John Beverly Oke (1928–2004)Author(s): James E. GunnSource: Publications of the Astronomical Society of the Pacific, Vol. 117, No. 829 (March2005), pp. 221-226Published by: The University of Chicago Press on behalf of the Astronomical Society of the PacificStable URL: http://www.jstor.org/stable/10.1086/428367 .

Accessed: 25/05/2014 21:19

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.

.

The University of Chicago Press and Astronomical Society of the Pacific are collaborating with JSTOR todigitize, preserve and extend access to Publications of the Astronomical Society of the Pacific.

http://www.jstor.org

This content downloaded from 194.29.185.71 on Sun, 25 May 2014 21:19:37 PMAll use subject to JSTOR Terms and Conditions

221

Publications of the Astronomical Society of the Pacific, 117:221–226, 2005 March� 2005. The Astronomical Society of the Pacific. All rights reserved. Printed in U.S.A.

Obituary

John Beverly Oke (1928–2004)

James E. Gunn

Princeton University Observatory, Peyton Hall, Princeton, NJ 08544; jeg@astro.princeton.edu

Received 2004 December 16; accepted 2004 December 22; published 2005 March 11

Fig. 1.—Photograph of John Beverly “Bev” Oke taken in 2003 Decemberby Jim Hesser.

On March 2 last year, the world and astronomy lost a gentlegiant. John Beverly “Bev” Oke died of heart problems whileenjoying a very active retirement in Victoria, Canada, wherehe had a guest appointment at the Dominion AstrophysicalObservatory (DAO). He had retired to his native Canada in1991 after a career of 34 years as the dominant force in in-strumentation in optical astronomy at Caltech.

Bev was born in 1928 in Sault St. Marie, Ontario. He wasthe son of a United Church minister and spent his childhoodmostly in small backwoods Ontario towns as his father movedfrom parish to parish. In one of these towns, Sundridge, hefrequented the library, which had been moved from a room insomeone’s home to the jail, which also housed the local bank,whose building had burned down some years before. He reck-oned that the location was OK, he said, because “I do not thinkanyone ever was put in the jail.” He began to show greatpromise in mathematics and science very early, and by the timehe was 12 was an avid reader. He attended high school inWelcome, Ontario, during the war, where he developed an en-during interest in engineering and machines and fed a thrivinginterest in music—one of his first serious projects was to builda phonograph turntable from scratch.

Bev went to college in 1945, just as the colleges were calledupon to absorb the enormous influx of young men returningfrom the war. At the time, he was interested in engineeringphysics, particularly as it applied to architecture, but decidedto pursue a more general program and was admitted to studyat the University of Toronto in mathematics, physics, and chem-istry. This was soon narrowed to mathematics and physics, andlater just physics, but he took a number of astronomy coursesas an undergraduate. He worked during the summers of 1949and 1950 at the Dominion Observatory in Ottawa, beginningwhat would be a long spectroscopic career, and decided topursue astronomy.

He took a Master’s degree at the University of Toronto in1950, under Ralph Williamson. His thesis showed convincinglythat the Sun was burning hydrogen on the weak-interactionmediated proton-proton cycle, which had just been postulated,instead of the then-assumed universal CNO (then just CN)cycle. One of his fondest memories was the apology later of-fered by Martin Schwarzschild, who poo-pooed the idea of thep-p chain as being at all important, when Bev told him about

it during a very influential visit by Martin to Toronto. Twoyears later, the result was universally recognized to be true andenormously important, and the gentle master apologized to Bev,who was by then a junior graduate student at Princeton,Schwarzschild’s institution. For those who knew Martin, onemay be sure that the student emerged from this experienceglowing with pride.

