Project Phoenix is definitely in one-site mode. Our second telescope at Jodrell Bank justsouth of Manchester, England sprung a busted wheel girder about a week ago,and is laid up for repairs. Badnews. But while having this secondinstrument is extraordinarily valuable, its not quite essential.

Jane gestures casually toward the glowing Phoenix workstations, carpeted with data. When wedesigned this stuff, a decade ago, everyone thought the universe would bequiet. To find extraterrestrial transmissions, all we had to do was point thetelescope at likely stars and search the spectrum for narrow-band signals. Just like in the movies.

In fact, the universe isquiet, especially at the microwave frequencies used for SETI. Unfortunately, Puerto Rico isnt. This isnta slam against the Isle of Enchantment.No place on Earth is free of man-made interference. No place.Even if youre in a deep valleycurtained by mountains, noisy satellites will sail overhead, rainingradio signals on your head and your antenna.

One way to beat this rap is to use two telescopes, farapart, which are simultaneously trained on the same patch of sky. Most earthlyinterference will show up at one site, but not both. And those signals that do can usually be sorted out on the basisof their frequencies. This is the scheme normally used by Project Phoenix: allpromising signals are checked out at both Arecibo and Jodrell Bank.

But when youre working with only one antenna, like we arenow, youve got to be inventive. Itsan old problem. In 1992, when theformer NASA SETI program began observing at Arecibo, our receivers were cloggedwith so many signals it was obvious that a reliable one-scope method was neededfor separating ET from AT&T.Fortunately, theres a way to do this.Terrestrial interference is generally so strong that it doesnt reallymatter where you point your antenna. You can pick up interfering signals in alldirections. An extraterrestrial signal,on the other hand, is going to be weak. Finding one requires aiming your scopeat the star system from which its coming.

So separating the wheat from the chaff is simple: Point atthe star of interest, and note all the signals. Then train your scope on another patch of sky a few degrees away a so-called comparison field and tally up the signals there. Any signal thats in the first group and notthe second is a candidate transmission from deep space, ace, ace.

In 1996, Jane Jordan turned this idea into a hunk ofhappening software. She made itpossible to have a two-star observing mode for Project Phoenix. For fiveminutes, the telescope is trained on Star 1, collecting data. Then thetelescope is slewed a few degrees to Star 2, and data are collected again. Eachstar serves as the comparison field for the other: vive la difference. After each cycle, the receivers are tuned upthe dial a bit, and the process repeats.

Its not as efficient as two-site observing because of theconstant nodding back and forth, and its not as speedy at recognizing sometypes of interference. Its also notnearly as automated, which gives both Jane and observers (such as your faithfulcorrespondent) cranial pain. But you gowith what youve got, and so far this observing run, its been one-siteobserving all the way.

In another three years, the new Allen Telescope Array willbe gearing up for SETI observations. Itwill initially be capable of checking out three stars simultaneously, so eachcan serve as an instantaneous comparison field for the others. Jane Jordan saysthis will make her head feel a lot better.