C ALIFORNIA

INSTITUTE OF

TECHNOLOGY engineeringE& andS science

Volume LXIV, Number 1,    

IN THIS ISSUE

An ice-cold snowflake

A lukewarm universe

A long, hot summer? Peering through the dust that enshrouds our galaxy, the Two Micron All Sky Survey (2MASS) provides a clearer view of the center of the Milky Way than ever before seen. The actual center in this mosaic of more than a million stars is the bright reddish spot near the upper left. An article describing more of what this survey saw at near-infrared wavelengths begins on page 20.

2 ENGINEERING & SCIENCE NO . X     engineering& and science C a l i f o r n i a Institute

of Technology Volume LXIV, Number 1, 2001 E S

2 Random Walk 10 Morphogenesis on Ice: The Physics of Snow Crystals — by Kenneth G. Libbrecht There’s more to snowflakes than meets the eye.

20 A Billion Points of Heat — by Jane Dietrich Cool stars and hidden galaxies emerge from the dust in a highly accurate, near-infrared sky survey.

28 Economic Policy in the Information Age — by Simon J. Wilkie The California power crisis shows what can happen when you apply 19th-century economic principles to a 21st-century economy.

Reviews — QED, Oxygen, Ahead of the Curve On the cover: Small snow 37 crystals, each about one Obituaries: Jeffrey Scot Banks millimeter in diameter, 40 grow on the ends of thin Faculty File ice needles in Ken 42 Libbrecht’s lab. For more on how an infinite variety of shapes can come from a

very simple molecule, see Engineering & Science (ISSN 0013-7812) is published Ted Jenkins the story on page 10. quarterly at the California Institute of Technology, 1200 President of the Alumni Association East California Boulevard, Pasadena, CA 91125. Annual J. Ernest Nunnally subscription $10.00 domestic, $20.00 foreign air mail; Vice President for Institute Relations single copies $3.00. Send subscriptions to Caltech 1-71, Robert L. O’Rourke Pasadena, CA 91125. Third class postage paid at Pasade- Associate Vice President for Institute Relations

na, CA. All rights reserved. Reproduction of material con- STAFF: Editor — Jane Dietrich tained herein forbidden without authorization. © 2001, Managing Editor — Douglas Smith California Institute of Technology. Published by Caltech Writer/Editor — Barbara Ellis and the Alumni Association. Telephone: 626-395-3630. Contributing Writers — Robert Tindol, Mark Wheeler PICTURE CREDITS: Cover, 10, 14–19 — Ken Libbrecht; inside Copy Editors — Emily Adelsohn, front cover, 20–27 — 2MASS; 2 — Robert Tindol; 4, 11 — Mark Michael Farquhar, Elena Rudnev Garcia; 4, 7, 10–11, 12, 14–16 — Doug Cummings; 6 — Bob Business Manager — Debbie Bradbury — Susan Lee Paz; 7 — Mika Nyström; 8 — Nathan Litke; 9 — Erik Winfree; Circulation Manager Photographer — Robert Paz 28, 34–35 — Edison International; 30, 31 — Doug Smith; 37 — Graphic Artist — Douglas Cummings Craig Schwartz; 38 — Ken Jacques; 40–41 —Shannon Banks; 43 — Joe Umbro; inside back cover — Margaret DePrima Visit Caltech on the Web at http://www.caltech.edu Random Walk

F LY L IKE AN E GYPTIAN

When you think about building the Egyptian pyramids, you probably have a mental image of thousands of slaves laboriously rolling massive stone blocks with logs and levers. But as one Caltech professor is demon- strating, the task may have been accomplished by just Although blustery winds added some excitement—the chief flyer, kite- four or five guys who flew the surfing instructor Eric May (middle of picture, dark shorts), got lifted off stones into place with a kite. the ground on the first attempt—the obelisk went upright in just 25 On June 23, Professor of Aeronautics Morteza (Mory) seconds on the second try. Graff (plaid shirt) handles a control line. The Gharib (PhD ’83) and his obelisk was provided by Daniel Correa (center, foreground) of Incablock, team raised a 6,900-pound, which makes concrete blocks. 15-foot reinforced-concrete obelisk into a vertical position in the desert near Palmdale, using nothing more than a kite, a pulley C ALIFORNIA : W INE , M OVIES , A VOCADOS , AND N OBEL L AUREATES system, and a support frame. The team eventually hopes to show that even 300-ton mon- uments—not to mention the far-less-massive building Not only does the State afternoon of October 24. by members of the Swedish blocks of Egypt’s 90-odd of California have the fifth President David Baltimore royal family and the Swedish pyramids—could have been largest economy in the world (physiology/medicine ’75), ambassador. It’s preceded by raised with a fraction of the (having recently passed up Rudy Marcus, the Noyes Pro- a morning symposium at effort that modern researchers France), it also has the largest fessor of Chemistry (chemis- UCLA, which has a few have assumed. concentration of Nobel laure- try ’92), and Ed Lewis, the Nobel laureates of its own. Gharib, whose primary ates. And, because the year Morgan Professor of Biology, On October 25 a Centennial research interest is the nature 2001 marks the 100th anni- Emeritus (physiology/medi- Celebration luncheon will be of fluid flow, has been work- versary of the Nobel Prize, cine ’95) will introduce held at the California Science ing on the project since local California will be celebrating speakers at the symposium. Center in downtown Los business consultant Maureen and paying tribute to its (Ahmed Zewail, the Pauling Angeles, chaired by Gayle Clemmons contacted Caltech laureates and to the signifi- Professor of Chemical Physics Wilson, the state’s former two years ago. Clemmons cance of science and technol- and professor of physics, who first lady and a member of had seen a picture in Smith- ogy in the state. Which no won the chemistry Nobel in Caltech’s board of trustees. sonian magazine in 1997 of doubt has something to do 1999, will unfortunately be On October 26, the celebra- an obelisk being raised, and with its economy. out of town.) Other Caltech tion moves northward to San thought that the Egyptians Of California’s 94 Nobel- speakers at the event include Franciso, where there will be could have used kites to ists, Caltech counts 27 among Richard Andersen, the Bos- a symposium at the Explora- accomplish the task more its faculty and alumni (that’s well Professor of Neurosci- torium. A banquet at City easily. All she needed was actually 28 prizes, since Linus ence, and Professor of Physics Hall follows on the 27th. an aeronautics expert with Pauling won two). Three of Andrew Lange. A reception The Nobel Centennial the proper credentials to the four faculty members still at the Athenaeum will follow. is also sponsoring an essay field-test her theory. It is a active on campus will be Caltech’s symposium is contest for junior- and senior- credit to her determination taking part in the Nobel only a portion of the Califor- high-school students and that the tests are occurring— Centennial Symposium in nia Nobel Prize Centennial establishing Centennial with no scientific or archaeo- Beckman Auditorium on the 2001, which will be attended Scholarships. ■ logical training, she has

2 ENGINEERING & SCIENCE NO . 1     F LY L IKE AN E GYPTIAN

P UT S OME C ESIUM IN Y OUR T ANK managed to marshal the idea if the ancient Egyptians efforts of family, friends, actually moved stones or and fellow enthusiasts. anything else with kites and Even today, moving heavy pulleys, but Clemmons has stones without power equip- found some tantalizing hints Gasoline averaging $3 per (sulfuric acid), and a normal

ment is quite labor-intensive. that the project is on the gallon? Oil drilling in an salt, such as K2SO4 (potassi- In 1586, the Vatican moved right track. On a building Alaskan wildlife reserve? A um sulfate). Solid acids can a 330-ton Egyptian obelisk to frieze now displayed in a need to relax air quality stan- conduct electricity at rates St. Peter’s Square. Lifting the Cairo museum, there is a dards? It seems the long- similar to polymers, they stone upright took 74 horses wing pattern in bas-relief that term future of fossil fuels is don’t need to be hydrated, and 900 men. For Gharib, does not resemble any living bleak. One promising solu- and they can function at the idea of accomplishing bird. Directly below are tion scientists have been temperatures up to 250 °C. heavy tasks with limited several men standing near studying is fuel cells, which They are also typically inex- manpower is appealing from vertical objects that could have their limitations too. pensive and easy to manufac- an engineer’s standpoint be ropes. And she has dis- But in the April 19 issue ture. But until now solid because it makes more logis- covered that a brass ankh— of Nature, Caltech Assistant acids had not been examined tical sense. “It’s one thing long assumed to be merely Professor of Materials Science as fuel-cell electrolytes to send thousands of soldiers a religious symbol—makes Sossina Haile reports on a because they dissolve in to attack another army on a a very good carabiner for new type of fuel cell that water; worse, they can lose battlefield,” he says. “But an controlling a kite line. may resolve these problems. their shape at even slightly engineering project requires Gharib next plans to raise Unlike an automobile elevated temperatures. To everything to be put precisely a 10-ton stone, then perhaps engine, where a fuel is burned solve these problems, Haile into place. I prefer to think a 20-ton one. Eventually, and expanding gases do the and her colleagues operated there were relatively few they hope to receive permis- work, a fuel cell converts the fuel cell at a temperature people involved.” sion to raise one of the chemical energy directly into above the boiling point of The concept Gharib devel- obelisks still lying in an electrical energy. Fuel cells water, and used a solid acid,

oped with SURF (Summer Egyptian quarry. “The whole are pollution-free, and silent. CsHSO4 (cesium hydrogen Undergraduate Research approach has been to down- The fuel cells in today’s sulfate), that is not very prone Fellowship) student Emilio grade the technology,” Gharib prototype cars are usually to shape changes. Graff, a senior in aeronautics, says. “We first wanted to based on polymer electrolytes. The next challenges, says is to build a simple tower show that a kite could raise a (An electrolyte is a nonmetal- Haile, are to reduce the elec- around the obelisk, with a huge weight at all. Now that lic substance that conducts trolyte’s thickness, improve pulley system mounted some- we’re raising larger and larger electricity.) Polymer electro- the catalyst’s performance, what forward of the stone’s stones, we’re also preparing to lytes must be humidified in and, most importantly, pre- tip. That way, the base of the replace the steel scaffolding order for the fuel cell to func- vent the reactions that can obelisk will drag the ground with telephone poles and the tion, and can only operate occur upon prolonged expo- for a few feet as the kite lifts steel pulleys with windlasses over a limited temperature sure to hydrogen. Still, she the stone, and it will be quite like the ones that may have range. Thus these fuel cell says, solid-acid fuel cells are stable once it reaches the been used on Egyptian ships. systems require many auxilia- a promising development. vertical. If the obelisk were In fact, we may not even need ry components and are less “The system simplifications raised with the base as a a kite. It could be we can get efficient than other types that come about [in compari- pivot, the stone would tend along with just a drag chute.” of fuel cells. son to polymer electrolyte to swing past vertical and fall Steady winds of up to 30 Haile’s laboratory has fuel cells] by operating under the other way. The kite rope miles per hour are not unusu- developed an alternative essentially dry and mildly is threaded through the al in the areas where the pyra- type of fuel cell based on a heated conditions are tremen- pulleys and attached to mids and obelisks are found. so-called “solid acid.” Solid dous. While there is a great the obelisk’s tip. A couple The wind in Palmdale gusted acids are chemical com- deal of development work

of fliers steer the kite with to over 20 m.p.h., although a pounds, such as KHSO4 that needs to be done before guide ropes, moving it in 12-m.p.h. breeze would have (potassium hydrogen sulfate), solid-acid-based fuel cells can figure-eights for maximum sufficed. ■—RT whose properties are interme- be commercially viable, the sustained lift. diate between those of a potential payoff is enormous.” ■ Of course, no one has any normal acid, such as H2SO4 —MW

    ENGINEERING & SCIENCE NO . 1 3 2 0 0 1 — A M ARS O DYSSEY

As the Red Planet looms regions that are warmer than large in the midnight sky this their surroundings—poten- summer, another JPL space- tial “hot springs,” at least craft has set out to meet it. by Martian standards. The 2001 Mars Odyssey The GRS (Gamma Ray lifted off from Cape Canaveral Spectrometer), provided on April 7, and is slated to by the University of Arizona, arrive on October 24. Its will scout out 20 chemical primary mission, to run from elements, including silicon, January, 2002 through July, oxygen, iron, magnesium, 2004, is to map the composi- potassium, aluminum, cal- tion of the planet’s surface, to cium, sulfur, and carbon, on detect water and shallow sub- the Martian surface. A pair of surface ice, and to study the neutron detectors in the GRS radiation environment. The package (built by the Los spacecraft carries three pri- Alamos National Laboratory mary instruments. and Russia’s Space Research THEMIS (Thermal Emis- Institute) will allow the sion Imaging System) is a amount of hydrogen present visible/infrared camera in the top meter of soil to be provided by Arizona State calculated, which in turn acts University that by day will as a proxy for determining map the distribution of the potential amount and dis- minerals, particularly those tribution of Mars’ water ice. that can form only in the And MARIE (Mars Radia- presence of water. At night, tion Environment Experi- it will look for active thermal ment), led by NASA’s John- In mid-May, the Division of Geological and Planetary Sciences commemo- rated the 75th anniversary of its founding (as the department of geology and paleontology under John Buwalda) with a reunion. About 100 alumni came for two days of talks by division faculty on their current research, as well as other sorts of celebrating, including this spectacular cake with a very literal “ring of fire.” The following month, on June 24, Bob Sharp (BS ’34, MS ’35), the Sharp Professor of Geology, Emeritus, celebrated his 90th birthday. Sharp has been a member of the faculty since 1947 and was chair of the division from 1952 to 1968.

Not featured at the GPS reunion was the fine vintage at right, bottled for the neigh- boring Division of Physics, Mathematics and Astronomy in honor of SIRTF (the Space InfraRed Telescope Facility; see page 26). Infra Red wine comes in three “wave- lengths”—3.6 microns, 30 microns, and 160 microns—indicating a range from light bodied to full bodied. It’s reportedly “pretty good.” Professor of Physics Tom Soifer (BS ’68), director of the SIRTF Science Center, has enough Infra Red in his wine cellar to last until the satellite’s launch in 2002.

4 ENGINEERING & SCIENCE NO . 1     son Space Center, will assess the radiation hazards to future human explorers. Cos- mic rays emitted by the sun and other stars can trigger cancer or damage the cen- tral nervous system; similar radia-tion monitors have been flown on the Space Shuttles B INOCULAR V ISION : and on the International “F IRST F RINGE ” AT THE K ECKS Space Station, but none has ever ventured beyond Earth’s protective magnetosphere, which shields us from much of this radiation. Astronomers’ vision got a cally, giving information The orbiter is also designed whole lot sharper on March on its temperature, pressure, to act as a communications 12, when at 10:40 p.m. and atmospheric composition. relay for future Mars landers, Hawaiian Standard Time An interferometer has to including JPL’s pair of Mars the W. M. Keck Observatory align the incoming signals so Exploration Rovers, to be became an interferometer. that their peaks and troughs launched in 2003. ■ The two 10-meter Keck match up to within a very telescopes atop the summit small fraction of a wave- of Mauna Kea successfully length. This is easy enough pooled the light received with radio waves, which are a from a star in the constella- centimeter or more in length, tion Lynx known as HD but a very tough challenge A R EALLY N E AT D ISCOVERY 61294, attaining what astron- with light, where the waves omers refer to as “fringes.” are measured in millionths Interferometry, which has of a meter. And it’s further long been a staple of radio complicated by atmospheric astronomy, means that the turbulence, which causes the signals from two or more star (and its thousandfold JPL’s NEAT (Near-Earth Asteroid Tracking) program’s newest telescopes are combined dimmer companion) to shim- telescope, the just-refurbished 1.2-meter-diameter (48-inch) to create a virtual telescope my disconcertingly. Each Oschin telescope at Caltech’s Palomar Observatory, officially whose dish—or in this case image wanders around on its bagged its first asteroid on May 16. Better yet, the catch was a mirror—is the size of the detector independently from Potentially Hazardous Asteroid—one of about 300 now known separation between the its twin at the other tele- whose orbit crosses Earth’s. Provisionally named 2001JV1, it is instruments, enabling you to scope, continuously altering about 0.7 kilometers (0.4 miles) in diameter, so it could leave a discriminate between objects the baseline separation. nasty welt. But don’t get out the tinfoil hats or the beach that are exceedingly close The Keck Interferometer umbrellas just yet—“potentially hazardous” means it would together. In this case, the sends an image of the target have to be significantly deflected from its current orbit to do telescopes are 85 meters (93 star (or a bright, nearby guide us any harm. yards) apart, and the goal is star) from each telescope to a Since its inception in December, 1995, NEAT has found to see warm, Jupiter-sized fast-readout infrared camera roughly 100,000 asteroids, including about 100 near-Earth planets orbiting around called the Keck Angle Track- asteroids; NASA’s goal is to find all of the estimated 700 to nearby stars directly, rather er that sends commands back 1,500 asteroids larger than 1 kilometer (0.6 mile) that approach than inferring their existence to adaptive-optics systems to within 48 million kilometers (30 million miles) of Earth, and to from the wobbles their gravi- compensate. The two star do so by 2008. About 500 have been detected so far. The vast tational tugs induce on their images are thus kept centered majority of these are harmless, but a tiny percentage have orbits parent stars. This doesn’t on a fiber-optic line that feeds that could eventually put them on a collision course with Earth. mean we’ll be able to take another fast readout IR cam- The Oschin telescope, built in 1947, has been used for two the planet’s picture, but with era called the fringe tracker. landmark sky surveys, the second of which was completed in luck, the distinction between The fringe tracker adjusts the 2000. Its half-million-dollar upgrade, sponsored by NASA/JPL, the star and the planet will be “fast delay lines” that bounce has turned it into a fully automated facility with a computerized clean enough that the planet the light through a system of pointing system and a state-of-the-art CCD camera. ■ can be studied spectroscopi- adjustable prisms and mirrors