Bev went to Princeton to work on his Ph.D. in the fall of1950 and did his thesis with Schwarzschild on another epochalidea, that red giants were evolved stars with helium cores and

This content downloaded from 194.29.185.71 on Sun, 25 May 2014 21:19:37 PMAll use subject to JSTOR Terms and Conditions

222 GUNN

2005 PASP,117:221–226

hydrogen envelopes. He forged a very strong friendship at thattime with Allan Sandage, who was then a postdoc at Princeton,having recently graduated from Caltech. During those years, itwas the custom of Lyman Spitzer and Martin Schwarzschildto spend several months in the fall at Mount Wilson, workingon observational projects. In his third year, Bev joined Spitzeron the trip, working on one of Lyman’s projects on interstellarberyllium, and on one of his own, Of stars. On this trip, hemet and became acquainted with Jesse Greenstein and beganthe first of a long and fruitful series of collaborations with him.He also met Francis Moseley, an early manufacturer ofx-yrecording instruments, which was a somewhat fateful meetingboth for Bev and for this writer.

Bev finished his degree in 1953 and accepted a lectureshipat David Dunlap Observatory and the University of Toronto.In his early years there, his interest in machinery evolved intoa lifelong love affair with telescopes and instrumentation. Atthis time, he built one of the first photoelectric spectrum scan-ners, and thus began a consuming interest in quantitative as-tronomical spectrophotometry, his contributions to which areamong his most important achievements and enduring legacies.

In 1954 a love affair of an entirely different kind began,which also lasted the rest of his life. He met Nancy Sparlingin the fall of 1954, and they were married in August of 1955.In 1957 their first child, Chris, was born, and a year later cameKevin, their second.

That year (1958) would also bear the first fruit of Bev’sacquaintance with Jesse Greenstein. Jesse was trying hard tobuild the best astronomy department in the world at Caltech,which was already a major force in science but was just be-ginning its program in astronomy—and what a beginning, withthe largest and unquestionably the best telescope in the world,the 200 inch (5 m; later Hale) reflector. Jesse well knew howimportant technology and proper instrumentation were goingto be in the coming years and had been very impressed withBev’s early work both in theory and in instrumentation. Heasked Bev to join the small department, which at that timeconsisted of Greenstein, Guido Muench, Maarten Schmidt, andFritz Zwicky. The attraction of the active astronomical com-munity in Pasadena (Caltech and the Mount Wilson observatorystaff were one close but sometimes fractious family, and AllanSandage was at Mount Wilson) and the prospect of developinginstruments for and doing research with the great telescopewere irresistible. Nancy was also enthusiastic about moving toCalifornia. In the summer of 1958, they drove to Pasadena tobegin what was to become a remarkably productive and im-portant association with Caltech and Palomar that would lastfor more than 30 years.

Bev and Nancy settled in Altadena, a community just northof Pasadena. They had a daughter, Jennifer, in 1961, and theirlast child, Valerie, in 1966. Bev was very much a family manat an institution noted for workaholism, but he was by no meansless productive than his colleagues—just happier than most.His interactions with his peers and especially with students

were always pleasant and reassuring. He was one of the young-est members of the faculty, which probably made relationshipswith students easier for him than for his older colleagues, buthis personality played a crucial role as well. He had an amazingfacility for making students feel at ease. Being invited to theOke’s home, either for an evening of music and conversationand one of Nancy’s wonderful dinners, or for the occasionalparty, was a very special and pleasant occasion, and a reas-suring one for graduate students and postdocs in the thick ofcompetitive research. Bev showed that it was indeed possibleto do stunningly good science and still have a life.

I went to Caltech in 1961 as a graduate student and verysoon developed a special relationship with Bev, since I wasalso keenly interested in instrumentation. We were kindred spir-its, and Bev proved to be incredibly supportive. For instance,he had developed a machine to convert photographic trans-mission to incident intensity using one of Moseley’s poweredx-y tables and a German sensor that was servoed to follow acharacteristic curve drawn on graph paper in conducting ink.It was a remarkable device that worked very well, but wasagonizingly slow. At the time, I was taking a course in solid-state electronics from Alvin Tollestrup of the physics depart-ment and had an idea for how to make a settable nonlinearfunction generator to do this job more easily and much faster.Although Bev had no knowledge of my ability to build things,he immediately found some financial support for me to try it.Fortunately it worked well, and thus began a long collaborationthat later became very close when I returned for a facultyposition at Caltech in 1969. But this experience of mine wasby no means unique; a fair fraction of the students workedwith Bev, and their experiences were in all cases good. Bev’smethods were a bit unorthodox, and he was famous for beinga man of very, very few words. But there were usually enoughto make it known that an approach was silly and that a slightlydifferent path might just yield the answer. And so small, mid-course corrections were made, and the result was almost in-variably excellent.

