the Kunene River that would be inundat-ed by the dam’s reservoir, two days after leaving the last tarredroad. Our 4×4 truck is packed with everything from jerricansof gasoline (the closest gas pump is a day’s drive away) to cas-


The situation surrounding the proposed dam on the KuneneRiver can be viewed as a microcosm of dam projects around

the world that are affecting indigenous peoples.

Himba mother and child glisten red from a coating of butterfat mixedwith the iron ore ocher. Like other adult women, the mother shaves herforehead and twists her hair into multiple braids that she daubs witha mud mixture. The heads of infants are shaved until they are weaned.

Not until we stand on a ridge overlooking the KuneneRiver—which forms part of the border between the southern Af-rican nations of Angola and Namibia—does tribal leader JakatungaTjiuma comprehend the immensity of the proposed dam. “Lookthere,” I tell him with the help of an interpreter, pointing to a distantnotch in the river gorge that a feasibility study says would be the mostlikely site of the wall of concrete. “That’s where the dam would be.”



es of bottled water, spare tires, emergency medical supplies,camping gear, and small gifts of tobacco, sugar and blankets.Tied to the top of our vehicle is a brand-new bicycle—the pay-ment requested by our Himba translator, Staygon Reiter, in ex-change for his services, although how he will use it in this in-hospitable landscape I don’t know. He has asked specificallythat the bicycle come equipped with a carrier basket largeenough to hold a goat.

Much of our journey is bumpy, jerky and slow as we at-tempt to follow the rough track while swerving to avoidwashouts and potentially tire-puncturing rocks. More thanonce we get stuck in sand while trying to cross a dry riverbed,our tires spinning and squealing until we jump out to deflatethem a bit or to stuff branches behind them for traction. At onepoint we stop to look at a particularly large scorpion in ourpath; I comment that I’ve seen smaller lobsters.

The settlement at Epupa Falls, where we camp, is a kind ofcrossroads, a no-man’s-land where Namibian Himba mix withtheir Himba relatives from across the river in Angola and withother tribes such as the Herero—to whom the Himba are close-ly related—as well as with the Zemba, Thwa and Ngambwe.There is a modest thatched church built by missionaries; a tinybut deluxe safari camp; a corrugated-metal store that sellsmostly bags of cheap tobacco, maize meal, and tepid Coke,Sprite and Fanta; and a community-run campsite where visi-tors like us can pitch a tent under the palmlike omerungu treesfor 50 Namibian dollars (about US$6) per night. Scarcely anypeople live at the settlement permanently: the Himba come fora few weeks or months at a time and build temporary hutswhile they attend funerals, divide inheritances, sell cattle, con-duct other business, and visit with friends and relatives.

Our first stop is to meet Chief Hikuminwe Kapika at hiscompound near Epupa Falls, which is part of the territory hecontrols. It is immediately clear that Kapika—who is one ofroughly a dozen Himba chiefs—is sick of talking about the pro-posed dam with outsiders but eager for us to appreciate the im-portance of his rank. From his shock of grayish hair and weath-ered face, I guess him to be in his 70s, although Himba don’thave a calendar system, so they usually don’t know the year inwhich they were born. He keeps us standing beside his whitemetal camp chair (the only one in his compound) swatting fliesfrom our faces as I try to catch his attention long enough to an-swer my questions. Several times during our interview he spits


MAP

BY

SUSA

N C

ARLS

ON

The Kunene River forms the northwest border between Angola andNamibia; the Himba live in the rocky, arid region known as Kaokoland(top). Tribal leader Tjiuma (middle) points to the spot (dam site 1 onmap) the Namibian government favors as the most economical placefor the proposed dam. The location is downstream of Epupa Falls (bot-tom), which would be inundated by the reservoir expected to back upbehind the dam wall. The flooding would eliminate the omerungu palmtrees that provide fruits the Himba depend on in times of drought. TheAngolan government prefers a site farther downstream, in the BaynesMountains (dam site 2), that would necessitate the renovation of an-other dam, which was damaged in the country’s civil war.

K A O K O L A N D

BAYNES MOUNTAINS

DAM SITE 1

DAM SITE 2

EPUPA FALLS

NAMIBIA

ANGOLA

ATLA

NTI

COC

EAN

OPUWO

RUACANA•

•

NAMIBIA

ANGOLA

BOTSWANA

SOUTHAFRICA

ZAMBIA

Kunene River

•

WINDHOEK•


through a gap in his front teeth created in his teens when, inkeeping with Himba tradition, his lower two central incisorswere knocked out and the top two filed to create a V-shapedopening. He makes a point of demonstrating what a busy manhe is by continuing to sew a black fabric loincloth and inter-rupting our translator to correct a group of rowdy children.

Eventually Kapika tells us that he vehemently opposes theproposed dam. He is afraid that the people who will come tobuild it will steal the Himba’s cattle—not an irrational fear, be-cause the Himba were nearly wiped out at the end of the 19thcentury as a result of cattle raids by the Nama tribe, which livesto the south. And cattle theft continues today. He is also wor-ried that the newcomers will take valuable grazing land, whichthe Himba are careful not to overuse. Family groups move theirhouseholds several times a year so that extensively grazed re-gions can grow back. The area around Kapika’s compound il-lustrates the need for such conservation: the cattle and goatshave eaten everything green they can reach, leaving the bushesand trees top-heavy with scraggly growth overhanging trunkslike lollipop sticks.

Himba leaders also object to the dam because it would floodhundreds of graves, which play a central role in the tribe’s reli-gious beliefs and social structure. In times of crisis, family pa-

triarchs consult their forebears through special ceremonies atgrave sites, and graves are often used to settle disputes over ac-cess to land. Acreage is owned communally, but each permanentsettlement is guarded by an “owner of the land,” usually the old-est man of the family who has lived at that place for the longesttime. When deciding who should be able to graze their cattle ina particular area, Himba compare the number of ancestors theyhave buried there. They ask, “Whose ancestral graves are old-er, ours or theirs?”

Kapika says the Himba will resist and fight “with stones andspears” if the Namibian government tries to build a hydro-electric dam at Epupa Falls. “I’m a big man,” he tells us. “I’m aman who can stand on his own.”

Dammed If They Dohow do you describe a megadam to someone who hasnever seen electricity? Or a building more than one story high?The dam planned for Epupa Falls would rise 535 feet—only 15feet shorter than the massive Grand Coulee Dam in Washing-ton State. It would generate 360 megawatts of electricity perday and cost more than US$500 million to build.

A dam was first proposed near Epupa Falls in 1969, whenNamibia was South West Africa, a territory of South Africa. Theidea went nowhere, but it was revived in 1991, a year afterNamibia’s independence, when Namibia and Angola commis-sioned a feasibility study to evaluate such a scheme. The study

considered two sites for the dam: Epupa Falls and a spot in theBaynes Mountains farther downstream. It concluded that Epu-pa Falls made more economic sense, but Angola has favored theBaynes site in part because building a dam there would meanthat the country would also get funds to renovate a dam on anAngolan tributary that was damaged during the civil war. Thatcost is one reason the Baynes site would be more expensive.

When the study’s consultants first came to discuss the in-tended dam with the Himba, the tribal leaders initially had noobjections, thinking it was going to be a small earthen dam likethe ones they built to help water their cattle. The degree of mis-communication took a while to become apparent. Margaret Ja-cobsohn of Integrated Rural Development and Nature Con-servation, a Namibian journalist turned anthropologist whoworked on the social impact part of the feasibility study, recallsa telling incident a few months into the process. She went to vis-it a Himba family compound near Epupa Falls and began ask-ing their views about the proposed dam. Oddly, they didn’tseem to know anything about it, even though the Namibiangovernment had told her that they had been informed. As shefinished her questionnaire, a family member asked her to helpthem with a mysterious piece of paper they had received sometime before. When the man brought an ocher-smeared enve-

lope out of his hut, she recognized it as a letter about the damin English that they had never even opened. After she translat-ed it for them, an old man of the family shook his head and said,“You’re talking about the great death of the Himba.”

