[214] The water pumped out by the steam-engine employed by Signor Rosa was always good clear drinking water, and had all the appearance of coming from an aqueduct; it was sent along an open channel the whole length of the Colosseum, and made a great swamp round the arch of Constantine for several months in the spring of 1875, the outlet for it being made on the outer side of the drain under that arch, which had been constructed by the Municipality about 1866. At last it was observed that an aperture into that drain might easily be made on the inner side, and so avoid the swamp in the road, and washing the foot of the arch. Soon after this the steam-engine was stopped on account of the enormous expense of it, after it had gone on for more than a year, always pumping out good drinking water, and sending it to waste.
[215] This article will be found in the Annales de l’Institut de Correspondance Archéologique, 1873.
[216] Near the Porta Maggiore, and under the arches of Nero. There are two large reservoirs close together in this vineyard, probably the Gemelli of Frontinus.
[217] Descending from the reservoir into the old specus under the arches of Nero.
[218] The Aqua Marcia was brought again into Rome by a new water company in the year 1870, after several years of very arduous work, and overcoming many difficulties. The company who made this great aqueduct consisted chiefly of English shareholders, with a mixture of French and Italian. It was long under the direction of the late Mr. Shepherd, an Englishman, whose loss was much regretted in Rome, and latterly under a Belgian engineer.
[219] See Frontinus de Aqueductibus, c. 5.
[220] There is a large and deep reservoir for it near the arch of Dolabella, under the garden of the Villa Mattei, now called the Villa Cœlimontana, and from thence it passed, still underground, to the cliff of the Cœlian, opposite to the Aventine. A short tunnel was made from this deep reservoir to a Nymphæum under the Cœlian, now S. Stefano Rotondo, called of Septimius Severus. The ruins of it are shewn in 895; a specus has been traced from one to the other.
[221] See Frontinus de Aqueductibus, c. 6.
[222] Ibid., c. 125, and 93.
[223] Frontinus de Aqueductibus, c. 6.