Bev became involved in instrumentation almost immediatelyupon arrival at Caltech. Art Code had just left, after buildinga spectrum scanner much like the one Bev had built at Dunlap.Bev inherited this machine and modified it to incorporate thebest of both his and Code’s scanners. After an ill-fated attemptto use this instrument in the east arm of the Hale telescope—after much effort, it turned out to be no more sensitive thanwhen used on the 100 inch Mt. Wilson telescope, owing to thevery poor throughput to the east arm—he began work on ascanner for the prime focus of the Hale telescope that wouldtake full advantage of its aperture. But it soon occurred to Bevthat the ideal way to record spectra was to use multiple detectorsto look at many or all spectral regions at once. Photographicplates of course did this, but they were very inefficient andalmost impossible to calibrate with the desired accuracy. Tech-nology was advancing very rapidly during this period, andexcellent photomultipliers and pulse-counting electronics were

This content downloaded from 194.29.185.71 on Sun, 25 May 2014 21:19:37 PMAll use subject to JSTOR Terms and Conditions

JOHN BEVERLY OKE 223

2005 PASP,117:221–226

Fig. 2.—Bev, circa 1997, holding one of the early CCDs.

becoming available, driven at least in part by the enthusiasmof Bev and a few others and aided by the attraction that as-tronomy had for the technological public. Almost as importantas sensitivity to the astronomer is dynamic range. The standardstars about which enough is known accurately to serve as cal-ibrators are inevitably very bright compared to the incrediblyfaint objects one is studying. One therefore needed very fastand quiet pulse-counting electronics to deal with both regimes.

Bev brought these ideas and goals to a brilliant conclusionwith the design and construction of his multichannel scanner.It used 32 photomultipliers matched to the particular spectralregion each looked at, and so it was much more that 32 timesfaster than any single-channel instrument, and it covered thewhole spectral range, from the ultraviolet atmospheric cutoffto the last whimper of photocathodes at about 1mm. The pho-tomultipliers were aided by a simple and elegant invention inthe form of a half-aluminized hemisphere optically coupled toend-on photocathodes, so that almost all the photons reflectedby the metallic cathodes were returned for another chance atabsorption. This dramatically increased the quantum efficiencyof the tubes. The wavelength coverage of the instrument wasalmost two octaves, accomplished by splitting the output of thegrating into the first-order red and second-order blue by a di-chroic filter. The multichannel scanner was almost certainly thefirst astronomical instrument to use this now common tech-nique. Pulse counting, which allowed the detection of individ-ual photons, had taken the place of sensitive DC amplifiers andallowed much higher accuracy measurements to be made.

At the very faintest levels, the sky is overwhelminglybrighter than the object one is studying. Bev and Ed Dennison,a talented astronomer/electronics engineer whom Bev had per-suaded Caltech to hire to take advantage of technological ad-vances, developed the first chopping pulse counter used inastronomy, in which the object and a piece of blank sky nearbyare alternately observed by the instrument several times a sec-ond, and the pulses from the detector are steered to separate

counting channels synchronously. By differencing these signalsand swapping the sky and object apertures at intervals, onecould accurately measure the weak signal from the object, re-moving the much larger sky signal. This had already beenincorporated into the single-channel prime focus scanner andnow was built into the multichannel. At the bright end, fastelectronics were needed to accommodate the very high pulserates encountered when measuring bright standards. Bevworked with industry to develop the Solid State Research (SSR)pulse amplifier/discriminator, which had about a 10 ns pulseresolution and excellent pulse-height discrimination, and hasbecome a standard throughout astronomy.