    ENGINEERING & SCIENCE NO . 1 5 F ROM P UNCH C ARDS TO P ALM P ILOTS

to align the waves, while simultaneously compensating Caltech’s computer-science for the earth’s rotation. A option turned 25 this year. peak occurs in the fringe At Caltech, a birthday party tracker when the paths are means a symposium, so there identically matched, and a were two days’ worth of minimum occurs when the speakers on fields where Cal- paths are different by one-half tech has left its mark: chip the wavelength of the light. design, parallel supercomput- When the peak-to-minimum ers, networking, and com- ratio exceeds a certain thresh- puter graphics, as well as a old value, fringes have been peek at what may lie beyond seen. silicon. Here are some HD 61294 was the first highlights. of about 20 stars that were Option cofounder Ivan locked on to during the Sutherland (MS ’60), then engineering run, which Jones Professor of Computer consisted of the first halves of Science, now vice president the nights of March 12 to 14. of Sun Microsystems, talked The fringes would last for up As new graduates sweltered under a broiling sun and parents and other about the early days. In the to 10 seconds at a time, and Caltech tradition, the fledg- for about 10 percent of the wellwishers huddled under a sea of parasols, speaker Gordon Moore, PhD ling program had to “pick total duration that each star ’54, former chair of Caltech’s board of trustees and Intel Corporation, one thing and do it well.” was tracked, which varied provided some advice from snowboarding: “Stay low and be confident as They opted for VLSI, or from 10 to 30 minutes. Now you move forward.” At Caltech’s 107th commencement on June 15, 204 Very Large Scale Integration, the challenge is to fine-tune because cofounder Carver the system to lock onto the new owners of BS degrees, 120 MS degrees, one Engineer, and 159 PhDs Mead (BS ’56, MS ’57, fringes for long enough to were ushered out into the real world. PhD ’60), Moore Professor make useful measurements, of Engineering and Applied a teething period that is Science, Emeritus, was one of expected to take the rest of its fathers and would shortly this year. Limited science write The Book on the sub- operations, including looking ject. VLSI, which allows you at these planets and the dust to put an entire circuit—or, rings from which planets these days, millions of tran- condense, may begin early sistors—on a silicon chip, is next year. now such a basic part of life The Keck Interferometer that it doesn’t merit a second is funded by NASA’s Origins thought. But back then, it program, and is a collabora- was revolutionary, and the tion between Caltech, the notion that grad students University of California, could actually design, build, and JPL/NASA. ■—DS and test several generations

6 ENGINEERING & SCIENCE NO . 1     the moment it finishes, rather The Adoption of Technology: The S Curve vs. The Perception of Technology: The Hype Curve One of the symposium’s sessions than flooding the network Market Development Stage: was a panel discussion on Invention Infection Growth Saturation Maturity when the clock says “send.” Such chips also draw less “Entrepreneurship and Computer Sci- Initial euphoria... Now there are s real buyers... power—as much as a quarter ence,” at which Phil Neches (BS Now it is the of the power consumed goes mainstream... ’73, MS ’77, PhD ’83), founder of Tera- to running the clocks. Asyn- data Corporation, compared chronous chips are inherently / Expectation faster, because the chip’s the standard business model’s speed is the average of all the S-shaped curve of market pene- processors’ speeds instead of enetration ...don’t we miss the tration to the EKG-like “hype good old days? the speed of the slowest one. et P ...but it hasn’t taken curve” of expectations typically over the world. The downside is that since Mark ...but it doesn’t work yet. each processor is primed to found in the technology sector. The Buyers Are: accept a message all the time, Heat- Visionaries Early Middle Late Seekers Majority Majority Majority Laggards any “glitch,” or spurious signal, will be interpreted as Time data. (Synchronous systems are immune to this, as only a glitch at the exact moment of chips in time to write ing the data up into streams anticipated for such a mon- that data is due would be a thesis was even more so. and assigning each stream to ster, he replied, “Really cool taken seriously.) Clever At the dawn of the com- its own region or regions; all video games.”) communications protocols are puter age, said Sutherland, the regions can then chew on Asynchronous chips, in needed to keep the data real logic elements (in the form of pieces of the problem simul- which each circuit operates and to manage the data flow vacuum tubes) were expensive taneously without having to at its own pace rather than in general. In fact, Martin and unreliable. Wires, on the talk to each other very much. to the tick of a master clock, titled his talk “Delays Have other hand, were cheap and This reduces the torrent of offer another way around the Dangerous Ends”—a quote reliable. Today, logic (transis- data through the wires (both wiring bottleneck by leveling from Henry VI—and was tors) is cheap and very reli- within and between chips) to out the communications going to subtitle it, “A able, but wires are expensive a manageable flow, alleviating traffic, said Professor of Shakespearean Approach to and very, very bulky. But the traffic jams that would Computer Science Alain VLSI Design,” but “decided “we’re still tied to the mind- otherwise occur as more and Martin. Each asynchronous it would not be appropriate set of 1950s programming, more components are packed processor forwards its results for a Frenchman to do so.” using detailed instruction on a chip. Dally is building sets. Instead, we need to put a one-teraflops (trillions of the programmer in charge of floating-point mathematical moving the data around. operations per second) Addition is simple. Getting machine that fits on a single the operands to the adder is shelf and draws less than a the hard part.” He drew an kilowatt of power. In con- analogy to another techno- trast, the teraflops supercom- logical transition: “What puters up at Lawrence Liver- we’re doing now is copying more National Laboratory are the Roman stonemason’s the size of Beckman Institute arched bridges in wrought Auditorium and use many iron. We need to begin megawatts of electricity. building suspension bridges.” “These days, most of the Most of the rest of the power consumption isn’t in speakers talked about these the processing, it’s in data suspension bridges. Stan- movement.” By 2011, he ford’s William Dally (PhD predicted, using such parallel ’86) discussed designing stream architectures, we chips based on a “parallel could have five teraflops on In the lower left corner of the picture above is a silicon potato chip. It’s stream architecture.” Such a chip and “a machine that also the world’s first asynchronous microprocessor—a 16-bit chip with chips are broken up into won’t require its own [power] many self-contained regions, substation.” A handful of roughly 23,000 transistors, designed in Martin’s lab in 1989. It ran at a each with its own memory those chips could get you 100 then-respectable 17.5 MHz with a 5-volt power supply, drawing 230 and processor units. The idea teraflops, easy—a staggering milliwatts of power. But it will also run happily, albeit some 500 times is to handle as much of the amount of computing power. more slowly, on the 0.9 volts and 40 microwatts obtainable from a nice, computation as possible as (When asked what home or locally as possible, by break- business applications he fresh, juicy potato, as Mika Nyström (MS ’97, PhD ’01) demonstrated.

    ENGINEERING & SCIENCE NO . 1 7 spacecraft. “NEAR [the Chandy calls “screen scrap- Near-Earth Asteroid Rendez- ing”—reading data from vous mission] sent back some someone else’s Web display 160,000 pictures. But the that isn’t formatted the way data product the scientists your computer likes it, but really wanted was a 3-D map interpreting it anyhow and of the asteroid. What if the replying appropriately. spacecraft could do the Warming to that theme, advanced processing and you Hewlett-Packard’s Rajiv could just download the map? Gupta (MS ’87, PhD ’91), That would be equivalent to a protégé of Chandy’s and a 106 data compression.” Mead’s, talked about having The new frontier lies in a chip in your car determine linking computers that need that your transmission is to share some data, but don’t about to blow and automati- want to get too intimate, in cally using your cell phone to networks that evolve as the make an emergency appoint- task dictates. Mani Chandy, ment at the closest dealer- Ramo Professor and professor ship, displaying directions to of computer science, has come it on your in-dash GPS unit, To which an audience mem- together. Caltech has been up with one such system, and booking a taxi to be there ber replied, “So is it a tragedy a pioneer in distributed com- called Infospheres, which he when you arrive so that you or a comedy?” puting since the early ’80s, plans to donate to humanitar- still get to your Very Impor- It was actually a history, when Charles Seitz (a CS ian agencies doing disaster tant Meeting on time. “You beginning with the 1979 faculty member until 1994, relief. Red Cross field work- don’t know or care how the Caltech Conference on VLSI, when he left to found Myri- ers could use their palmtop other components work, or where concurrent processing com, a high-speed network- computers to coordinate food even whose they are. You emerged as a discussion topic; ing company) wired together and clothing shipments, for just need their output. The Martin’s lab built the world’s a boatload of off-the-shelf PC instance, by tapping into industrial revolution removed first asynchronous micropro- processor boards to attain various donors’ inventory people as the bottleneck in cessor a decade later. A dem- supercomputer performances systems, railroad and airline the production of goods; this onstration chip built in 1998 at Radio Shack prices, and schedules, and the truckers’ removes people as the bottle- ran four times faster than the Geoffrey Fox (then professor dispatch centers. “Even neck in the production of commercial two-million- of theoretical physics and poorer countries like India services.” transistor chip it was mim- dean of educational comput- are heavily wireless,” says Yaser Abu-Mustafa (PhD icking, and his lab still holds ing, now at Florida State) Chandy. “Technology leap- ’83), professor of electrical the record for the fastest used it to tackle a gnarly frogs, so they’ve skipped engineering and computer working asynchronous cir- quantum-field problem right past telephone lines.” science, designs systems cuits. (He quoted Carver whose immensity had previ- The system employs what called neural networks that Mead: “There is nothing ously deterred all comers. more useless than a fast Called the Cosmic Cube circuit that doesn’t work.”) because its processors were He’s trying to make them connected like the vertices even more efficient. “If of an n-dimensional cube, you track how the variables it spawned an industry. move—the atomic structure In the early ’90s, JPL of the program, if you will— Senior Scientist and Caltech and then do a spectral analy- Faculty Associate (at CACR, sis to see which operations Caltech’s Center for Advanced ‘cluster,’ according to some Computing Research) Thom- measure, you can optimize as Sterling, then at NASA Schröder’s lab develops methods for compressing and manipulating 3-D geo- your design based on the Goddard, and his colleagues clustering.” Martin, Dally, went one better by using metric information. Once properly encoded, you can treat the data like any and a host of others are creat- commercial network technol- other kind of signal, filtering out noise or changing some aspect of the object. ing programming and chip- ogy and operating systems as Here, pushing up two critical grid points at the corners of the design tools that automati- well to create Beowulf—the cally deal with the complex model for today’s teraflops mouth puts a smile on someone’s face. Schröder sees geometry as the details, leaving the humans computers. Sterling foresees fourth wave of multimedia, following on sound, still photos, and video, all of free to draw the big pictures. a petaflops (quadrillion flops) which are now routinely transmitted over the ’net. Fields as diverse as Another session covered machine by 2010. He pro- archaeology and biomedicine will benefit, and let’s not forget the how you link devices, poses using Beowulfs—on asynchronous or otherwise, a slightly smaller scale—in ubiquitous e-catalogs.

8 ENGINEERING & SCIENCE NO . 1     learn from hint and example tant Professor of Computer Single-walled carbon nanotubes are rather than having to be Science Andre DeHon, a essentially soda straws whose walls programmed. For years, silicon wire slated for produc- are one carbon atom thick. Two Abu-Mustafa has been tion around 2005 is about perpendicular tubes, suspended training them to predict 100 nanometers, or 200 trends in the foreign- silicon atoms, wide; about two billionths of a meter apart, exchange market—a task 44,000 square nanometers can flex under the influence of chosen because “it is rich will be needed to encode static cling to encode zeroes and in data, is very noisy, and for one bit of information. But ones (top). Chips built from arrays which there is no mathemati- a variety of other technologies cal model,” thereby inadvert- offer the possibility of build- of such junctions (bottom) would ently launching the discipline ing molecular wires and be more compact than silicon ones. of computational finance. “If switches one nanometer Computer architect DeHon is I get a good system, you’ll wide; a bit might occupy 400 hear about it. If I get a very square nanometers. DeHon’s collaborating with Harvard chemist Charles Lieber to try to build good system, you won’t hear own work centers on carbon After Lieber, et al, Science, volume 289, page 95, about it,” he joked, going on nanotubes, which with the July 7, 2000. practical circuits this way. to note that the results have, right electrostatic charge will in fact, been published. flex to form binary (on/off) Jim Kajiya, a CS professor logic elements. “This will from 1978 to ’94 and now require a paradigm shift from of Computer Science Leonard Science and Computation assistant director of research the top-down methods of Schulman. The NSF recently and Neural Systems Erik at Microsoft, spoke of the bulk carving and etching to established an Institute for Winfree (PhD ’98), described emergence of computer bottom-up strategies for self- Quantum Information at the first steps toward com- graphics as a medium in assembly, taking into account Caltech, to which Schulman puting using DNA molecules its own right. “Applied the specific characteristics of and several colleagues, whose structures encode a computer sciences extend individual atoms.” Even including one of Caltech’s “program,” and whose self- human capabilities: robotics, more exotically, quantum two most recent MacArthur assembly into an array reads our muscles; memory and computing will attempt to “genius” Fellows, Assistant out the “answer.” Look for computation, our brains; and exploit the bizarre properties Professor of Physics Hideo more on Mabuchi and Win- graphics, our imagination. So of quantum-mechanical sys- Mabuchi (PhD ’98), belong. free in a future issue. ■—DS are graphics just for games for tems to solve problems that And finally, Caltech’s other 14-year-old boys?” No. The ordinary computers can’t MacArthur Fellow, Assistant confluence of computer vision crack, said Associate Professor Professor of Computer and computer graphics becomes computer video— malleable, editable objects you can manipulate on your screen. Peter Schröder, pro- fessor of computer science and applied and computational mathematics, picked up the theme in describing his 3-D modeling research. How do you store and transmit geo- metric information so that a coarse but usable rendition of the object shows up almost immediately, with progres- sively finer detail filling in Left: A DNA computer would use “tiles” that interlock like puzzle pieces. afterwards? How do you search a collection of shapes In this example, the computer counts upward from 1 (in binary numbers) for common themes—to find as the tiles fall into place, beginning from the one labeled “S” in the all the animals, perhaps, or to bottom right corner. Above, left: The first step is to create a lexicon of recognize a face? And can rigid, two-dimensional tiles that act as logic elements by binding to one you put digital objects that were scanned as separate another only in specific configurations. Here, four bits of single-stranded entities together, reassem- DNA self-assemble into double-crossover units that can take either of two bling a vase from its shards? forms, “A” or “B.” A and B, in turn, fit together only one way to form a The final session looked at repeating pattern. Above, right: An atomic-force microscope scan of the means of computing beyond silicon. According to Assis- actual A-B crystal. The B tiles stick up higher, giving it a corduroy look.

    ENGINEERING & SCIENCE NO . 1 9 Morphogenesis on Ice: The Physics of Snow Crystals

by Kenneth G. Libbrecht Nearly everyone is familiar with snow crystals— tiny flakes of ice, remarkably crafted into intricate six-fold symmetric patterns, common icons of ski sweaters and holiday shopping malls. But exactly how do snow crystals develop into such intricate, symmetrical shapes? Where is the sculptor that creates these miniature masterpieces of frozen water, quite literally out of thin air? In my lab at Caltech we have been trying to answer these questions, and more general ones relating to the materials science of crystal growth and the under- lying physics of morphogenesis, or pattern formation. And what we’re finding is that there’s a world of interesting and sometimes puzzling physics in snow crystals and how they form, much of which derives from the unique surface properties of ice. Many people think that snow crystals are made from frozen raindrops, but they’re not—rather they grow from water in the gaseous state. In the cold atmosphere above us, this water vapor condenses directly to form solid ice. The process begins with some nucleus, typically a tiny dust grain, to which water molecules can easily attach. If the humidity of the air is above 100 percent (the air is then said to be supersaturated), water mole- cules freeze onto the dust nucleus, forming a tiny piece of ice, which subsequently grows into a snow crystal as more water molecules condense out of the air. The intricate patterns we see are the result of just how water molecules selectively attach to the growing crystal. The actual ice in a snow crystal is no different from that in an ordinary ice cube, and, at the molecular level, we understand the physics quite well: quantum mechanics tells us how two hydrogen atoms and one oxygen atom form a A Caltech-designed snow single water molecule, and how water molecules crystal heads a flurry of line up and fit together to form a hexagonal crystal "wild" ones caught by lattice (opposite). The basic molecular structure of ice was first proposed by Linus Pauling in 1935, photographer Patty and was one of the early triumphs in the applica- Rasmussen. tion of quantum physics to materials science.