Lifeways of the Himbathe himba are one of the last tribes of traditional peoplewho are generally self-supporting and fully or partially isolat-ed from global society. Anthropologists find them particularlyinteresting because they observe a system of bilateral descent.Every tribe member belongs to two clans, one through the fa-ther (a patriclan) and another through the mother (a matriclan).Tribes that practice bilateral descent are rare: besides the Him-ba, the custom occurs among only a few peoples in West Africa,India, Australia, Melanesia and Polynesia.

Each Himba patriclan is led by the oldest man in the fami-ly. Sons live with their fathers; following marriage, daughtersleave to join their husband’s family’s household and become amember of that patriclan. But the inheritance of materialwealth—in the Himba’s case, primarily cattle—is determinedby the matriclan. Accordingly, a son does not inherit his father’scattle but his maternal uncle’s instead.

Bilateral descent is particularly advantageous for tribes thatlive in precarious environments, such as the drought-prone re-gion of the Himba, because during a crisis it allows an individ-ual to rely on two sets of relatives spread over different areas.


How do you describe a megadam to someone who has neverseen electricity? Or a building more than one story high?



The system could also play a role in alleviating inbreedingamong Himba livestock. Various patriclans have taboos pro-hibiting their members from owning cattle or goats of a cer-tain color or coat pattern. When cattle are born that violate apatriclan’s taboos, they must be swapped with nonoffendingcattle from another patriclan.

The religion of the Himba is also organized according tobilateral descent and is practiced through an individual’s patri-clan. Himba believe in a god-creator, but that entity is very re-mote from human affairs and can be petitioned only by in-voking dead paternal ancestors to act as intercessors. Thetribe’s religious observances center on holy fires that were ini-tially kindled at the graves of ancestors and are maintained bythe leader of each respective patriclan in his family compound.

The holy fire is small, often just a smoldering log sur-rounded by several rocks. It is always located between theopening of the headman’s hut and the corral where the cattleare penned at night. That area of the compound is consideredsacred: strangers cannot cross between the holy fire and thecorral or between the holy fire and the headman’s hut without

first asking permission. Traditionally, the headman keeps thefire going during the day as he sits by it to commune with hisancestors about any problems facing the family. At night, theheadman’s wife takes an ember of the fire into the main hut;in the morning, the ember is taken outside to the hearth again.

The Himba are also intriguing to anthropologists as sub-jects of rapid social change. One way in which this change ismanifesting itself is in patterns of dress. Many more Himbamen than Himba women have adopted Western clothing andhairstyles. At Epupa Falls, where Himba occasionally have con-tact with outsiders, a Himba man can be seen one day bare-chested and wearing a Himba apron-skirt and jewelry, and thenext day dressed in pants and a shirt. Few young men therewear the “bachelor ponytail” that is traditional for unmarriedmen, and even fewer married men follow the custom of not cut-ting their hair and of covering their heads with a cloth. And itis extremely rare to find a Himba man at Epupa Falls who wearsocher: indeed, many wash daily in the Kunene River using soap.

Himba women, however, are much more conservative intheir dress. Even at Epupa Falls, most of the women go bare-

breasted and wear traditional apron-skirts made of calfskinsor goatskins; they smear themselves liberally from head to toeevery morning with the ocher-butter mixture and almost nev-er use water to wash. Young girls wear their hair in two thickbraids that drape over their foreheads and faces, whereaswomen have a cascade of long, thin braids, each of which theycoat with a mud mixture that dries to a hard shell.

According to anthropologists, Himba women are not mere-ly clinging passively to their traditional dress: they are activelyrejecting change because it is the only way they can maintaintheir prestige and value. Himba men occasionally earn moneydoing menial jobs or selling livestock, but Himba women havenot had such opportunities. By preserving their ocher-coveredbodies, braids and calfskin skirts, Himba women are engagedin what modern anthropological theory calls “change throughcontinuity” or “active conservatism.” “Remaining apparent-ly traditional can be a strategic—and rational—response tomodern events,” Margaret Jacobsohn says.

The recent report by the World Commission on Dams de-clares that tribal peoples such as the Himba, whether they are

actively conservative or not, often get caught between a damand a hard place. Such projects have “inadequately addressedthe special needs and vulnerabilities of indigenous and tribalpeoples,” the report concludes, adding that the effects of a damon local peoples are “often not acknowledged or consideredin the planning process.” It calls for improving existing waterand energy facilities rather than constructing new megadamsand stipulates that sponsoring countries and internationallenders base their decisions to build new dams on agreementswith affected communities.

But in February the World Bank said it would use the com-mission’s guidelines only as “reference points” rather than asbinding procedures for financing large dam projects. A groupof 150 nongovernmental organizations from 39 countries—

including Namibia—countered in March with a letter to WorldBank president James Wolfensohn to reconsider that stanceand to place a moratorium on funding new dams until the bankimplements the commission’s guidelines. The organizationsare requesting that the bank conduct independent reviews ofplanned and ongoing projects and set up procedures for pro-viding reparations to people harmed by earlier dams. In theletter, they insinuate that the World Bank helped create theWorld Commission on Dams in 1998 with the World Conser-vation Union–IUCN only “to deflect opposition or to buy time.”Unless the bank amends its position, they write, they “may beless inclined to engage in future...dialogues with the WorldBank.” According to the commission, the bank has provided


The recent report by the World Commission on Damsdeclares that tribal peoples such as the Himba often get caught between a dam and a hard place.

Young Himba man living at a settlement near Epupa Falls shows theresult of contact with other cultures. Besides the thick necklace tra-ditional to the Himba, he also wears colorful necklaces from the Zem-ba tribe and a Western tracksuit jacket. Himba women have beenmore reluctant to change their traditional attire, perhaps becausethey seek to preserve their identity.


an estimated $75 billion for 538 large dams in 92 countries.So what is the case for a dam at Epupa Falls? Jesaya Nya-

mu, Namibia’s minister of mines and energy, emphasizes thathis country currently imports 60 percent of its power fromSouth Africa and needs to pull the plug as a matter of nation-al sovereignty. “No one seems to see our need for independentpower,” he laments.

Ensconced in the deep upholstery of a sofa in his cabinetminister’s office in Windhoek, Namibia’s capital, he labels theforeign environmental groups that oppose the dam as meddlerswith a double standard: one for their own industrial countriesand another for countries they consider untouched and exotic.“The whole of Europe and America is dammed,” Nyamu says.“These people live in their own countries on hydropower.”

Indeed, according to a trade group of dam builders, the In-ternational Commission on Large Dams, the U.S. has the sec-ond-largest number of large dams (higher than 90 meters) inthe world, after China. And the American experience withdams and indigenous peoples is less than laudatory: the GrandCoulee Dam inundated the lands of Native Americans fromthe Colville and Spokane tribes and ruined their salmon fish-ery. The tribes sued for reparations in 1951, but the govern-ment took 43 years to settle the lawsuit. In 1994 the tribes ac-cepted a $54-million lump sum and $15 million per year aslong as the dam produces electricity.

But Katuutire Kaura, president of Namibia’s main oppo-sition party, the Democratic Turnhalle Alliance/United Demo-cratic Front Coalition, contends that another dam on theKunene River is “absolutely not necessary.” An existing damthat was built in the 1970s upstream at Ruacana is runningat less than 20 percent capacity, he points out. And the recentlydiscovered Kudu gas field off Namibia’s southern coast is es-timated to contain 20 trillion cubic feet of natural gas—morethan enough for Namibia’s needs. “The Kudu gas field can lastus 25 to 30 years,” Kaura asserts. Shell Oil and the Namib-ian government are currently working to tap those gas fields.

Kaura adds that the Himba will reap few of the dam’s ben-


Namibia’s minister of mines and energy says that his countrycurrently imports 60 percent of its power from South Africa.

“No one seems to see our need for independent power.”

