PART VIII.
THE AQUEDUCTS.

THE
ARCHÆOLOGY OF ROME,

BY
JOHN HENRY PARKER, C.B.
Hon. M.A. Oxon., F.S.A. Lond.;

KEEPER OF THE ASHMOLEAN MUSEUM OF HISTORY AND ARCHÆOLOGY, OXFORD;
VICE-PRESIDENT OF THE OXFORD ARCHITECTURAL AND HISTORICAL SOCIETY,
AND OF THE BRITISH AND AMERICAN ARCHÆOLOGICAL SOCIETY OF ROME;
MEMBER OF THE ROYAL ARCHÆOLOGICAL INSTITUTE,
MEMBRE DE LA SOCIETÉ FRANÇAISE D’ARCHÉOLOGIE,
HONORARY MEMBER OF THE ROYAL INSTITUTE OF BRITISH ARCHITECTS,
AND OF VARIOUS ARCHÆOLOGICAL SOCIETIES, ENGLISH AND FOREIGN.

THE AQUEDUCTS.

OXFORD:
JAMES PARKER AND CO.
LONDON:
JOHN MURRAY, ALBEMARLE-STREET.
1876.

THE AQUEDUCTS.

PREFACE TO THE AQUEDUCTS.

It is impossible to understand the archæology of Rome without studying the Aqueducts. In every part of the City, and of the country round it, there are remains of them; they are frequently mistaken for something else, and called by other names, misleading those who have not given attention to the subject. They are necessarily mixed up with the Thermæ, for most of the aqueducts were made to bring water to those great establishments. There was also a reservoir of water supplied by the aqueducts under each of the palaces, larger houses and villas, as may be seen in many parts of the Palatine, under the palaces of the Cæsars. In their original state the Aqueducts must have been among the grandest objects in Rome, and the most conspicuous in all directions. The principal approach from the east passed between two fine arcades of the aqueducts, one carrying three of them, the other two. These two fine arcades were not more than a hundred yards apart; all the great roads from the eastern side were brought into this space, and certainly, for the last mile into Rome, must have had one of these arcades on either hand: to the left, or south, the Claudian arcade, fifty feet high; to the right, the Marcian, thirty feet high.

For the last half-mile, the Claudian arcade was also the boundary of the palace gardens of the Sessorium, the residence first of the Kings and afterwards of one branch of the imperial family (that of Verus or Varius), who resided there for more than a century. That portion of the arcade still exists for a considerable extent, forming the northern wall of that garden, and at the same time part of the wall of the city, Aurelian having adopted it, and incorporated it with his great wall. This portion is thoroughly shewn in my Photographs, as it is a very important part of the antiquities of Rome, and illustrates many points. It was the place where the principal aqueducts entered Rome, and the whole ground is full of remains of them with their reservoirs and filtering-places. The inner side of the wall is the part necessarily shewn, because the outer side is concealed by the arcade of the Aqua Felice, excepting in places where the specus of the Claudia and the Anio Novus appear above it. Just outside of this garden, at the west end of it, is the great foss which separated the Sessorium, or fortified palace, from the other fortifications of the City, and in it are remains of the two great reservoirs that were probably the Gemelli or Twins of Frontinus, made originally for the Appia and the Anio Vetus side by side, and used afterwards to receive the surplus water of the later aqueducts on a higher level. We then have a considerable number of the Arches of Nero, following a straight line for another mile to the west end of the Cœlian Hill; where remains of the great reservoir are on a level with the specus carried on the top of this arcade, fifty feet above the ground in that high situation, from which the water was distributed in all directions, first on other arcades, three of which branched off from this point, and afterwards in metal pipes when it was sub-divided; but for the main supply no metal pipes then to be had were large and strong enough to bear the pressure of a stream of water four feet deep and two feet wide, running at the rate of five or six miles an hour, if not more. The direct line went on over the Palatine to the Capitol, passing over the Forum Romanum on the bridge of Caligula, of which, also, some remains are shewn in my Photographs; the other two branches from the great reservoir over the Arch of Dolabella, went one to the right to the Colosseum (to supply the Stagna there), the other to the left to the Aventine, to supply the private house of Trajan and the Thermæ of Sura.

The Porta Maggiore stood at the end of the long vista between the two great aqueducts, and was itself made out of two of the arches of the Claudian arcade, in the last of the angles that occurred at every half-mile. The penultimate one was at the eastern end of the Sessorium palace gardens, where this water entered Rome, and then, after passing the angle to the north, with the usual piscina and castellum aquæ, or reservoir and filtering-place, turned again to the west, as far as that gate where the Claudian arcade terminated. From this point the Marcian turned to the north upon the great bank of earth which formed the outer defence of Rome in that part, and continued along it for another half-mile, as far as the Porta di S. Lorenzo, and beyond it to the Prætorian Camp; then turning again to the west, it went across the great inner foss between the outer bank, the great agger of Servius Tullius, which formed the inner line, where remains of it were found near the railway station in 1871, with inscriptions on two of the cippi, stating that the three aqueducts passed there. It is not quite correct to say that the arcades were carried there, because where the ground is high the specus are carried underground. This is the case between the Porta Maggiore (the Porta Esquilina of Frontinus) and the Porta di S. Lorenzo (the Porta Viminalis of the same). The specus is seen on arches close to each of the gates; but, between the two, it passed underground through a sort of hillock, about midway between the two gates, as is mentioned by Frontinus. The ground is high again near the Porta Chiusa at the Prætorian Camp, and in the great bank or agger of Servius Tullius.

These two great arcades are those chiefly known to visitors to Rome as The Aqueducts; but the underground aqueducts are at least equally interesting when understood. The Anio Vetus being near the surface can be traced all over Rome; the Appia being very deep, is not so easily traced, and can only be seen with certainty near its source and its mouth. There are other arcades nearly as fine as these; but, as they do not come so near Rome, they are less known; in the parts nearest to the City they are either underground or destroyed. One of them is that called Alexandrina by Fabretti; but the original parts of it are of the time of Hadrian and Trajan. This runs from the source near Gabii and Labicum, now la Colonna, to the place called Cento-Celle. Possibly some of the hundred cells or vaults found there and supposed to have been all tombs, were really reservoirs of water. At every great villa there are always remains of these great cisterns for the supply of water, and this is about three miles from Rome, where a great villa of Hadrian was situated. This arcade extends for miles across the country between the two great roads, one from Gabii, the other from Preneste (now Palestrina). Another fine arcade is seen on the road to Albano, at the Tor di Mezza Via (or half-way house). This has been ascertained to be the Aqua Aurelia, which goes to the villa of the Quintilii on the Via Appia, and from thence into Rome to supply the Thermæ of Commodus and Severus, of which also remains have been found; the latter part of its course is chiefly underground.

It was very difficult at first to ascertain to which of the aqueducts each of the remains belonged that were seen on all sides; and I found it necessary to follow each aqueduct up to its source, and then down to its mouth, in order to ascertain this. This work has taken me some years, in connection with the other branches of the general subject of the Archæology of Rome. It has now been done, with one exception,—the aqueduct made to bring water to the Thermæ of Diocletian on the Viminal,—which has not been traced; but there is great probability that it was a branch from one of the older aqueducts, probably brought across from the great reservoir where the Nymphæum of Alexander Severus was situated, on which the Trophies of Marius were hung. This is on high ground, and at a short distance only, so that it would be very convenient for the purpose. We see that there are several branches from that point, and one of them probably supplied these great Thermæ, for which purpose more than one aqueduct would be tapped. It may be that the curious sort of tower reservoir—long supposed to have been a tomb, a short distance to the south of the Trophies of Marius—was the one for these Thermæ; in either case, these reservoirs were supplied from the great old aqueducts.

I may almost say that wherever the aqueducts are visible, they can now be seen in my Photographs, and in some places the remains have been destroyed since these were taken, during what is called the Restoration of the City Walls. Perhaps the finest and most interesting part of the Aqueducts are the great cascades at the source of the Anio Novus above Subiaco, in the bed of that river, situated in some of the finest scenery in the world. Among later works the ingenious manner in which the bed of the small river Almo is made use of to carry the water of two other mountain streams, the Aqua Crabra and the Marrana, the water of which never fails; and the manner in which the tunnel of the Aqua Julia has been used again in the twelfth century, are the most curious and interesting. This had hitherto escaped observation, and was not easily traced. The Aqua Felice was unfortunately carried out in a very rude manner; the plan of the Pope was a good one, but was spoiled by the ignorance of the engineer.

