Life Stories

The Triumph of Geometry over Geography

Inscribed upon that grand design, the surface of the earth, are the marks of human occupancy. Patterns resulting from humankind’s activities, although individually not of the great scale of natural features, in aggregate give to certain areas their most distinctive character. And of all these human designs, one of the most widespread, if not the most important, is the subdivision of land for purposes of settlement and taxation — the cadastral survey. Once laid down the boundaries of land division become part of an inheritance to be accepted or modified by later generations.


Air travelers see the effects of cadastral surveying instantly when they look out the window. Two seemingly similar pieces of land appear quite different because of the way they have been laid out. The subdivision on one piece of land looks highly patterned and systematic, rectangular in shape and oriented in cardinal directions, with roads and boundaries spaced at specific intervals. By contrast, subdivision on another piece of land, even for the same uses and even when lines have been drawn straight, appears haphazard, its boundaries and markers apparently fixed arbitrarily, creating a patchwork of properties.


Most of the world’s land areas have been subdivided unsystematically, and this method — or lack of method — partly explains the complicated pattern of property and civil boundaries in many older regions of the earth. Such surveys rely on natural features, including streams or landforms as well as other individual and often ephemeral objects, such as rocks and trees. Early American surveying followed this course, with the majority of the land in the original thirteen English colonies surveyed in a rather unsystematic manner, known variously as “metes and bounds” or “indiscriminate location.” Such cadastral surveying methods were also used in eastern Texas as well as in smaller areas originally surveyed under French or Spanish control, often with some modifications.


That all changed in America with the implementation of the famous Land Ordinance of 1785, based on a report from a committee chaired by Thomas Jefferson and passed by Congress on May 20 of that year. The landmark act created the United States Public Land Survey (USPLS), which contained detailed instructions for systematic land subdivisions in the western regions of the United States and established a benchmark for the surveying of land in the public domain. A number of historians have lauded it as one of the most important acts in the young nation’s life; it resulted, over time and with subsequent modifications, in the most extensive area of the world over which uniform systems of land survey extend. 


A systematic method of the land subdivision is not unique to the United States. Aerial photographs and excavations have revealed the existence of gridiron patterns in early town plans in the Indus Valley. The Romans used a rectilinear cadastral survey in Europe and North Africa, known as a centurion, employing a groma, a surveying instrument with four arms set at right angles to one another, with which they established the corners of the rectangular plots. Systematic methods of the land subdivision were also employed on the polders of Holland during the seventeenth century and later. However, in no other country or territory does a uniform cadastral survey, with an overall plan, cover such a large area as in the United States. 


The essential features of the USPLS were worked out in the Old Northwest. Its foremost features included a survey prior to settlement, the orientation of survey lines, the township unit, and the section.


In the rectangular survey of the United States, two fundamental lines govern an original survey tract. The first, running north-south, is called the principal meridian. The second, running east-west, is called the principal base or parallel. These lines meet or cross at right angles. Other surveyed lines, called range lines, run parallel to these principal lines, spaced at six-mile intervals. (Jefferson preferred a metric division of hundreds, but eventually the measurement in English miles prevailed.) The resulting six-mile squares form townships, each with a unique numerical designation. Each township is further subdivided into thirty-six sections, each a square mile in size (640 acres) and numbered 1 to 36. (Revenues from the sale of one section in each township were earmarked for education.) By this system, a parcel of land can be located by reference to its section, township, and range numbers within the original survey tract.


A land beyond established settlements in the United States was to be subdivided according to these principles. However, before a program of rectangular surveying could begin, the states had to relinquish some of their land claims to the federal government under the policy of the public domain. Six of the former thirteen colonies had such claims. Moreover, the USPLS underwent further modification in 1796, including the numbering of sections, before it was adopted for all subsequent cadastral surveying. 


As a result of negotiations with the states and these later changes to the rectangular surveying method, a number of earlier unsystematic surveys remained intact and different versions of systematic surveying existed in adjoining areas. Several of the six states retained some of their western lands, including the Virginia Military District (VMD) in central-southwest Ohio, which had been awarded to Revolutionary War veterans and which had been surveyed unsystematically by the state. Thus, Ohio had not only areas of different systematic cadastral surveys, including a large region known as Northwestern Ohio (NWO) — the last major part of Ohio to be settled and subdivided in the manner later to be used for the rest of the country — but also a large district that was subdivided by metes and bounds. 


These forms and effects are clearly visible by comparing two hundred-square- mile areas that are remarkably alike in their physical geography. The first is a tract in NWO surveyed beginning in 1819 prior to sale and settlement, as required by the provisions of the USPLS. By contrast, the original survey of an area of the VMD was completed over a period of several years beginning in circa 1810, with no restrictions as to the shape of plots and with later fundamental survey units being fitted in between previously surveyed tracts and preferred riparian areas. In both areas, subsequent subdivisions (e.g., properties) are described in terms of parcels of lands as originally surveyed. In the case of the USPLS, these terms are townships and sections; in the case of the VMD, originally numbered surveyed units.


