THE REGIONAL GEOGRAPHY OF SOUTH CAROLINA 

 ILLUSTRATED BY CENSUS STATISTICS 



By Eolaxd M. Harper 



Plate 29 



In the Journal of School Geography (afterwards called the Journal 

 of Geography) for January and March, 1898, there is a 14-page de- 

 scription of South Carolina by Prof. L. C. Glenn — a former resident 

 of that state, and a North Carolinian by birth — which is a good ex- 

 ample of what can be done by a careful observer with the proper 

 geographical point of view, without reference to census reports or 

 other previous literature. 



The present paper supplements Prof. Glenn's bj' dividing the 

 same state into natural regions and illustrating some of the contrasts 

 between them by means of statistics, mostly taken from the 13th 

 United States Census (1910). The main object of this study is to give 

 readers an idea of the vast amount of geographical information that is 

 buried in census reports and going to waste, as it were, for lack of 

 geographers sufficientl}' interested in that kind of work to dig it out 

 and put it together by regions. Most of the same kinds of ratios here 

 worked out are indeed given in recent census reports for whole states, 

 but as the average state is divisible into at least half a dozen regions, 

 differing in all sorts of ways, the contrasts between them are pretty 

 effectually concealed by the common practice of using state averages. 



Description is here reduced to a minimum, to avoid unnecessary 

 duplication of what has already been written by Professor Glenn and 

 others. And in order to economize space only a few of the many pos- 

 sible kinds of statistics are used, but those few are believed to be 

 among the most significant. By utilizing more of the census tables, 

 and also by going back to earlier censuses, the length of this article 

 could have been increased man}' times witliout sacrificing much of its 

 geographical interest. Statistics could also have been obtained from 

 other sources than census reports; for example, the relative areas of 

 diflf'erent soil texture classes from government soil surveys, and the 

 relative abundance of different trees or other |)lants from the writer's 

 field notes. But these lines have not yet hecn worked up with a degree 

 of tiioroughness at all comparable with a census, and they umy there- 

 fore be left out of consi(l<T;iti<)ii at present. 



[ 1"5 ] 