The new aqueduct, the Aqua Marcia-Pia, restoring the celebrated Aqua Marcia to use in Rome, is a work deserving of high commendation. It is much to be regretted that my lamented friend the late Mr. Shepherd, to whose energy and perseverance we are chiefly indebted for this, did not live to see the completion of his work. Much credit is also due to Signor Moraldi, the originator of the scheme, to whom the Company still pay a premium, which he well deserves. His map[1] of the Aqua Marcia, of which he kindly gave me a photograph, was of great service to me, as far as it went; but I saw it was necessary to go further, and include all the Aqueducts on the eastern side of Rome. I was fortunate in meeting with Dr. Fabio Gori, who is a native of Subiaco, and has been interested in the Aqueducts from his boyhood. He shewed me that Signor Moraldi had not gone quite far enough, and that the real spring of the Marcia is about a mile further from Rome than the one he supposed to be so, which was a subsidiary spring, though of equally good water. At the original source the stone specus was found, having been long concealed by being a foot or two under water. I saw it, and stood upon it, and had a photograph made, so that there could be no mistake, and the engineer of the company also saw it, and carried his aqueduct to that point; so that the real ancient Aqua Marcia now comes into Rome again, and is getting rapidly into general use, being much the best drinking water. The water supply of ancient Rome has long been a subject of interest, and can now be more perfectly understood than it ever could before. The series of Photographs of them are a thorough illustration of their history, such as could not have been made before that art was invented.

CONTENTS OF THE AQUEDUCTS.

AQUEDUCTS.—LIST OF PLATES.

CHAPTER IV. PART I.
THE AQUEDUCTS.

In treating of the Aqueducts we have a trustworthy guide in a writer who flourished in the time of the Emperors Nerva and Trajan, namely, Sextus Julius Frontinus[2]. While he informs us of what improvements he made during the time he had the charge of these important public works, he also gives in his treatise a historical account of the several changes which had been from time to time made in the means of supply of water to the Imperial City, which kept pace with the growing wealth and population of Rome. When he died they had probably reached perfection, and were justly the admiration and surprise of all travellers[3].

His treatise is well worth examination, not only from an archæological point of view, but as suggesting also many curious enquiries as to the engineering abilities which the Romans possessed, compared with those exhibited at the present day. In elucidating, however, the architectural antiquities of the city of Rome, it will be necessary, as far as possible, to limit the extracts from his treatise De Aquæductibus, to those portions which refer either directly to existing remains, or which indirectly explain them, by pointing out the principle on which the several Roman aqueducts seem to have succeeded each other.

Frontinus tells us in his fourth chapter, that for 441 years after the building of the city, or until B.C. 312, the people were content with the water which they could draw from the Tiber, or from wells[4] or from springs, of some of which the memory was in his time still held sacred and honoured, for they were thought to afford health to the sick[5].

Passing to his own time, he says, “there now flow into the city:”—

Frontinus describes the above nine aqueducts in order, giving details as to the source of the water in each, the quantity, and the distribution. Such extracts as are calculated to throw light upon the existing remains, are given in the course of the following remarks. Seven other Aqueducts were added after the time of Frontinus at different periods, of which an account will also be found in the second part of this Chapter.

I. The Aqua Appia (A.U.C. 441, B.C. 312).

“The Aqua Appia was brought into Rome by the censor Appius Claudius Crassus, afterwards called Cæcus (the blind), who also caused the Via Appia to be constructed from the Porta Capena to the city of Capua[7].”

“The Appian stream rises in the Lucullan fields on the Via Prænestina, between the seventh and eighth milestone, and about 780 paces (or about ¾ of a mile) off on the left-hand side. The channel from its source to the ‘Salinæ,’ which is a spot near the Porta Trigemina, measures 11 miles, 190 paces (about 300 yards) in length. For 11 miles, 130 paces, (about 200 yards,) the channel runs underground, and for 60 paces (about 100 yards) it is carried above ground on a substructure and arcade (in the part) nearest to the Porta Capena[8]. At the ‘Spes (Specus) Vetus,’ (the old specus or conduit,) on the confines of the Torquatian and Pallantian gardens, a branch called the Augustan was added to it by Augustus as supplementary, whence it received the name of ‘Gemelli[9]’....”

“The distribution of the Appian water begins at the bottom of the Clivus Publicii at the Porta Trigemina[10]; the stream having passed beneath the Cœlian and Aventine Hills.”

“... The measure at the head of this could not be obtained because it consists of two channels. At the Gemelli, however, which is under[11] the Spem (Specum) Veterem, (the old conduit,) where it joined with the Augustan branch, I found the height or depth of the water to be 5 ft., the width 1 ft. 9 in., which makes the area 8 ft. 9 in.: ... it should be noticed that the water in many parts of the City was observed to be lost, that is, by trickling away. Moreover we found some of it intercepted by illegal pipes within the City; without the City, owing to the pressure, that which was underground (at the head 50 ft.), received no injury[12].”

The Lucullan fields, in which are the sources of the Aqua Appia, are nearly due east from Rome, on the bank of the river Anio, at about six miles from the present gates of Rome, and three quarters of a mile off the Via Collatina, now a cart-road. This is the old Via Prænestina, which went through Collatia, now called Lunghezza, about two miles higher up the river Anio, at the point where the smaller river Osa falls into it. In this part the old paved road remains for some distance in use, both before and after the entrance to Collatia, and was the more direct road; the present carriage-road, now called Prænestina, was the Via Gabina, and went through or near to Gabii, but was not so direct a line to Præneste or Palestrina as the one through Collatia. The old road passes close in front of the ancient stone quarries, on the bank of the Anio, in which one of the springs or reservoirs of the Aqua Appia is situated; the pavement of it remains, and is now used as a foot-path only.

There are several springs or sources for this aqueduct, and also several ponds to collect the rain-water, made in the clay, the water from all of which was collected into a central reservoir cut in the rock, with a large well over it. There are also seventeen wells visible in two lines, eight in one and nine in the other, converging and meeting in the central reservoir. The soil is clay upon stone. The specus is here a large tunnel cut in the rock, and the wells descend into it. They may be distinguished from above by the bramble-bushes which grow over and cover each of the openings, and do not grow on the other part of the field; each well is also protected by a wooden rail, to hinder cattle from falling into it. The water still runs through in some parts. Near to this reservoir is a tomb of very early character, similar to what are usually called Etruscan[13] tombs, cut out of the rock.

As the length of its course was rather more than eleven miles, while the distance by the Via Prænestina direct was only between seven and eight, it is clear that it did not follow that road. It was carried to the south towards the Via Labicana by a winding course, according to the levels of the ground. The Aqua Virgo, whose source is in the same Lucullan fields as the Appia, is carried to the north by another winding course of a greater length, to supply the northern parts of the city.

The course of the old Via Prænestina[14] has undergone but little change, and the 7th milestone is soon measured. A little beyond it, and 780 paces off on the left of the road (i.e. on the northern side), we are brought to a point[15] where the sources of the Aqua Appia are still to be seen in caves formed by ancient stone quarries, near a tenement called La Rustica. One cave is of a triple form, with three springs, and the soil above the stone is much mixed with clay. These three springs meet at the mouth of the cave in one stream, which runs in the old specus cut in the rock, but open at the top, and having the appearance of a mere country ditch, half hid by the grass and weeds in the winter, and in the summer months entirely concealed by them. This specus runs across a low meadow for about a quarter of a mile. It crosses the line of the Aqua Virgo, the respirators of which cross this meadow in an opposite direction; the sources of the Virgo are about a mile further from Rome, on the same road. It then enters the rock again in a tunnel, and at this point is the first castellum or reservoir, with a large circular well cut out of the rock, which forms the vault over it. There are the grooves for sluices at each end of the reservoir, and an opening has been made into the tunnel a little beyond this reservoir. The course of the aqueduct is then in a tunnel entirely underground, in the direction of a series of ancient caves, through which, or by which, it must pass, but being underground it cannot be seen. These caves are evidently an ancient stone quarry, but of later character than the one first mentioned. In the earliest one, at the source, the stone has been split off the rock, not cut; in the other caves, the stone has been well cut, and of these many are to be seen, the openings being large square apertures like doorways, but side by side in the face of the scarped cliff, not following each other: a practice common in quarrying, where the circumstances require it, down to the present day. Between the chambers are massive square or oblong piers, left to support the vault and the earth above. At the bottom of one of these chambers is water running from another spring, and the shepherds state that there is an opening from it to the specus of the aqueduct; but as the bottom is filled up to the depth of several feet with broken stone, it cannot be seen. Some of the chambers of this ancient quarry have been used as a burying-place, probably for the neighbouring town of Collatia. The graves are cut out in the side of the walls, like the loculi, and there are also chambers like the cubicula of a catacomb. The tombs of this necropolis of Collatia are very ancient, perhaps anterior to the Kings of Rome, and certainly not posterior to the Republic. This stone has been cut out in large square or oblong blocks, such as are usual in work of the time of the Kings; it is probable that Appius Claudius made use of a quarry which he found, and that the ground was not excavated for the purpose of the aqueduct only.