10 ENGINEERING & SCIENCE NO. 1  Build a better snowflake, and the world will shovel a path to your door.—KGL

It’s this hexagonal symmetry of the ice crystal the air can readily attach to these rough that is ultimately responsible for the six-fold surfaces, and they grow relatively symmetry of the snow crystals that fall from the quickly. The facet planes are special, sky. But just how do molecular forces, acting at however, in that they tend to be the subnanometer scale, control the shape of a smoother on a molecular scale, with snow crystal ten million times larger? This same fewer dangling bonds. Water molecules question applies to all crystals that form facets— cannot so easily attach to these smoother flat surfaces that define the crystal shape—such as surfaces, and so the facets grow more the mineral specimens shown below. slowly. After all the rough surfaces have Facets have an interesting history in human grown out, what's left are the slow- society, which has led to some confusing expres- growing faceted surfaces. sions in our language. Many minerals in the earth A big problem in crystal growth, and grow into beautiful faceted shapes, and in early not just for ice, is that quantum mechanics cannot times these crystals were quite valuable. Since really tell us how fast a given surface will grow. It Snow crystals are six-sided nice mineral specimens are rare, people naturally could in principle; but in practice, for essentially started carving facets into other materials, particu- all real surfaces, the problem is exceedingly because water molecules larly glass (in which the molecules are randomly difficult. Supercomputer simulations can produce bond to form a hexagonal arranged, in contrast to the regular arrangement of molecular models of crystals, but typically these ice crystal lattice, above. atoms in a crystal). So now one can go to Macy’s elaborate models are not useful for modeling and buy a piece of fine “crystal”—which is in fact growth rates, because the thermal motions of glass—with facets cut into it. Diamonds and molecules are very fast, while crystal growth is other gemstones are crystalline materials, but here quite slow. A typical timescale for molecules in a again the facets are usually of human origin. An crystal to jiggle back and forth is on the order of amusing recent development has come from the picoseconds (10–12 seconds), whereas the timescale superstition that quartz (and other) crystals possess for crystal growth is far longer—typically micro- mystical healing powers, which has greatly seconds (10–6 seconds) or more. increased the demand for attractive mineral Of course we don’t need to keep track of every specimens. Demand begets supply, and lately I’ve molecular wiggle, and over the past five decades seen “fake” quartz crystals in stores—real quartz, many excellent theoretical techniques have been but with artificially made facets, cut to look like developed to describe the statistical mechanics The crystals below show natural faceted quartz. I can’t help but speculate behind crystal growth. Much of this has been the natural faceting that is that fake facets must diminish the healing powers, driven by commercial concerns, as semiconductor also a feature of snow but I haven’t explored the matter further. crystal growth is the foundation of a very, very big crystals. From left, blue Getting back to the natural world, the reason business, and ever more sophisticated electronic many crystals grow into faceted shapes is simply and optical devices require the growth of ever halite, (genuine) quartz, because some crystalline surfaces grow more more complex layered crystals. However, despite tourmaline, and fluorite. slowly than others. And this in turn arises from this huge, the molecular structure of the crystal. For commercially example, if we imagine beginning with a small, funded research round ice crystal, we would find that the surface effort, crystal was quite rough on a molecular scale, with lots of growth remains dangling chemical bonds. Water molecules from largely an

 ENGINEERING & SCIENCE NO. 1 11 cold and very dry, snow crystals in this environ- ment grow extremely slowly and typically don’t become very large. Slow growth is the key factor for making simple snow crystals, and these tiny gems—in snowfall circles called “diamond dust”—probably floated through the Antarctic air for hours before growing as large as the thickness of a human hair. Not exactly ski sweater material, c axis but these examples show the simple faceting that results from the underlying hexagonal symmetry of the ice crystal.

, p. 21, © 1994 by the American Geophysical Union. In more hospitable climates, snow crystals tend basal facet to grow much more quickly, and have a greater variety of forms. Simple prisms can grow into long, thin columns, which are usually hollow. Atmospheric Halos Faster-growing columns branch into clusters of long, thin, needlelike crystals. Sometimes a W. Tape, Tape, W. snowfall can consist entirely of these columnar and needlelike crystals—quite painful! On another a axis day, a snowfall could drop a preponderance of empirical science, even for simple and well-studied platelike crystals: small examples are often just The basic ice crystal shape, materials like silicon. And we’re now finding that simple hexagonal plates, but larger ones tend to ice has its own very different and quite fascinating a hexagonal prism, can be have more structure, such as sectored plates or story to tell, mainly because, unlike silicon, it has stellar dendrites (“dendrite” means “treelike,” seen in very simple a very high vapor pressure. To tell this story, we describing the branched structure of these crys- crystals that fall at the must first see what ice crystals growing from tals—these are the ones that come to mind when South Pole, above. water vapor look like. thinking of snow crystals). Platelike crystals are We can learn a great deal simply by observing quite common in the wild, and they are the the great bounty of snow crystals that appear out biggest specimens—often several millimeters of thin air, in snowfall. across and easily seen with the naked eye. The earth’s atmosphere is Permutations of the simpler shapes give us a not a bad laboratory for nearly infinite variety of complex snow crystal the study of snow-crystal shapes—dendrites with sectored plate extensions, structure, and it’s all for sectored plates with dendritic extensions, plates free. Natural snow with platelike extensions, and even such exotica as crystals exhibit a “tsuzumi” crystals, which consist of columns remarkable variety of capped with plates (named after a Japanese drum crystal shapes. The with a similar appearance). Some examples of the photos on the first two main types are shown on the opposite page. pages of this article show Twelve-sided snowflakes (left) can also be found a collection of beautiful floating through the air, and are not even that snow crystals that fell uncommon. These are actually two separate from the skies over platelike crystals joined together at the center, one Wisconsin, and these rotated 30 degrees relative to the other in a process beauties demonstrate called crystal “twin- nicely the intricate ning.” For the sake of structure that most of us anatomical correctness, associate with snow- however, I must point flakes. (A brief note on out that although eight- meteorological terms: a snow crystal refers to a sided snowflakes are Twelve-sided snow crystals single ice crystal, while snowflakes are clumps of often seen on holiday snow crystals that stick together and fall to earth wrapping paper, on ski form via a twinning as little puffballs). While such fancy specimens sweaters, and even on mechanism and are are clearly the favorite of snow-crystal photogra- soup cans (right), they actually fairly common. phers, there are many other common shapes that are never found in (Photo, P. Rasmussen) tell us about the physics of snow crystals. The nature. most basic ice-crystal shape is a hexagonal prism, Snow-crystal watch- which has two “basal” facets and six “prism” facets ing is incredibly easy, and can be quite fascinating. (see above). Very simple crystals like these are Simply pack a magnifying glass on your next ski actually quite common, and the photo shows some trip, or whenever you might encounter some examples that were collected at the South Pole. snowfall (a small pocket-size fold-up plastic model Because the conditions at the Pole are both very is perfect—you don’t need the bulky Sherlock

12 ENGINEERING & SCIENCE NO. 1  erra

, ed. I. Sunagawa, T

Morphology of Crystals Part B Part Morphology of Crystals

T. Kobayashi & T. Kuroda, Kobayashi & T. T. Scientific Publishing Co., 1987.

Snow crystals don’t just Holmes variety). On a good day you can see some One of Nakaya’s early problems was that it was come in the ski-sweater beautiful crystals—and as with gemstones, very difficult to grow individual snow crystals out photographs rarely do them real justice. You may of supersaturated air—the usual result was a variety. Plate-like forms find yourself echoing the sentiments of Thoreau, mixed-up jumble of crystals that formed what was (top, from left to right) who remarked, “How full of the creative genius is essentially a frost. Nakaya experimented with include a simple plate, the air in which these are generated! I should many thin fibers on which he tried to grow sectored plate, stellar hardly admire them more if real stars fell and individual snow crystals—cotton fiber, silk fiber, lodged on my coat.” metal wire, and even spider’s web. Nothing dendrite, and endless So how do we make sense of all these different worked, until he finally happened upon rabbit variations. Columnar forms snow-crystal shapes? The first real scientific steps hair, and on this he was able to grow single range from simple hollow were taken by Japanese physicist Ukichiro Nakaya crystals that were a great deal like natural speci- in the 1930s. Nakaya trained as a nuclear physi- mens. (We tried this briefly in my lab, but columns (far left) to thin cist, but on graduation found that no suitable jobs without much success—I suspect we were using needles (middle). Capped were available in his field. He eventually took a the wrong kind of rabbit!) columns, or tsuzumi crystals, professorship at Hokkaido University in northern Nakaya discovered that snow crystals grow in can also be seen (right). Japan, even though they had no nuclear physics different shapes, or morphologies, depending on department. Undaunted, Nakaya turned his the conditions in which they grow—in particular attention to snow crystals, which were available on the temperature and supersaturation of the air. locally in great abundance. After observing and Detailed measurements by Nakaya, and subsequently categorizing natural snow crystals in great detail by others, were used to produce a snow crystal for the first time, Nakaya then developed tech- “morphology diagram” (next page) showing the niques to grow artificial snow crystals in his crystal shapes that grow under different conditions. laboratory, so that he could study their properties This interesting diagram tells us a great deal under controlled conditions. about snow-crystal physics. It shows that the

Above, Nakaya in his laboratory. Right, part of his classification of natural snow crystals into 7 major groupings subdivided into 41 morphological types. Reproduced by permission from Snow Crystals: Natural and Artificial by U. Nakaya, © Harvard University Press, 1954.

 ENGINEERING & SCIENCE NO. 1 13 Snow crystals grow into different shapes depending on the temperature and supersaturation levels in the clouds in PlatesColumns Plates Columns and Plates 0.3 which they form. The morphology diagram, left, shows that some temperature ranges produce columns, others plates. Shapes get larger and more intricate the higher the super-

3 saturation levels. (Adapted from a diagram by Y. Furakawa.) Needles Dendrites 0.2 Sectored Dendrites plates Water saturation Hollow columns

Columns 0.1 Plates Thin plates Supersaturation (g/m ) Solid In my lab we’ve been working on a series of plates Plates quantitative measurements to try to figure out in Solid prisms more detail what’s really behind all this. The 0 physics we’re exploring is the statistical mechanics 0 –5 –10 –15 –20 –25 –30 –35 of crystal growth, which in turn depends on the Temperature (C) detailed surface properties of ice. Ice is a better material than you might think for doing basic materials physics—it’s a relatively simple sub- stance, and, because of its environmental impor- tance, ice has been studied using practically every experimental and theoretical tool known, so that the material itself is well characterized. Ice also basic shape of a growing crystal depends mainly on turns out to be very convenient to work with in temperature: plates form at around –2 °C, columns the lab—the stuff is cheap, it freezes at an easily at –5 °C, plates again at –15 °C, and either accessible temperature, and as a chemical it poses columns or plates below –25 °C. The appearance no health hazards whatsoever (aside from drown- of structure depends more on the level of super- ing, which usually isn’t much of a problem). saturation, and therefore on growth The basic idea for our experiments is quite rate: when the humidity is high, simple: we grow snow crystals under controlled rapidly growing columns become conditions, and measure the crystal dimensions as feathery needle crystals, and a function of growth time. In these experiments hexagonal plates grow into stellar we usually grow only very tiny crystals, smaller dendrites. An especially remarkable even than the Antarctic ones. Smaller crystals thing about snow crystal growth is grow mainly as simple prisms, which makes the the way the shape, or morphology, growth much easier to model theoretically. After changes back and forth between some roundabout mathematical machinations plates and columns several times as we’ve been able to infer the growth rates of the a function of temperature. And the different ice surfaces as a function of temperature, changes are large: within a few supersaturation, and other conditions. And as you degrees the morphology changes from very long, The world’s largest stellar narrow, needle-like crystals (at –5 °C) to very thin, dendrite? The crystal above flat, platelike crystals (at –15 °C). This is pretty grew to a diameter of bizarre—other materials don’t change their growth morphology with temperature nearly as more than 1 inch in 1.5 much as ice does. So what’s going on? hours. Right: a hollow Two main factors produce what we see in the column, looking like a morphology diagram: the intrinsic growth rates of small shot glass, grows on the crystal facets, and diffusion. Let’s look at crystal facet growth first. When the prism facets the end of an ice needle. grow more slowly than the basal facets, we get columnar crystals; when the basal facets grow more slowly than the prism facets, we get platelike crystals. So the morphology diagram already tells us that the prism and basal facets grow at different rates, and both rates depend sensitively on temperature.

14 ENGINEERING & SCIENCE NO. 1  Ice crystals, left, grown at controlled temperatures and The quasiliquid layer on ice has been studied for humidities in a crystal growth chamber, show the same a long time (it was first proposed by Faraday in the 1850s to explain some unusual properties of variations in size and shape predicted by the morphology ice), and it seems to play a role in many disparate diagram on the opposite page. From the top, with two environmental phenomena. Lightning, for examples for each group: small plates formed at –2 °C with example, is known to arise from collisions between ° ice particles in clouds—tropical clouds with no ice low supersaturation; hollow and filled columns at –5 C seldom produce lightning. The charge transfer with moderately high supersaturation; large plates and during these collisions is thought to depend on stellar dendrites at –15 °C with high supersaturation; and the details of the ice surface structure, and thus on small, fairly thick plates at –30 °C with low super- the quasiliquid layer. On a different front, some of the chemistry that takes place in high-altitude saturation. clouds, and which is important for ozone deple- tion, also occurs on the surface of ice crystals, and is affected by the presence of a quasiliquid layer. And back down on earth, it’s the quasiliquid layer might expect from looking at the morphology that makes ice especially slippery and helps snow diagram, the results are a bit odd. stick together in snowballs. The growth rates of the basal and prism facets The quasiliquid layer can affect crystal growth depend strongly on supersaturation as well as on in a number of ways. At low temperatures, when temperature, and some recent measurements show the layer is gone, the facet surface can be very that the rates even depend strongly on the back- smooth, and so grows slowly. It’s said that the ground gas in which the crystals are grown. Snow surface has a high nucleation barrier, since crystals grown in pure water-vapor conditions (i.e., molecules on the smooth surface don’t have much in a vacuum chamber containing no gases other to hold on to. At higher temperatures, however, than water vapor) do not grow into the same thin when the quasiliquid layer first starts to form on a plates or long needles that we see in air, but rather facet surface, molecules begin to jostle loose, so into more nearly isometric simple prisms. Air is that the surface cannot be so smooth any more, relatively inert, and helium gas is even more so, and the growth rate shoots up. But then, at still but both nevertheless slow the surface growth higher temperatures, the quasiliquid layer is so rates substantially, even after factoring out the thick it starts to look like a real liquid, and the effects of water-vapor diffusion. solid/quasiliquid boundary can itself become We still don’t understand why ice does all of smooth, like that between a solid and its melt. what it does, which is why we continue studying Thus there is a nucleation barrier at the solid/ it. But one feature of ice that almost certainly quasiliquid interface, and the growth rate goes plays a big role in its crystal growth is a phenom- down again. There’s considerably more to the enon called surface melting. For any crystal, the story, but you get the idea. surface molecules are not as tightly bound as the When you consider that the properties of the molecules deep inside, since surface molecules quasiliquid layer, and how they depend on don’t have so many neighbors to hang on to. temperature, can be different for the different ice Thus, the surface molecules can sometimes jostle facets, it becomes plausible that ice growth shows loose while still remaining attached to the solid, the unusual characteristics that it does. But the forming what’s called a quasiliquid layer (see devil is in the details, and our current research is diagram). This layer disappears at very low temperatures, when it’s so cold that the mole- cules don’t have much quasi- jostle in them, and liquid layer gets thicker at higher temperatures, eventually increasing to effectively infinite thickness at the melting point. ice Surface melting is crystal present on a large variety of materials, but the physics of this phenomenon is partic- A quasiliquid layer forms on the surface of ice when water molecules at the edges of the ularly puzzling for high-vapor-pressure crystal jostle loose from the tightly bound inner lattice. The thickness of this layer depends materials like ice. on temperature and is generally different for the basal (left) and prism (right) crystal faces.

