From the piscina and reservoir on the cliff of the Cœlian another branch went to the south, over the spring of the Camenæ (?) 692, and near that of Egeria (?); it was carried over the Porta Metronia, and on the bank of the City Wall as far as the Porta Latina.

Another fine arcade of the time of Nero (No. 1317) leads to the Nymphæum before mentioned, where the Trophies of Marius were hung. The elevation shews that this water must have come from the highest of the aqueducts, the Anio Novus; and the remains of the reservoir near the Porta di S. Lorenzo, supposed to have been for the Aqua Julia, being on high ground, may have been for this branch, which conveyed water to the Thermæ of Titus.

X. Sabatina Trajana.

Great works for the aqueducts were carried on in the time of Trajan. One great work of his time was to bring water from the lake Sabatina to the top of the Janiculum. Augustus had previously brought water from that lake, supplementary to his aqueduct, from the lake Alseatina (VII.), to supply his Naumachia in the Trastevere; but the Aqueduct of Augustus was on the lowest level, that of Trajan on the highest. It does not appear that Trajan made use of the old specus of Augustus; but his aqueduct was afterwards made use of by the engineers of Pope Paul V. for the Aqua Paola, although they also brought a branch to it from the Alseatina, as Augustus had done.

That of Trajan is chiefly subterranean, and has been described under the head of the Alseatina (VII.), but nearer Rome it is above ground, and is carried on an arcade against the wall of the Villa Pamphili-Doria, near the Porta di S. Pancratio. Both the arcade and the specus are faced with the opus reticulatum of that period. (664* is from a drawing, 1065 from nature.) In some parts, one side of the specus has been cut away (1063). Just on the outside of that garden a large castellum aquæ of this aqueduct has been made into a farm-house, and in the yard of that house a branch from it can be seen (665*), apparently for the purpose of irrigation; or, as some think, this was formerly the point of division, one branch going to the Vatican, the other to the great fountain on the Janiculum (960), above S. Pietro in Montorio. The division now takes place at a short distance from this point. Procopius, writing in the sixth century, admires the enormous quantity of water brought by this aqueduct to the highest point in Rome; as it descends the hill, it turns the wheels of the flour-mills. After it arrived at the low level of the ground in the Trastevere, the respirators of the pipes for this aqueduct are carried in tall pyramids resembling chimneys (540). Part of the arcade and specus rebuilt by Paul V., near the garden before mentioned, is shewn in 1064, with the inscription of A.D. 1609 above it.

XI. Hadriana (?), Trajana (?) or Alexandrina (?).

The next great work of the period of Trajan or of Hadrian, on the eastern side of Rome, was probably begun in the time of Trajan. It brought water from springs under Labicum, now La Colonna, the same that is now brought for the Aqua Felice. The water from several springs was collected in a central reservoir, on which an inscription of Hadrian was found by E. Q. Visconti in the eighteenth century. This is between Pantana and Gabii, in the valley under La Colonna (1540). There are several other reservoirs of the time of Trajan or Hadrian along the line (1637, 1638). This aqueduct was considered by Fabretti to be of the time of Alexander Severus, and since his time it has usually been called the Aqua Alexandrina. It may have been partly rebuilt and brought into use again in his time, after having been choked up with stalactite, one of the springs used proving to be a petrifying spring.

Near the sources this arcade is low and much damaged, and the specus where it remains is nearly filled up with stalactite (1541, 1542). In some parts of the line the stalactite has all the appearance of a petrified cascade, and is evidently formed by the water oozing out and dripping and petrifying as it fell (1436). Further on there is a fine arcade for it across the country in the direction of Cento-Celle; and in some places the arcade is double to raise the specus to the necessary level, as in the Arches of Nero (1428, 1429). A portion of the arcade, where it is broken off, is seen in 1427, with the tower of Cento-Celle in the distance. Another portion of this fine arcade is shewn in 1640. It is of two periods; the upper part is of the third century, and may have been rebuilt by Alexander Severus (as has been said). At Cento-Celle the ground is high, and the aqueduct passes underground for some distance, along the side of the road towards Rome.

About a mile nearer to Rome, there is a branch aqueduct from the foot of the Marcian arcade, in the direction of the Mausoleum of S. Helena, which Fabretti considers as part of the same aqueduct; but it is difficult to see upon what grounds. There is a fine arcade here also for about a quarter of a mile; but it is of the time of Constantine, and I have not been able to trace any connection between this and the other. This arcade has been originally double, and the lower one only now remains, with a modern specus made upon it (555 and 556); but the water now flows from the Marrana, at the foot of the arcade of the great aqueducts, which were here on higher ground, and runs down upon this arcade to the garden and small monastery of S. Peter and Marcellinus, at the Mausoleum of S. Helena, called the Torre Pignattara, from the earthenware pots of which the vault was made.