Himba women and children in a traditional family compound sit bytheir fires during the early morning (top) to warm themselves afterthe chilly desert night. The women wear erembes, pleated rabbit-skinhats that signify that they are married. Teenage boys (middle) tendthe family’s cattle as the animals graze during the day. Cattle are asign of wealth among the Himba: when a rich man dies, his family of-ten slaughters dozens of his cattle, whose skulls are used to deco-rate the man’s grave (bottom) as a sign of status. Such graves—

roughly 160 of which will be underwater once the dam is built—arethe sites of important tribal rituals.


efits while paying high costs. They are not qualified to workon the dam, so it will not bring them jobs. They are also un-likely to get electricity from the project. Electricity did not cometo the residents of Opuwo, the town closest to the Ruacana Dam,until 1994, more than 20 years after it was built. In the mean-time, a dam at Epupa Falls would destroy the Himba’s liveli-hood. It “will dislocate the Himba to the margins of societywhere they cannot survive,” predicts Phil Ya Nangoloh, exec-utive director of Namibia’s National Society for Human Rights.

In a way, the dam will take the river away from the Himbaand confer its benefits to people outside Kaokoland. Accordingto the World Commission on Dams report, “Dams take a setof resources . . . generating food and livelihood for local peopleand transform them into another set of resources . . . providingbenefits to people living elsewhere. There is a sense, therefore,in which large dams export rivers and lands.”

Toward a Struggle?one morning when Tjiuma comes to our camp to share acup of coffee, I ask him what he really thinks will happen if thegovernment goes ahead with its plans for the proposed dam. Iknow he is no stranger to combat, having been drafted as atracker to fight on South Africa’s side during the war for inde-pendence. As we gaze over the Kunene River in the still of theearly morning, he admits that the Himba have a plan for resis-tance. More than 50 of the Himba headmen were in the mili-tary during the war, he says, and they still have their old .303rifles in their compounds.

A week later, when I visit the minister of mines and energyin Windhoek, I tentatively ask him what the Namibian gov-ernment would do if the Himba resist with violence. His re-sponse is chilling: “We know them; they cannot do anything.If they try anything, we will neutralize them, of course. But Idon’t think it will come to that.”

Carol Ezzell is a staff editor and writer.

M O R E T O E X P L O R EHimba: Nomads of Namibia. Margaret Jacobsohn, Peter Pickford and Beverly Pickford. Struik Publishers, Cape Town, 1990. Published in the U.S. by Appleton Communications, 1992.

Dams and Development: A New Framework for Decision-Making. World Commission on Dams, 2000. Available at www.dams.org/report/

For information on dam projects around the world, including the proposedKunene River dam, visit the International Rivers Network site at www.irn.org


Social change is already coming to the settlements near Epupa Falls.A U.S.-based Christian church operates a mission near the river, wherechurch members routinely baptize new Himba converts (top). A smallshop (middle) sells staples such as maize meal but also does a briskbusiness in cheap liquor and beer. Himba sometimes loiter outside theshop to beg money from the occasional tourist to buy alcohol; the areaaround the shop is littered with empty bottles. Some of the bottles endup being used by other Himba to carry water (bottom). The young girlfilling her bottle in the Kunene River wears the two forward-hangingbraids traditional for preadolescent females.



A Low-Pollution

EngineSolution

CLEAN-BURNING, SPARKLESS-IGNITION AUTO ENGINES MAY OFFER THE BEST

CHANCE OF MEETING NEW EXHAUST EMISSIONS STANDARDS by Steven Ashley

nyone who has driven a car, truck or motorcycle thatis at least 20 years old is probably familiar with whatengineers call after-run, an annoying and now rarecondition in which an engine keeps turning over fora few seconds after the ignition is off. But today thefundamental fuel-burning process that causes after-run is firing the interest of the automotive industry.

Known as homogeneous-charge compression-ignition com-bustion, or HCCI, the process could provide the basis for a newclass of low-emission, high-mileage power plants. Many com-bustion engineers believe that HCCI-based piston engines willbe able to deliver the good fuel economy of diesel engines with-out the diesel’s high emissions of nitrogen oxides and soot.

Faced with increasingly stringent governmental pollutionstandards as well as the realization that practical and afford-

able fuel cell technology (in large production volumes) is stillmany years away, researchers at the world’s major automak-ers and diesel-engine manufacturers are working to determinewhether HCCI technology will be technically and economical-ly feasible. If so, power plants based on this new combustionmode might serve as a potential bridge technology between to-day’s high-emission diesel- and gasoline-fueled piston enginesand tomorrow’s ultraclean fuel cell power trains.

Standing-room-only attendance at the technical sessions onHCCI combustion at the recent Society of Automotive Engi-neers (SAE) 2001 World Congress in Detroit is one strong in-dication that the homogeneous-charge compression-ignitionengine could be the next big thing in the car industry. Anotheris the marked rise in the numbers of technical papers written onthe topic, says Dennis Assanis, professor of mechanical engi-


JOSÉ

CR

UZ

ACopyright 2001 Scientific American, Inc.

neering at the University of Michigan at Ann Arbor and direc-tor of the university’s Automotive Research Center. “Since1995, when only a few HCCI papers were published, we’veseen what seems to be an exponential increase,” he reports.

Yet another sign of HCCI’s newfound status is growing re-search support from the U.S. Department of Energy, which be-gan funding its study in 1997. The Partnership for a New Gen-eration of Vehicles, an R&D consortium involving govern-ment, industry and university scientists and engineers devotedto advanced automotive technology, has established a $3-mil-lion, four-year academic research program on the novel com-bustion process. At the same time, industry and academic re-searchers in the field have prepared a report to the U.S. Con-gress about the technology. Interest is also great in Japan, whereengineers who pioneered the exploitation of HCCI call it activethermo-atmosphere combustion, and in Europe, where it isknown as controlled auto-ignition.

Burning Lean and CleanHCCI CAN BE THOUGHT OF as a crossbreed technologycombining attractive aspects of both conventional gasoline anddiesel engines to achieve good fuel economy and near-zero ex-haust emissions. Roughly speaking, internal-combustion enginesfall into four categories, defined by the degree of mixing in thecharge of fuel and air in the cylinder and by how this charge isignited. The familiar gasoline engine, in which a premixed, thor-oughly commingled fuel-air charge is set aflame by a spark plug,falls into the spark-ignited, homogeneous-mixture classification.The diesel engine is an example of the compression-ignition, het-erogeneous-mixture type: fuel is sprayed into the cylinder dur-ing the piston’s compression stroke, and turbulent flow partial-ly mixes it with air until the rising temperature induces burning.The gasoline direct-injection engine, in which injected fuel par-

tially mixes with air until set alight by a spark, is considered aspark-ignited, heterogeneous engine. As its name indicates, thehomogeneous-charge compression-ignition engine is a fourthtype: it uses a thoroughly premixed charge of fuel and air thatis compressed by a piston until it self-ignites.

Because the amount of burning fuel is low in comparison tothe volume of air inside an HCCI engine, the combustion tem-peratures stay relatively low. That means the engine producesonly small quantities of nitrogen oxide and dioxide (collective-ly, NOx). Also, because the charge is well mixed and does notcontain excess fuel, the combustion generates only smallamounts of sooty particulates. Engine efficiencies are high be-cause the HCCI combustion process allows the use of high,diesel-like compression ratios (generating more power per unitof fuel burned) and because, like diesels, HCCI engines canmeet load demands without the use of intake throttling, thuseliminating so-called breathing losses. In addition, if properlyengineered, these power plants could burn nearly any availablehydrocarbon fuel or even hydrogen.

Automotive engineers and combustion researchers are pur-suing this alternative form of the internal-combustion engine be-cause current propulsion configurations are not likely to meetupcoming environmental regulations that will place severe lim-its on exhaust containing greenhouse gases—mostly carbon diox-ide—and other pollutants, including NOx, particulates, carbonmonoxide and unburned hydrocarbons. The gasoline engine runstoo hot and is too fuel-inefficient to make the grade. Poor fueleconomy translates into excessive production of carbon dioxide,whereas high-temperature combustion yields too much NOx. Incontrast, the fuel-efficient diesel engine generates too much NOx

and particulate emissions to satisfy soon-to-be-imposed pollu-tion standards. Although catalytic after-treatment systems offerhope that exhaust emissions from these conventional power JO

SÉ C

RU

Z


Gasoline Engine Diesel Engine


plants can be cleaned up after the fact, there is no surety that thiscan be accomplished affordably. Direct-injection gasoline enginesoffer better fuel economy, but their NOx and hydrocarbon emis-sion levels are not much lower than their conventional gasolinecounterparts. As such, they would still require the use of sophis-ticated exhaust after-treatment systems, which, in turn, need newgrades of low-sulfur gasoline to avoid poisoning their catalysts.

