Starting with the modern data set, the historical boundary changes were cartographically interpreted from "Map Guide to the U.S. Censuses, 1790-1920" (Thorndale and Dollrahide, 1983). Boundaries were deleted and area points corrected. Some of the boundary changes that occurred after 1920 were not fully explained in the reference book. This data was collected from state historical societies, state land boards, county assessors and historical USGS published maps. The historical boundary positions are not based on legal land descriptions and may not be accurate on a large scale.
The county population data was retrieved from the US Census web site as plain ASCII text for the six states in the region. A Unix shell script program was used to compile tables of the entire region for all the decennial data sets. The population tables were joined to the Arc/Info polygon attribute table using the state and county FIPS code as an Info relate item.
The square miles for each county was calculated based on the area field in the polygon attribute table. The population density was calculated by dividing the total population by the square miles for each county. The square miles and population density data may not be accurate due to cartographic generalization of the boundaries and map scale.
The classification scheme and visualization graphics was compiled using ArcView GIS software. The classifications used are exponential and slightly modified from natural breaks. Postscript files from ArcView layouts were converted to GIF images for the web page.