From this point the course was entirely underground, therefore no traces are visible without some difficulty. Incidentally, as we have seen, Frontinus notes, that at the head the water ran as much as fifty feet below the surface, and was thus protected from being fraudulently diverted. It could not have taken a very direct line, as the levels of the crossing valleys must have necessitated here and there a circuitous course; and moreover this is clearly implied by Frontinus, who says that the total length from the source to the “Salinæ” was more than eleven miles, while the distance in a direct line is less than nine miles. As it had no piscina[16], it flowed on in an uninterrupted course to the spot where it was distributed for the public service.

One branch seems to have entered Rome under the line of the Claudian arcade, near the Porta Maggiore[17], and its course within the city is not difficult to follow, since for certain purposes access has been made in more than one place into the old specus, part of which is used to carry the lead pipes of the Aqua Felice to the Hospital of the Lateran and other places. From the surface where one of the shafts is situated to the bottom of the channel, is a depth of twenty-five feet. No water runs along it at present, except in the metal pipes from the Aqua Felice, belonging to the modern system.

The direction of this channel, after it enters Rome, happens to be easily seen above, as it is marked throughout its course, from the eastern end of the Cœlian to the Arch of Dolabella, by the fine arches of the Neronian branch from the Claudian Aqueduct erected almost over it. It is very probable that one of the laws (referred to by Frontinus as having been in existence long before his time) respecting aqueducts was in force from the commencement, namely, that there should be no buildings of any kind within ten feet on either side of the aqueducts; thus an open space was probably existing at the time when Nero wished to carry water to his magnificent reservoir at the north-western end of the Cœlian, and of this he availed himself in rearing the splendid arcade which is called by his name. It will be remarked, however, that the brick buttresses actually came down to the flat roof of the original specus of the Appian. The top of that specus was raised to a greater height by the engineers of Sixtus V., and in this enlargement the underground supports of the Neronian arches are cut through to give headway to the aquarii passing along the channel. They have however to stoop their heads as they pass under each buttress.

The old tunnel specus was discovered in the time of Sixtus V., (Felice Peretti), filled up by clay deposit to the depth of three feet, or half the height of the specus, as in other places, and the builders found it easier to raise it, for the aquarii to go along it, than to clear it out. They therefore knocked away the flat tile-covering of the time of Nero, and raised a vault of rubble-stone three feet higher.

From the great subterranean reservoir, still in use for the water of a spring, near the Arch of Dolabella, to the edge of the western cliff of the Cœlian is but a short distance, and here the specus seems to run in a bank dividing the vineyard or garden of the Villa Mattei from that of the monks of S. Gregory, on which a wall has been built. At this point subsequent alterations, and especially those under the Emperors Nerva and Trajan, have buried to a great depth the actual remains of the original aqueduct; but this later work occupies the same line, and must be described. We first come to a large piscina or filtering-place on the cliff of the Cœlian, immediately above the Porta Capena. This piscina is divided into two parts, one above the other; but both are below the top of the cliff, and are faced with brick and reticulated work of the time of Trajan. There are remains of two specus running along against the face of the cliff; of the upper one the lower part, or pavement, of Opus Signinum, only remains, and this runs down in a sloping direction from the upper reservoir to a lower one, a little to the south of it, which is very extensive. This lower reservoir consists of several parallel chambers, through which, or rather in front of which, another specus runs at a lower level; this appears to be horizontal. This specus goes on in the direction of the ruins of a building of various periods, possibly the remains of the Ædes Camenarum, and passes then to the south in the direction of the Thermæ of Severus and Commodus.

Another branch was evidently afterwards made from the upper reservoir going towards the north for a very short distance, merely for one of the usual angles; then, turning again to the west, in the valley below, there is another large castellum aquæ, or reservoir of five chambers, all as usual oblong and side by side. This is on lower ground, and is built upon the wall of Servius Tullius, where it crosses the valley; the upper part of this reservoir has been made into a house for the gardener. The underground part of the chamber nearest to the Cœlian is built of the large square blocks of tufa usual in the time of the Kings, and belongs to the fortifications of the Porta Capena, over the Via Appia, which here passed close under the Cœlian; the other chambers appear, from the bricks, to be of the time of Trajan. From this point to the Piscina Publica, the line of tall brick piers of Trajan can be distinctly traced by the existing remains, passing across to the other side of the road and of the stream here close to it[18]. The ruins of the Piscina Publica, as rebuilt by Trajan, remain visible under the corner of the Pseudo-Aventine, near S. Balbina, and from this the water was again distributed in different branches. At this point, let it be observed, and only at this point, is the valley which divides the Cœlian from the Aventine sufficiently narrow to admit of agreement with the direct and clear assertion of Frontinus, that here alone, throughout the whole course, was it carried on a substructure and on arches for a distance of one hundred yards.

Although there are no remains of the specus now visible at this spot, because the gate has been destroyed, there can be little doubt that the channel for the water was carried over the southern gate of the city according to its extent at that time. This gate was called the Porta Capena; and as the Appian aqueduct was allowed to fall into decay, it gave rise to the descriptions both of Martial and Juvenal, who describe it as wet or moist[19].

The excavations made in 1868 and 1869 under the direction of Mr. Parker, with the help of the British Archæological Society of Rome and the Roman Exploration Fund, have clearly shewn the specus of the Aqua Appia and those of two other aqueducts carried upon the agger of Servius Tullius across the valley from the Cœlian to the Aventine, with branches to the left running into the subterranean chambers of the Piscina Publica. These underground chambers are of the time of the Republic, the walls are built of rubble stone as usual at that period, and there are small openings through these walls for the circulation of the water, although the upper part has been rebuilt in the time of Trajan, as the remains of the wall are faced with the brickwork usual in his time. The lowest specus is cut out of tufa rock under the wall of Servius Tullius, which is built of the usual large blocks of tufa.

The principal branch went underground to the cave reservoir at its mouth, on the level of the quay of the Marmorata between that and the Salaria, just outside of the old Porta Trigemina; the old agger in which that gate was situated still forms the southern or lower boundary of the wharf of the Salaria. Part of the course of this earliest specus, that of the Appia, can be traced and seen in a subterranean stone quarry nearly under S. Sabba. The specus is six feet high and two feet wide, and it is filled up to nearly half its height by solid clay, evidently the deposit left by the water. The old specus, long after it had been out of use, seems to have been employed by the quarrymen: as it was a tunnel in the tufa rock just high and wide enough for a man to walk in, by cutting away one side they made it wide enough for a horse and cart to carry the stone, and they raised the vault as high as was convenient for the purpose of making an entrance into the quarry. The series of wells descending into the old specus from the gardens above remain at intervals, with notches for steps cut in the rock to enable a man to go up and down when required. The course of the specus is cut off by the road to the Porta Ostiensis, but probably that road was originally carried over it[20] at the crossing; it then passed through another large subterranean quarry nearly under S. Prisca, to the cave reservoir at its mouth[21].

Before arriving at the two large reservoirs just outside of the garden of the Sessorium, now Sante Croce, the specus of the Appia must have passed by another smaller reservoir at the same low level, near the ruins of the apse of a hall, miscalled the temple of Venus and Cupid. This seems likely to have been the point at which the branch specus, coming from the north, entered Rome, and it was then carried on to the two large reservoirs outside this garden, supposed to have been the Gemelli[22].

Below the “Salinæ” or salt warehouses on the bank of the Tiber, and near the “Porta Trigemina,” the water was “distributed.” This was also close under the Clivus Publicii, or the slanting zig-zag road leading up from the wharf to the top of the hill.

So far the general course can be traced; but the exact point of entrance into Rome could not be fixed without excavations, which have not as yet been made. There are, however, some data given by Frontinus which should not be overlooked, as they bear incidentally upon the course of some other of the aqueducts.