Existing alternative propulsion technologies seem to providelittle or no respite in the search for an environmentally and eco-nomically sound personal transportation technology for thenear-term future. Because of the energy storage limitations ofelectrochemical batteries, electric cars now offer insufficient dri-ving range, at best about 160 miles. The electric car’s overall en-vironmental fitness is, moreover, dependent on the cleanlinessof the method employed to generate electric power. Widespreaduse of coal-burning power plants, for example, reduces the elec-tric car’s total system cleanliness substantially. Today’s hybridelectric vehicles, which combine downsized internal-combustionengines with batteries and electric motors, produce less pollu-tion than current cars and trucks, but auto companies must sub-sidize their price tags by as much as $10,000 apiece to makethem cost-competitive with conventional vehicles. As a result,automotive engineers are searching for ways to improve the in-ternal-combustion engine so that it can carry the industrythrough the forthcoming transition to ultraclean, next-genera-tion propulsion systems.

Taking the LoadHCCI COMBUSTION has been studied by many inventors andengineers under many different guises during the past century.But according to Paul Najt, spark-ignition-engine group man-ager at the General Motors Research and Development Centerin Warren, Mich., the modern investigation of this unique com-bustion mode actually began in the late 1970s. At that time, aresearch team led by Shigeru Onishi of Nippon Clean EngineCompany in Japan reported that it had been studying what mem-bers called active thermo-atmosphere combustion in two-strokeengines. “Rather than avoid this natural mode of combustion,Onishi said: ‘Let’s try to exploit it,’” says Najt, who was an en-gineering school graduate student in the early 1980s. He and hisfellow students took up the challenge but soon found that theengine controls of the day could not manage the difficult task ofcontrolling the auto-ignition process as engine speeds and loadswere varied. Unfortunately, that problem remains unsolved.

“The HCCI process really works well in the laboratory ona dynamometer when all the engine components are in thermalequilibrium,” explains Thomas Asmus, senior research execu-tive at DaimlerChrysler Liberty and Technical Affairs inRochester Hills, Mich. “But when you add a load and try tomake the engine do work, as it would in a vehicle, it just tendsto slow down and stop completely. If you add more fuel so itcan handle the increased load, it tends to start knocking veryseriously.” Nearly all HCCI experts mention how an engine

The homogeneous-charge compression-ignitionengine could be the next big thing.

CROSSBREED CONCEPT—The homogeneous-chargecompression-ignition (HCCI) engine can be considered ahybrid of the spark-ignition or gasoline engine and thecompression-ignition or diesel engine. Gasoline enginesuse a premixed or homogeneous fuel-air charge that isplaced in the cylinder and then ignited by a spark plug,causing a hot flame front to sweep through the charge. In diesel engines, fuel is injected into the cylinder duringthe piston’s compression stroke, where it partially mixeswith the air (producing a heterogeneous mixture) until the rising temperature induces self-ignition.

An HCCI power plant would combine aspects of both—

the premixed fuel-air charge of the gasoline engine andthe pressure-heated auto-ignition of the diesel—to pro-duce a low-temperature, even-burning combustionprocess that yields good fuel efficiency as well as lowemissions of nitrogen oxides and soot in the exhaust.

HCCI Engine



running in HCCI mode can easily “run away on you,” pro-ducing a tremendous pounding noise that will eventually de-stroy the engine hardware.

The problem is twofold, Asmus says. First, HCCI combus-tion occurs extremely quickly. Once the temperature in the en-gine cylinder is sufficiently high, the premixed fuel-air mixtureignites all at once. “For use in a practical engine, we need thecombustion to have a smoother, more extended heat-releaseprofile,” he explains. For maximum efficiency, designers preferfor the ignition process to begin at 10 to 15 degrees of crankangle before the piston reaches top dead center and then the re-mainder to continue afterward. If the burn starts too early, the

hot gas is exposed to the cylinder walls for too long and heatis lost, cutting efficiency. If it starts too late, the hot combustiongases do not undergo full expansion and therefore do not im-part maximum work to the piston.

The second problem with HCCI, Asmus says, revolvesaround the fact that “there’s no triggering event, like a spark orfuel injection, which are the handles we use for controlling thetiming of the burn event” in conventional engines. To keep theprocess under control as speed and load changes occur (operat-ing conditions that engineers call transients), the engine has tomake very rapid adjustments from one cycle to the next. Cur-rently no one knows exactly how to accomplish that reliably andaffordably. “With HCCI,” he points out, “it’s not clear whatwill serve as a strong, robust lever for phasing the burn.”

Najt concurs: “The benefits of HCCI are clear; the difficul-ty is in controlling HCCI combustion. Right now it’s a ques-tion of technology execution. All kinds of control concepts that20 years ago seemed a bit over the edge are being considered.It may be that, relative to fuel cell technology, these kinds oftechnologies and the extra cost to implement them don’t seemas much of a stretch.”

These still unproved engine-control technologies are man-ifold. They include variable-valve actuation, in which hot resid-uals—embers from the previous burn—are inducted into thecylinder to control the phasing of the next burn cycle. Variable-valve-timing systems are based on “camless valve trains” op-erated by electromagnetic, electrohydraulic or piezoelectric ac-tuators rather than mechanical cams. To get the needed cycle-to-cycle response-rate resolution, however, the valves have tomove extremely rapidly, and that kind of operation is difficultto keep up over the lifetime of an engine.

Another possibility, Dennis Assanis says, is to use variable-compression-ratio systems, which involve changing the volumeof the combustion chamber, and so the compression ratio, onthe fly. This effect can be accomplished by opening and clos-ing engine valves at the appropriate time or by installing pis-tons that alter their height in accordion fashion in reaction topressure changes, a radical concept that is under investigation

Steps toCompression-Ignition

OPTICAL ENGINE—Special laboratory apparatus at Lotus Engineering in England allows researchers to see inside a test-engine cylinder as HCCIcombustion, or what they call controlled auto-ignition, proceeds. D

ON

LAW

AN

D J

EFF

ALL

EN

(p

hot

ogra

ph

)

HOT-ZONE COMBUSTION—Computersimulations produced at LawrenceLivermore National Laboratory model theclimbing temperatures inside the cylinderof a Volkswagen TDI diesel engine as thepiston rises and compresses the air. Thetemperature distribution inside an HCCIengine is crucial because chemical kinet-ics, rather than turbulence, tend to controlthe combustion process, and chemicalkinetics are sensitive to temperature. Thehottest zones burn first. Colder zones maynot fully react, producing hydrocarbon andcarbon monoxide emissions. The coolestareas are where fuel contacts the walls.



JOSÉ

CR

UZ

(ill

ust

rati

on);

SAL

VAD

OR

M.

ACE

VES,

DAN

IEL

L. F

LOW

ER

S AN

D J

OE

L M

ARTI

NE

Z (s

imu

lati

on)


at the University of Michigan in cooperation with Ford MotorCompany and Federal Mogul Corporation.

Yet another approach being looked at by many researchersis to add some inhomogeneity (variation in local density andtemperature) into the fuel-air mixture to extend the duration ofthe burn. Of course, “that’s like playing with the devil,” Assa-nis notes. “You’re giving up some homogeneity [thus boost-ing pollutant output] to get a smoother heat release.”

“The problem with these advanced concepts,” Najt says,“is that they introduce a whole series of mechanical-complex-ity and cost issues. The use of two unproved technologies oftensquares the difficulty of the original problem. And the cost ofintegrating them into a viable package may be too great.”

Handling High Loads and HybridsEVEN IF THESE OBSTACLES are surmounted, another over-all concern must be addressed. Because the fuel-air mixture hasto be lean (a low fuel fraction in relation to the volume of air)to obtain the emissions benefits, HCCI is suitable only for lightand medium loads and speeds. To handle higher loads andspeeds, more fuel needs to be added to the mixture, but doingso would raise combustion temperatures and eliminate much ofthe environmental benefit. Therefore, HCCI combustion wouldprobably be used in what engineers call dual-mode engines. Athigh engine loads, the system would switch from auto-ignitingHCCI mode to spark ignition (in gasoline-fueled engines) or tostandard fuel injection (in diesel engines).

