Put the United States of 1790 next to the United States of 1860 and the difference is not really about numbers. It is about direction. In 1790, slavery is an Atlantic phenomenon: dense along the tidewater of Virginia and Maryland, darkest in the rice country of coastal South Carolina and Georgia, thinning out fast as you move inland. Seventy years later the deep colour has swung southwest, across Alabama and Mississippi and into Louisiana and east Texas, following a band of soil most Americans have never heard of.

That band has a name. The Black Belt of Alabama and Mississippi is a crescent of dark, heavy prairie soil weathered from Cretaceous chalk — laid down when the Gulf of Mexico reached much further north than it does now. It grows cotton exceptionally well. By 1860 the counties sitting on it had some of the highest concentrations of enslaved people anywhere in the hemisphere; in a handful of Mississippi Delta counties, more than nine residents in ten were enslaved. A geological formation from a hundred million years ago is legible on a nineteenth-century census map, and later on twentieth-century maps of voting behaviour, because of what was planted on top of it.

Two changes did most of the work. Eli Whitney's cotton gin, patented in 1794, made short-staple cotton — unlike the long-staple variety confined to the Sea Islands — commercially viable across the interior. And in 1808 the United States banned the international slave trade, closing the Atlantic supply just as demand in the new cotton lands was accelerating.

The result was not a smaller system but an internal one. Between roughly 1790 and 1860, on the order of a million enslaved people were moved from the Upper South to the Deep South, mostly by sale. Historians call it the Second Middle Passage. Virginia, Maryland and Kentucky became exporters of human beings to Alabama, Mississippi and Louisiana, and trading firms advertised in newspapers and kept pens in Alexandria, Richmond and Natchez. The shift on the map is that trade, rendered in colour.

By 1860 the census counted just under four million enslaved people in a national population of about 31 million. Cotton accounted for more than half the value of American exports, and the United States supplied most of the world's raw cotton, the input for Lancashire's mills. Estimates of the market value of enslaved people as property in 1860 commonly run near three billion dollars — more than all American railroads and factories combined. This was not a marginal or dying institution.

The map at the top of this letter made all of this visible to the people running the war. In September 1861 the US Coast Survey published a map of the southern states shaded county by county by the share of the population held in slavery, compiled from the 1860 returns. It sold for forty cents, proceeds to sick and wounded soldiers. It was among the earliest American attempts to map statistics rather than terrain, and it worked: darkness meant the plantation core, pale meant something else entirely.

Lincoln used it heavily. Francis Bicknell Carpenter, who spent six months in the White House in 1864 painting First Reading of the Emancipation Proclamation, put the map into the lower right corner of the canvas and recorded that the president consulted it often. What he was reading, in part, was the pale areas. Western Virginia is light — and split away to become West Virginia in 1863. Eastern Tennessee is light, and stayed stubbornly Unionist. The southern Appalachians and piney uplands are light, and produced desertion and open resistance to Confederate conscription.

That is the quiet argument the map makes. The Confederacy was not uniformly committed to the thing it was fighting for, and the variation was geographic and predictable. Later, as emancipation became Union policy, the same shading indicated where the largest populations were that the proclamation would reach, and where Black troops might be recruited in numbers.

A choropleth also hides a great deal. Percentages flatten variation: a county at 40% might hold a few vast plantations and many households with none. Urban slavery in New Orleans or Richmond — skilled, often hired out — looks nothing like the Delta, yet takes the same colour. And the figures come from a census that recorded enslaved people as unnamed tallies under an owner's name, which is much of why tracing African American ancestry before 1870 is so hard.

Still, the two panels together do something few maps manage. They show an institution moving — not expanding evenly outward, but sliding across a continent toward better soil, dragging a million people with it.

Sources: US Census Bureau, 1860 Census of Population; US Coast Survey, "Map showing the distribution of the slave population of the southern states of the United States, compiled from the census of 1860" (1861), Library of Congress; Susan Schulten, "The Cartography of Slavery and the Authority of Statistics," Civil War History (2010); Steven Deyle, Carry Me Back: The Domestic Slave Trade in American Life; Encyclopedia of Alabama, "Black Belt"; Francis B. Carpenter, Six Months at the White House with Abraham Lincoln (1866).

You're reading Lines & Legends, the Forummapping newsletter — the lines that move the world, and the legends behind them. Follow @forummapping on X for daily history maps, and forward this email to a friend who loves history.