First, it should be remarked, that the water of the Appia was augmented by an additional stream. This was not accomplished till the time of Augustus. The source of this latter[23] was on the same side of the Via Prænestina, but a little nearer to Rome (near the sixth milestone) we are told, and its course was, like the main stream, entirely underground. It was more direct, as its length was less than six miles and a-half, while the original stream, by its windings, required about eleven miles to complete the same distance[24].

It joined, however, the Appia at the “Spes (Specus) Vetus,” and at the Gemelli[25], which was under (infra) the “Old Specus,” it could be measured. This is an important landmark; it is well at once to observe that the usual interpretation of the passage as referring to a “Temple of Spes,” and the statement that a temple occupied the site of some ruins which are marked thus in a few maps, (in most, Templum Veneris et Cupidinis) does not seem to fit the circumstances. It is proposed to read in both these two instances “Specus Vetus.” The expression occurs, altogether, five times in Frontinus, and it will be most satisfactory therefore to consider the passages together, which will be done more conveniently in connection with the next aqueduct described.

The Torquatian and Plautian, or Pallantian gardens, seem by the context to have occupied a spot on either side of the point of junction of the two streams. The Torquatian gardens are not mentioned elsewhere, and with respect to the Plautian[26] (which is only conjectural, and has the authority of neither of the early MSS.), it has been suggested to read Pallantian, which gardens are mentioned twice elsewhere by Frontinus in connection with the Specus Vetus, besides once in such a way as to imply the same[27].

The Torquatian gardens probably occupied the space between the outer wall and the modern “Via di Porta Maggiore,” and were probably those of Titus Manlius Torquatus, a member of an important patrician family; Livy mentions at least three generations of that name. This ground shews numerous remains of aqueducts and reservoirs, and here also is the fine building of the third century called the Minerva Medica, which was probably the nymphæum of some thermæ of that period[28]. The Pallantian gardens is a natural name for those belonging to the Sessorium, or the Sessorian fortified Camp and Palace[29], and occupied the low ground near it. They would therefore be to the south of, but adjoining to, the Torquatian gardens.

II. The Anio Vetus (B.C. 272).

“Forty years after the Appian aqueduct had been completed, and in the year A.U.C. 481 (B.C. 272), Marcus Curius Dentatus, who bore the office of censor with Lucius Papirius Cursor, contracted to bring into the city the water of the Anio, which is now called the Anio Vetus, from the spoils taken of Pyrrhus[30].

“The Anio Vetus takes its rise beyond Tibur (now called Tivoli), outside the gate, at the twentieth milestone, where it parts with some of its water for the use of Tibur. Its course has in length, in consequence of the difficulties caused by the levels, 43 miles. Of this for 42 miles and 779 paces, the stream is underground, and for 221 paces (about 350 yards) on a substructure above ground.

“Near the fourth milestone within the new road (infra novum) the Anio Vetus crosses between the arches from the Via Latina into the Via Lavicana, and has (there?) a reservoir (or piscina). Thence just within the second milestone it gives a part of its water into the specus, which is called the Octavian, and this comes in the direction of the New Road to the Asinian Gardens, whence it is distributed through that neighbourhood; but its direct channel, following the specus (or conduit) coming within the Porta Esquilina, along the Spes (Specus?) Vetus, is drawn off for distribution in the city, through the high channels[31].

“It held the sixth place in height, but would have been sufficient for even the higher parts of the city, if it had been carried across the valleys and the low ground, where requisite, upon substructure, arches, and buttresses[32].”

Passing on to the second aqueduct, namely, the stream taken from the River Anio (and which was called the Anio Vetus to distinguish it from a later diversion from the same river), we find that Frontinus is also very distinct in many of his details; a few only he leaves to conjecture.

The point where a branch was taken from the river Anio can be traced[33]. There was a reservoir by the side of the river, of which some ruins remain; but it was nearly destroyed a few years since by some peasants in their ignorant and too eager search after hidden treasure which they expected to find there. The specus is cut in the rock as a tunnel in the cliff of the valley of the Anio, just below the level of the road; it follows the line of the cliff, of the river, and of the road: for they are all the same, the only opening through the rocky mountains of limestone in this part, obviously made by the river itself. It follows this line as far as the valley called the “Valle degli Arci,” or of the Arches, where three later aqueducts cross the river on arcades about two miles above Tivoli. At this point several aqueducts are visible crossing the valley on arcades, each on a separate arcade, not three on one arcade and two on another, as near Rome. The one nearest to Tivoli is the Marcian, and at the foot of one of the piers of the arch, which here crosses the road, the specus of the Anio Vetus is visible, partly underground, but the upper part above ground, on the right-hand side of the road in going from Tivoli. It then is carried in a tunnel through the hill, but appears again on the other side with a large reservoir, and runs gradually downwards in a winding course round part of the hill.

It is not easy to distinguish to which of the aqueducts belongs any one of the numerous reservoirs, the ruins of which are conspicuous objects on the roads up to Tivoli on the other side towards Rome, especially along both sides of that called the “Promenade of Carciano,” for the distance of about three miles from Tivoli. The specus and reservoir of the Anio Novus are on a natural terrace above that road, and the Marcia below it; these have passed through Tivoli, winding round that end of the hill. The Claudia passed upon a great arcade the valley of the Arci, and another valley in the direction of Gerocomio. The Anio Vetus was carried in a tunnel through the hill, and appears on the other side near Tivoli, at a considerably lower level than the others. The road passes across the specus, and is made upon the surface of the vault, which is visible about four miles and a-half from Tivoli. A little further on there are two large reservoirs belonging to it, near a modern villa called Gerocomio, built by the Cardinal Santacroce in the year 1579, now a farm-house only, but on the site of an ancient villa, of which some remains are built into the walls. One of these reservoirs or piscinæ (?) appears to be unaltered, the buttresses remain and apparently the vault; but it is covered with herbage and shrubs, which conceal it. The other has been turned into a cottage or an out-house of the villa, and this is of later date than the other. The Anio Vetus, the Marcian, the Claudian, and the Anio Novus, all have to be conveyed down this lofty hill from the high ground on which the remains above mentioned are situated, to the valley below. The river Anio itself rushes straight down the celebrated cascades by a fall of about a hundred feet; but the channels of the aqueducts follow a winding course, which allowed of the descent towards Rome being very gradual.

This road or promenade of Carciano is rather high on the side of the hill above the villa of Hadrian, looking towards Rome. The dome of S. Peter’s is distinctly visible from it. The road is on a ledge on the side of the cliff, and the aqueducts run along the steep side of the hill or cliff, until they come to the valley of a small stream winding down to the Anio; along the cliff of this they are then carried. The Anio Vetus being at a considerably lower level, is more difficult to trace than the others; but it seems clear that after the later aqueducts had arrived at the point where the Anio Vetus emerges from the hill, they all followed the same line, winding along with the small stream, and their specus or channels cut in the cliffs as far as they could be made available. The Anio Vetus has been discovered at the left hand of the modern road to Rome, and to the right of the promenade of Carciano. It passed, upon a lofty bridge called the Ponte di S. Antonio, over a torrent, with the Marcia. Afterwards upon another bridge called Ponte delle Mole di S. Giovanni, and over the Ponte Lupo it went with the others. Near Gallicano (or the ancient Pedum) are two cippi of Augustus in two wells of his specus; one is in the country called Le Sette, and another at Obrego[34] dell’ Ermito. The others have all been followed, and an account of them will be found under their respective heads.

In order to avoid the many small valleys occupied by streams which run parallel to each other from the hilly ground on the south, down to the river Anio on the north, (the general course of which river is west and east,) the course was kept along the higher ground, and in fact wound round the heads of those valleys in order to retain a level, gradually becoming more and more depressed, till, by the time it reached Rome, the base of the specus (according to the computation of Piranesi), was 55 ft. above that of the Appian[35]. The whole course is underground, as has been said, except for 221 paces (about 360 yards); for this short distance it was carried on a substructure above ground. It is reasonable to suppose that this exception to the subterranean course was, as in the case of the Appian, within the present boundary of the city.

Before it reached Rome, two circumstances have especially to be noted. Just within the fourth milestone we are told it had a reservoir and piscina or filtering-place, and at the second milestone it parted[36] with some of its water, which was conveyed to Rome in a separate channel.