 ENGINEERING & SCIENCE NO. 1 15 How snow crystals get their arms. This photo series of a growing crystal shows how a hexagonal sectored plate develops six distinct arms. The hexagonal growth is eventually unstable because the corners of the plate stick farther out into the supersaturated air and thus collect more water molecules. On the nanoscale, below, what appears to be a straight facet actually contains many molecular steps. Diffusing molecules are more likely to hit the corners, but molecules are more likely to stick where the surface is rough, in between the corners.

aimed at reaching a better understanding of just the Mullins-Sekerka instability, that goes far in what’s going on. This is all pretty basic stuff—the explaining why snow crystals grow into such structure of ice surfaces, and how it affects crystal intricate structures. growth—but we are surprisingly ignorant of even If you imagine that our newborn snow crystal is these fundamental issues, not only for ice but also a hexagonal plate, you can see that the corners of for most real surfaces. the crystal stick out a bit farther into the air than The second big factor affecting snow crystal the center faces in between, making it a bit easier shapes is the diffusion of water molecules through for water molecules to diffuse to them. The corners the air, and here at least we know the underlying should therefore grow a bit faster than the rest of physics fairly well. A growing crystal captures the crystal surface. However, as the corners grow, water molecules from the air right around the they leave molecular steps behind them, which crystal, and additional water molecules have to make the center faces a bit rougher than the make their way through the surrounding air smooth corners (see diagram above). molecules to reach the surface. Diffusion tends to Since rough surfaces grow more quickly than hinder crystal growth, and this produces what smooth ones, the overall result is that the center turns out to be a fairly common effect, known as faces grow just as fast as the corners, so that the

Snow Crystal Growth and "The No-Two-Alike Conjecture"

Nucleation around a dust particle

The growing snow star on the opposite page was Grows to hexagonal prism, since smooth facets grow most slowly created by cycling it between –15 °C, to develop dendritic spikes, and Simple plate unstable as crystal grows larger... corners sprout arms –12 °C, to develop sectored plates, mirroring the way natural snow crystals form Crystal moves to different temperature... plates grow on arms complex shapes as they move through different Crystal moves through many different temperatures... temperature zones in a each change causes new growth behaviors on arms snow cloud, right. Complex history Complex crystal shape Each arm experiences same history Symmetry No two paths similar No two alike

16 ENGINEERING & SCIENCE NO. 1  A normal ice dendrite overall shape of the crystal doesn’t change as it grows. And since the molecular steps are very (right) grew at a steady 3 tiny, the crystal continues to look like a simple microns per second until hexagonal prism as it grows. the application of 1400 This goes well for a while, but to continue this volts. Then the growth shape-preserving growth, the center faces become rougher and rougher as the crystal grows larger, developed into a thin until eventually they become so rough that they needle morphology (inset). can’t be any rougher. At this point the center faces These electric needles have can no longer keep up, and the corners really do been clocked at up to a grow faster, with the end result that our hexagonal snow crystal sprouts six tiny arms. As the arms staggering 200 microns per themselves grow longer, they too come under the second. influence of the Mullins-Sekerka instability, and can sprout their own side branches. In the end we find that a very complex crystal shape results from a rather simple physical phenomenon. Growth instabilities like these are quite common in nature, and are responsible for a great deal of pattern formation, in the biological world as well as the physical world. the local temperature and supersaturation experi- With these tools, we can now can pretty much enced by the crystal, since, as we’ve seen from the explain why there’s such a rich variety of sym- morphology diagram, snow-crystal growth is metrical snow-crystal shapes in nature, or what I extremely sensitive to environmental conditions. like to call the “no-two-alike conjecture.” After a It’s important to note that the local conditions snow crystal is born, it quickly grows into a small are essentially the same for each arm on a tiny hexagonal prism, with the facet surfaces growing snow crystal, so that the arms all grow in the same more slowly than the other surfaces. Then, as the way. Then, as this growing crystal travels through crystal grows larger, the Mullins-Sekerka instabil- regions of the atmosphere with different tempera- ity often kicks in, causing the corners to sprout tures and supersaturations, its growth will change arms. Exactly how fast the arms grow depends on with the conditions, and the arms will all change their growth in unison. The final crystal shape can be very complex, reflecting the complex path the crystal followed through the atmosphere. Yet since the arms all grow at the same time, they tend to look alike, so that the crystal has a sixfold symmetric appearance. And since no two crystals follow exactly the same path through the sky as they fall, each grows into a slightly different shape. So we end up with a myriad of complex, symmetric patterns, with no two alike. In principle, one could look at a snow crystal and decipher the conditions under which it grew. Or, to quote Nakaya, “Snow crystals are hieroglyphs sent from the sky.” Another area we’ve been exploring in my lab is that of using high electric fields to enhance and control the dendritic instabilities in crystal growth. We started by growing normal dendritic crystals like the one in the photo above, which are very easy to grow at –15 °C when the supersatura- tion is high. The tip of such a crystal grows at a steady 3–4 microns per second, and side branches appear in a semiregular pattern as dictated by the Mullins-Sekerka instability. Dendritic growth like this has been studied for quite a while (it has applications in metallurgy and the structure of alloys), and we know how to calculate the simpler properties of the growth, such as the steady-state tip velocity, which is the rate at which the tip elongates.

 ENGINEERING & SCIENCE NO. 1 17 Electric needles, left, were grown at –5 °C with the application of 2000 volts and chemical vapor additives to induce growth along the c-axis of the crystal. When the voltage was removed and the needles moved to –15 °C, plates grew. Right: needle growth at –5 °C.

The theory we developed to explain the physics of all this suggested that electrically enhanced growth would work even with nonpolar molecules (water is highly polar), and we recently demon- strated this in the lab by growing iodine needle crystals. Metallic needle crystals have also been grown from certain metal-carbonyl vapors using related techniques, and it’s tempting to speculate that diamond needles could be produced one day using electrically enhanced growth (although I suppose it’s too late to corner the market in phonograph needles). To make life a bit more interesting, we began In our lab, we have a special fondness for growing these crystals on the end of a wire growing “c-axis” electric needles of ice (think tiny connected to a high-voltage power supply. As we ice versions of a standard hexagonal pencil). For a turned on the voltage, the electric field near the long time we found that growing this particular tip of the growing crystal polarized the water type of needle crystal was a hit-or-miss proposi- molecules in the surrounding air, and these tion; some days it worked well, some days it didn’t polarized molecules were attracted to the tip. This work at all. At first we thought we were having enhanced the normal diffusion rate, and the tip contamination problems from solvent vapors velocity increased accordingly—all nicely accord- floating around in our chamber (the laboratory ing to theory. But when the tip velocity was about version of air pollution), so we decided to give the twice as fast as normal, whoosh! The tip just took chamber a good, long bake to clean it out. To our off, growing as much as 200 microns per second— surprise, when we started our experiments again, greased lightning in the crystal-growth game! we couldn’t get any good needles at all—not one. The crystals grew into long, thin needles, with tip It turned out we weren’t being hurt by contamina- radii as small as 100 nanometers. tion in the chamber; we were actually being helped

A designer snowflake. This crystal was grown at –14 °C, except that at approximately periodic intervals it was moved to –7 °C for just a few seconds. Each move induced the development of side branches at the tips of the growing arms. A movie of this can be seen at snowcrystals.net.

18 ENGINEERING & SCIENCE NO. 1  A gallery of designer snow stars—tiny ice flowers on thin ice stems.

by it. So we tried adding various solvent vapors, at chemistry, all working in concert to create these concentrations of only a few parts per million in tiny, filigreed ice sculptures that fall down from the air, and Voilà!—we were soon producing the sky. Snow crystals are not only beautiful to beautiful c-axis needles again, very reliably. Acetic look at, but they also teach us about surface acid, the main ingredient in vinegar, seems to physics, the statistical mechanics of crystal work especially well, and we’re still puzzling over growth, and the intricacies of pattern formation exactly what physical and chemical mechanisms processes in nature. ■ are controlling the needle growth. A wonderful feature of c-axis needles is that after Growing up in Fargo, North Dakota—where it's growing a long ice needle, we can turn off the said there's 10 months of snow and two months of poor PICTURE CREDITS: voltage and grow a normal platelike snow crystal sledding—may have inspired Ken Libbrecht’s current 10, 14–19 — Ken Lib- on its end. Most of this work is done in our interest in snow crystals, but for most of his early career brecht; 11 — Mark vertical diffusion chamber with a convenient he worked at the other end of the temperature spectrum, Garcia; 10–11, 12, 14–16 — Doug temperature gradient—warm at the top, cold at investigating the internal structure of the sun. A Cummings the bottom. By moving the needle from –5 °C, Caltech graduate (BS ’80), he returned to join the where the c-axis needles grow best, to –15 °C, faculty in 1984, and has been Executive Officer for where plates grow best, we can grow a beautiful Physics since 1997. Nowadays, as well as investigating snow crystal on the end of the needle, like a tiny the physics of snow crystals, Libbrecht is involved in ice flower on a thin ice stem. advanced detector development for LIGO, the Laser We’ve developed these techniques to the point Interferometer Gravitational-wave Observatory. Check where we can now create “designer” snow crystals, out his award-winning Web site, snowcrystals.net, for growing shapes of our own choosing by controlling much, much more on snow crystals, including how to the humidity and temperature of the growth, and make your own in the kitchen, and how to help locate the some examples are shown in the photos on these snow crystal capital of the world. This article is pages. It’s something of a new art form—miniature adapted from a Watson lecture given in January 2001. ice sculpture. Instead of cutting away material, we design and fabricate using the natural rules of pattern formation. Stellar dendrites, sectored plates, hollow columns can all be made relatively easily, and now we’re even learning to control side branching and other features to create more complex and unusual snow-crystal shapes. So we see that snow-crystal growth is governed by some sophisticated physics, mathematics, and

Right, two photos of a growing stellar dendrite, taken 5 minutes apart, show the development of side branching. The ice needle on which this c-axis star was grown is hidden from view below it.

 ENGINEERING & SCIENCE NO. 1 19 20 ENGINEERING & SCIENCE NO. 1  A Billion Points of Heat by Jane Dietrich

Infrared radiation—that segment of the electro- National Aeronautics and Space Administration magnetic spectrum with wavelengths longer than (NASA) and the National Science Foundation visible light, between visible light and micro- (NSF) funded the Two Micron All Sky Survey waves—can pierce through the dusty universe to (2MASS), a collaboration between the University disclose quantities of unseen stars and galaxies; it of Massachusetts and Caltech, that just finished reveals cool stars that radiate heat but no visible gathering its data in February 2001. light, failed stars, and stars just being born in The first survey was also a Caltech undertaking. gaseous clouds. Although water vapor and carbon Back in the early 1960s, Bob Leighton and Gerry dioxide in the atmosphere absorb much of the Neugebauer aimed a telescope equipped with infrared, satellites above the atmosphere, such as infrared detectors to scan as much of the sky as Left: The Orion Nebula, or the Hubble Space Telescope and the Infrared could be seen from Mount Wilson—about 70 Messier 42, one of 2MASS’s Astronomical Satellite (IRAS), have extended our percent of it. Leighton designed the 62-inch sight spectacularly. But there’s also one window, telescope’s epoxy reflecting dish based on a prin- greatest hits (which at about 2 microns in what is called the near ciple he had first observed in his mother’s mop required only 10 minutes infrared (the infrared extends from wavelengths of bucket as a child, and built it in the back of his of observing time). The 1 to 300 microns), where waves can pass relatively office (see E&S, No. 4, 1998). The infrared detec- Trapezium Cluster in the easily through the atmosphere and be detected tors were samples donated to Neugebauer by a from the ground. And at 2 microns, you’re seeing friend in the defense industry; they had been center contains more than light given off directly by stars, whereas at longer developed after World War II for the heat-seeking 3,000 bright, hot stars, the infrared wavelengths you see starlight that has guidance system of the Sidewinder missile. Their densest concentration of been absorbed and reradiated by dust glowing in Two-Micron Sky Survey (TMSS), published in the space between stars. 1969, noted 20,000 infrared sources and cataloged young stars in our solar Astronomers first surveyed the sky through the about 5,700 previously unseen celestial objects. neighborhood. Even 2-micron window in the 1960s. But dramatic “We thought it was a fun thing to do,” says younger, protostellar, advances in detector technology in the last 20 Neugebauer, now the Robert Andrews Millikan objects appear in the small years have made it possible to detect objects more Professor of Physics, Emeritus. (Leighton, BS ’40, than 80,000 times fainter than those discovered by PhD ’47, the Valentine Professor of Physics, red region at the top. The the first 2-micron sky survey, and in the 1990s the Emeritus, died in 1997.) In an era when infrared bright, wispy clouds come from molecular hydrogen in dust-scattered starlight, detectable at two microns.

In April 1964, Bob Leighton’s 62-inch two-micron infrared telescope (right) prepares to move into its “dome” on Mount Wilson, which is shown under construction in September 1963 at far right. The carpenter on the left is Jerry Nelson, ’65, who later went on to design the 10-meter Keck Telescope’s segmented mirror.

 ENGINEERING & SCIENCE NO. 1 21 astronomy was starting to grow rapidly, they also hand-me-down, of defense research (for sensing “happened to be there first,” according to Leigh- heat from Earth rather than the sky). In their ton’s oral history. Leighton’s telescope is now a transfer from the military to astronomy, such piece of history enshrined in the Smithsonian infrared detectors made possible the Hubble Space Institution. Telescope’s spectacular high-resolution images of Infrared astronomy has long since left behind infant stars forming in glowing, billowing clouds seat-of-the-pants technology and missile leftovers. of space dust. 2MASS, says Neugebauer, “is everyone’s dream of Each telescope has been producing a 20- how you should do something, really do it right.” gigabyte tape per night, and once a week for the The completely automated twin telescopes—one past four years a stack of these digital linear tapes for each hemisphere—were designed and con- has been arriving (via Fed Ex) at the Morrisroe structed by the University of Massachusetts team Astroscience Laboratory, home of Caltech’s Infra- under astronomer Michael Skrutskie, principal red Processing and Analysis Center, or IPAC. Top: The 2MASS 1.3-meter investigator of the sky survey. At 51 inches (1.3 Altogether there are hundreds of tapes, amounting telescope in the southern meters) they’re a bit smaller than Leighton’s to 25 terabytes (1012 bytes) of data. For compari- hemisphere sits high in the telescope; smaller telescopes with large fields of son, the photographic plates of the 1998 Palomar Andes at the Cerro Tololo view are better suited for covering the larger optical sky survey (this was also the second one, swaths of the heavens needed for a survey of the the first having been done in the 1950s), when Interamerican Observatory. whole sky; huge mirrors like the 10-meter Keck digitized, comprised 3 terabytes of data, already Inside the dome (above) Telescope, which gather faint light from the edges several hundred times more than what was the camera, with its of the universe, peer through a pinhole in com- collected by IRAS, which was at that time one of parison. The northern-hemisphere instrument, the largest collections of data ever. How do you massive arrays of detectors on 8,550-foot Mount Hopkins (even at the 2- take such an incomprehensibly huge data set and and filters for three micron window, less atmosphere is better) in turn it into something comprehensible: high- wavelengths, is attached to Arizona, started sweeping the sky in June 1997 resolution images and an accurate, reliable catalog the back of the telescope. and finished its observations in November 2000; of half a billion objects? You turn it over to IPAC, its twin at the Cerro Tololo Interamerican Obser- which was created to analyze the IRAS data. vatory, at about 7,000 feet in the Chilean Andes, Caltech had stayed involved in infrared astron- started up in March 1998 and sent its last data to omy after Leighton and Neugebauer’s pioneering Caltech this past February. Each telescope’s cam- work. Neugebauer, in fact, was U.S. cochairman era scanned strips of sky running 8.5 arcminutes of the joint science working group (which also wide and 6 degrees long before moving on to the included scientists from the Netherlands and the next strip, or “tile.” The whole sky is tiled by United Kingdom) for IRAS, which was launched 59,650 such strips. in 1983 and collected data for 10 months. IRAS The heart and soul of the telescopes are their was managed for NASA by the Jet Propulsion sensitive detectors, which represent the greatest Laboratory, but when it came to analyzing the leap in technology. Where the Mount Wilson mountains of data—a task that was to take several model contained eight detector elements, a years—it became more convenient to move the 2MASS telescope camera sports three arrays of project to campus as IPAC because of JPL’s access 64,000 such elements, each with a filter for a restrictions, especially for noncitizens, and because particular wavelength. Like the earlier telescope, of the great interest on campus in the data and the new arrays are also a legacy, if not an actual their scientific interpretation. In the southwest

22 ENGINEERING & SCIENCE NO. 1  Below: The whole sky, as seen in a 2MASS three-color, composite image. The Milky Way, edge-on, stretches almost all the way across, its center the bright bulge. Filaments of dust cut through the galactic plane, while the Magellanic Clouds, our nearest galaxy neighbors, are visible just below it at lower right.

corner of campus, behind Braun gym, and lying low to appease its residential neighbors along The task fell to Roc Cutri, who, as task leader as Arden Road, the astroscience laboratory has been well as project scientist, actually does live in both in operation since 1986, largely a mystery to the cultures. Cutri oversaw the design and implemen- rest of campus. tation of the automated software pipeline, called “IPAC is unique as a piece of campus with 2MAPPS, which stands for 2MASS Production special connections to JPL,” says George Helou, Processing System. To convert the raw digital IPAC’s director. Although the funding comes data into calibrated images in three colors and to from NASA, and the lab carries a bit of JPL extract the quantitative information of the “flavor,” all of the lab’s approximately 80 staff position and brightness of each object, Cutri and members are employees of the Division of his team wrote 300,000 lines of computer code— Physics, Mathematics and Astronomy. “IPAC not the usual pastime for an astronomer, but Below: The Mount Hopkins straddles the two cultures,” Helou says, “the intellectually challenging all the same, says Cutri. telescope begins its night’s science and research culture of campus and the The three colors represent three different work surveying the project and engineering culture of JPL. And wavelengths in the near infrared: the J band at 2MASS combines the science and engineering a wavelength of 1.25 microns, the H at 1.65 northern sky. in unique ways.” microns, and the Ks at 2.17. (The infrared’s 2MASS is the first digital, electronic survey of photometric “bands”—the natural windows the sky at relatively high resolution. While the astronomers make use of—follow a rather loose recent Palomar Sky Survey turned to JPL’s Arti- alphabetical scheme; the longer wavelengths after ficial Intelligence group to develop a program to K continue in a more regular fashion: LMNOP.) turn its photographic images into computer- Classes of astronomical objects can vary in bright- digestible data for cataloging, 2MASS was born ness at the different wavelengths, so simulta- digital. What this project needed was robust neously observing them at all three wavelengths software to turn its terabytes into images, as well makes identifying and interpreting them easier, as into useful and reliable catalog information. says Cutri.