This instrument was a workhorse at the 200 inch for manyyears, doing work on the recently discovered quasars (its cov-erage of the near-infrared allowed investigation of visible spec-tral lines in moderate-redshift quasars and established the con-tinuity of quasar properties even to the very highest redshifts),and faint and very high redshift galaxies and radio galaxies. Itprovided accurate data on faint white dwarfs, helped to estab-lish the equality of the redshifts of several quasars and faintgalaxies associated with them (the cosmological nature of thequasar redshifts was still disputed at the time), elucidated thenature and origin of supernova spectra, investigated the just-beginning study and unification of the AGN/quasar phenom-enon, and produced seminal work on cataclysmic variables. Inmany of these areas, Bev also led the scientific efforts, andindeed this was characteristic of Bev’s work all his life. Heviewed instruments as tools to do his research, not as ends inthemselves, and he was lucky enough to be working in an eraand environment in which instrumentation could be such anintensely personal and private matter. Bev would be remem-bered as a superb researcher even if he had not been an in-strument builder—but of course he could not havedone hisresearch without his instruments, nor could most of his col-leagues, who could not and did not build them, have donetheirs.

At about this time (the late 1960s), it became clear that theability to understand the spectral energy distributions of astro-nomical objects was beginning to be limited not by the accuracyof the measurements but by the accuracy of the calibrationmeasurement. There were no objects in the sky for which theemergent energy distributions were sufficiently well known.Bev’s work in this area, with his then-postdoc associate RudySchild, is among his most important contributions; it involvedfor him learning entirely new technological techniques (and adeep appreciation of the difficulty of the problem) and resultedin two calibration systems, the second of which, a slightlyrefined version of the first (1969), I was lucky enough to workwith him on. This was in 1979, and the system is only nowbeing substantially improved, largely through vast improve-ments in model atmospheres, not new observations. This workcontinued well into theHubble Space Telescope era, and Bevwas one of the chief contributors to the set of spectrophoto-metric standards forHST, and also to the “gold standard” set

This content downloaded from 194.29.185.71 on Sun, 25 May 2014 21:19:37 PMAll use subject to JSTOR Terms and Conditions

224 GUNN

2005 PASP,117:221–226

of faint subdwarfs that have been the fundamental standardsfor moderate and large telescopes for two decades.

In 1972 Bev was named associate director of the Hale Ob-servatories, the joint organization of Caltech and Carnegie thatran Palomar and Mount Wilson and was building the Las Cam-panas observatory in Chile. He held this position until 1978,when Maarten Schmidt became director (the first director fromthe Caltech side). Shortly thereafter, it was mutually decidedthat the financial interests of the two organizations were suf-ficiently different that a single management structure would notwork well any longer, and the close partnership was dissolved,although access to the respective institutions’s telescopes re-mained possible.

Meanwhile, Bev pursued yet more powerful instruments.Wonderful as the multichannel was, it was clear that some typesof work really needed much higher spectral resolution, andwhile this was possible with the multichannel by insertingmasks in front of the detectors, it became very time-consumingbecause of the loss of light and the many settings required tocover the spectrum. It was still 32 times faster than a singlechannel, but not fast enough. A number of detectors were beingdeveloped at the time that had the sensitivity of photocathodesbut that had many hundreds or thousands of resolution elementsin one or two dimensions, and it was clear that the future laywith this kind of device. Bev had already worked with imagetubes, in which the photoelectrons from a cathode were ac-celerated by high potentials and focused either electrostaticallyor magnetically onto a phosphor screen, which could then bephotographed, but the output medium in this case was stillphotographic, with all of the attendant problems of stability,calibration, and limited dynamic range. One of the first workinglow-light-level devices that had an electrical rather than lightoutput was the SEC vidicon. This used an image-tube–like frontend, but the electrons were focused not on a phosphor but ona fragile potassium chloride target, which provided moderatecharge multiplication and could be read out, vidicon-fashion,by an electron beam. The target stored the charge reasonablywell at room temperature, and long exposures were possible,allowing enough signal to build up that the beam-reading noisewas not catastrophic. The detector was being developed for thewide-field camera for theSpace Telescope by Westinghouseand the group at Princeton under Don Morton and LymanSpitzer, and Bev arranged for tests at the Hale telescope, whichwere quite encouraging, providing one of the first high-reso-lution studies of the Lya forest in quasars.

