611] ( return) [ According to Robertson, there were twice as many slaves as free citizens.—G. Mr. Blair (p. 15) estimates three slaves to one freeman, between the conquest of Greece, B.C. 146, and the reign of Alexander Severus, A. D. 222, 235. The proportion was probably larger in Italy than in the provinces.—M. On the other hand, Zumpt, in his Dissertation quoted below, (p. 86,) asserts it to be a gross error in Gibbon to reckon the number of slaves equal to that of the free population. The luxury and magnificence of the great, (he observes,) at the commencement of the empire, must not be taken as the groundwork of calculations for the whole Roman world. “The agricultural laborer, and the artisan, in Spain, Gaul, Britain, Syria, and Egypt, maintained himself, as in the present day, by his own labor and that of his household, without possessing a single slave.” The latter part of my note was intended to suggest this consideration. Yet so completely was slavery rooted in the social system, both in the east and the west, that in the great diffusion of wealth at this time, every one, I doubt not, who could afford a domestic slave, kept one; and generally, the number of slaves was in proportion to the wealth. I do not believe that the cultivation of the soil by slaves was confined to Italy; the holders of large estates in the provinces would probably, either from choice or necessity, adopt the same mode of cultivation. The latifundia, says Pliny, had ruined Italy, and had begun to ruin the provinces. Slaves were no doubt employed in agricultural labor to a great extent in Sicily, and were the estates of those six enormous landholders who were said to have possessed the whole province of Africa, cultivated altogether by free coloni? Whatever may have been the case in the rural districts, in the towns and cities the household duties were almost entirely discharged by slaves, and vast numbers belonged to the public establishments. I do not, however, differ so far from Zumpt, and from M. Dureau de la Malle, as to adopt the higher and bolder estimate of Robertson and Mr. Blair, rather than the more cautious suggestions of Gibbon. I would reduce rather than increase the proportion of the slave population. The very ingenious and elaborate calculations of the French writer, by which he deduces the amount of the population from the produce and consumption of corn in Italy, appear to me neither precise nor satisfactory bases for such complicated political arithmetic. I am least satisfied with his views as to the population of the city of Rome; but this point will be more fitly reserved for a note on the thirty-first chapter of Gibbon. The work, however, of M.