The clearest and most detailed account of the origin of the Servian tribes is given by Dionysius iv. 14. 1 f.: “When Tullius had surrounded the seven hills with one wall, he divided the city into four parts, and giving to the parts the names of the hills—to one Palatina, to another Suburana, to the third Collina, and to the fourth Esquilina—he made the city to consist of four tribes, whereas up to that time it had comprised but three.... And he ordained that the men who lived in each of the four parts should not change their abode or give in their census elsewhere. The enlistment of soldiers also and the collection of taxes, which they were to pay individually to the treasury for military and other purposes, were distributed no longer among the three gentile tribes but among the four local tribes instituted by him.... [15. 1:] And the whole country he divided, as Fabius says,[274] into twenty-six parts, also called tribes, adding to them the four city tribes; but Venonius is authority for thirty-one rural tribes, which with those of the city would complete the thirty-five of our own time. Cato, however, who is more trustworthy than either of these two, says that all the tribes in the time of Tullius amounted to thirty, though he does not separate the number of parts” (into urban and rural). A great variety of opinion has arisen regarding the original number of the Servian tribes. Niebuhr[275] believed that Servius created in all thirty, afterward reduced by unfortunate war with the Etruscans to twenty. This view found supporters but was refuted by Huschke.[276] Those who rejected it generally agreed that Servius divided the city into four tribes and the country into districts, regiones, pagi.[277] Mommsen[278] gave a new phase to the theory of the subject by assuming that the four so-called city tribes, which all the sources agree in ascribing to Servius,[279] included the country as well as the city. According to this hypothesis Alba[280] and Ostia,[281] for instance, belonged to the Palatine tribe. His opinion has found wide acceptance.[282] Afterward changing his mind, he asserted that the four urban tribes were confined within the pomerium—a view which now seems to be established beyond doubt.[283] With this final position of Mommsen the creation of theories as to the number and limitations of the Servian tribes has not been exhausted; for against the view that Servius instituted only the four urban tribes may be placed that of Pais,[284] who assigns their origin to the censors of the year 304. The theory of Pais implies that the sixteen rural tribes which bore gentile names were far older than the four urban tribes.