Najt notes that the most straightforward application ofHCCI would probably be in a fuel-efficient hybrid-electric vehi-cle, which combines an internal-combustion engine, an electricmotor and a battery. When used in a hybrid configuration, theinternal-combustion engine, whether it is of the HCCI type ornot, operates in a smaller speed and load range. As this suitsclean-burning HCCI technology, this may turn out to be a per-fect match for the hybrid-electric configuration. “Yet it’s not clearthat HCCI will end up as the optimal engine for hybrids, becauseeven hybrids run through a fairly extensive speed and load range,so the engine must follow any increased load demands,” he com-

ments. This is the case because the energy storage capacity oftoday’s batteries is not large enough to provide all the extrapower for acceleration and hill climbing when it is needed.

The technical hurdles that must still be overcome have ob-viously done nothing to suppress enthusiasm for HCCI withinmuch of the engineering community. Some observers wonder,however, whether it will really prove to be the long-sought-after solution to the environmental/economic quandary. Everyfew years the auto industry focuses eagerly on a particular en-gine technology as the next big thing that might do the trick,Najt says. Less optimistic experts attending the recent SAE ses-sions called HCCI the latest boutique engine.

“In the mid-1980s it was the two-stroke engine, which did-n’t pan out,” Najt explains. “A couple of years ago it was thegasoline direct-injection engine, which, although it has been rel-atively successful, doesn’t seem to be a panacea.”

Many engine researchers expect that HCCI-based powerplants will be the first automotive engines designed “from theinside out.” In other words, using advanced computational mod-eling techniques, engineers will be able to explore the chemicalkinetics of fuel-oxidation and fluid-mechanics phenomena as-sociated with mixing and burning that control HCCI beforethey settle on an engine design. Nevertheless, a lot of good oldexperimental work with test engines will be needed before apractical, affordable engine can be developed.

“Clearly, it’s too early to tell whether HCCI will prove tobe successful,” Najt concludes. “It’s a high-risk technology.”Still, most engineers agree that it is worth exploring while we’rewaiting for something cleaner to come along.

Steven Ashley is a staff editor and writer.

M O R E T O E X P L O R ESociety of Automotive Engineers: www.sae.org/servlets/index/University of California, Berkeley: www.me.berkeley.edu/~mctai/hcci.html/Sandia National Laboratories:www.ca.sandia.gov/CRF/03_facilities/03_FacHCCI-SCCI.html/Lund Institute of Technology:www.vok.lth.se/CE/research/HCCI/i_HCCI_uk.html/

TEMPERATURE(kelvins)

1,100900700500


MOST MODERN BALLSare variations of two basic designs. Woundballs have a small, solid or liquid-filled rubbercore wound with rubber thread. Solid ballshave a large, solid rubber core and thin, polymerinner cover. Outer covers for either type may becomposed of one or more layers of urethane or anionomer resin such as Du Pont’s Surlyn. Some pros prefera soft, synthetic-balata cover for better spin control.


GOLF BALLS

The quest to engineer a golf ball that flies long andtrue dates back to 15th-century Scotland. Artisansthere tediously stuffed a wet leather pouch with a hat-full of boiled goose feathers, then stitched it shut. Thedrying feathers expanded while the leather shrank, toform a ball as hard as a rock. A skilled ball maker couldproduce only four featheries a day, relegating the gameof links to the rich.

Four hundred years later, in the 1840s, the guttiearrived. Craftsmen heated and molded gutta-perchagum (which was, and is, used in dentistry) from theMalaysian Palaquium tree into a solid sphere. Durableand inexpensive, the guttie brought golf to the mass-es. Golfers noticed, however, that new, smooth gut-ties did not fly as straight or as far as old, nicked ones.Ball makers began cutting, hammering or impressingvarious patterns of indentations into each ball’s sur-face, a practice that helped balls fly straighter andlonger. Because the science of aerodynamics was stillyoung, no one really knew why.

In 1898 Ohio golfer Coburn Haskell and B. F.Goodrich employee Bertram Work unveiled a rubbergolf ball. They wound rubber thread around a solidrubber core. Soon the balls were coated with balata,a strong, water-resistant latex from the tropical bul-ly tree. In 1908 English engineer William Taylor re-ceived a patent for an inverted bramble pattern ofevenly distributed circular depressions imposed onthe ball’s surface; these dimples reduced aerodynam-ic drag and enhanced lift.

By 1930 British and American golf associationshad standardized the diameter and weight of balls fortournament play. Manufacturers have tried all man-ner of dimple patterns since then. Today most ballssport around 400 dimples. Lately makers have beenexperimenting with two-tiered dimples in hopes offurther reducing drag.

Some golfers complain that better technologygives players too much aid. It does help duffers, butit hasn’t diminished the skill pros require. Wally Uih-lein, CEO of Acushnet, maker of the Titleist brand,notes that “in spite of space-age balls and clubs, theaverage score on the PGA Tour has improved but onestroke over the past 17 years.” —Mark Fischetti

Flight Control

WORKINGKNOWLEDGE

RUBBERCORE

URETHANECOVER

POLYMERINNERCOVER

RUBBERWINDINGRUBBER

WINDINGRUBBER

WINDINGRUBBER

WINDING

RUBBERCORE

Rubber/resin (2001)



➤ DIMPLE WARS In 1983 Titleist introduced the 384 Tour

ball with 60 extra dimples. Golfers said it went farther. Com-

petitors responded by advertising balls with “more dimples

for more distance.” But if this argument were taken to its

extreme, thousands of tiny dimples would re-create a

smooth ball, which tests show travels only half as far as a

dimpled one. A ball with 300 to 500 dimples works best, ac-

cording to Steve Aoyama, principal scientist at Titleist.

➤ GOT YOUR NUMBER Manufacturers paint different digits

on balls purely to help golfers in a party tell whose ball is

whose. The numbers have nothing to do with ball quality or

characteristics. Nevertheless, golfers will swear that they

hit a Top-Flite 2 longer than a 5 or that a Maxfli1 hooks less

than a 6. Superstition? In golf? Never.

➤ SMASH TEST Hye Precision Products in Perry, Ga., tests

the durability of newly minted golf balls’ cores, paint, even

logos in a closed, squat, 2,500-pound machine. The beast’s

air cannon fires 12 balls a minute at up to 300 feet a sec-

ond across a five-foot chamber into a hardened steel plate.

The same set of balls is continually and automatically col-

lected, reloaded, refired, and inspected for deformation

and surface condition.DID

YOU

KNOW

...

DRAG on a smooth, soaring golf ball iscreated by the difference between high airpressure against the front of the ball andlow pressure behind it, the latter caused bythe laminar separation of airflow.

DIMPLES cause turbulence in the thin layerof air against the ball, which reduces airflowseparation, creating more back pressureand thereby reducing drag.

BACKSPIN, imparted by the angle of a golfclub’s head, deflects airflow downward justas an angled airplane wing does. The result-ing upward reacting force gives the ball lift.

Brambled guttie (1890s)

ILLU

STR

ATIO

NS

BY

GE

OR

GE

RE

TSE

CK

; M

ICH

AEL

NE

VEU

X C

orb

is (

rub

ber

/res

in);

LE

O M

. K

ELL

Y, J

R.

The

Old

Ch

ica

go

Gol

f Sh

op (

rub

ber

, g

utt

ie,

fea

ther

ie);

AER

OD

YNAM

ICS

SOU

RC

E:

STE

VE A

OYA

MA

Titl

eist

Rubber (1898)

AIRFLOW

DRAG

LIFT

Featherie (late 1400s)


REVIEWS


BE

TH P

HIL

LIP

S

Dinos and DarwinJUST HOW IMPORTANT WERE THE DISCOVERIES OF DINOSAUR FOSSILS? BY CARL ZIMMER

Not long ago my friend George, who re-cently celebrated his second birthday, putsome of his vocabulary on display for me.“Tee-wex,” he said, and roared. “Tie-say-watops,” he said, and roared again.“Apatosaw-us.” Roar number three. It isa remarkable thing that children todaycan speak Latin, but more remarkablestill that the only Latin words they speakare the names of dinosaurs. I have yet tohear George or any other child say “Hal-lucigenia,” or “Ambulocetus,” or “Acan-thostega”—although they were as remark-able as any velociraptor.