This fourth milestone was clearly on the Via Latina, as Frontinus in the previous paragraph had referred to some piscinæ at the seventh milestone on the same road; and it appears, although the sentence is exceedingly corrupt, that the course of the Aqueduct left the Via Latina at this point, and crossed towards the Via Labicana “amongst the arches[37]” at the fourth milestone. The specus is here visible, and appears to be perfect; it is very near to the Torre Fiscale[38], between that and the Osteria, called the Tavolato (which itself appears to be made out of another, but a later, reservoir belonging to the aqueducts, as many of the houses in the Campagna have been). The vault of the old castellum aquæ or piscina is now covered with turf, but the side of it forms a sort of cliff like the edge of a quarry.

This place, where the Anio Vetus leaves the Via Latina, is near the great junction and crossing of the aqueducts, over which the tall medieval tower called the Torre Fiscale has been built. Here six aqueducts meet and cross each other. The Marcian arcade, with three of these, makes one of its many angles, and the lofty Claudian arcade, with two more, is carried over it. The expression that it passed amongst the arches is a very natural one, to any person who knows the locality, as there are many arches at this point. It then goes on to another angle and crossing of the great aqueducts, where there is a gate called Porta Furba, about two miles and a-half from the Porta Maggiore. There is a castellum aquæ of the time of Nero, with an ancient piscina under it, and a fountain of Sixtus V. by the side of it. At the second milestone it parted with some of its water into the specus called the Octavian, which enters Rome at the Asinian gardens, following the direction of the new road.

By referring to the maps it will be seen that the original line of the Via Latina united with the Via Appia within the outer wall and before reaching the old southern entrance of the city (the Porta Capena); but, in joining this latter road (the most convenient course to pursue in the then state of the fortifications), the Via Latina swerved rapidly to the south-west. Had it been continued in a direct line, it would have reached the Cœlian Hill, near the Porta Asinaria, as the Via Appia Nova still does, following the line of the old Via Asinaria.

At this second milestone also is another castellum aquæ[39], mentioned by Frontinus as two miles from Rome. This is near the Porta Furba; it is entirely buried, but the vault of it is not many feet underground. In the spring of 1871, some excavations were made under my direction in a large vineyard hard by, and another subterranean reservoir was found near the road to Tusculum and Frascati, with a specus cut in the rock going in the direction of that road, and apparently passing under it, on the line of a cross-way to the Via Appia Nova, a short distance only. On the other side of that is the “Albergo dei Spiriti,” near the junction with the Via Latina; and, in the garden at the back of that house, a specus was found in a stone quarry, the vault of which had fallen in and brought the specus to view. It seems to have passed underground along the southern side of that ancient Via for a short distance, and then crossed it to a piscina, of which there are remains at the foot of the bank on which the road runs in that part. It then goes along the edge of some higher ground, and for a short distance underground again towards a by-lane (diverticulum), parallel to the Via Latina; remains of a brick arcade can be seen on the bank of that lane, which is a deep foss-way, and goes on to Rome about a mile distant. It is cut by the railway before it arrives at the wall of Rome. It then passes through that wall and underground again as far as the arch of Drusus, over which it passes; and thence on an arcade, part of which only remains, to the great piscinæ at the back of the thermæ of Caracalla[40].

The Via Appia Nova was probably made in the time of Frontinus, and is the road which he calls Via Nova. The part nearest to Rome was previously called the Via Asinaria, and extended from the Porta Asinaria to the junction with the Via Latina, at three miles from Rome, which name was then dropped. The new road continues parallel to the Via Appia Antiqua as far as the eleventh milestone, and there forms a junction with it. Both of these roads are now open. The railway to Capua and Naples passes near to this point of junction. The short Via Lateranensis, going out of the Porta Lateranensis (excavated in 1868), ran into the Via Asinaria, and so joined the Via Appia Nova, which has tombs of the first century along the line, and none of any other period. Those which were of stone have been used as a quarry by the farmers to build the low walls that line the road on both sides, the foundations only being left in the banks; but those which were built of concrete faced with brick, would not pay for the trouble of destroying them, and have therefore been left standing in their places. One of these, a very fine one, faced with brick of the time of Trajan, with moulded pilasters, remains nearly perfect near the seventh mile, just where a path turns off to the left across the fields to the piscinæ, which are near the line of the old Via Latina in this part, about half-a-mile from the Via Appia Nova, and forming the carriage-road to Albano through Marino. It left Rome by the Porta Asinaria as it now does by the modern Porta S. Giovanni, which is close to the old gate, but at a much higher level. The Via Latina crosses it in a diagonal line, and runs nearly parallel to it as far as the Torre Fiscale, that is, for about a mile gradually diverging from it.

Wherever the exterior of the specus of the Anio Vetus is visible, it is faced with Opus Reticulatum. The reservoirs of this Aqueduct are of the same construction, and this may serve to distinguish those on the slope of the hill at Tivoli from the other aqueducts there.

“The water of this straight branch,” says Frontinus, “coming within the Porta Esquilina along the Spes [specus] Vetus, was carried down into the city in the high streams[41].”

This specus is on the high bank of the Tarquins, the outer and lofty line of defence on the eastern side of Rome. On this the wall of Aurelian was afterwards built. After its entry into Rome, the Anio Vetus was divided into several branches, in the same manner as the later aqueducts were.

The right-hand branch, in crossing the Campagna, appears to have run nearly under the Marcian arcade, which was afterwards built on the same line and nearly over it. As we approach near to Rome, there are remains of a reservoir built of large stones a little way down the road, about a hundred yards from the Porta Maggiore. It enters Rome under the wall, almost on the level of the ground; the upper part of the specus was visible in 1868, but in 1869 was studiously concealed by a modern brick wall. It is visible again inside the wall, on the other side of the road, going into the vineyard in which the Minerva Medica stands, while one branch went to the great reservoir near to it, westward of the gate.

Another branch passes along the bank under the wall of Aurelian, and is not visible again until it reaches the Prætorian Camp, where a portion of it was excavated in May, 1868, built of large stones of tufa, under the Porta Chiusa. It may then be traced all round the three sides of the Prætorian Camp, and near the north-east corner there was, in 1868, an opening into it, now closed by a modern wall. It is distinctly visible in several places, especially on the north side of the Camp, where the wall of Tiberius remains perfect, faced with the fine brickwork of his period, whereas the specus under it is faced with Opus Reticulatum, probably of the time of the Republic. The wall of Tiberius distinctly stands on the old specus. There was an opening into it here also in 1868; but this was also carefully walled up in 1869 under the influence of the Garibaldian panic, which had a bad effect upon the Roman authorities at that period. Remains of a reservoir or castellum aquæ were also found on the surface of the ground on the bank near the Porta Chiusa, with the present wall of Rome built right across it, but this part of the wall has been rebuilt of old materials, and is not exactly in the same line. There are remains of several other reservoirs on the bank at intervals outside the wall in the modern road, and in many places all along the eastern and southern sides of Rome.

Beyond this, near the Porta Nomentana, are the ruins of another reservoir or piscina on the surface of the ground, against the wall. From near the Porta Chiusa another branch went along the old road, which passed through that gate across the inner foss to the Thermæ of Diocletian, where a part of the specus and two cippi, with two inscriptions upon them, naming this aqueduct the Anio Vetus, were found in the year 1861, when the railway was made[42]. These cippi are now in the Vatican Museum.

Another branch went from the reservoir near the Porta Maggiore before mentioned, across the road into the bank on which the arches of Nero stand, near the Lateran, and passed under them apparently into the old specus of the Appia, which runs parallel to and nearly under the arches on the other side. Two small specus or stone pipes can be still seen (in 1872) passing obliquely into that bank; and, as they came from this large reservoir and piscina, (or from this direction,) they seem to have belonged to two aqueducts at different levels. A branch of the Marcian may have been brought to the same filtering-place for distribution, and the surplus water carried into the old specus at the lowest level, which was evidently used for receiving and carrying off the surplus water of all the other aqueducts in the same line.

The fifty yards outside the wall added to the three hundred yards between the Cœlian and the Aventine, make up the three-hundred-and-fifty yards above ground, mentioned by Frontinus. In this valley we found it again, in 1869, parallel to the Appia, sometimes on the agger or bank of Servius Tullius, which was used as a substructure for it, in other places on an arcade built up against the tufa wall of Servius Tullius, and faced with reticulated-work.

APPENDIX.
On Spē or Spc̄.