 ENGINEERING & SCIENCE NO. 1 23 Right: In the Flame Nebula, NGC 2024, part of the Orion Molecular Cloud Complex, the near infrared reveals a dense stellar cluster, comprising stars likely surrounded by the accretion disks that could be sites of where planets form. Below: The Cat’s Paw, or Bear Claw, Nebula (NGC 6334).

But before the identifying and interpreting begin, the data must be fed through the software pipeline. While the telescopes finished their observations this past February, it will take six months to a year to finish crunching through the raw data to produce the visual and quantitative results. The full set of images—the whole sky— and catalog will be published next spring. Analyzing the data will take decades; Cutri reckons there’s enough fodder for a 50-year legacy. “2MASS offers the statistical context to study a very large number of objects, and also the detail to follow up the interesting ones,” he says. In the meantime, half the sky has been acces- sible on line since March 2000, the single largest collection of astronomical data from a NASA mission. The final catalog will contain half a billion objects. The team actually tallied one and a half billion sources, but have winnowed down the number to those whose brightness and position can be guaranteed with great precision. “Astronomers need to be able to depend on the accuracy of the data,” says Cutri, who as project scientist is responsible for ensuring the scientific quality of the data. He compares it to a phone book, which would be useless if one out of ten digits were wrong. With 2MASS, extraterrestrials could phone home—confidently—to any of nearly half a billion stars. The best way to insure the scientific quality of the data is “to do the science yourself,” according to Cutri. He was a collaborator in one of the early discoveries to come out of 2MASS: brown dwarfs, cooler than any ever before seen. Brown dwarfs are stars that never ignited, and therefore give off no visible light. They do give off some heat, how- ever, which can be detected in the infrared. Already in 1998, in only one percent of what would be the total data, a team of 2MASS re- searchers led by Davy Kirkpatrick at IPAC found a number of these objects, which they named L dwarfs, revising a century-old alphabetic system

24 ENGINEERING & SCIENCE NO. 1  (even more mysterious than the designation of the sky. He and his group are creating software to wavelengths) in which stars are ranked from improve the accuracy of the matches. Brunner is hottest to coolest as OBAFGKM—with L now at also currently working with Microsoft Research on the end). Then came Caltech grad student Adam a Sky Server to merge data from 2MASS, the Palo- Burgasser’s discovery of still cooler, fainter, mar Sky Survey, and other surveys, approaching methane-rich brown objects called T dwarfs. the problems less from a computer-science angle (Make that OBAFGKMLT.) Together, L and T and more from the viewpoint of the scientific user. dwarfs, scarcely bigger than Jupiter and collapsing “We’re starting to move on to something of service Emissions from the under their own weight, are probably the most to the broader community.” Ultimately, the planetary nebula NGC 3132 populous stellar objects in our galaxy, outnumber- information-rich dataset of these combined digital ing real stars two to one, and some of them may surveys will be accessible on line in a National (above) are mostly in the also be our solar system’s nearest neighbors. Virtual Observatory. Ks band, which gives the Because the Milky Way itself contains so much Sometimes the overlap hasn’t been foreseen. ring around the wind- dust, it has been heretofore impossible to see Richard Ellis, professor of astronomy and director blown center its reddish neighboring galaxies in the wonderfully named of Palomar Observatory, is part of an international “zone of avoidance,” which extends to 30 degrees team conducting the 2-degree-Field Galaxy color. In the center is a on either side of the galactic plane. But 2MASS Redshift Survey (2dF) of a swath of the southern low-mass star like our sun, has cut through that murky zone and revealed sky, using a novel instrument built by Keith losing its outer envelope many more galaxies in our local universe. It has Taylor (now a member of the professional staff at also produced a high-resolution map of our galaxy, Caltech) at the Anglo-Australian Telescope near as it dies. including the galactic center, which at other Coonabarabran, Australia. The team is seeking to Top left: The remnant of a wavelengths is obscured by dust—and an exquis- measure the stellar population density and mass supernova (IC443) exhibits ite census of the Milky Way. distribution of galaxies on large scales. They a bright blue arc of excited One of the challenges of such an immense data mapped the positions of 170,000 galaxies, but set, says Helou, “is how to search through a billion optical wavelengths cannot accurately discern the iron in the J band and a sources to find the really interesting objects— stellar content of galaxies. Infrared radiation red ribbon of hydrogen in the unusual stars and quasars.” So perhaps the provides a much better handle on how many stars

the Ks band. greatest importance of 2MASS lies in overlapping are in a galaxy, because the infrared output is Top right: 2MASS reveals it with surveys at other wavelengths (for example directly proportional to the number of stars. But the very short X rays on one end of the spectrum until 2MASS, no infrared telescope had seen deep the nearby spiral galaxy and the much longer radio waves on the other), enough into space to get a fair sample. Combin- Maffei 2, which lies in the the “synergy,” as Cutri describes it, “of putting ing the digital catalogs of both surveys (a process “zone of avoidance” and data sets together. The value of combined surveys that was accomplished literally overnight) goes far beyond that of a single one.” matched redshifts/distances from 2dF with 2MASS was previously almost lost Postdoc Robert Brunner agrees. “The sum is luminosity/population data for 17,000 galaxies. in the dust. greater than the parts,” says Brunner. “You gain With that large a sample Ellis and his group could more from data by joining them together than determine how many stars there are in the uni- using them independently.” As project scientist verse per unit volume of space, and concluded that for the Digital Sky Project, he’s trying to mesh a the amount of mass in stars is only about 0.2 bunch of surveys at different wavelengths—a percent of the total mass needed to stop cosmic computationally challenging task matching expansion. “The overlap led to the most accurate billions of sources at different resolutions all over census of stars in the local universe,” says Ellis,

 ENGINEERING & SCIENCE NO. 1 25 The infrared shows up a Opposite page: The Carina, bright core in the radio or Keyhole, Nebula (NGC source Cepheus A. Its 3372) contains an massive young stars and unusually high concentra- molecular gas are tion of young massive stars completely invisible to and stars that are still optical wavelengths. forming.

“an achievement that neither survey ever imagined home page. This is not its only role in show when it was originally conceived.” business: the Star Trek Voyager TV series has also “It’s gratifying to develop a product that other snapped up some 2MASS images for space scientists can use,” says Cutri. And not just backdrops. scientists. 2MASS can turn anyone’s computer A topical session (“The Big Picture: Latest into a “desktop observatory” through IPAC’s Science Results from 2MASS”) at the American Infrared Science Archive (IRSA), the Web-based Astronomical Society’s meeting in early June, held system for accessing the databases. Besides the conveniently in Pasadena, highlighted the already catalog of data, about 5 million images in the substantial science that is emerging from the data digital Image Atlas will also be available on on half the sky. The session spanned a range of line. Just enter the name of your favorite galaxy topics including—besides brown dwarfs— or pick a point on the sky. You can start off at asteroids, stellar populations in the Milky Way http://www.ipac.caltech.edu/2mass/ and be sure to and the Large Magellanic Cloud (its nearest catch “2MASS Galactic Center: The Movie” on the neighboring galaxy), other nearby galaxies, distant active galactic nuclei, and the cosmic near-infrared background radiation. As of the beginning of June, 174 papers had been published using 2MASS data. Funding for 2MASS will continue for another year and a half, when the project will deliver its final products and then fold, says IPAC’s Helou. “Then we’ll find something else interesting and worth doing, something with a specifically Cal- tech point of interest.” Already well under way is SIRTF, the Space InfraRed Telescope Facility, a satellite to be launched in July 2002 to observe the sky at wavelengths from 3 to 180 microns. 2MASS data are critical in laying the groundwork for the mission, which is administered by JPL for NASA. There’s considerable Caltech scientific interest in this new venture, and the SIRTF Science Center will be located on campus, sharing the Keith Spaulding building with the remaining segments of Business Services that have not moved elsewhere. And IPAC will be analyzing the data. ■

The 2MASS team in the galactic center. Front row, from left: Schuyler Van Dyk, Diane Engler, Ron Beck, Eugene Kopan, Roc Cutri, Tracey Evans, William Wheaton, Robert Hurt, Sherry Wheelock, and Jeonghee Rho. Back row: Ken Marsh, Cong Xu, Howard McCallon, Tom Jarrett, Laurent Cambresy, J. Davy Kirkpatrick, John Gizis, and Raymond Tam. Not pictured: Brant Nelson, Helene Huynh, Tom Chester, and John Fowler.

26 ENGINEERING & SCIENCE NO. 1   ENGINEERING & SCIENCE NO. 1 27 The information revolution has had a profound impact on the economy but very little impact on economic

policy, which is largely still generated by 19th-century ideas.… Having a 21st-century economy based on

laws designed for the manufacturing sector is really quite ridiculous.

28 ENGINEERING & SCIENCE NO. 1  Economic Policy in the Information Age by Simon J. Wilkie

The United States is in the throes of the third These are the wrong questions. The important industrial revolution—the first one, of course, questions are really in the small details of how being the harnessing of mechanical power, and a market is structured. Second, I want to make the second one being the harnessing of electrical a case for the fundamental importance of abstract power. This one is the harnessing of information, economic theory—the type of arcane research, and has been sparked by the biggest capital invest- divorced from the real world, that we do here ment in the history of humankind. The U.S. at Caltech. If you don’t pay attention to these spent roughly $4 trillion on information technol- extremely mathematical, abstract models, you’re The standard economic ogy, broadly defined, from 1960 to 1994, and it’s bound to make disastrous policy mistakes. So model, in which many expected that we’ll be spending a trillion a year by minutiae are important, and the boring questions 2005, even with the current slowdown. Like the are really the interesting ones. I’m going to apply buyers and many sellers other industrial revolutions, it has taken 30 to 50 this perspective to the Microsoft antitrust case and compete with one another years for the results to show up. The productivity to California’s electricity-deregulation debacle, in the marketplace, works gains we’ve seen in the economy in the last few which is an endless source of fun until your bill well for tangible goods years have their origins in the early investments comes at the end of the month. that are just now starting to kick in. To see where we need to go, we first need such as bananas. But you The information revolution has had a profound to know where we are. The standard economic get into deep trouble impact on the economy but very little impact on model says we have a market in which many when you try to apply it economic policy, which is still largely generated buyers and many sellers compete with one another. by 19th-century ideas. The legal framework was Each buyer has a personal valuation for each good. to network commodities set by the Sherman Act and the Clayton Act, When you buy bananas, you know how much such as electricity. which were developed in the 1890s to bust trusts you’re willing to pay for any given quantity. And such as Standard Oil. The notion of regulating each seller knows his or her production costs, and utilities appeared in the early part of the 20th sets an individual minimum price accordingly. century, culminating in the Telephone Regulation The market adjusts supply to be equal to demand, Act of 1933, which established AT&T as a monop- as that old axiom says. This is summed up on the oly. Manufacturing, however, which was the next page, in the economist’s favorite diagram. dominant paradigm at the turn of the century, The vertical axis is price, and the horizontal axis is now less than 17 percent of our economy. It’s is quantity. The downward-sloping demand curve going the way of agriculture. Health care is now says that the lower the price is, the more of a almost 15 percent of the economy, and in a couple product people are willing to buy. The supply of years, it’s going to be bigger than manufactur- curve slopes upward, saying that the higher the ing. Having a 21st-century economy based on market price, the larger the quantity that sellers laws designed for the manufacturing sector is will want to sell. At the intersection lies the really quite ridiculous. equilibrium, which is the quantity (q*) that There are two main points I want to make. The is actually traded at the market price (p*). first is that minutiae are important in the design This competitive market has several desirable of economic institutions—that is, the details properties, the first of which is efficiency, or E. matter. And they matter a lot. Which is kind of That is, the market maximizes the sum of the ironic because most economic-policy debate is big- valuations minus the sum of the costs. This think debate: should we have a market or not? maximizes social welfare—the greatest good for Should an industry be regulated or unregulated? the greatest number. (However, it doesn’t mean

 ENGINEERING & SCIENCE NO. 1 29 Supply Price The competitive-market supply and demand curves as drummed into the heads of generations of business majors. The supply curve (more properly, the marginal cost p* curve) rises while the Demand demand (or marginal value) curve falls. They intersect at point (p*, q*), 0 q* Quantity and the shaded region is the social surplus, or value it maximizes your welfare. For that, you’d need to without needing an infusion of cash or goods in excess of cost, that do a weighted sum, where your weight was higher external to the system—unlike, say, Russia, which people get from trading in than anybody else’s.) The distance between the is kept afloat by large amounts of funds flowing in the market. marginal cost (the cost of producing one more from the World Bank and the United States. Or unit of the product) and the marginal valuation your kids might trade toys, but that only works (the price the consumer is willing to pay for one so long as there’s a perpetual infusion of new toys more unit) is the amount of what we call surplus, from the parents. But a normal adult market has or benefit, that people get from trading in the the miracle of always being balanced. The last market. Mathematically, the market starts at property, S, is the most important one—this quantity q = 0, and runs until q equals q*, at market is strategy-proof. People have no incentive which point the market price has dropped to p* to game the market. If I go to Von’s supermarket and the sellers are selling at cost, so they quit. to buy three pounds of bananas, and they’re a But they made a profit on all the previous sales. dollar a pound, I have no incentive to buy four pounds, and I don’t think, “Ha! I’ll fool The telephone system, the Internet, and the power grid are obviously networks. But HMOs are actually them and only buy two pounds!” If I networks, too.… A country club is a network. When you join the club, you get the use of their facilities; want three pounds, you also get to enjoy, or unenjoy, the company of the other members of the club. In fact, another word for I buy three pounds. There is absolutely “schmoozing” is “networking.” no benefit to me from strategic behavior. To sum up, the competi- And at p* the buyers, who have been buying the tive-market model has four really nice properties: product for less than their personal valuations, are it’s efficient, it’s voluntary, it’s balanced, and it’s paying as much for the last unit of the product as strategy-proof. That’s why policymakers tend to they think it’s worth, so they quit. But they got be opposed to monopolies and market regulation, a bargain on all the previous units. So the market which short-circuit the market’s functioning. is efficient because it maximizes the integral—the But the general paradigm for the 21st century shaded region between the two curves—without is a network-market model. Network markets are even knowing what that integral is. Adam Smith ubiquitous—the telephone system, the Internet, discovered the magic of this “invisible hand” a and the power grid are obviously networks. But couple of centuries ago. HMOs are actually networks, too. When you join The market has three other remarkable proper- a primary-care physician’s group, you’re also ties that we don’t talk about as much. First, signing up for the set of specialists affiliated with notice that there’s no Tony Soprano—we don’t that group. Banks and ATMs are networks. A need coercion to get people to use the market. country club is a network. When you join the Participation is voluntary, because until the club, you get the use of their facilities; you also market hits equilibrium and shuts down, every get to enjoy, or unenjoy, the company of the other buyer leaves with a bargain and every seller leaves members of the club. In fact, another word for with a profit. We call this property V. The next “schmoozing” is “networking.” The eBay Web property is B, for balance. Supply equals demand site is a network. Network markets work precisely

30 ENGINEERING & SCIENCE NO. 1  v

Marginal Things look quite different in a network market. Now the Value marginal-cost curve is flat, at next to nothing, while the product’s value increases as more people buy it. The marginal value, which is the derivative of the value, increases even faster until 100 percent of the population

Price owns the product and its value is maximized. This means Value that a monopolistic seller can name any price up to the maximum value, v, and people will still buy the product. In this case, a monopoly can be efficient, in the economic sense of the word. Marginal Cost 0 Quantity (% of population) 100

because of the mass of users that they attract. Even more troublesome, the demand curve In a network market, neither the supply curve slopes downward only when the quality of the nor the demand curve behaves as expected. A product is inherent and is independent of its quan- supply curve only slopes upward when the margin- tity. But network commodity’s quality is systemic; al cost, the derivative of the cost function, is it’s not inherent in the commodity itself, as it is in increasing. And the marginal cost does go up, for a banana. In the electricity market, if I decide to traditional commodities. Most people think of the flip on my air conditioner, it affects the voltage— marginal cost as going down as more units are pro- minusculely, but it affects the quality of the duced, but that’s only true up to a point. If I’m service that everybody else in my neighborhood growing bananas, I’ll cultivate my most fertile gets. And a network’s value to a user also depends land first. As demand grows, I’ll use increasingly on its quantity, but not on the quantity you buy as poorer land, and I’ll have to buy more fertilizer an individual—rather, the value depends on the and more water to produce a crop. The same is number, or the identity, of users. If I’m the only true of steel—if a blast furnace runs around the clock, labor costs will skyrocket. The foundry either has to hire more people or pay massive overtime. So there’s actually a dis-economy of scale. But let’s think about software for a moment: The marginal cost to Microsoft of me buy- ing an extra copy of their operating system is pretty much zero. They have the fixed cost of developing the product, and then the cost of burning one more CD is virtually nil. That’s also true of the Internet—the cost of setting it up was huge, and the cost of adding person with a telephone, it’s worthless. Its value an extra unit is minuscule by comparison. Most to me increases as other people buy it, because network-structured economies have this funda- then I have more people I can talk to. And if I mental problem that supply tends not to be were the only person in the country club, I upward-sloping. They really do make it up wouldn’t be willing to pay very much to join on volume. it. The value typically increases with the number