Dinosaurs have such a powerful gripon the public consciousness that it is easyto forget just how recently humans be-came aware of them. A two-year-old boytoday may be able to rattle off three di-nosaur names, but in 1824 there was onlyone dinosaur to be named, period. Theword “dinosaur” didn’t even exist until1842. Those confused early years, whenthe world was baffled by the discovery ofabsurdly enormous reptiles, representone of the most fascinating stories in thehistory of science.

One reason is that its cast is so extra-

ordinary. On thesouth coast of En-gland, Mary An-ning, a poor and uneducated beach-comber, spent 30 yearsdigging up giant marine reptiles andpterosaurs. Gideon Mantell, a shoemak-er’s son turned doctor, discovered thefirst dinosaur; he thought dinosaurswould make him rich, but they ultimate-ly destroyed his life. William Buckland,an Oxford geologist who tried to recon-cile giant extinct reptiles with Genesis,had a raft of eccentricities, including apenchant for keeping live hyenas andjackals in his college rooms. The time isripe for a book for the general publicabout these early paleontologists, andnow we have not one but two.

The Dragon Seekers is the work of apracticing paleontologist (ChristopherMcGowan is a senior curator at the Roy-al Ontario Museum and teaches at theUniversity of Toronto). The book istherefore filled with historical details thatmatter to a fossil hunter: the methods theearly fossilists used to extract bones from

cliffs, the squabbles over naming newspecies, the staffing of museums. It isbrief and pleasant, but for sheer narrativepleasure, I’d have to recommend Debo-rah Cadbury’s Terrible Lizard instead.Cadbury, a BBC television producer,turns what could have been just a stringof anecdotes into high drama. Much ofher success comes from her depth of re-search: she has scoured diaries, lettersand newspaper archives and can tell herstory in the words of the people wholived it.

For Cadbury, Gideon Mantell is thetragic hero of the early days of dinosaurhunting. Scrounging one quarry after an-other, he built up one of the finest privatefossil collections in the world at the time.Even when he had just a few scraps of di-nosaur bones, Mantell knew that he hadfound the remains of giant reptiles. Hedidn’t back down when the leading sci-entists of his day told him he had foundnothing but fish teeth and rhino horns.

But Mantell’s obsession with his fos-sils eventually left him bankrupt andalienated from his wife and children. Andjust when he began to earn scientific re-spect, he crossed paths with the ruthlessSir Richard Owen. Owen didn’t knowhow to dig up fossils, but he did knowhow to pluck the strings of academicpower. He managed to make himselfEngland’s authority on all life, both liv-ing and dead.

It was Owen, not Mantell, who in1838 was appointed by the British Asso-ciation for the Advancement of Science tosurvey the giant extinct reptiles of En-gland. In his report, he gave them the

THE DRAGON SEEKERS: HOW AN EXTRAORDINARY CIRCLE OFFOSSILISTS DISCOVERED THE DINOSAURS AND PAVED THE WAY FOR DARWINby Christopher McGowanPerseus Publishing, Cambridge, Mass., 2001 ($26)

TERRIBLE LIZARD: THE FIRST DINOSAURHUNTERS AND THE BIRTH OF A NEW SCIENCEby Deborah CadburyHenry Holt, New York, 2001 ($27.50)


name “dinosaurs” but mentioned Man-tell only in scorn. The dinosaurs liftedOwen on their colossal backs to heightsof fame and wealth. Mantell meanwhilefaded into obscurity, his fossils dispersedand forgotten.

Owen used dinosaurs as an argumentagainst evolution: if life progressedthrough time, it made no sense that theextinct dinosaurs were so much more im-pressive that today’s reptiles. He thoughtthat life unfolded over geologic time ac-cording to certain laws, but no one—noteven himself, it seems—really understoodhis ideas. Then Darwin blindsided Owenwith an elegant, powerful theory that en-compassed all of life, dinosaurs included.Owen ultimately became irrelevant. Thedinosaur sculptures that he had built forthe Crystal Palace exhibition are nowchipped, broken and beset with weeds.

My chief complaint with both booksis that they unintentionally raise an im-portant question that neither answers:The discovery of dinosaurs was unques-tionably fascinating, but was it ultimate-ly very important? I’m not so sure. Oneof the great triumphs of 19th-century pa-leontologists was their use of fossils asRosetta stones to decipher the global ge-ologic record. But dinosaurs were onlyone group of animals among many thatthey used. Darwin was certainly inspiredby fossils but not by dinosaurs. Instead itwas the giant ground sloths and other ex-tinct mammals he dug up in South Amer-ica that made him think about the con-nections between present and past life.The word “dinosaur” appears nowherein The Origin of Species.

In the 1940s George Gaylord Simp-son reinvigorated paleontology when heshowed that mutations and natural se-lection could account for changes in thefossil record. But he used horses and oth-er mammals as proof, not dinosaurs.Since then, paleontology has surged fromthe backwaters of evolutionary researchto the crest of the wave. Punctuated equi-librium, the causes and effects of massextinctions, plate tectonics’ role in the

origin of species, the coevolutionary armsraces between predators and prey—in allthese cases, it was fossils that showedhow evolution works. Rarely did thesefossils belong to dinosaurs; they were in-stead snails, plankton, pond scum and

other unglamorous creatures. The fossilsof dinosaurs are beautiful and spectacu-lar, but compared with other animals,they’re also very hard to find. A thousandfossil crabs can say much more abouthow evolution works than a single T. rex.


REVIEWS

Only in the past couple of generationshave paleontologists really started to usedinosaurs in cutting-edge research in evo-lution. Dinosaur paleontologists wereamong the first scientists to use thenewest methods of classification (knownas cladistics), and as a result, they’ve beenable to explore the incremental evolutionof birds from dinosaurs. Dinosaurs arebecoming important sources of informa-tion about how the breakup of conti-nents influences evolution, how changingthe rules by which embryos developshapes new anatomy and how asteroidscan trigger mass extinctions. The subtitleof The Dragon Seekers claims that thestudy of dinosaurs paved the way forDarwin, yet the reverse may be closer tothe truth.

Carl Zimmer is the author of Parasite Rex and At the Water’s Edge.

GLORIOUS ECLIPSES: THEIR PAST, PRESENT AND FUTUREby Serge Brunier and Jean-Pierre Luminet. Cambridge University Press, New York, 2000 ($39.95)An extraordinarily beautiful book, Glorious Eclipses guides us elegantly through the histo-ry of obscurations of the sun and moon—from the ancient Chinese belief that a dragon wasdevouring the daytime star through current theories and on to celestial events forthcom-ing in the next six decades. Some of the most startling images come from the days whenphotography was in its infancy and astronomers still made drawings of eclipses—the oneat the right, by Warren de la Rue,is the solar eclipse of July 1860,observed from Spain.

Published in time for June’seclipse across southern Africa,the book, by Brunier, formereditor of the magazine Ciel etEspace, and Luminet, a researchdirector at CNRS, is translatedfrom the French.

Books reviewed can bepurchased at www.sciam.com

THE EDITORS RECOMMEND


TECHNICALITIES

A wonderful childlike excitement filled thehalls of Scientific American on the daythat the Aibo ERS-210 arrived in themail. Clutching the box tightly, I rushedto the office of my colleague GeorgeMusser and shouted, “It’s here! It’s here!”George beamed like a kid on Christmasmorning and immediately tore the boxopen. Inside was a foot-long, three-pound machine that looked like a JackRussell terrier wearing gray plastic ar-mor. Sony Electronics had bred the ro-botic dog in its research laboratories inJapan; the name “Aibo” (pronouncedEYE-bo) means “pal” or “companion” inJapanese and is also a rough acronym forArtificial Intelligence Robot. For weeksI’d been begging Sony officials to let meevaluate the computerized canine.