For the supposed Temple of Spes, the ruins of an apse in the gardens of S. Croce, of “Venus and Cupid,” (as it is marked in most maps, and as “Speranza Vecchia” in others,) was fixed upon by Piranesi, who carefully examined all that he could with a view of mapping out his aqueducts, according to the knowledge possessed in his time. This building was no doubt a hall belonging to the Sessorian Palace. Others, again, have suggested the so-called temple of Minerva Medica, but this again is a nymphæum, or pantheum, and not a temple at all. Besides, a further difficulty lies in one being too far south, while the other is too far north.

Canina, in his account of the results of the excavations at the Porta Maggiore[43], and of the tomb of Eurysaces the baker there discovered, just outside of the gate, gives a plan in which he inserts a temple just inside the gate on the southern side, which he calls the Temple of Spes[44]. It is quite possible that this was a Temple of Spes. There certainly was a temple on the site indicated, where the modern guard-house stands; and during excavations carried on there, fragments of a temple of the time of the early Empire were found, consisting chiefly of a fine cornice of travertine.

When, however, the words of Frontinus have to be applied, the difficulties of the theory of his referring to a temple are increased. In the case of the expression last referred to, “following the specus (spes), or, according to one manuscript, the old specus, which is obviously the sense,” it is very difficult to imagine what circumstances there were in connection with a temple which could warrant the use of such words; and even in the instance mentioned, where we were noticing the Appian aqueduct, “the expression, the Gemelli, which is a place under the old specus (Spe̅s̅ Vetus),” is somewhat singular. Granting that a temple once existed just within the wall of the city (which, from the context, must have been its position if any) it is singular that he should use it as a landmark when describing the junction of two streams of water. The remaining three expressions are simply “at the old specus (Spe̅s̅ Vetus[45]).”

With these difficulties to contend with, it has been thought well to seek a different solution, and this is found in the reading of “Specum Veterem” for “Spe̅m̅ Veterem,” i.e., the “old specus.” With this reading, it naturally follows that it would refer to the old specus, or the specus of the Appia and the Anio Vetus; and it is singular that the first time it occurs in Frontinus, the Codex Urbinas[46], only second in authority to the Codex Cassinensis, has the reading “Anienem Veterem,” instead of “Spem Veterem.” What was the true reading of the Codex or Codices from which these two copies were made, it is impossible to say; they are the earliest we have, and it is clear from several other instances that the scribes did not copy with much knowledge of the matter in question: then it was easy to mistake Spc̄ for Spē. That it was the Specus Vetus which was meant, must rest therefore upon the circumstances which allow of its application to the passages named, and it remains to shew that this is the case.

In the instance under discussion, the water of the Anio is said to flow along its own specus, and therefore it would not be probably possible to find an interpretation more suitable as far as this case is concerned. The conduit of the water of the Anio Vetus had “emerged,” as the other aqueducts emerge, near the Porta S. Lorenzo, from the higher ground between the Porta Maggiore and that gate: it must consequently have been carried on a substructure from that point on the outer bank of the original fortifications of Rome, that is, on the high bank on which the aqueducts were carried, and on which the wall of Aurelian was afterwards built, to the inner bank, on lower ground faced with the tufa wall of Servius Tullius. The names of the gates are matters of dispute, and are quite immaterial; the levels of the ground decide the question[47].

In another place, Frontinus says[48] that several streams of water, and first the Marcia, were carried to the Aventine from the specus (a spe-cu), that is from the old specus he had before mentioned on the Cœlian, and obviously from the west end of the Cœlian. The Temple of Spes of Canina and others at the Porta Maggiore, is at least a mile from the Aventine. The only way of giving an intelligible meaning of the passage is that the author refers to the old specus he had before mentioned, as leading along the Cœlian to the Porta Capena. In the excavations made in 1868 and 1869, on the line of the wall of Servius Tullius from the Cœlian to the Aventine, the conduits of three specus were found, two of which must have passed over the arch of the Porta Capena, in order to cross the Via Appia, there a deep foss-way. Two conduits were seen in each of the pits that were dug at intervals along the line, and at the junction with the Piscina Publica under the Aventine they were all three perfect; the lowest one is there cut in the tufa rock under the wall, the other two are on the wall, and partly cut out of it.

Piranesi has preserved a sketch of the specus of the aqueduct, which he supposed to be the Anio Vetus upon its substructure; but he gives no clue as to the exact spot whence that sketch was taken[49]. In its character the masonry is very similar to that of the Marcian, but there are minor differences sufficient to shew that it belongs to an earlier age.

Those who have paid attention to the manner in which ancient books have been transmitted to modern times before the invention of printing, and who are familiar with the use of records and of other medieval manuscripts or transcripts, well know how full of abbreviations they are, and how difficult it often is for the editor to fill up these abbreviations, if he does not happen to know the word intended. The name of specus for the conduit of an aqueduct was essentially a technical word. The first transcribers of the text of Frontinus were not Romans, and did not know the term: hence they filled up the abbreviation spē, or perhaps originally spc̄, with spem or spei, instead of specum or specûs. The same thing may have occurred in the text of Lampridius (as mentioned in a note on a preceding page).

The Abbot of the monastery on Monte Cassino (now a public school and public library) has kindly given me tracings of all the passages in that manuscript in which this abbreviation occurs, and I have had them reproduced by photography and phototype on the page annexed. Opposite to this the same passages are given in the Italic characters, and a few words that are necessary to complete the sense in Roman characters. This is followed by an English translation of these passages, and by some extracts from Livy and other authors in explanation.

FACSIMILE FROM THE MANUSCRIPT AT MONTE CASSINO[50].

FRONTINUS DE AQUÆDUCTIBUS.
EXTRACTS REFERRING TO THE ABBREVIATION SPE̅S̅ FOR SPECUS.

I.

iungitur ei ad s .. em ueterem[51] in confinio ortorum torquatianorum

et ... novum ramus Augustae, hac tres ... ad Viminalem usque

portam deveniunt.

I. 5.

II.

ibi rursus emer

gunt prius tamen pars julie ad spē ... veterem excepta, castelli celii

montis diffunditur.

I. 19.

III.

partem tamen sui claudia prius

in arcus que vocantur neroniani ad spē ueterem transfert hi directe per

Caelium montem juxta templum divi Claudii terminantur.

I. 20.

IV.

rectus vero ductus secundum

spē veniens intra portam exquelinam in altos rivos per urbem diducitur.

I. 21.

V.

ad gemellos tamen que locus infra spē ueterem, ubi jungitur

cum ramo Augustae.

II. 64.

VI.

sed postquam nero im

perator claudiam opere arcuato ascus[52] (?) excepta usque ad templum

divi claudii perduxit, ut inde distribuetur.

II. 76.

VII.

quibus nunc plures aque et imprimis

marcia reddita amplo opere a spē in auentinum usque perducitur.

II. 87.

Translation of the Extracts from the Treatise of Frontinus on the Aqueducts, containing all the passages in which the abbreviation Spē occurs.

I. [The Augustan branch] is joined to it [the Aqua Appia] at the old Specus [or old Spes (?), temple of Hope], in the border of the Torquatian gardens[53]. I. 5.

II. [The three, Marcia, Tepula, Julia] there emerge again; first, however, part (of the water) intercepted from the Julia is poured at the old Specus (or at the old temple of Spes[54]), and so into the reservoirs on the Cœlian Hill, near the temple of the divine Claudius. I. 19.

III. First, however, the Claudia transfers part of its [water] into the old Specus [or at the temple of Spes], on the arches called Neronian[55]. I. 20.

IV. But the direct conduit [of the Marcia, &c.], passing by the old Specus (?) of the Appia (at a higher level, or following the old Hope?)[56], coming within the Porta Esquilina, [the water] is drawn off in the high streams through the City. I. 21.

V. [The Aqua Appia], however, at the Gemelli[57], which place is below the old Specus [or below the old temple of Hope (?)], is joined with the Augustan branch. II. 64.

VI. But after the Emperor Nero had carried the Claudian (water), which he diverted, on arched work to the temple of Claudius [or the Claudium], at the specus[58] (or at the Temple of Hope?), that it might be distributed thence. II. 76.

VII. By which many waters, especially the Marcia, being supplied in great abundance, are conducted from the specus[59] (or from the temple of Hope?) on to the Aventine. II. 87.

Passages in which the Temple of Spes occurs in Livy.

