l. iii. p. 88;) and these numbers are ascertained by the agreement of the texts of the different Notitioe. Nardini, l. viii. p. 498, 500.] [Footnote 72: See that accurate writer M. de Messance, Recherches sur la Population, p. 175-187. From probable, or certain grounds, he assigns to Paris 23,565 houses, 71,114 families, and 576,630 inhabitants.] [Footnote 73: This computation is not very different from that which M. Brotier, the last editor of Tacitus, (tom. ii. p. 380,) has assumed from similar principles; though he seems to aim at a degree of precision which it is neither possible nor important to obtain.] [Footnote 7311: M. Dureau de la Malle (Economic Politique des Romaines, t. i. p. 369) quotes a passage from the xvth chapter of Gibbon, in which he estimates the population of Rome at not less than a million, and adds (omitting any reference to this passage,) that he (Gibbon) could not have seriously studied the question. M. Dureau de la Malle proceeds to argue that Rome, as contained within the walls of Servius Tullius, occupying an area only one fifth of that of Paris, could not have contained 300,000 inhabitants; within those of Aurelian not more than 560,000, inclusive of soldiers and strangers. The suburbs, he endeavors to show, both up to the time of Aurelian, and after his reign, were neither so extensive, nor so populous, as generally supposed. M. Dureau de la Malle has but imperfectly quoted the important passage of Dionysius, that which proves that when he wrote (in the time of Augustus) the walls of Servius no longer marked the boundary of the city. In many places they were so built upon, that it was impossible to trace them. There was no certain limit, where the city ended and ceased to be the city; it stretched out to so boundless an extent into the country. Ant. Rom. iv. 13. None of M. de la Malle's arguments appear to