By his own admission, Jeremy Bernstein set himself a formidable task in attempting to characterize the preeminent and multifaceted Bell Laboratories, the research and development arm of what used to be our monopoly telephone system. In constructing a "kind of mosaic of people that, when assembled, would make a coherent sample of the whole," Bernstein, a physicist and science writer for The New Yorker, has succeeded in illuminating both the science and the human organization behind this unique laboratory.

The "Three Degrees Above Zero" of the title reflects only the content of the fourth and last section of the book, an account of Robert Wilson's and Arno Penzias' serendipitous discovery of the remnant birth cry of the Big Bang explosion -- microwave background radiation with a uniform temperature of 3 degrees Kelvin. In another section, "Bits," Bernstein explores computer science and mathematical research at the Labs. "The Solid State" recounts the development of the transistor in the 1940s and the progression of its offspring, integrated circuitry on silicon microchips. "Telephony" at Bell is the art and science of the telephone, and in the chapter by that name we pick up inklings of the communications revolution yet to come.

The content of the science being done at Bell Laboratories is in a way the least important subject of this book. Bernstein emphasizes the lives and personalities of the researchers and tries to explain how the interdisciplinary and relatively unstructured environment of the Labs has made their work bloom. Indeed, there is a direct relationship between these researchers and the progress of 20th-century science as it has exploded from decade to decade.

Even the layman not particularly interested in the science and technology of Bell's research will enjoy some of the personal sketches Bernstein provides: the world-class mathematician who put himself through graduate school by performing in a circus with a trampoline troupe (and he doesn't even look like a mathematician!); the electrical engineer who escaped the abortive Hungarian revolution of 1956 and went on to do pioneering work on human stereoscopic vision after tackling a mundane engineering problem. Then there's the advanced chess playing computer developed at the Labs, "Belle," which once duplicated one of the most surprising and remarkable moves in chess lore (the sacrifice of a queen on the 17th move by Bobby Fischer) and went on to win the game.

Nobel laureates abound at Bell Labs -- seven in the laboratory's history -- and Bernstein is quite taken with them, perhaps too much so. In any organization of more than 25,000 people (18,000 after divestiture) with many millions of dollars to invest in the best people, it is not surprising to find so many Nobelists. Of course we all admire them, but the attention given to their honors is a minor but curious flaw of the book.

But beneath this otherwise delightful romp through the halls of a unique research organization is a serious question. What potential effect will AT&T's divestiture have on the quality and direction of future Bell Labs' work? Generally optimistic views expressed by those whom Bernstein interviewed are given with the caveat that they could be missing the mark. Up to now about 10 percent of Bell research has been directed not to immediate applications but the exploration of nooks and crannies far afield. Often this kind of research has led to practical applications a decade or more later. How will a more competitive corporation regard this kind of exploratory research?

Bernstein draws attention to the impressive statistic that in the last year before divestiture, Bell Laboratory's basic research budget (8.3 percent of its total research) was between 10 and 15 percent of the entire budget of the National Science Foundation. He draws no conclusion as to what direction the Labs will now take, but states firmly that "if the divestiture changes Bell Lab's fundamental character -- the special mixture of basic and applied science, of long-term and short-term research, of science and engineering -- then the United States will have lost one of its greatest technological assets."

Readers who themselves may hold scientific or engineering positions and are familiar with the foibles of typical technical organizations will be impressed with some of the nuggets Bernstein has dug up. From one engineer: "We have a slogan here at Bell Laboratories, 'Either you do something very useful or you do something very beautiful.' " The best description of the Labs is quoted by Bela Julesz, the Hungarian engineer: "It is like a big baroque organ. If you are interested in playing one-finger accompaniments on a baroque organ, then you shouldn't be at Bell. But if you have something to do where you have to pull out every register, then this is the place to do it." Don't all stampede now to the personnel office.