 ENGINEERING & SCIENCE NO. 1 31 Von Neumann thought he was the smartest person in the world, so he couldn’t understand why he kept

losing at poker. Being von Neumann, he decided to figure it out, and he developed the theory of strategic

interaction between individuals.

of users, but not necessarily. If you have a cable Internet, I’d like a high-speed connection better modem, your downloading speed is divided by than a low-speed connection, for instance—and the number of users who share your connection. it depends on U, the particular group of users. Now you have a positive benefit proportional to It might depend positively on U, as in a telephone the total number of users, plus a negative benefit network; or it might depend negatively on U, as proportional to the number of users who live on in the cable modem example, or in a swimming your block. The net benefit, if you’ll pardon the pool—the more people in the water, the larger pun, of having your neighbor in the system may the negative impact when I jump in. be negative. We call such attributes externalities, We can describe our four goals mathematically. because now the product’s value is external to the Efficiency requires us to choose the A and the U product itself. that maximizes the sum of our individual welfares, So we have to throw our beloved picture out minus the cost of providing network configuration and go to a more abstract framework to analyze A. Voluntariness means that if we pay for the

network markets. We use game theory, a math- network by charging each user an amount, Ti, ematical technique developed in the 1940s by which could be a flat fee or a function of some John von Neumann and Oskar Morgenstern at sort, the benefit you get minus the money you pay Princeton. Von Neumann thought he was the has to be nonnegative. In other words, everybody smartest person in the world, so he couldn’t comes out ahead from being part of the network, understand why he kept losing at poker. Being or at least breaks even. And the network itself

von Neumann, he decided to figure it out, and breaks even—that’s balance. The sum of the Ti’s he developed the theory of strategic interaction has to equal the cost of the network. If the Ti’s are between individuals. Game theory lay dormant insufficient, we need an external infusion of cash until the 1970s, when it was seized upon by econ- or assets. If they exceed the cost, we have to omists starting to get interested in the economics decide what to do with the surplus. And finally, of information. In fact, a lot of the pioneering strategy-proofness says that your benefit from

work on game theory in economics was done (A,U) minus Ti has to be at least as great as the here at Caltech by some of my colleagues—Matt benefit you could get by lying, by manipulating Jackson, John Ledyard, Dick McKelvey, Tom the system to induce some other (A´,U´) minus

Palfrey (PhD ’81), and Charlie Plott. When your Ti´ for that different choice of network condi- we model the market as a game, we ask: can we tions and users. For instance, if I’m in a rural area, design an economic mechanism, like we would it might be very expensive to connect me to the design an engineering device, that has the attri- network. But if I lie and say I have an enormous butes we want and solves the problem we want to value, then it’s still efficient to connect me, so it’s solve? Is it mathematically possible to construct in the interests of rural users to manipulate their such a thing? values upwards to ensure that they’re connected. The network math works like this. We have The only way to stop that from happening is to

a set of alternatives, which we call A’s. For the make their Ti’s extremely high. In fact, U.S. electricity grid, each A would be a possible topol- policy is exactly the opposite—we subsidize rural ogy of the network: the capacities of the transmis- users to help them connect to the network. sion lines, how they are connected, and so on. The Game theory has led to several relevant theo- benefit I get from being in this network depends rems. The fundamental one, published indepen- on the choice of A—if we’re talking about the dently in the early 1970s by Allan Gibbard (then

32 ENGINEERING & SCIENCE NO. 1  as long as nobody else is gaming the market. If everyone else is playing fair, the system enforces fair play on my part. But if a group of people collude and try to game the market, they can do it. This local strategy-proofness is called the Nash equilibrium, because mathematician John Nash developed the idea at Princeton in about 1950. So the state of the art in network models is that we can get three out of four. It’s mathematically impossible to achieve all four. This means that we have to tailor each market to the particular charac- teristics of the network it serves—there is no one- size-fits-all optimal policy. And yet, our economic It’s a holdup in both senses—it’s an impediment to policy is still largely driven by the standard, com- petitive-market goal of four out of four. But in your use of the network, and it’s highway robbery. fact, when we look at the best way to handle dif- ferent network markets, we may arrive at conclu- sions that are polar opposites of each other, as I’ll demonstrate with a couple of examples—Microsoft and electricity. at Chicago, now at Michigan) and Mark Satter- Microsoft’s operating system is a network exter- thwaite (BS ’67) at Northwestern, stated that it’s nality, because the more people that use it, the impossible to find a mechanism that satisfies our more products are developed for it, and the more four requirements that it be efficient, voluntary, benefit you get from it. (I use a Mac, myself, so strategy-proof, and balanced. However, the I’m denied a bunch of software that other people approach was so abstract that the possibility have; but for some reason I get more benefit from remained that for some class of network models having a Mac, and fewer friends, than other people one could, in fact, have all four. This hope was do from enjoying more friends and cheaper prod- dashed in 1979, when Harvard’s Jerry Green and ucts. Go figure.) J. J. Laffont (Laffont is now at the University of The problem with this network is a really subtle Toulouse) revisited the issue. Their work was done one, but it’s very interesting. It’s what economists in the context of providing a public good, such as call the holdup problem, and it occurs when you building a bridge, but applies to networks as well. have a network made of different components that It says that no general network-market model can are priced separately. Imagine that you’re in New satisfy all four of our desiderata. York City and you want to travel down to Wash- Caltech’s Matt Jackson, then at Northwestern, ington, D.C., to protest some issue. You jump in in collaboration with Salvador Barberà at the your car and you get on the New Jersey Turnpike, Universitat Autonoma de Barcelona, found that which is a toll road. You drive through Jersey to if we’re willing to chuck out our beloved, slavish the Delaware Pike, you pay a second fee to Dela- devotion to efficiency, we can come up with a ware, and you get to Washington. Here’s the mechanism that will satisfy the other three holdup problem: suppose it’s worth a dollar to you requirements. We won’t need the Sopranos, to take the trip, and the Jersey Turnpike charges people won’t game the system, and it requires no 50 cents. A dollar minus 50 cents leaves 50 cents. external infusion of funds. However, the mecha- You drive down the road and you get to the Dela- nism looks a lot like price caps, which makes ware Turnpike. What if the Delaware Turnpike industry very nervous. This result was also hits you for a buck? They have you over a bar- obtained independently by Hervé Moulin (then at rel—you’ve already spent 50 cents, but you’re Duke, now at Rice) and Scott Shenker (a computer closer now than ever and it’s still worth a dollar scientist then at Xerox PARC, now at Berkeley, to you to finish the trip. So you go on to Wash- who was interested in network protocols). And ington, you eat the other 50 cents, and you mutter a theorem by Ted Groves at UC San Diego says, to yourself, “What a rip-off. I’ll never do that when applied to this context, that we can keep again.” So it’s a holdup in both senses—it’s an efficiency while getting strategy-proofness and impediment to your use of the network, and it’s voluntary participation, if we’re willing give highway robbery. up balance. Now, imagine I’m Microsoft and I’ve got a nice And finally, several other people and I have little monopoly going, with all these peripheral shown that we can get efficiency, voluntary partic- products adding value to my network, and then ipation, and balance if we’re willing to give up somebody new comes along with something as strategy-proofness as a global concept and replace essential as my operating system. Suddenly, in it with a local concept. That is, instead of it not order to get the full benefit of your computer, being in anybody’s interest to game the market you have to buy my product plus this other guy’s ever, it’s not in my interest to game the market product. But you have to buy my product first.

 ENGINEERING & SCIENCE NO. 1 33 A browser isn’t much good without an operating areas and thwarted it in others, so the overall effect system to run it on—at the moment. This means on efficiency isn’t clear. We don’t know what the that the industry overall faces a holdup problem. efficient policy actually is, and it might not be Microsoft knows that if it charges a high price for that the efficient policy is the best policy. For its product, then the browser company, which now example, if the efficient mechanism’s not balanced, has a captive market, can also charge a high price. the social cost—the flow of money in or out of the But then nobody will buy either product. So the system—might outweigh the benefits. It’s a very first firm in—I hate to use this expression—the complex issue, and we need to spend a lot of time value chain really is threatened by the firms farther modeling the minutiae of the industry in order to down the line. In the big picture, it might actual- get the right solution. ly be mathematically efficient for the first firm in Let’s move on to the mother of all mess-ups: electricity “deregulation.” I’ve got that in quotes because people usually think of deregulation as Let’s move on to the mother of all mess-ups: electricity “deregulation.” I’ve removing regulations, but this “deregulation” produced a new set of rules the size of a phone got that in quotes because people usually think of deregulation as removing book. Electricity is another pervasive network regulations, but this “deregulation” produced a new set of rules the size of a externality. Electric power follows Kirchhoff’s law, as you may remember from Phys 1, so we phone book. don’t know where the individual electrons are going but we know systemically what’s going to happen. Electricity users are very sensitive to line to kill off the second firm and integrate the fluctuations in voltage—I have a set of expensive two products. Predatory pricing—selling the tube amps in my stereo at home; I’m really unhap- second good for free, or below cost, in order to kill py when they blow. And there are lots of comput- a competitor—is illegal, but it solves the holdup ers containing lots of business records in lots of problem. Maybe Microsoft should be allowed to offices. So it’s essential that the quality of the decide whom to subsidize and whom to kill, given service is held constant; that is, the voltage fluctu- that it already has the operating-system market ations must be kept within tolerable levels. tied up. If Microsoft had a viable competitor in The way this was traditionally dealt with was that market, that might not be true, because then by having a monopoly; the monopoly solved the people would have an alternative route to make systemic problems; we regulated the monopoly. the journey. But a monopoly could charge a high price and be Current federal policy is driven by the idea that inefficient. Under regulation, it turns out it was we want to break Microsoft’s monopoly because still inefficient. Under the old regulatory system, monopolies are bad. We outlaw predatory pricing, we had balance—the system broke even; the price because predatory pricing enhances monopoly. regulations ensured that. We had voluntary We want to have open systems, open network participation. The Supreme Court ruled in the platforms, to guarantee the largest amount of Hope Natural Gas Company case of 1904 that access and the largest amount of product develop- a regulated firm was entitled to a fair return on ment. But in the network model, none of those its investment. So the utilities weren’t coerced— things can be shown to always be efficient. Micro- Edison voluntarily sold electricity under regula- soft’s strategy has encouraged innovation in some tion, and made money as a result. The argument

34 ENGINEERING & SCIENCE NO. 1  on the consumer side is a bit more subtle: nobody and SDG&E should just have increased their forced me to be part of the network. In theory, I demand and paid the higher price. But the cost could have always gone “off the grid” and put they could pass on to the consumers was fixed photovoltaic panels on my roof, or a windmill in by the retail price caps, so they didn’t want to my backyard. It wouldn’t be cheap, but I could do that—they’d take a bath if they did. (As you do it, and as electric bills spiral upward, a number know, they took a bath anyhow, but I’ll get to of people are. Or I could have renounced my TV, that in a moment.) On the other hand, if they microwave oven, air conditioner, computer, etc. reduced their demand a little bit, it would push and lived like the castaways on Gilligan’s Island. It the unfilled consumer demand into those last- wasn’t likely to happen, but nobody was stopping ditch markets where the ISO would have to cover me. And, finally, the price caps made the system it. The ISO divides its cost between the users, so relatively strategy-proof. We had three out of the logic was this: I could buy an extra dollar’s four, and the downside was that we lost efficiency. worth of electricity today, but if I don’t, the ISO Deregulation was enacted under political con- will buy it tomorrow. It will cost the ISO two straints, so they went for four out of four. The bucks, but if the ISO divides that equally among theorem says you can’t do it—unfortunately, us, two over three is less than a buck. Unfortu- nobody read the theorem in policyland. They nately, this only works if I’m the only person who intended to lower prices, so consumers would does it. If two people do it, the cost becomes two benefit and would join voluntarily; they were people times two dollars divided by three people, going to induce efficiency by relaxing producer which is more than a buck; if all three do it, every- price controls, so producers would join voluntarily, one winds up paying two bucks. There’s a strong too; and the mechanism was set up to break even, incentive to be the first to act, even though the so it had balance. And they relied on competition advantage you get is fleeting, because if you are to make the mechanism strategy-proof. honest you are guaranteed to lose unless everyone So what happened? There’s actually a sequence else is equally honest. (This particular scenario is of markets. The so-called day-ahead market, for a staple of game theory, and is called the Prisoner’s delivery tomorrow, is by the hour: 10 o’clock, Dilemma, because it was originally couched in 11 o’clock, 12 o’clock, and so on. The day-ahead terms of two cellmates given the opportunity to market matches expected supply and demand. rat each other out in exchange for a lighter sen- But say the next morning it turns out that the day tence.) So when the price went up, the declared is going to be hotter than forecast, and people are demand—the amount the utilities said they going to crank up their air conditioners. So there’s wanted to buy—went down, and the excess was the morning market, which is for same-day deliv- pushed into the residual markets. Suddenly the ery in 15-minute intervals, to fine-tune supply and ISO, which was intended to do the transactions demand. And as the delivery deadline approaches, needed to suppress the last little fluctuations in there are many more markets: a market for spin- the system, was buying 15 percent of the power. ning reserves—people being paid to keep their It was never meant to do that. And it was allo- generators running in case they’re called upon; cating costs in a way that was completely non- for nonspinning reserves—people who have their strategy-proof. plants fired up, but the generators aren’t turning; The supply shock—that price jump—was set and so on. There’s this whole hierarchy of markets up, again, because there was no balanced, strategy- based on how quickly a particular plant can be proof mechanism. The miracle is that it took a called on to produce. Then, at the very last year for the flaw to become apparent. Anyway, moment, there’s a market that forces supply when supply was withdrawn, the same thing to be equal to demand. happened—the demand was forced into the last- The later markets are run by the Independent minute markets, where you can charge almost System Operator, or ISO, a sort of quasigovern- anything and the ISO has to pay it. All at once, mental operation that pays whatever price has to the scheduled maintenance time, the downtime be paid to maintain our constant voltage. The when generators were removed from the system, ISO’s mandate is to keep the lights on at all costs. roughly doubled. Maintenance outages are a mat- The ISO loses money, but it has to break even ter of public record—whenever a generator is out because the system has to be balanced. So it allo- for part of a particular day (the data don’t track cates its cost to the users, which in this case in- duration) the operator has to file a report. Genera- clude the Big Three utilities: Southern California tors were going on the fritz left and right—on Edison, Pacific Gas & Electric, and San Diego Gas some days we had 30 percent outages. And one & Electric. But it doesn’t instantaneously know study estimates that the producers’ profits went who caused the excess demand, because it’s a sys- up by $6 billion. Once again, you can’t get four temic problem, so the cost is shared via some rule. out of four. If we’d settled for three out of four, In May 2000, for reasons that I’ll talk about we never would have had this problem. shortly, prices on the day-ahead market jumped So the combination of giving the producers through the roof. Consumer demand was up, so if an incentive to withhold supply, magnified by the mechanism was strategy-proof, Edison, PG&E, an incentive for the buyers to withdraw demand

 ENGINEERING & SCIENCE NO. 1 35 Forced or Planned Outages the lowest bid. So it’s to your advantage to bid a low price, because you’ll get paid the highest price. Then there’s another market for the next 100, and the process repeats. Now, if you try to withhold supply, you take yourself out of all the markets except for the last one, so your action benefits you only on the last 100 megawatts— unlike the exisiting situation, where you would Percent affect the price of all the megawatts sold. And of Generating 6/17/2001 by taking yourself out of the previous markets, 6/18/2001 Capacity 6/19/2001 you lose all the business transacted therein; if 6/20/2001 6/21/2001 demand is less than you predict and the market 6/22/2001 6/23/2001 never gets to the 100 megawatts you’re holding 6/24/2001 6/25/2001 out, you’re only hurting yourself. The more 6/26/2001 markets, the better this mechanism works— within computational reason, of course.