The last and best of this sort of detector was the SIT vidicon,which was much like the SEC but used a silicon target. Thecharge multiplication was very large, and achieved essentiallyphoton-limited sensitivity. The detectors needed to run cold,but this was not a problem; Bev had developed excellent vac-uum-jacketed Dewars for photomultipliers, and these werereadily adapted for SITs. Bev and I designed and built a low-resolution spectrograph around the SIT vidicon, which wasquite successful. This instrument brought a new and delightful

personality into our little group, in the form of Barbara Zim-merman. Barbara had been a programmer in the physics de-partment for some years, and minicomputers, which were apassion of hers, were just beginning to be used in astronomyfor data acquisition, storage, and instrument control. She cameto the astronomy department, learned FORTH, and was so goodthat we simply did not worry about data acquisition and re-duction any more.

But technology moved on. Photocathodes reached at mostabout 10% quantum efficiency, and photoconductors like sil-icon in the visible and near-infrared were near 100%. Jim West-phal had pioneered the use of silicon vidicons in astronomyand later was first to use SITs; Bev and I basically used histechniques. The silicon vidicon was like a SIT, except that thetarget was illuminated by the incoming light, not electrons.Thus, the device had very high quantum efficiency but terriblenoise properties, because of the beam noise. Clearly, oneneeded a quieter way to read the charge signal. The answer,of course, was CCDs, which began to appear in the mid-1970s.CCDs moved the electron signal essentially noiselessly to anamplifier, which could be very good or not so good, dependingon the expertise of the designer and fabricator, but the bestCCDs had noise equivalent to only 10 or so electrons. Theyessentially were and are the perfect detectors.

Bev’s next instrument was a dichroic-split double spectro-graph, which could accommodate a variety of detectors on itsblue side and had a permanent CCD detector built into a vac-uum Schmidt camera on its red side. The various vidicon cam-eras on spectrographs had sometimes been better, sometimesworse, for various reasons, than the multichannel scanner, butthe new double spectrograph was better at everything, and itremained the preeminent moderate-resolution spectrograph atPalomar for many years. The multichannel was thus retired.

One of the innovations from the double spectrograph thathas become part of observational astronomy’s lexicon was theability to use a multislit mask to take spectra of more than oneobject at a time. The technique essentially sped up observationsof a set of objects nearby in the sky (which is very commonin work on clusters of galaxies and stars) by a factor that wassimply the number of objects observed at once. Thus, for manyproblems, the combination of CCDs and multislits resulted ina telescope that was a thousand times more powerful than ithad been in the one-spectrum-at-a-time photographic era.

The multislit idea reached a kind of logical limit with thedevelopment of f-ratio preserving optical fibers; hundreds ofthese fibers could be placed in the focal plane to match thelocations of objects of interest and brought together to form aslit at the entrance of a spectrograph. Bev designed such aninstrument for the Cassegrain focus of the 200 inch telescopeand was successful in raising funds for its construction; thus,the very successful Norris spectrograph, with 180 fibers, cameinto being. Bev recruited a new group of young people to workwith him on this project, most notably Don Hamilton and JohnCromer, and enticed Mike Carr, a very capable instrument de-

This content downloaded from 194.29.185.71 on Sun, 25 May 2014 21:19:37 PMAll use subject to JSTOR Terms and Conditions

JOHN BEVERLY OKE 225

2005 PASP,117:221–226

signer/builder who had begun at Caltech in geophysics, to workwith him. While almost infinitely flexible, fiber spectrographsadd to the normal optical problems of spectrographs the daunt-ing mechanical problem of placing the fibers sufficiently ac-curately in the focal plane that the light from the target objectsin fact enters them with little loss. For the Norris, Bev solvedthis problem by combining the tiny lenses necessary at theentrance to the fibers (the f/16 Cassegrain focal ratio was tooslow to use in fibers directly) with tiny prisms and magnets tomake a kind of magnetic puck that, when placed accurately ona steel surface, would locate that fiber stably and accurately.A robot then had to be constructed to pick up, locate, and placethe 180 pucks, and do so accurately and rapidly enough thatseveral fields with different fiber configurations could be ob-served during a night without losing excessive time to recon-figuring the fibers. All of this worked brilliantly.