“At the time when this disaster happened, Caius Horatius and Titus Menenius were in the Consulship. Menenius was immediately sent against the Etruscans, elated with their victory. He also was worsted in battle, and the enemy took possession of the Janiculum; nor would the City, which besides the war was distressed also by scarcity, have escaped a siege (the Etruscans having passed the Tiber), had not the Consul Horatius been recalled from the country of the Volscians. So near indeed did the enemy approach to the walls, that first the engagement was at the temple of Spes, in which little was gained on either side; again at the Porta Collina, in which the Romans gained some small advantage, and this, though far from decisive, yet by restoring to the soldiers their former courage, qualified them the better to contend with the enemy in future[60].”

“At Rome a dreadful fire raged during two nights and one day; everything between the Salinæ (or salt wharf) and the Porta Carmentalis was levelled to the ground, as were the Æquimælian and the Jugarian streets. The fire catching the temples of Fortuna, of mater Matuta and of Spes, on the outside of the gate, and spreading to a vast extent, consumed a great number of buildings, both religious and private[61].”

“After this, in pursuance of a decree of the Senate, and an order of the people, an assembly of election was held by the city prætor, in which were created five commissioners for repairing the walls and towers, and two sets of triumvirs: one to search for the effects belonging to the temples, to register the offerings; the other to repair the temples of Fortuna and mater Matuta within the Porta Carmentalis, and likewise that of Spes on the outside of the gate, which had been consumed by fire the year before[62].”

“He agreed with contractors for building a theatre near the Temple of Apollo, and for embellishing the Temple of Jupiter in the Capitol, and the columns around it; he also removed from those columns the statues that stood incommodiously before them, and took down the shields and military ensigns of all sorts which were hung upon them. Marcus Fulvius made contracts for more numerous and more useful works—a haven on the Tiber, and piers for a bridge across it, on which piers Publius Scipio Africanus and Lucius Mummius, censors, many years after, caused the arches to be erected; a court of justice behind the new bankers’ houses, and a fish-market, surrounded with shops for private sales; also a forum and porticus, on the outside of the Porta Trigemina; another porticus behind the dockyard, and one at the Temple of Hercules; also a temple of Apollo Medicus, behind that of Spes, near the bank of the Tiber[63].”

It will be observed that all these passages apply to the well-known Temple of Spes near the bank of the Tiber, of which there are considerable remains now in the church of S. Nicolas in Carcere, and do not apply to a Temple of Spes at the Porta Maggiore.

The word specus is used by Vitruvius in the sense of a covered water-course:—

“But if there should be mounds in the middle between the walls and the fountain-head, it must be so contrived that the water-channel (specus) be dug under the earth, and poised on the top[64].”

And at a later period by Hirtius:—

“Alexandria is almost wholly undermined with water-courses, and has a specus extending to the Nile, by which water is conveyed into private houses[65].”

III. The Aqua Marcia (B.C. 145).

“127 years afterwards, that is, from the building of Rome 608 years, when Servius Sulpicius Galba and Aurelius Cotta were consuls, as the aqueducts or conduits (ductus) of the Appian and Anio were much decayed by age, (and also intercepted fraudulently for private purposes,) the business of repairing and reclaiming the said aqueducts was entrusted by the senate to Marcius, who was then acting as Prætor. And because the increase of the population of the city seemed to demand a more ample supply of water, instructions were given to him by the senate that he should carefully examine how far there were other streams which he might be able to bring into the city[66].

“He therefore restored the two old conduits, and introduced a third, which he caused to be erected with ‘squared stones,’ and larger aqueducts, and carried through them the water which he had obtained for the public service[67]. Hence it received the name of the ‘Marcian’ from himself, as the author of it.

“The Aqua Marcia has its origin on the Via Valeria at the thirty-sixth milestone, three miles off in the diverticulum (or cross-road), on the right-hand to those going from Rome. On the road to Sublacum, now called Subiaco (Via Sublacensis) also (which was paved for the first time under the Emperor Nero), at the thirty-eighth milestone, for the space of two hundred paces on the left-hand side the water lies like a pond, bubbling up in innumerable springs from beneath the stony hollows, and is very green in colour.

“The length of the course from its head to the city is 61 miles, 710 paces;—by an underground channel 54 miles, 247½ paces, on structure above ground 7 miles, 463 paces. Out of this, in many parts away from the city, in the upper part of the valleys, it is carried on arched substructure for 473 paces; nearer the city, from the seventh milestone, on a substructure for 528 paces. In the rest of the work it is carried on an arcade for 6 miles, 472 paces[68].”

“The Marcian ranks fifth in height, and is at its head even in level with the Claudian[69].”

“In the year of the building of the city, 719 (i.e. B.C. 44), Agrippa repaired the three aqueducts—the Appian, the Anio Vetus, and Marcian—and took care to supply the city with many fountains[70].”

“[Temp. Nervæ, (A.D. 96)] the Marcian having been enlarged was carried across from the Cœlian [i.e. its specus] to the Aventine[71].”

IV. The Aqua Tepula (B.C. 126).

“In the year 627,” writes Frontinus, “after the building of the city, when Plautius Hypsæus and Fulvius Flaccus were consuls, the censors, Cneius Servilius Cæpio and Lucius Cassius Longinus, took care to bring into Rome and the Capitol the stream called the Tepulan, from the Lucullan Fields (which some call the Tusculan[72]).

“The Tepula has its source on the Via Latina at the tenth milestone, two miles off on the right of those going from Rome. Thence it was brought by a separate channel into the city[73].”

V. The Aqua Julia (B.C. 34).

“Afterwards Marcus Agrippa collected the natural waters of another stream, at 12 miles from the city on the Via Latina, (2 miles off on the right of those going from Rome,) and so intercepted the stream of the Tepula. To the newly-acquired water the name of Julia was given, from the finder of it; nevertheless the distribution was so divided that the name of Tepula was retained[74].

“The course of the Julia runs for the length of 15 miles, 426½ yards. In work above ground 7 miles; out of this in parts nearest to the city from the seventh milestone (it is carried) on a substructure for 528 yards; the rest on arched work for 6 miles, 472 yards[75].”

III., IV., V. The Aquæ Marcia, Tepula, and Julia.

“Of these [Aquæ], six within the seventh mile, on the Via Latina, are taken up into covered piscinæ, where, as though breathing again after their course, they deposit mud. The Julia, the Marcia, and the Tepula, are joined there; of these, the Tepula, (which had been intercepted, and joined to the stream of the Julia,) now receives from the reservoir of the same Julia its proper quantity, and flows out in its own channel, and under its own name.

“These three are carried from the reservoirs on the same arcade.

“The highest of them is the Julia, lower the Tepula, then the Marcia. These come down towards the Viminal, running together beneath the ground, on the same level as the Collis Viminalis, as far as the gate. There they again emerge.

“First, however, a part of the Julia is, at the Spes Vetus, taken out and distributed in the castella on the Cœlian Hill; but the Marcia, after the Pallantian Gardens, throws off part of its water into a stream called the Herculanean. This conduit through the Cœlian, being of no use for the houses on the hill because at too low a level, comes to an end above the Porta Capena[76].”

The Piscinæ.

To this point, Frontinus tells us, the three aqueducts tend[77], while from this they are carried on the same arcade into Rome. The ruins of these remain visible. Some of them are situated a little way off the south side of the Via Latina, others on the east side of the Via Appia Nova. Of these, two belonging to the Claudia and Anio Novus are subterranean, and are now only to be distinguished as mounds of earth, looking like tumuli. Others are above ground, near that part of the Via Latina [now the road to Frascati and Tusculum], and close to the Torre di mezza via, or half-way house from Rome to Frascati, just beyond the sixth milestone of the modern road. Others are at or near the Villa of the time of Hadrian, called Sette Bassi (which is supposed to be a corruption of Septimius Bassus), near the same point. All of these are between seven and eight miles from the City; they are the chief landmarks in tracing the course of the three aqueducts now to be explained, and each of the three comes from its own separate source. The Marcia, the lowest on the arcade, has its origin at the greatest distance from Rome. The Tepula and Julia have their sources comparatively close to the city. The Marcia takes its rise from the Simbrivine Hills, as far removed beyond Tibur to the east as Tibur is from the city, while the latter two find their way from springs in the volcanic region around the lake of Albano.

III. Marcia.

The source of the Marcia is plainly visible, the exact description of Frontinus pointing to the spot without leaving room for doubt. In one of the numerous little valleys which run down on the north side of the River Anio, feeding this stream with their rivulets, the Marcian has its rise. The lake of S. Lucia[78] in that valley is so called from a small village situated some distance up the slope; along the bottom of this valley the Via Valeria passes. This lake is usually, but erroneously, considered as the source of the Aqua Marcia.

