We next come to the question of distribution. On examining the registers, Frontinus says that he found that there were assigned to the nine aqueducts 12,755 quinaries; but it appeared that the distribution amounted to 13,470 quinaries. To account for this discrepancy, the whole were re-measured. In the annexed table the measures of each aqueduct are given separately, and by adding together the results of each, as calculated from the data which Frontinus gives of the water measured at the head, the total, as the table shews, comes to 24,413 quinaries. The fact was that in process of time some of the channels had fallen into decay, and the water was wasted. In other and more numerous cases, it had been abstracted by persons for their own use without authority; in addition to which there were no doubt errors in the computation, which had crept into the books, and if we are to believe Frontinus, these errors had been turned to the profit of his predecessors. In the table of distribution will be seen the number of quinaries distributed outside the wall of the city; next, the Regiones which they serve within the city; thirdly, the measure of water so distributed within the city, and the number of _castella_, or reservoirs, used for that purpose. These reservoirs, at least such as remain, will be described in the several Regiones in which they occur. That over the arch of Dolabella on the Cœlian Hill, and the so-called Castellum Aquæ Juliæ on the Esquiline, are the most prominent. Many, no doubt, were of much smaller size, and merely large cisterns, as the total number was 247. The supply of 13,470 quinaries is thus accounted for. Outside Rome 3,164 quinaries were distributed, inside 10,306; of that outside Rome 1,718 quinaries were used “in the name of the Emperor,” that is, to supply the imperial villas and gardens, &c., while 2,345 were charged to private persons. Inside Rome 1,707 were charged to the Emperor for his palaces, &c., while private persons used 3,847 quinaries, leaving 4,401 quinaries, which were for the “Public Service.” These, together, account for the 9,955 distributed within Rome. In the details of the distribution, we learn how much water was employed in the palaces, &c., in the service of the Emperor and his household (_nomine Cæsaris_), how much was carried into houses for the use of private persons, and how much was used in public buildings and the public reservoirs or fountains which were established in all convenient positions, and generally accessible to the population.