Big Five Actually, I think the best mechanism is to

All Others have a spot market with the no-S solution I’ve just described, coupled with the ISO using long- term contracts for power reserves that are awarded has proved to be disastrous. The system can’t be by what I call the Teacher’s Pet method. Brownie balanced—it’s running a giant deficit of about $15 points are given to producers who consistently In the current market, billion, and the taxpayer has to pick up the bill. have the lowest prices on the spot market, or who when supply is reduced We have managed to achieve none out of four. have the best reliability record—i.e., the fewest Since we can attain three out of four, there at the sellers earn a higher maintenance outages. Then if two bidders come least four possible solutions. We could renounce E in at the same price, the one with the most brown- price on every megawatt (efficiency). That is, we can go back to a regulated ie points wins. Many government agencies already sold. Thus, when a gener- monopoly, or we could introduce price caps. A lot do this—the Defense Department, for example, ator goes off line, the of people are crying for that, because, well, things puts performance incentives into its contracts, and were bad in the old days but they weren’t this bad! bases future awards on the contractor’s history of sellers that own many Or we could give up on V (voluntariness)—the cost overruns and so forth. This system could also generators make higher state could seize the power plants through be used to reward whistle-blowing companies by profits. The above data eminent domain, and force them to sell power to giving them major brownie points in the next from the ISO shows a us at a fixed price. There have been a lot of calls round of contracts. However, as I said before, for that, too. Alternatively, we could abandon B when the ISO enters into long-term contracts, week’s worth of generator (balance). If we got rid of the balance require- the balance requirement goes out the window. outages reported by the ment, we could assign long-term contracts for the So I call this the no-BS solution. so-called Big Five delivery of a specified amount of power based on These kinds of issues are going to get even big- our best guesses for demand. We know we can producers (AES, Duke, ger and more complex as our economy increasingly engineer the awarding of these contracts in a becomes a network of networks. So we really need Dynergy, Mirant, and strategy-proof manner, per Ted Groves’s theorem I to sit down now, and figure out the arcane details Reliant) compared with mentioned earlier. However, then the ISO will not of how these markets work, in order to head off ■ the smaller, independent always break even, and the taxpayer will have to future missteps. foot the bill. And the imbalance could be large, producers. in which case we’re no better off than we are now. My preferred solution is to design a better market; that is, we relax S (global strategy-proof- ness) and go for local strategy-proofness by giving people the correct incentives. In the previous mar- PICTURE CREDITS: ket design, all the units of electricity that were bid 28, 34–35 — Edison for less than the market-clearing price—the price International; 30, 31 — of the lowest unsuccessful seller (if you arranged Doug Smith all the bids from lowest to highest, the lowest Assistant Professor of Economics Simon J. Wilkie unsuccessful seller would be the first seller whose earned his B. Comm. from the University of New South bid was not taken)—were sold at the market- Wales in 1984, and his MS and PhD from the Univer- clearing price. We could stand that on its head, sity of Rochester in 1988 and 1990, all in economics. by breaking up the market into a set of smaller He was a member of the technical staff at Bellcore, where markets for each unit of capacity—per 100 mega- he specialized in game theory as applied to telecommuni- watts, say. So there’s a market for the first 100 cations markets, before coming to Caltech as an instruc- megawatts, and the market sets a price. But then tor in economics in 1994. He was made an assistant what we do is we award the sale—at the market- professor the following year. This article is adapted clearing price—to the generator who submitted from a Seminar Day talk.

36 ENGINEERING & SCIENCE NO. 1  Ralph Leighton’s R e v i e w s Tuva or Bust!, and recruited Parnell (previously best known for his adaptation of The Cider House Rules) as playwright. The play consists of Feynman talking—sometimes on the telephone, with his wife, friends, colleagues, and doc- tors, as well as with a student (the only other character in the play), but mainly directly to the audience—during a day and evening near the end S CIENCE ON S TAGE of his life. Feynman/Alda talks mostly Alan Alda plays Richard Feynman in QED at the Mark Taper Forum (Caltech about himself: his interests, physicists were consulted on the blackboard.) his past life, his future—his science? We do get some, especially in the first act, but by Jay A. Labinger Certainly science pervades our it is hardly integral to the Administrator, contemporary world, and play. We are treated to a Beckman Institute QED, a play by Peter Parnell, equally certainly the theater number of platitudes about must reflect that world to science; we are told, but hard- produced by the Center Theatre stay relevant. But it is not ly ever shown, how excited Group at the Mark Taper Forum, clear whether the aesthetic scientists are about their Los Angeles, The 20th century has given and intellectual demands of work. Alda tries to illustrate us a number of encounters the two spheres are compat- what doing Feynman’s kind March 10–May 13, 2001 between the spheres of science ible—the encounter might be of physics might be like by and the theater—Brecht’s more of a collision! means of an example from Galileo (1939) and Dürren- At the very least, the play- chess, not from science. On matt’s The Physicists (1962), wright tackling a science- the occasions when real sci- Oxygen, a play by Carl Djerassi and to cite just two examples. As related theme will have prob- ence is presented, it is at a Roald Hoffmann, we transit to a new century lems to solve. How much of level way over a nonphysi- produced by the San Diego and millennium, the fre- the scientific content must cist’s head, as when Alda quency and impact of such the audience understand, for starts sketching Feynman Repertory Theatre at the Lyceum plays seem to be decidedly the play to be fully effective? diagrams on a blackboard, Theatre, San Diego, April 2–7, 2001 on the increase. The most For example, scientists’ moti- explaining them in terms of (published by John Wiley & Sons, visible has certainly been vations might well appear virtual photons and the like. 2001) Michael Frayn’s Copenhagen, incomprehensible to an au- This combination of vague an account of a World War II dience that doesn’t appreciate generalities and arcane com- meeting between Niels Bohr the significance of their scien- plexities, with little in be- and Werner Heisenberg, tific work. On the other tween, has the effect (whether which has enjoyed lengthy, hand, one of the more basic intended or not) of marginal- sold-out runs in London and rules of theater is “show, don’t izing the scientific theme. New York, despite (or, possi- tell.” How can that signifi- The audience is encouraged to bly because of, though that cance be adequately commu- take in what’s easy and tune doesn’t seem very likely) a nicated, without violating out what’s hard, never chal- heavy dose of dense quantum that rule, and risking a com- lenged to work at making mechanics. David Auburn’s plete breakdown of rapport? sense of unfamiliar ideas. Proof, about a mathemati- Two plays with science Perhaps the clearest indica- cian’s daughter, recently won connections have recently tion of how little is expected the Pulitzer Prize for best premiered in Southern is that every time (it seemed drama of 2000. Tom California. The first, QED, like dozens, though I suppose Stoppard’s Arcadia (my per- features Alan Alda portraying it was only three or four) Alda sonal favorite in this genre), the late Caltech physicist says “quantum electrodynam- featuring themes of chaos Richard Feynman. Appar- ics” he turns to the audience theory and thermodynamics, ently Alda himself was the and repeats “QED.” Couldn’t has been a continual success, prime initiator of the project, they trust the audience to with frequent revivals, since having been impressed by the figure out the title’s signifi- its first appearance in 1993. dramatic potential of Feyn- cance after the first time? Should this surprise us? man’s life as depicted in The net result is that Feyn-

    ENGINEERING & SCIENCE NO . 1 37 man the character is not a toine Lavoisier, the first to material) but their solution to scientist with a personality; understand what oxygen the problem is not entirely he’s just a personality who really is? All three? satisfying. happens to be a scientist. Interwoven with the con- In an interview with a San QED may well appeal to temporary action is an ac- Diego paper before the many—it does afford the op- count of a (fictional) 1777 premiere, Djerassi claimed portunity to spend some time meeting of the three chem- that their writing about “a with an entertaining persona ists, invited to Sweden by part of our culture which we (though how much of that King Gustav III to decide did not have to absorb” was is Feynman, and how much who should get credit. Each an advantage; but it may have Alda, is not easy to ascertain). of the three is assigned his also been somewhat of a dis- But the problems of dealing advocate on the committee, advantage, making them a bit with a scientific theme in a whose arguments in favor of less sensitive to the needs of play have not been solved in their candidates echo not only an audience that is unfamiliar any way, merely evaded. those made by the candidates with that culture. Similarly, Oxygen is a different matter. on their own behalf but also the contemporary chemists The playwrights are two well- sad stories about priority are not so compelling char- known chemists, Nobel claims and professional jeal- acters as one might wish. laureate Roald Hoffmann and ousy among the advocates They are obviously meant to National Medal of Science themselves. This resonance be seen as passionate about awardee Carl Djerassi. (Both is nicely reinforced by having their science, which carries are also well known outside of a single actor play each over to the positions they chemistry as prolific authors candidate-defender pair; take during the committee’s of fiction, nonfiction, and temporal scene shifts are deliberation, but we aren’t poetry.) The premise of signaled by minor costume really shown where such Oxygen is that the Nobel changes. Another resonant passion might come from. Foundation has decided to device is the inclusion of a Perhaps the authors, as pas- institute a new program of young historian of science sionately committed scien- “retro-Nobels,” recognizing writing her dissertation on tists themselves, thought it work done before the estab- “Women in the lives of 18th would be obvious? lishment of Nobel Prizes at century scientists” as secre- It seems likely that Oxygen the beginning of the 20th tary to the Nobel com- was influenced by Arcadia: century. A committee for the mittee; the wives attend and the two plays exhibit certain retro-chemistry award quick- play important roles at the similarities (beyond the ly zeroes in on the discovery 1777 meeting, especially scientific themes), most of oxygen as a worthy subject Mme. Lavoisier. prominently the use in both for the award. But who Evading the playwright’s of alternating time frames. If In the San Diego Repertory Theatre should receive it? Carl dilemma is not an option here the latter is more successful as Wilhelm Scheele, a Swedish as it was in QED: the scien- a dramatic event (which it is), production of Oxygen, Lou Seitchik pharmacist, who was appar- tific content is central to the there is no shame in that for (left) stars as Priestley, Randall ently the first to obtain a dramatic argument. Lavoisier Hoffmann and Djerassi— Dodge (center) as Lavoisier, and Jeff sample in the laboratory? was the first to understand Stoppard is, after all, one of Anthony Miller as Scheele. Joseph Priestley, the first to the role of oxygen in phe- the leading playwrights of publish his findings? An- nomena such as combustion our time. But possibly there and rusting, thereby over- is an instructive message, that throwing the phlogiston one must be wary of being too theory in which both Scheele close to one’s subject. Oxygen, and Priestley devoutly be- much more than QED, illus- lieved. Unless one appreci- trates both the potential ates the significance of that, problems and rewards of the priority dispute makes dramatizing science. Let’s little sense. So somehow it hope that Hoffmann and must be explained, without Djerassi, and others as well, squelching the drama by a will keep on trying. ■ descent into didacticism. Hoffmann and Djerassi try Jay Labinger, a chemist and hard to steer between the two member of the professional staff, looming cliffs (at one point is coeditor of The One Culture? they interpolate a stylized A Conversation About masque, performed by Science, which has just been Lavoisier and his wife, to published by the University of communicate some of the Chicago Press.

38 ENGINEERING & SCIENCE NO . 1     the United States.” Fast and sentence. His second mar- two-fisted. riage is recorded, but without Having your biography any close-up detail. written while you are still Nonetheless, there are areas alive, the English poet Philip of private life that are rele- Larkin said, is like being vant to the personal evolution measured up (still breathing) of someone so extraordinarily by the undertaker. I am not distinguished as Baltimore. a scientist (even Crotty’s Most careers, even “a life in accessible explanations about science,” follow the rule As recombinant DNA and retro- the twig is bent, so grows viruses are sometimes a bit the tree. Childhood—the beyond me). But I am a formative years—is impor- biographer. And it is the tant. Baltimore’s first 20 A L IFE IN S CIENCE problems of the biographer’s years are summarized here in craft that primarily interest three pages. Crotty gives us me in Crotty’s enterprise. a luxuriant and protracted It is difficult to write word picture of the Swarth- “authorized” biography about more campus (“the dogwood the living. Baltimore evi- trees and a thousand rhodo- by John Sutherland, dently sanctioned this book, dendrons bloomed, carpeting Visiting Professor of Lit- although the interviews he the campus with red, cream, erature gave his biographer seem to and pink petals”) but nothing have been singularly un- about the Baltimore home or revealing. Punches have to be even whether there were pulled when dealing with a siblings other than a brother What, one wonders, goes living subject. If they’re not, briefly mentioned. What did through David Baltimore’s authorization and “permis- his parents do for a living? mind when, having picked sions” are yanked. And the We learn that Baltimore up his espresso from the Red libel lawyers are in the wings and Francis Ford Coppola Door (as I’ve seen him do), (you can’t, as every biogra- “were the two-man tuba he browses the Caltech pher knows, libel the dead). section in the Great Neck Bookstore and passes a rack But for the reader the plea- High School Marching Band” stacked high with two books sure in biography is, essen- (a curious fact for which one Ahead of the Curve: about David Baltimore: Dan tially, voyeuristic. We want is profoundly grateful). But Kevles’s The Baltimore Case to see what makes the person I can find nothing in this David Baltimore’s Life in Science (Norton, 1998) and now, “tick.” To do that, you have biography about Baltimore’s 270 pages three years later, Shane to take the back off the watch father. His mother is credited University of California Press, 2001 Crotty’s biography. and do some prying. with being his lifelong in- $29.95 Everything has happened In Baltimore’s case, bio- spiration. But she has no fast in Baltimore’s life. As graphical prying is further index entry, nor does she Crotty records, he believed, in discouraged by the fact that make any real presence in his early 20s, that if one did he is, manifestly, someone the narrative. not make one’s mark by 30, who values and protects his There is nothing, apart there would be no mark for privacy. Crotty has been from one throwaway refer- posterity to admire. He got careful not to trespass. So ence, to Baltimore’s Jewish- his Nobel aged (if that’s the much so, that at times he ness. As someone born in word) a prodigiously youthful seems to be complicit with 1938, he may conceivably as a 37. His career hit a wall, it his subject in veiling what young man have encountered seemed, in his 50s with “the biography normally conceives prejudice. He is now, we case.” Now, phoenixlike, he its responsibility to uncover. learn, “unreligious.” Was his is arisen to lead Caltech into This is not to say that one upbringing secular, or did he the New Millennium. “I live wants National Enquirer or lose his faith? in the future, not in the past,” “blackwash” revelations (not The lack of personal back- he is quoted as saying. Hav- that there would be any here). ground is tantalizing. More ing achieved so much, he has One can respect Crotty’s de- so, since there are fleeting still, it seems, much to cision not to press on personal allusions to important aspects achieve. Nor will they be but (in this context) irrele- of his adult personality passed ordinary achievements. He vant aspects of his subject’s down from his family. There is, he believes, “the only life. Baltimore’s first mar- is, for example, a parenthetic functioning scientist who is riage and divorce, for ex- reference to the Baltimores running a major university in ample, are dealt with in a sympathizing with, for two

    ENGINEERING & SCIENCE NO . 1 39 generations before David, O b i t u a r i e s “leftists and socialists.” In his thirties, we learn, Baltimore “hated Nixon,” and thought his “War on Cancer” a sham. At this period of his life (when he was doing his most exciting scientific work) Baltimore “disdained capital- ist society” and declared himself “an anticapitalist.” J EFFREY S COT B ANKS When did his views change? 1958–2000 Or have they? One of the more interesting human subplots to the narra- tive is Baltimore’s impas- sioned resistance to the Vietnam War (had he been born five years later, Canada might have been able to claim him as its most dis- billionaires. Accommodations Pandora’s box his genius has tinguished scientist). Balti- must be made: with institu- opened. more’s truly eloquent and tions, with the state, and This is an interesting study idealistic outburst against the with “capital.” As a young of a fascinating and important ineffable John Dingell during scientist, Baltimore appar- man. But, as biography, “the case” reminds one of ently believed that if funds Crotty’s book stimulates an nothing so much as those were needed for his kind of appetite it signally fails to gallant dissidents who stood science it should ideally be satisfy. There remain enig- up publicly to denounce from the taxpayer (“the only mas. For instance: the best HUAC and McCarthyist way to do research was on prose in the book is Balti- purges, 40 years earlier. government money”). more’s (I would point to (Crotty, incidentally, handles But when he made his pact the witty summary of his this episode very effectively.) with a large institution (with “education in irrationality,” The aspect of Baltimore’s the ultracapitalist name, in his inauguration address intellectual character that Rockefeller) did he have any at Caltech, quoted here as emerges most clearly is that twinges of “radical, leftist” epilogue). Baltimore is a he is a loner. As a young conscience? brilliant scientist, yes. But scientist he was a self-made What went through Bal- he is also a highly cultivated man. His alma mater will timore’s mind, in August man, with a love of theater, take no pleasure in Crotty’s 1980, when Jack Whitehead jazz, art, and literature. We book. Baltimore, perhaps its offered Baltimore a research do not learn from this book most famous living alumnus, institute? He who sups with how he became that unusual is quoted as saying: “At the devil should use a long man. The posthumous Swarthmore the teaching of spoon? Or, this is the only biography will doubtless tell biology was poor—at best. way forward for research, such us. And, by the time it The courses were really as that into molecular biol- comes along, there will, for generally bad.” But perhaps ogy, which needs unimagin- a certainty, be much, much genius needs to be left alone, ably large sums of money? more for the biographer to to grow at its own rate in These are questions that the record. ■ its own peculiar way. For reader (legitimately, I think) students like David Balti- asks. This biography gives more, bad courses are the hints, but no answers. best courses. Would under- There is much to applaud John Sutherland taught at graduate education at Cam- in Crotty’s book. I found his Caltech, in the Division of the bridge, MIT, or Caltech have expositions of Baltimore’s Humanities and Social Sciences, crushed the original genius research for the layman (as a from 1983 to 1992, and has out of him? layman) admirably compre- visited quarterly since. He has Late-20th-century, labora- hensible. Crotty is good on written biographies of Sir Walter tory-based science cannot be the ethical problems raised Scott and Mrs. Humphry Ward done at the highest level by by gene research, and Balti- and is currently writing the “loners.” It costs too much. more’s (sensible, one appre- authorized biography of the poet Few biologists are born hends) thinking on the Stephen Spender.

