In the millet zone where there was much work to be done the slaveholdings were in many cases very large and the control relatively stringent; but in the banana districts an easy-going schedule prevailed for all.

The general scale of slaveholdings in the tobacco districts cannot be determined prior to the close of the American Revolution; but the statistics then available may be taken as fairly representative for the eighteenth century at large.

Ann Arundel, Charles, and Prince George, the ratios among the slaveholdings of the several scales, according to the United States census of 1790, were almost identical with those just noted in the selected Virginia counties, but the non-slaveholders were nearly twice as numerous in proportion.

Prosperity from tillage, and especially from dairying and horse-breeding, caused the rise in that neighborhood of landholdings and slaveholdings on a scale more commensurate with those in Virginia than with those elsewhere in New England.

These reports from some two-score scattered localities enumerate 2456 slaves, about one-third of the total negro population of the specified age; and they yield unusually definite data as to the scale of slaveholdings.

These two featuresthe multiplicity of slaveholdings and the virtually uniform pettiness of their scaleconstituted a régime never paralleled in equal volume elsewhere.

The scale of cotton-belt slaveholdings rose not because free labor was unsuited to the industry but because slaveholders from the outside moved in to share the opportunity and because every prospering non-slaveholder and small slaveholder was eager to enlarge his personal scale of operations.

The Ordinance of 1787, however, though it was not strictly enforced, made slaveholdings north of the Ohio negligible from any but an antiquarian point of view.

The scale of slaveholdings in middle and western Tennessee, while superior to that in Kentucky, was never so great as those which prevailed in Virginia and the lower South.

The extreme effects of this are illustrated in Issa-quena County, Mississippi, and Concordia Parish, Louisiana, where in 1860 the slaveholdings averaged thirty and fifty slaves each, and where except for plantation overseers and their families there were virtually no non-slaveholders present.

In the six counties of Dallas, Greene, Lowndes, Macon, Perry, Sumter and Wilcox, for example, the average slaveholdings ranged from seventeen to twenty-one each, and the slaveholding families were from twice to six times as numerous as the non-slaveholding ones.

Again, the scale of slaveholdings was in some degree a measure of social rank, and men were accordingly tempted by uneconomic motives to increase their trains of retainers.

Slaveholdings by colored townsmen were likewise fairly frequent.

Evidence of slaveholdings by colored freemen occurs also in the bills of sale filed in various public archives.