In the past few years, toy companieshave introduced an entire litter of elec-tronic pets: Poo-Chi the Interactive Pup-py (from Tiger Electronics), Tekno theRobotic Puppy (Manley Toy Quest) andRocket the Wonder Dog (Fisher-Price),to name just a few. But Aibo is the mostadvanced pooch of the new breed. Sonyunveiled the first-generation Aibo ERS-110 in 1999, producing a limited run of5,000 robotic hounds priced at a howl-ing $2,500 each. They sold out almost in-stantly, prompting Sony to rush 10,000similar units to market. The improvedsecond-generation ERS-210, a relativebargain at $1,500, has also won thehearts of well-heeled customers since itwent on sale last November.

What can explain such intense devo-tion? Certainly part of Aibo’s appeal is itssophisticated technology. Each mechan-

ical mutt possesses nearly as much com-puting power as a typical desktop PC.Tucked into the rear of the dog’s torso isa 200-megahertz microprocessor with 32megabytes of main memory. (Like somepeople, Aibo has its brains in its posteri-or.) Sensors scattered throughout thedog’s anatomy relay signals to the mi-croprocessor, thus mimicking a flesh-and-blood nervous system. To simulate asense of vision, for example, a transceiv-

er in Aibo’s head shines an infrared beamat nearby objects to gauge their distance;this information helps the robot avoidwalking into walls. Aibo can perceive col-ors as well: in its muzzle is a complemen-tary metal oxide semiconductor (CMOS)imager that gives the dog a 40,000-pixelview of the world.

Aibo’s software is equally impressive.Sony designed the robot so that a varietyof programs, each encoded on a memorystick, can be inserted in the dog’s belly. Ifyou choose the Aibo Life program, for in-stance, the robot initially acts like a new-born pup and gradually learns how towalk and do tricks. But if you want a ful-ly trained pooch right away, simply pop


PH

OTO

GR

APH

S B

Y K

EN

KAR

P

Kibbles and BytesHOW MUCH IS THAT ROBOTIC DOGGY IN THE WINDOW? BY MARK ALPERT

MOVE OVER, ROVER, and let Aibotake over. The robotic dog from SonyElectronics walks, whines and wags its tail just likea real pooch. But be warned: this computerizedcanine will take a bite out of your wallet.


in the memory stick containing the Hel-lo Aibo program. The catch is that thememory sticks are not included in the$1,500 price; each costs about $90. Ifyou throw in a spare battery charger andassorted extras, the total price can easilyrun over $2,000.

Perhaps the most intriguing aspect ofAibo’s software is that the robot is imbuedwith canine instincts. We got our first lookat Aibo’s mercurial character just secondsafter George took the dog out of the boxand turned it on. (For the purposes of ourreview, Sony had equipped the machinewith a fully charged battery pack and theHello Aibo memory stick.) In response toGeorge’s enthusiastic petting, the mecha-nized mongrel nipped one of his fingers.The bite didn’t hurt—Aibo has no teeth,just a motorized plastic jaw—but just thesame, George gave the junkyard robot afirm rap on the top of the head. Aibo hasa touch sensor there, and giving it a sharptap admonishes the creature: the dog isprogrammed to avoid the behavior thatled to the reprimand. Aibo’s tail droopedas a sign of contrition, and red LEDsgleamed forlornly behind the visor of thedog’s helmetlike head.

By this time several other members ofthe magazine’s staff were peeking intoGeorge’s office to see what all the fusswas about. Mark Clemens, one of our as-sistant art directors, cheered up Aibo bygently stroking the touch sensor on itshead. Green LEDs behind the visor flasheda happier expression, and a miniaturespeaker in the dog’s mouth warbled ajaunty sequence of high-pitched notes.Then Aibo began walking across theroom, stopping every now and then tosurvey its surroundings.

It was fascinating to watch the robotscramble across the carpet. Each of itslegs has three motors, one in the knee andtwo in the hip socket, allowing the ma-chine to maneuver around obstacles skill-fully. As the robot steps forward, it shiftsits weight from side to side to keep itsbalance. Although Aibo’s herky-jerkymovements have the restless, playful qual-ity of a dogtrot, they also seem vaguely

menacing. On seeing Aibo for the firsttime, many of my colleagues remarked,“It’s so cute!” But on reflection severaladded, “It’s also kind of creepy.”

Aibo’s neatest trick is chasing the pinkball that comes with the dog. Sony has pro-grammed the robot to become terrifically

excited when its CMOS imager detectsthe colors pink or red. As soon as Aibospots the ball, the LEDs in its tail and vi-sor light up and the miniature speakertrumpets the familiar “Charge!” tune.The robot rushes toward the ball, whir-ring like mad, until it senses that it’s in the


right position. Then it lifts one of its frontlegs and gives the ball a kick. Aibo scansthe room to see where the ball went andruns after it again, sometimes butting thething with its head instead of kicking it.

Some of Aibo’s other feats, however,failed to awe me, perhaps because the ro-bot’s lifelike behavior had unduly raisedmy expectations. The machine has twomicrophones in its head and voice-recog-nition software that can respond to sim-ple commands. If you tell Aibo to say hel-lo, it will either bow or wave its paw. Ifyou say, “Let’s dance,” the robot willplay a snatch of music and do a four-legged jig. But very often I had to shoutthe commands a few times before the dig-ital dog would pay attention. Because Ai-bo is designed to be moody, it won’t per-form tricks unless it’s in the proper frameof mind. Sony says the occasional disobe-dience makes the machine more lifelike,but I just found it annoying. If you shell

out $1,500 for the cur, it should at leastshow a little respect.

Clearly, my Aibo needed to go toobedience school. I opened the dog’sstomach cover and replaced the Hello Ai-bo memory stick with the Aibo Life pro-gram. An old robot can’t learn new

tricks, I thought, so I needed to workwith a puppy. But when I turned on themachine, it just lay on the floor with itslegs splayed, making pathetic mewlingnoises. As it turns out, the newborn Aiboneeds at least 40 hours of training to de-velop into a fully functional adult.

I might have undertaken thetask if I didn’t have a wife, a kid ora job. But instead I turned back tomy computer screen while the babyAibo sprawled under my desk, cry-ing its little LEDs out. In the end, Ihad to put the poor creature tosleep. Although the robot is a tech-nological marvel, it didn’t hold myattention for very long. Mind you,this is definitely a minority opin-ion—tens of thousands of customershave each plunked down a small

fortune to adopt Aibo. I suspect that itappeals to people who have a lot of timeon their hands.


TECHNICALITIES

AIBO’S BEST TRICK is chasing and kicking a pink ball across the room. Unfortunately, themechanical mutt isn’t programmed to fetch slippers.


PUZZLINGADVENTURES


JOH

N M

cFAU

L

Five people are in a room. One is a consistenttruth teller, who will answer every questiontruthfully, and four are alternating liars,who will alternate between telling the truthand lying: true, false, true, false, true, false,. . .Unfortunately, you do not know whetherthe alternating liars will start by telling thetruth or not. In fact, an alternating liar maynot decide whether to answer your firstquestion truthfully or falsely until he hearsthe question. But after his first answer, hemust alternate truth with lies. You alsoknow that everyone in the room knowswho the consistent truth teller is.

All the suspects look equally untrust-worthy. You, however, are to determinewhich one is the consistent truth teller. Youare allowed only two questions, althoughthey need not be yes/no questions. You havefive people to choose among. You must ad-dress each question to one person (althoughboth questions can be addressed to the sameperson), and you will be answered by thatperson only. Can you do it? (To figure outhow to approach the problem, see thewarm-up puzzle for three people, usingyes/no questions, illustrated at right.)

For extra credit, here’s an even tricki-er problem: How many questions wouldyou need if there were seven people in theroom?

Dennis E. Shasha, professor of computerscience at the Courant Institute of New York University, creates and solvespuzzles for a living.

Alternating Liars BY DENNIS E. SHASHA

Answer to LastMonth’s Puzzle: The fish vandal mustbe in bungalow G.Because the vandalwalked eachpathway exactlyonce, he couldn’thave ended up in anyof the bungalowswith an even numberof pathways: eachtime he visited one ofthese bungalows, hehad to use onepathway to enter andanother to leave.Only D and G aretouched by an oddnumber of pathways.The vandal enteredthe resort from thelagoon side, so hemust have firstvisited bungalow Dand then made hisway to G.