40 ENGINEERING & SCIENCE NO . 1     In 1996 Jeff Banks (right) received the National Academy of Sciences Award for Scientific Reviewing, presented by John Ferejohn, Munro Professor of Political Science at Stanford and senior fellow of the Hoover Insitute (Ferejohn taught at Caltech from 1971 to 1983). Banks’s name appears on the wall with those of awardees from previous years.

Jeffrey Scot Banks, profes- the Humanities and Social thinking, honesty, and clever- and return to Caltech, he sor of political science, died Sciences, likened the empty ness, I think of Jeff,” said claimed, Banks would have December 21 at the age of 42 seat at the conference to the Porter. “And if you worked substantially changed polit- of complications of a bone- hole in a pilots’ formation or with him, you knew you were ical science and political marrow transplant. the empty barstool. Ledyard in for a lot of laughs and fun.” economy at the university. After earning his PhD from welcomed everyone and intro- In addition to numerous “There are some people with Caltech in 1986, Banks left to duced the other speakers— academic papers (“It’s re- whom everyone identifies as a join the faculty of the Uni- Banks’s teachers, colleagues, markable how much he ac- friend,” he said. “Jeff was one versity of Rochester and re- and students, who offered complished in such a short of those people.” turned to Caltech in 1997. remarks in roughly the span of time,” said McKelvey), John Duggan, PhD ’95, He taught and did research in chronological order in which Banks coauthored a book, who followed Banks’s path to the general field of political each speaker had encountered Positive Political Theory I: Col- the University of Rochester, theory, including political his career. Ledyard was lective Preference, with David where he is now associate economy, game theory, and slightly out of chronological Austen-Smith. Austen- professor of political science social choice. He made sig- order himself, arriving as a Smith, now professor of and economics, described him nificant contributions to a professor at Caltech just as political science and econom- as a “really deep thinker, a field of political science Banks was finishing his PhD. ics at Northwestern Univer- careful and rigorous thinker, characterized by the use of Born in San Diego, Banks sity, spoke of Banks’s “evan- and he challenged you to be formal mathematical and graduated from UCLA in gelical zeal” for political also.” deductive methods to model 1982. Richard McKelvey, the science. “To Jeff, doing “He just loved research so political behavior—behavior Wasserman Professor of Polit- research was sheer pleasure.” much,” said Duggan. “His such as strategic voting, bar- ical Science, recalled hearing He also noted Banks’s easy energy and enthusiasm were gaining, coalition formation, of this “really smart UCLA disposition and enthusiasm infectious, and that made and jury decisions. student” who had applied to and his ability to introduce working with him so much A conference in his honor Carnegie Mellon for graduate lines from the movie This Is fun.” Banks was productive was held on campus April 7; school. McKelvey set about Spinal Tap into seminar even when he was ill, Duggan students and colleagues pre- explaining to Banks why he presentations. added, and left several papers sented papers that drew on should come to Caltech in- In 1986, Eric Hanushek that will be published post- Banks’s work on the role of stead. He did, earning his (now at Stanford’s Hoover humously. “In the profession, incomplete information in PhD in 1986 with a thesis on Institution) was chairman of our debt to him is great.” models of political processes. “Signaling Games: Theory the economics department at Banks received numerous Those colleagues, who had and Applications,” with Mc- the University of Rochester awards and recognition for his gathered from around the Kelvey as his thesis adviser. and found himself for the first work. From 1989 to 1994, country, also joined friends David Porter, currently on time in competition with the he was a National Science for a memorial service in the staff of the Economic political science department Foundation Presidential Dabney Lounge to remember Science Laboratory at the for a faculty appointment. Young Investigator, and re- and celebrate the life of Jeff University of Arizona, first This led to Banks’s unique ceived the National Academy Banks in a less scholarly met Banks as a fellow grad- appointment and ultimately of Sciences Award for Scien- fashion. uate student. “When I think tenure in two departments. tific Reviewing in 1996. In John Ledyard, professor of about the wonderful qualities “He was a natural success at 1996 he was also elected a eonomics and social sciences of Caltech, namely, cutting- Rochester,” said Hanushek. fellow of the Econometric and chair of the Division of edge research, innovative If it hadn’t been for his illness Society.

    ENGINEERING & SCIENCE NO . 1 41 “Jeff was my teacher and my thesis adviser,” said Faculty File Daniel Diermeier, who studied with him at Roches- ter. Diermeier explained that the German term for thesis adviser is Doktorvater. “There’s truth in this con- cept,” he said, “which, as a father and a teacher now myself, I appreciate more. Teaching is about creating H ONORS AND A WARDS someone who is then creative in turn. We grow into our research.” Diermeier, who is now the IBM Professor of Regulation and Competitive Practice at Northwestern University, also appreciated the American informality he met at as well as the big ones.” John Abelson, the Beadle Rochester. “All my previous On behalf of Banks’s family Professor of Biology, has been teachers had the same first (which includes sons Bryan, elected to the American Phil- name: Professor Doktor. And 15, and Daniel, 13), his wife, osophical Society. now here was ‘Jeff’ in his Shannon, thanked all those Paul Bellan, professor of sneakers.” Two qualities who had come that day “not applied physics, has received made him unique, said just to mourn his passing but one of two 2001 SPD Popular Diermeier: “the deep joy” to celebrate his life.” She Writing Awards, given each that radiated from him and thanked the anonymous year to a professional scientist his deep commitment to bone-marrow donor “who and to a science writer or allowed the extra time” and journalist by the Solar Physics research. Professor of Biology Pamela Banks was diagnosed with also all those in the audience Division of the American leukemia in 1995 and under- who had signed onto the Bjorkman is one of 72 American Astronomical Society. went a bone-marrow trans- bone-marrow registry because scientists elected this year to mem- Roger Blandford, the Tol- plant in the summer of that of her husband’s illness. “He bership in the National man Professor of Theoretical year. In 1997, he returned to fought long and hard to stay Astrophysics, was named the Caltech as professor of politi- with us.” She also presented Academy of Sciences (NAS). She’s Tetelman Fellow at Yale for cal science. He became exec- to Ledyard and to Larry the first woman out of a total of 2001; he delivered the Tetel- utive officer for the social Rothenberg, director of the 67 living Caltech faculty members man lecture in mid-February. sciences in 1999, a post in Wallis Institute at the Uni- elected to that honor. Bjorkman, In June, he traveled to which his dry wit and calm, versity of Rochester, framed Munich to give the Siemens easygoing nature, as well as copies of Banks’s Presidential who is also executive officer for Lecture. his knowledge of voting Young Investigator Award biology and an investigator with Mory Gharib, professor of theory, helped smooth many and a photo of him receiving the Howard Hughes Medical aeronautics and faculty mem- the National Academy of ber in bioengineering, was meetings. Institute, has been a member of Professor of Economics Sciences award. invited by the American Matt Jackson came to Caltech In closing, Ledyard stated the Caltech faculty since 1988. Her Association for Thoracic at the same time that Banks that the new seminar room in research focuses on molecules Surgery to give the Honored returned. “He wanted every- Baxter Hall would be named involved in cell-surface recognition, Speaker address to the 81st one to enjoy life,” Jackson in Banks’s memory and also AATS conference on May 8 said. “He could always see announced the creation of the particularly molecules of the in San Diego. He discussed the humor in a situation.” In Jeff Banks Memorial Seminar immune system. the challenges and rewards of his professional life, he taught Fund. Contributions to the applying bioengineering prin- others to “sweat the details; fund may be sent to Susan ciples to space exploration. details matter.” And even Davis, Caltech 228-77, William Goddard, the when his health deteriorated, Pasadena, CA 91125. Checks Charles and Mary Ferkel “he came in, taught his should be made out to the Professor of Chemistry and courses, met with graduate California Institute of Applied Physics, has been students, and kept doing the Technology. ■ selected by the Southern day-to-day things, no matter California Section of the how difficult. He made a big American Chemical Society difference in the small things to receive the Richard C.

42 ENGINEERING & SCIENCE NO . 1     Shri Kulkarni, the MacArthur Professor of Astronomy and Planetary Science (left), and Ahmed Zewail, the Pauling Professor of Chemical Physics, have been elected to the Royal Society, established in England in 1660, the world’s oldest scientific academy in continuous existence.

Tolman Medal. The medal entitled Probability, the provide research support to omy and Planetary Science, was formally awarded at the Philosophy of Religion, and the most promising young has been invited to give this Athenaeum on April 19. the Philosophy of Science. faculty members in the early year’s Sackler Lecture at Alan Hajek, associate pro- Linda Hsieh-Wilson, stages of their academic Princeton University’s de- fessor of philosophy, has assistant professor of chemis- careers in the chemical and partment of physics. received a $10,000 grant try, has been selected to life sciences.” This year Andrew Lange has been from the Center for Theology receive a 2001 Beckman marks the 10-year anniversary named the Marvin L. Gold- and the Natural Sciences. He Young Investigators award. of the program, which is berger Professor of Physics, will develop a new course The award program “helps funded by the Arnold and effective July 1. This title Mabel Beckman Foundation, replaces that of professor of an independent, nonprofit physics. foundation established in Eliot Meyerowitz, professor 1977. of biology and chair of the Philip Hoffman, professor biology division, has been of history and social science, named a Wilbur Lucious has been named a Fellow of Cross Medal winner for 2001 the John Simon Guggenheim from Yale University. Memorial Foundation. His Paul Messina, director project for the fellowship of the Center for Advanced period will be “The Role Computer Research, received of Crises in Economic and the Distinguished Associate Financial Development,” award from the U. S. Depart- on which he will collaborate ment of Energy for his with UCLA professor of achievements in computa- economics Jean-Laurent tional science and for his Rosenthal. Together with contributions to the DOE’s Gilles Postel-Vinay, Hoffman Stockpile Stewardship pro- Of 50 awards presented by the American Chemical Society at its April and Rosenthal coauthored gram, designed to ensure the meeting, more than 10 percent went to Caltech faculty members. From Priceless Markets: The Political safety and reliability of the left: the Nakanishi Prize to Jack Roberts, Institute Professor of Chemistry; Economy of Credit in Paris, nuclear weapons arsenal. the ACS Award for Creative Advances in Environmental Science and Tech- 1660–1870 (University of John Preskill, professor of Chicago Press, 2000). theoretical physics, has been nology to Michael Hoffmann, the Irvine Professor of Environmental Science; the Wolfgang Knauss has been invited to be the 2002 Herbert D. Brown Award for Creative Research in Synthetic Methods to Bob named the Theodore von Lorentz Chair at the Univer- Grubbs, the Atkins Professor of Chemistry; the ACS Award in Polymer Chemis- Kármán Professor of Aero- sity of Leiden. Described as nautics and Applied Mechan- try to David Tirrell, the McCollum-Corcoran Professor and professor “the most prestigious visiting ics, effective April 1. This professorship in the Nether- of chemistry and chemical engineering; the George C. Pimentel Award in title replaces that of professor lands,” the chair since its Chemical Education to Harry Gray, the Beckman Professor of Chemistry; and of aeronautics and applied founding in 1955 has been the ACS Award for Creative Innovation to John Baldeschwieler, the Johnson mechanics. held by 10 Nobel Prize Shri Kulkarni, the Mac- winners. Professor of Chemistry and professor of chemistry, emeritus. Arthur Professor of Astron- Richard Roberts, assistant

    ENGINEERING & SCIENCE NO . 1 43 professor of chemistry, has foreign associate of the Royal received a Presidential Early Astronomical Society “in Career Award for Scientists recognition of her inspiring and Engineers “for his inno- leadership and outstanding vative combinatorial method service to the promotion of of selecting and designing astronomy.” protein motifs that specifi- Wallace Sargent, Bowen cally recognize biologically Professor of Astronomy, has important RNA structures.” been selected as the fourth The award recognizes out- Icko Iben, Jr., Distinguished standing young professionals Astronomy Lecturer at the F ACULTY B OARD at the outset of their indepen- University of Illinois at C HAIR E LECTED dent research careers, provid- Urbana-Champaign, where ing up to five years of grant he will deliver a public lec- support. He has also been ture, give a joint colloquium selected as an Alfred P. Sloan to the astronomy and physics Research Fellow. departments, and interact For the first time, the George Rossman, professor with faculty, staff, and stu- faculty board chair will be of mineralogy, has been se- dents. The lectureship brings occupied by a woman, lected to receive the Mineral- world-renowned astronomers Marianne Bronner-Fraser, the ogical Society of America’s and astrophysicists to the Billings Ruddock Professor of Dana Medal, which recog- university. Biology. Melany Hunt, pro- nizes “continued outstand- Edward Stone, the Morris- fessor of mechanical engineer- ing scientific contributions roe Professor of Physics, and ing, was voted vice chair, and through original work in director of the Jet Propulsion Ned Munger, professor of the mineralogical sciences.” Laboratory from 1990 until geography, emeritus, secre- David Rutledge has been May 2001, has received tary, relieving Ward Whal- named the Kiyo and Eiko NASA’s Distiguished Service ing, professor of physics, Tomiyasu Professor of Medal. emeritus, after a 16-year tour Electrical Engineering. Keith Taylor, a member of of duty. Effective April 1, this title the professional staff in as- Bronner-Fraser studies the replaces that of professor of tronomy, has received the development of neural crest Alexander Varshavsky, the Smits electrical engineering. Astronomical Society of the cells in vertebrate embryos. These cells emerge from the Professor of Cell Biology, has been Anneila Sargent, professor Pacific’s 2001 Maria and Eric of astronomy and director of Muhlmann Award, for his neural tube shortly after named the co-recipient of the the Owens Valley Radio “unique contributions to neurulation and migrate to 2001 Wolf Foundation Prize in Observatory and of the astronomical instrumentation various parts of the body to Medicine. He shares the $100,000 Interferometry Science at various observatories.” ■ establish diverse cell types Center, has been elected a such as neurons, glia, and prize with Avram Hershko of the pigment cells. Investigating Technion, awarded for their what dictates the pattern of discovery of the “ubiquitin system cell migration and what of intracellular protein degradation determines the type of cell Recipients of this year’s ASCIT Teaching Awards are they become could elucidate and the crucial functions of this Oscar Bruno, professor of applied mathematics; George the causes of birth defects and system in cellular regulation.” The Cheron, lecturer in Russian; Kjerstin Easton, grad student cancers associated with neural Wolf Prize was established in 1978 in electrical engineering; Glenn George, lecturer in crest cells, and suggest ways computer science and electrical engineering; Loren to prevent them. to promote science and art for the Hoffman, undergraduate; Dirk Hundertmark, Todd There was another first in benefit of mankind. Varshavsky Instructor in Mathematics; Edward McCaffery, visiting this year’s election: the voting was also recently elected to the professor of law; Thomas Neenan, lecturer in music; and was conducted on line. As American Philosophical Society. Charles Peck, professor of physics. befits one of the partners in The Graduate Student Council’s Teaching Awards went the Caltech/MIT Voting to Hans Hornung, the Johnson Professor of Aeronautics; Technology Project, an Julia Kornfield, associate professor of chemical engineering; electronic ballot was used, and Brian Stolz, assistant professor of chemistry. Recipients set up by project member of the GSC Mentoring Awards are Agustin Colussi, senior Michael Alvarez, associate research fellow in environmental engineering science, and professor of political science, Brian Stolz. and Marionne Epalle, com- munications specialist in Engineering and Applied Science. ■

44 ENGINEERING & SCIENCE NO . 1     Office of Gift and Estate Planning

When the late Charles earlier years at the Courant For information contact: M ATHEMATICS DePrima and his first wife Institute and the NYU Annemarie built their dream R ETREAT Mathematics Institute. He house in The Sea Ranch (a was a dedicated teacher Chris Yates, JD bucolic California coastal during his years at Caltech, Susan A. Walker, CFP community about two hours and observed that there were north of the Golden Gate), very few special talks or Carolyn K. Swanson they considered the possibil- seminars in mathematics Office of Gift and Estate Planning ity that it might someday designed for undergraduates. California Institute of Technology become a retreat for Caltech’s So, before his death, he and Mail Code 105-40 mathematics faculty. In fact, Margaret made a gift to DePrima envisioned it as a Caltech to endow an under- Pasadena, California 91125 smaller version of Oberwol- graduate mathematics lecture phone: (626) 395-2927 fach, a mathematics retreat in series, which bears his name. fax: (626) 683-9891 Germany’s Black Forest that His good friend and colleague he visited several times dur- Jack Todd, professor of [email protected] ing his 40 years as a profes- mathematics, emeritus, www.gep.caltech.edu sor at Caltech. Annemarie delivered the first DePrima passed away in 1984 after Lecture in 1991. only four years in their new Above: Annemarie DePrima at the home; she and Charles had no children. Sea Ranch house. Charles and Margaret met Right: Charles and Margaret in 1985 and married in 1987. DePrima. Together they finalized the trust originally conceived by Charles and Annemarie, naming Caltech to receive the Sea Ranch property upon the death of the survivor. Charles DePrima died in 1991. Al- though she never knew Anne- marie, Margaret has kept their dream alive by making their Sea Ranch home avail- able one month each summer for use by the mathematics faculty. DePrima was a distin- guished mathematician who had become acquainted with Albert Einstein during his

    ENGINEERING & SCIENCE NO . 1 45 Engineering & Science NON-PROFIT ORG. U.S. POSTAGE PAID California Institute of Technology PASADENA, CA PERMIT NO. 583 Pasadena, California 91125