In the 1980s, however, it was already clear that Palomar wassoon to be eclipsed by newer, larger telescopes, and after arocky beginning, a new Caltech collaboration was undertakenwith the University of California system to build the 10 mKeck telescope on Mauna Kea, Hawaii. With 4 times the col-lecting area of the 200 inch Hale telescope, and with a sitelocation that had darker skies and average seeing at least twicea good as Palomar, the Keck promised speed gains of the orderof factors of 20–25 for very faint objects. Since CCDs alreadyhave quantum efficiencies near unity, and since fiber and largemultislit spectrographs were already using the multiplex ad-vantage to essentially its full extent, the only ways to go muchfainter were either to go to space to take advantage of thedarker skies and the absence of seeing degradation or to buildgiant telescopes on the ground. Both of these paths were taken,of course, with theHST and with Keck, Magellan, MMT-II,and the VLT.

Bev emerged from early negotiations on the Keck projectwith the responsibility for its workhorse optical instrument, theLow Resolution Imaging Spectrograph (LRIS), which as itsname implies is a combination imager and spectrograph. In-struments for telescopes as large as Keck are engineering chal-lenges that are very much more difficult than for smaller tele-scopes. Bev tackled the LRIS challenge with much the sameteam as he had used for Norris, with the notable addition ofJudy Cohen, who was responsible for the software. The in-strument works very well and has made crucial contributionsto the enormous success of the Keck telescope. Objects forwhich poor-quality spectra might be obtained with all-nightexposures at the Hale telescope can be obtained at a very highquality in less than an hour with Keck/LRIS, and the largeangular field allows as many as about 30–50 slits to be usedin a multislit mask. A new era for astronomy had arrived inwhich astronomers at long last had instruments powerfulenough to spectroscopically observe galaxies near the epochof their formation, using giant telescopes on the ground andwith images fromHST. Among the many accomplishmentsmade using this instrument included the first studies of very

high redshift ( ) galaxies, the first spectroscopic proofz p 3–4that the mysterious gamma-ray bursts were in fact at cosmo-logical distances, and the demonstration by means of spec-troscopy of very distant quasars that the universe became reion-ized at about redshift 6.

Bev was also heavily involved in the design of the telescopeand its enclosure and contributed centrally to the final domedesign (I should note that the excellent mechanical and thermalproperties of the dome are also critical ingredients in the successof Keck).

But with very large projects like Keck inevitably comeslarge, ponderous management and committee structures, inwhich individual scientists and instrument builders have verymuch less freedom than in previous times. Bev was from ageneration in which the idea for an instrument, the fund-raising,the design, and construction were centered in one individual.Bev built instruments to further his own scientific goals, as didmost of the builders of the time. His goals were largely con-cordant with or even further reaching than those of his col-leagues—he certainly understood the technological issues, ca-pabilities, and limitations far better than they did—and so theybenefitted directly from the instruments he designed and built.Bev was incredibly generous of his efforts and time, but didnot necessarily take kindly to ill-informed requests from hiscolleagues to incorporate this or that difficult (or impossible)frill into an instrument he was building to further their ownresearch interests. In addition, he and his counterparts at otherobservatories held all the cards; nobody else could do whatthey did. With the coming of Keck andHST, all of thischanged—the capabilities and broad outlines of instrumentswere determined by committees, and the builders workedwithin quite constrained envelopes of direction, schedule, andbudget. Astronomical instrumentation had become Big Science.

