It was James Cornell who--as information officer at the Smithsonian Astrophysical Observatory in Cambridge, Mass.--issued the press release that rocked the archaeology/astronomy worlds in 1963: a resident astronomer had fed some data into a computer, coming up with the idea that Stonehenge was an ancient astronomical observatory, the stones aligned so that particular lines of sight marked the summer and winter solstices. A hub-bub ensued--delightfully recaptured here--and, indeed, only in the last few years has the infant science of ""archaeoastronomy"" begun to enjoy respectability. So Cornell, a smooth and knowledgeable writer, now offers a roundup of what recent scholars have deduced about the observational talents of ancient cultures. Understandably, there may be a bit more about Stonehenge than the reader cares to know. And occasionally Cornell rattles on too long about right ascensions and foresights and backsights. For the most part, however, he presents solid and interesting information for the general reader. Among the sites discussed (often charmingly introduced by Cornell's eyewitness descriptions): TeotihuacÃ¡n, Chichen ItzÃ¡, Monte Alban, Tikal, and other noteworthy Mesoamerican ceremonial centers (where buildings or pyramids now seem laid out in ways that are obviously tied to lunar or solar cycles); Amerindian mounds, the Chaco Canyon petroglyphs; and African sites--with some amusing rebuttal of the dizzy theory that the Dogon people of Mali were once visited by extraterrestrials who told them about Sirius' dark companion. Plus--appraisals of the prowess of Egyptians, Babylonians, Mesopotamians, the Chinese, Cambodians, and even the Polynesian and Micronesian navigators. General conclusions? Well, people in temperate climates tended to be horizon-conscious, noting the rising and setting of celestial objects against natural or man-made markers; tropical people were more zenith-oriented, the Chinese pole-oriented; and the Babylonians looked to the Zodiac. True, Cornell is hardly the first to remind us that ""preliterate"" does not mean unsophisticated; and there are dull patches here. But this up-to-date survey makes a sturdy, occasionally fascinating, case for the impressive accuracy and cleverness of our time-and-space-measuring forebears.