These units exhibit a remarkable permanence, as illustrated by cadastral maps of a small area of the VMD separated in time by almost a hundred years. The property lines show change, but the fundamental survey units by which they are described endure. Federal surveyors were required to interrupt the USPLS lines when they encountered prior cadastral surveys, as happened in areas surveyed earlier by the French and Spanish. Native Americans had tribal, not individual, land ownership, which made it easier for Euro-Americans to gain control of the land as cadastral surveys progressed across the country.


A great many cultural and geographical features are influenced directly or indirectly by the cadastral survey. These include the boundaries of states, counties, civil townships, school districts, and properties as well as the sites of roads, bridges, residences, and other structures. A glance at political divisions in the United States reveals more state boundaries fixed by cardinal orientation (north-south and east-west) in the West than in the Atlantic region. This orientation is even more pronounced at lower administrative and political levels.


Detailed studies in Ohio and other areas show that the lower in the hierarchy of places, the more important the type of survey becomes to those occupying the land. For example, there is much more litigation in the VMD tract than in NWO. In NWO property boundaries are typically fractions of sections: 1/4 (160 acres), 1/8 (80 acres), and 1/16 (40 acres), each usually contained within a single section. In the VMD, by contrast, many of the properties are in more than one fundamental survey unit. Properties near county or township lines are often divided between two (or more) of these administrative units, creating problems in taxation and school districting, etc. Properties in NWO, where a property is normally contained within a single section, do not usually suffer such problems. Furthermore, in such areas lost boundaries can usually be reconstructed more easily.


Roads are another feature to be considered in relation to the cadastral survey. In systematically surveyed rural areas, such as NWO, nearly all of the local roads have been placed along cardinally oriented section lines. In the VMD roads usually divide, rather than border, fundamental survey units and even individual properties. The latter method has both advantages and disadvantages. Perhaps the greatest difference between the two areas is where and how roads cross streams. Because public roads in the VMD have been sited to minimize major stream crossings, most of these crossings are usually minor highway culverts. In the NWO area, governed by a gridiron pattern of highways, most of these crossings are larger and more expensive general highway bridges. Other factors, such as traffic usage and accident rates, are also affected.


A number of other aspects of the relationship between the type of cadastral survey and land use are important, including erosion. The practice of plowing in a manner suggested by the orientation of the USPLS lines rather than following the contours of the land is said to have contributed to dustbowl conditions in the 1930s. More subtle is the sense of disorientation experienced by some people brought up in areas of systematic surveys when they travel in places characterized by unsystematic (“crazy-quilt”) survey patterns. On the other hand, perhaps most people find the aesthetics of an unsystematically surveyed landscape more appealing than areas surveyed under the general constraints of the USPLS.


These and many other factors can be appreciated by studying county maps and atlases. More than four thousand commercially published atlases, as well as generally later and less elaborate plat books of U.S. counties, were produced over a period of a hundred and fifty years. Of the approximately three thousand counties in the United States, more than half have no such specific atlas, while others have been covered several times by privately produced atlas maps. A typical U.S. county atlas will have a map of the state and a map of the county as well as individual maps of the county’s civil townships, usually one to a page. Such maps show roads, schools, and farms, property boundaries and acreage, and the names of the owner or owners.


Publishers typically gleaned the data for these atlases from manuscript maps kept in the county courthouse. Because these maps are revised constantly, they are self-liquidating records, whereas the published county maps and atlases show a snapshot in time. For those willing to pay for the privilege, many county atlases also contain lithographic pictures of families and their properties.


In his great American novel, Raintree County, in which the county atlas is a central motif, the author Ross Lockridge, Jr. describes the contents of such an atlas in graphic detail. Lockridge writes that the Illustrated Historical Atlas of Raintree County“was a very substantial piece of work” and “remarkable for its illustrations.” Most telling of all, however, was how such atlases were an important part of the American ethos: “Fifteen years ago Mr. Shawnessy had bought a copy of the Atlas for ten dollars and put it on the parlor table with the Family Bible and the Photograph Album.”


Norman J.W. Thrower is Professor Emeritus in the Department of Geography at the University of California, Los Angeles, and a frequent contributor to Mercator’s World. He has been interested in cadastral surveys for more than forty-five years, or most of his professional life.


Further Reading


It has been my privilege to know a number of scholars who have worked and published on different aspects of cadastral surveys and associated graphics. Space does not permit a full and proper citation of their works, but I mention them here by name and encourage readers to consult their writings: Michael Conzen, Clara Le Gear, Ronald Grim, Ralph Ehrenberg, Hildegard Johnson, Roger Kain, George Kish, Francis Marschner, William Pattison, Edward Price, John Rooney, Richard Stephenson, James Walker, and Wibur Zelinsky.


Listed below are several of my own works on the subject:


“Cadastral Surveys and County Atlases of the United States.” The Cartographic Journal 9, No. 1 (1972): 43-51.


“Cadastral Survey and Roads in Ohio.” Annals of the Association of American Geographers 47 (1957): 181-2.


Maps and Civilization: Cartography in Culture and Society. 2nd edition. Chicago: University of Chicago Press, 1999.


Original Survey and Land Division: A Comparative Study of the Form and Effect of Contrasting Cadastral Surveys. Association of American Geographers, Monograph No. 4. Chicago: Rand McNally & Co., 1966.


“The County Atlas of the United States.” Surveying and Mapping 21 (1961): 365-73.


Illustrations courtesy of Norman J.W. Thrower