Like most of us, Bev adapted to the new regimen, but wasnot very happy with it. He also felt strongly that Caltech waslosing the focus and vigor that Jesse Greenstein, who had retiredfrom the chairmanship in 1972 and from the department in1979, brought when he built the department. The enormousburden of preparing for Keck weighed heavily on the smalldepartment, and Bev and Judy, who were responsible for LRIS,bore a very large part of that burden. So even before LRISwas delivered, Bev and Nancy were beginning to think aboutretirement; Bev not from astronomy but from the frenetic ac-tivity and responsibility that had been part of his life for sucha long time. They also longed to return to Canada, so in 1989they bought a house in Victoria and moved there in late 1991.Bev took a visiting appointment at DAO, and Nancy becameheavily involved in the local museum. Bev became the instru-mentation editor forPASP shortly after the move, and he re-mained very active—according to one of his much youngercollaborators on a largeHST cluster project, he was much moreactive than they were. He contributed greatly to Canadian as-tronomy, which had been a fond wish of his for a very longtime, and he taught instrumentation courses at Victoria. Canada

This content downloaded from 194.29.185.71 on Sun, 25 May 2014 21:19:37 PMAll use subject to JSTOR Terms and Conditions

226 GUNN

2005 PASP,117:221–226

joined the consortium led by Caltech that studied the feasibilityof a 30 m telescope, and Bev threw himself into thinking aboutand designing instruments for it, in particular a modular low-resolution, very high efficiency spectrograph consisting ofmany identical, relatively simple systems. He also remained aforce in the continuing attempt to establish an accurate spec-trophotometric calibration system.

Bev’s instrumental legacy is huge; much of the superb sci-ence that emerged from Palomar in its heyday was accom-plished with his instruments (and no small part of it by Bevhimself and his close collaborators and students), and Keck’sLRIS was for a very long time the mainstay of the spectacularwork on very faint galaxies, quasars, and stars. He saw in hislifetime the first real glimmers of appreciation, in the form ofappointments and awards, for young instrument makers in as-tronomy. He felt very keenly that these indispensable talentswere not properly appreciated or recognized, and he workedvery hard to change the intransigent sociology of the field. Thisseems finally, slowly, and painfully to be happening. It wouldnot have done so without leaders like Bev Oke, who were bothsuperb scientists and superb instrumentalists.

But Bev’s instruments are not all he left. Some of his sci-entific accomplishments have been mentioned in passing above,but heading a long list would be his early work on the proton-proton chain and the structure of red giants, his absolute spec-trophotometric calibration work, early decisive work on thevariability of AGN and radio galaxies, the first accurate effec-tive temperature determinations for stars using spectrophotom-etry and line profiles, work on a broad variety of white dwarfproblems with various collaborators, including finding the massof Sirius B and accurate temperatures and surface gravities,and showing that the mysterious BL Lac objects were linelessquasars associated with galaxies, studies of dynamics and prop-erties of the globular cluster systems in Virgo, surveys for andstudies of distant clusters of galaxies with the Hale, Keck, and

Hubble telescopes, discovery of the UV-dominant hot star pop-ulation in elliptical galaxies, and an explanation of the mys-terious spectra of Type I supernovae. Many of his collaboratorson these projects were students and postdocs who benefitedfrom his wisdom and leadership and working style and wenton to become powerful presences in their own right in theastronomical community. (I will not even attempt a list here,for fear of missing and slighting someone, but they are verymany, and without exception they learned about the art of livingas well as the art of astrophysics from Bev).

Astronomy and the world are very much the poorer for hispassing.

Addendum from the author: I would like to acknowledge anumber of contributions from Lori Lubin concerning Bev’spost-“retirement” career and his involvement with the ThirtyMeter project. Both Lori and Jill Knapp very kindly did carefulproofreading. I would also like to extend warmest thanks toNancy Oke, who made available to me a family memoir thatBev wrote a few years ago and that provided most of the detailsof his private life, and many of his scientific life, that appearin this piece. ——Jim Gunn

Addendum from the PASP Editors: PASP readers will recallthat Bev Oke served as the Instrumental Editor of the journalfor over a decade—all as a volunteer! He provided us withexcellent help, support, and advice, and was also a wonderfulfriend and colleague. Bev put great thought and care into allhis dealings with authors and referees. Although he steppeddown from his formal duties as an editor in the spring of 2003because of ill health, he continued to referee papers and consultwith us right up until the time of his death. We greatly misshis cheerful presence in Victoria.

—— Anne Cowley and David Hartwick

This content downloaded from 194.29.185.71 on Sun, 25 May 2014 21:19:37 PMAll use subject to JSTOR Terms and Conditions