Web Solution:For a peek at theanswer to thismonth’s problem,visit www.sciam.com

WARM-UP PUZZLE: Who is the consistent truth teller?

THREE PEOPLE

THREE YES/NO QUESTIONS

1) “A, are you a consistent truth teller?” YES



2) “A, are you a consistent truth teller?” NO

1) “A, are you a consistent truth teller?” NO

3) “A, is B a consistent truth teller?” YES

2) “A, is B a consistent truth teller?” YES

A is the consistent truth teller.

A B C

A B C

A B C

A B C

A is an alternating liar, but he is tellingthe truth this time. His next responsewill be a lie, so you ask:

A is an alternating liar, and his nextresponse will be a lie, so you ask:

C is the consistent truth teller.

C is the consistent truth teller.


ANTIGRAVITY

Musky, floral, “etheral,” camphoraceous,minty, pungent and putrid. No, these arenot the names of the Seven Dwarfs beforethey went Hollywood. Rather they arethe seven scents by which NASA cali-brates and certifies the nose of GeorgeAldrich every three months. If everythingis up to snuff, George Aldrich smells.

At the White Sands Test Facility in LasCruces, N.M., Aldrich and other volun-teers put their noses to the grindstone andsmell materials destined for spaceflight.First, rigorous instrumental analysis de-termines whether any gases coming offthese objects are toxic or carcinogenic.(Coincidentally, Aldrich does these testsas well, in his role as a chemistry labora-tory technician.) Once the gases aredeemed safe, Aldrich and a four-colleague“odor panel” sniff the stuff. They decidewhether the object would make an astro-naut die for an open window, which is ex-actly what would happen 200 miles up.

Aldrich is the dean of about 25 smell-ing employees at White Sands, rangingfrom secretaries to engineers, who rotatein groups of five to lend NASA their noses.He started smelling back in 1974 andhad lived through a record 743 “smellmissions”—over 100 more than anyoneelse as of early April, when we spoke.

When a sample comes up for review,Aldrich and his co-panelists smell it andthen rank the rankness from 0 to 4. “0 isa nondetect,” he explains, “1 is barely de-tectable, 2 is easily detectable, 3 is objec-tionable, and 4 is get-me-out-of-here.” Of-ficially, 4 is “offensive.” One may wonderwhy NASA has not switched to some high-tech electronic nose or to the supersensi-

tive snouts of dogs. “We’re a screening testfor astronauts, and they’re human,”Aldrich points out, “so they want to usehuman subjects.” Not to mention the factthat some dogs might rate a dead squirrelon a sizzling Albuquerque highway as a 2,with the additional comment, “Needs onemore day.” Electronic noses are gettingbetter [see “Plenty to Sniff At,” by MiaSchmiedeskamp, Scientific American,March], Aldrich notes, “but they can’t tellyou if an odor will leave an aftertaste inthe back of your throat.”

An aftertaste was just one bad thingabout some Velcro straps that got hurriedup to the space station in February with-out the all-important olfactory evalua-

tion. The smell, reminiscent of your fin-gers after slicing onions, almost made theastronauts launch their lunch. “One ofthe astronauts opened the bag that thesestraps were in,” Aldrich says. “It stunk.He zipped the bag right back up and putit back on the shuttle. They brought it tous, and we basically agreed. It stunk.”(The nefarious funk most likely emanat-ed from the compound used to bond thebacking to the straps.)

Aldrich recently nosed his way intobeing a judge at the annual Odor-EatersRotten Sneaker contest, held in Vermontin March. The winner this year was un-questionably 11-year-old Rebekah Fahey,also of Las Cruces. To anyone smelling a

New Mexican fix, Aldrich explains thatFahey was truly committed to victory:“She’d worn the same socks for threemonths. It just completely knocked meback. I asked for oxygen.”

Some men are born great, some be-come great, and some have greatnessthrust up their noses. “Everybody thinksthat I have this big schnozz,” Aldrich re-marks. “People say, ‘Thank goodness Itook a bath today.’ But I tell everyone Imay be average to slightly above average.”Well above average is his willingness to en-dure, ad nauseam, people in effect asking,“Does this smell funny to you?” Thanksto him and his fellow smellers, astronautsin space can breathe a little easier.


ILLU

STR

ATIO

N B

Y M

ATT

CO

LLIN

S; P

HO

TOG

RAP

H B

Y FR

ANK

VE

RO

NSK

Y

NostrildamusIN SPACE, NO ONE CAN HEAR YOU GAG. WHICH IS WHY YOU NEED PREDICTIONS ABOUT WHETHER SPACE STUFF STINKS BY STEVE MIRSKY

They decide if the object would make an astronaut die for an open window, which is

exactly what would happen 200 miles up.


ENDPOINTS

Mark Ford, fabrication development manager at AFG Indus-tries, Inc., explains:

Tempered glass is about four times stronger than ordi-nary, or annealed, glass. And unlike annealed glass, whichcan shatter into jagged shards, tempered glass fractures intosmall, relatively harmless pieces. It is therefore used in envi-ronments where human safety is an issue. Applications in-clude side and rear windows in vehicles, entrance doors,shower and tub enclosures, racquetball courts, patio furni-ture, microwave ovens and skylights.

To prepare glass for the tempering process, it must firstbe cut to the desired size. (Any fabricating operations, suchas etching or edging, that take place after the heat treatmentrun the risk of reducing the strength of the glass or causingproduct failure.) The glass is then examined for imperfec-tions that could cause breakage at any step during temper-ing. An abrasive—such as sandpaper—takes sharp edges offthe glass, which is subsequently washed.

Next, the glass undergoes a heat treatment in which ittravels through a tempering oven. The oven heats the glassto more than 600 degrees Celsius. (The industry standard is620 degrees C.) The glass then undergoes a high-pressurecooling procedure called quenching. During this process,which lasts just seconds, high-pressure air blasts the surfaceof the glass from an array of nozzles in varying positions.Quenching cools the outer surfaces of the glass much morequickly than it does the center. As the center cools, it tries topull back from the outer surfaces. As a result, the center re-mains in tension, and the outer surfaces go into compression.

This compression is what gives tempered glass itsstrength. Glass in tension breaks about five times more eas-ily than it does in compression. Annealed glass will break

at 6,000 pounds per square inch (psi). Tempered glass gen-erally breaks at approximately 24,000 psi.

Another approach to making tempered glass is chemicaltempering, in which chemicals exchange ions on the surfaceto create compression. But because this method costs farmore, it is not widely used.

For answers to many other questions, visit Ask the Experts(www.sciam.com/askexpert).

Q

Scientific American (ISSN 0036-8733), published monthly by Scientific American, Inc., 415 Madison Avenue, New York, N.Y. 10017-1111. Copyright © 2001 byScientific American, Inc. All rights reserved. No part of this issue may be reproduced by any mechanical, photographic or electronic process, or in the form of aphonographic recording, nor may it be stored in a retrieval system, transmitted or otherwise copied for public or private use without written permission of the publisher.Periodicals postage paid at New York, N.Y., and at additional mailing offices. Canada Post International Publications Mail (Canadian Distribution) Sales AgreementNo. 242764. Canadian BN No. 127387652RT; QST No. Q1015332537. Subscription rates: one year $34.97, Canada $49, International $55. Postmaster: Sendaddress changes to Scientific American, Box 3187, Harlan, Iowa 51537. Reprints available: write Reprint Department, Scientific American, Inc., 415 Madison Avenue,New York, N.Y. 10017-1111; (212) 451-8877; fax: (212) 355-0408 or send e-mail to [email protected] Subscription inquiries: U.S. and Canada(800) 333-1199; other (515) 247-7631. Printed in U.S.A.

How is tempered glass made?—Fred Mueller, Torrington, Conn.

PAT

RIC

K H

ARD

IN


Documents

The Paradox CORONA Sun’sof the Hotthe-eye.eu/public/concen.org/Scientific American...through a gap in his front teeth created in his teens when, in keeping with Himba tradition,