EVEN ATLAS, the earth-shouldering Titan after whom Gerardus Mercator christened his first collection of maps 400 years ago, would feel the burden of these two weighty books. In fact, Mercator would scarcely recognize them as examples of the cartographic science he pioneered. Yet they exemplify some of the present achievements and uses of mapmaking--the compression of political, cultural, and geographic facts into their spatial relations--and one of them makes a signal contribution to the history of American political life.

The Historical Atlas of United States Congressional Districts is the culmination of a project begun in 1936 by the Historical Records Survey of the WPA. Those who conceived it as model relief work for scholars helped lay the basis for modern statistical scholarship into congressional voting patterns--research that has revealed, among other things, the cohesiveness of political parties on legislative issues, the geography of shared interests, historic traditions of voting, and the social and cultural bases of American politics. The author and editor of this volume, Kenneth Martis, has written an illuminating, if somewhat technical, review of the history and application of congressional roll-call analysis, of which he has included a striking sample map of a roll- call vote on water reclamation projects in 1958. He has also provided succinct histories of congressional districting laws, methods of election and apportionment, and district boundary setting.

The atlas is thus a treasure trove of previously inert information, most of it never gathered together before, now brought to life. It is this that makes it a major work of historical reference. Never previously, for example, has there been compiled from legal documents the boundary descriptions of all House districts ("Northwest on Munn Rd. to Rocky River Dr., north and west on Rocky River Drive extended along the southern boundary of 1970 Federal Decennial Census Tract 1609 to the Rocky River"). The list of boundaries for every congressional district, including urban districts, since 1789 alone makes the atlas indispensable for reference and research. It also identifies, again for the first time, the district affiliations of all representatives since the first Congress and lists each representative by name. Scholars, congressional representatives, and others will now be able easily to identify the exact areas served by each congressman since 1789, trace the history of representation in each district and the congressional history of every spot in the nation, and map the geographic distribution of many congressional and political phenomena.

Yet while a boon to historical scholarship, this atlas is also much more. For in the United States, power lies upon the land. Constitutional provisions--those that preserve the states and apportion representation among them--have created a unique geography of politics. American communities are constituents, not only of counties, states, and regions, but of arbitrary and alterable jurisdictions--congressional districts--fashioned by state legislatures after each census. These districts have created the most inconsistent and capricious political relationships of American politics, perhaps of any political system in the world. In fact, congressional districts have no history, and to my knowledge none has ever been written, because districts are temporary artifacts, laid out in space, of demographic tides and partisan struggles. The Gerrymander of 1811 is simply the most notorious symbol of this adventitious hooping together of lesser civil divisions to meet constitutional requirements of House representation. Gerrymandering goes on, as it will continue to go on, despite the Supreme Court's 1964 decision in Wesberry v. Sanders and subsequent rulings, which mandate that each district within a single state contain roughly the same number of people. Only a system of representation uncoupled from population would rid us of gerrymandered districts, and only then might congressional districts endure and have their own histories.

Stable districts, however, would pose a threat to the system as we know it. The republican form of government was meant by the Founding Fathers to assure the representation of the people in their local situations; and the method of apportionment by population was meant to be a check upon the size of the House. Had congressional districts possessed enduring boundaries from the outset, it is conceivable that they would have gained their own identities, loyalties, and histories, analogous to those of the states, in which case they might have come inadvertently to be permanent geopolitical fetters upon the national government, rather than accessories of it. As it happens, their very insubstantiality, so clearly exposed by this atlas, led to the same result. For congressional districts--whose boundaries have been drawn to protect political fiefdoms or to assure the representation of racial, ethnic, and occupational groups--have allowed these separate interests to offset each other and prevent the rise of a permanent majority, acting through the state, that would ignore the rights of those in the minority. Thus gerrymandering has been one of the bulwarks of our liberties.

That this collection of maps provokes such reflections is a tribute to its clarity and comprehensiveness. Its only serious limitation is its omission of political party identification for each representative, the result being that these maps cannot be read for politics. Though promised for a subsequent volume, from which a demonstration map of the political party distribution of the 46th Congress (1879) is included, party affiliations should have been mapped here too for greater economy and utility. A third volume, mapping the major roll-call votes in Congress since 1789, is planned to complete this distinguished project.

If the congressional atlas uses standard land maps to plot American political facts, This Remarkable Continent displays better the virtuosity and complexity of modern mapmaking. It is a kind of cartographic sampler of American and Canadian society and culture, neither comprehensive nor coherent nor organized with a convincing rationale. Introductory notes are spare, captions inconsistent and often confusing. Yet the atlas will be catnip to those drawn to the exotica of mapping.

Part of the appeal of its 390 maps and figures, some prepared especially for this volume, most drawn from other sources, lies in the sheer variety of map design anthologized by the editors. Readers are likely to enjoy the maps for what they reveal of the promiscuous concerns to today's cultural geographers. In addition to maps of such standard cultural features as language stocks, land ownership, word incidence, religious patterns, and voting trends, the curious will also find spatial figures of places in which sliced tomatoes are eaten for breakfast; the distribution of gay bars in the United States in 1978; regional variations in the names of big-bun sandwiches; Anglo-French fiddling styles; and the comparative interregional migration of Pennsylvania and Texas football players.

What all this adds up to is hard to say. The atlas mirrors the intellectual disarray of contemporary popular culture studies in North America, a field characterized by snippet scholarship rather than by governing themes and large views. The maps reveal what can be done by thematic cartography, not what to make of them or the value of presenting information in this way. Nor, in contrast to the author of the congressional district atlas, do the editors offer much assistance. What, for instance, is the significance of the fact, mappd here, that those of lower social status in Decatur, Illinois patronize nearby taverns while those of higher standing travel to other and more distant drinking spots?

Unlike a great anthology of poetry, we cannot be sure that we have here the best, the classic, the most representative, or the most enduring examples of cultural maps. We are not asked to reflect on the overall significance of the data mapped or of the designs, some of them spectacular, by which they are presented. What might thus, through the application of more coherent standards of selection and explanation, have been a genuine contribution to historical understanding is instead a pleasant, and expensive, diversion, rather patronizing of the general reading public to which the editors address it.