The exact position of the source is about two miles from the village of Marano, but on the other side of the river, on the right of the valley to one looking towards the mountains, and there are still at times pools of water forming here from the springs which emanate from the overhanging hills. The water now falls into two or three rivulets, which run at the bottom of the valley into the river Anio.

The aqueduct, however, when it reached the high road from Rome to Subiaco, along the north bank of the Anio, turned abruptly to the west, followed the course of the road back towards the city, chiefly passing the further or hill side of it, but winding somewhat according to the nature of the ground. After following the road for some seven or eight miles, it crossed the river (close to the monastery of S. Cosimato) and then pursued for some six or seven miles the southern or right bank of the river Anio. Here, at a spot scarcely more than a mile from Tivoli (Tibur), the course of the aqueduct left the line of the river, and wound its way in a south-westerly direction towards the piscinæ before referred to; at times on a substructure, and in a few places, where the valleys were deep, on arches, but for by far the greater distance beneath the surface.

The line of the Marcia can be clearly traced along its whole length. The principal source, called Acqua Serena, is a beautiful spring gushing out from the mountain under the present carriage-road, about seven miles and a-half before arriving at Subiaco, clear as crystal and very abundant; it forms a small lake in the valley, and fully realizes the description of Martial. The old specus, which had long been concealed by being a foot or two under water and overgrown with weeds, was brought to light in 1869 by the engineer of the Aqua Marcia Pia, in making his new works for the restoration of this beautiful water to use in Rome. He drained the lake a little, and by that means made the specus visible[79]: this specus is then carried to the cliff of the valley of the Anio (into which the water from the small lake falls), and then follows the course of the valley and the river, until it comes to the monastery of S. Cosimato, a mile from Vico-Varo: here it crosses the river on an arcade of brick. It there leaves the present road, and then is carried in the cliff on the other side of the river until it arrives within two miles of Tivoli, where it again crosses the river (here a deep ravine) on a fine arcade parallel to that of the Anio Novus, and at a very short distance from it. There are very fine and picturesque ruins of both these arcades in the valley called the Valley of the Arches. It then continued in the cliff by the side of the present road, which is on a ledge of the rock or hill overlooking the Anio, as far as the old town of Tibur and the Cascades, behind the present town of Tivoli: to avoid these it follows a serpentine course, winding round the end of the hill, and has a fine reservoir about half a mile beyond the town, on the side towards Rome, below the level of the Promenade of Carciano, by the side of which are considerable remains of a large reservoir or piscina, with the specus running into it and from it. This large building is divided down the middle by an arcade, into two nearly equal parts. Some of it is faced with Opus Reticulatum of a peculiar pattern. On the exterior, there are remains of niches and fountains, a villa of some importance having been built at this spot to take advantage of the abundant supply of excellent water[80]. The specus then runs near the cliff below the level of the road to another extensive reservoir and piscina about a mile further on. There are considerable remains of a large building erected on a steep slope just under the cliff, with no upper wall, as the cliff itself supplied its place; but there is on the lower side a fine wall of considerable extent and height, built of large square or oblong stones of the construction called Cyclopean Masonry[81], in this instance differing little from that of the Walls of the Kings in Rome, except that the stones are rather larger, the building-material being not tufa, but a kind of calcareous stone of the country, dug on the spot. This wall has been supposed to be a portion of the fortifications of the ancient city of Tibur, but for this there does not appear to be the slightest evidence or probability. It is simply the natural construction of the material at hand, and therefore the cheapest wall that could be built for the purpose. No cement is used, because none was required; these large blocks of stone require none, and some chippings are used to fill up interstices, as usual when no cement is used. Again the wall turns the corner at both ends; it is not part of a large wall but is complete in itself, as the facing of one side of the reservoir and filtering-place, the interior of which is built of the usual concrete of rough stone and mortar, and lined with cement of the kind which holds water, Opus Signinum, used for all the aqueducts. The end is faced with Opus Reticulatum, and there are remains of niches against the wall at intervals. It was also more ornamented, because it was intended to be the one seen chiefly; the other side, being near the edge of a precipice, would only appear from a distance, and the large stones were therefore more effective in that situation. There is no reason to doubt that the whole was built together at the time that the Marcian Aqueduct was made, and it was probably restored by Augustus. It may be doubtful whether some caves in the cliff, which formed part of the reservoir, were natural or were cut, and the stone dug out from them. It is altogether a very picturesque and interesting structure. There may be a question also whether there was not a branch from this reservoir to the Villa of Hadrian at the foot of the hill, perhaps a mile lower down.

The specus of the Marcia is visible at the Ponte di S. Antonio over that of the Anio Vetus; and again, but alone, at the Ponte di S. Pietro, and it passes the Ponte Lupo with the Anio Novus, Claudia and Anio Vetus. It then continues, chiefly underground, to the great piscinæ before mentioned, and thence on the arcade into Rome.

After reaching the City, the branch mentioned by Frontinus, c. 87, as being carried across the valley from the Cœlian to the Aventine in the time of Nerva, seems to be the one found during the excavation in 1868, passing over the Porta Capena at a higher level than the Appia, but still at a much lower level than the lofty arcade of Trajan, of which only the bases of the series of piers crossing the valley now remain. The aqueducts following this line had all to cross the Via Appia, here a foss-way, on the arch of the Porta Capena. This southern branch of the Aqua Marcia is probably the one that can be traced along the side of the cliff of the Pseudo-Aventine, after it had been repaired and brought again into use, for it must be remembered that at one time, as Frontinus says, it ended at the Porta Capena.

The conduit and arcade were rebuilt by Augustus, as recorded on the inscription on his arch at the Porta S. Lorenzo, and in the sixth decree of the Senate, on the subject of the aqueducts. This branch arcade of Augustus is there expressly named as distinct from the others, all needing repairs at the time of this edict. The consuls are charged to see to “the repairs, at the expense of the city, of the streams, conduits, and arches, of the Julia, Marcia, Tepula, Anio; also of those streams and arcades which Augustus Cæsar had rebuilt.”

A branch of the Marcian aqueduct was carried along the agger into the Prætorian Camp. Some leaden pipes were found there in 1742, with an inscription upon them, recording that they were of the time of the Emperor Macrinus (A.D. 217). This probably indicates either a renewal of the pipes, or an additional supply of water. The garrison in that camp was twelve thousand men, and a large supply of water must have been required for their use[82].

The excellent qualities of the Marcian water are mentioned by several of the classical authors, and were celebrated for a long period; they were known in England in the time of Shakespeare, as appears from the following passage in Coriolanus, Act ii. Sc. 3—

Brutus loquitur.—“What stock he springs of,

The noble house of the Marcians; from whence came

That Ancus Marcius, Numa’s daughter’s son,

Who, after great Hostilius, here was king:

Of the same house Publius and Quintus were,

That our best water brought by conduits hither;

And Censorinus, darling of the people,

And nobly named so, being censor twice,

Was his great ancestor.”

There is a slight anachronism here. Coriolanus lived in B.C. 489, and the Marcian aqueduct was constructed in B.C. 145, more than three hundred years after his death. Marcius Censorinus also, who was twice Censor (the only Roman who filled that office twice), lived B.C. 294.

In the thirteenth century, the church of S. Bibiana is incidentally mentioned as being near the arcade of the Marcian aqueduct[83]. In some excavations made in the year 1871, a portion of the Marcian arcade, built of the usual large squared stones, was shewn, passing under some high ground to the north of the Porta Maggiore, within the wall of Aurelian, in a direction to join the bank on which that wall is built, and passing between the Minerva Medica and the wall in that part. This is very near the church of S. Bibiana.

IV. Tepula.

The sources of the Tepula and the Julia are in the valleys on either side of the promontory on which the modern town of Marino stands, the ancient Castrimœnium (under the village of Rocca di Papa). That of the Julia is on the south side, and almost close under the crater now the Alban Lake. The Tepula rises near the bottom of the valley which comes down from the hills in the neighbourhood of Grottaferrata, and along which the Via Latina passes. This lies somewhat nearer to Rome than the source of the Julia; but the Julia joined it before it had advanced far, and thus the expression of Frontinus, “Marcus Agrippa intercepted the Tepula[84],” is explained. The waters, therefore, flow into the same series of reservoirs and cisterns which received the Marcian after they enter Rome.

Under the account of the Julia, the Aqua Crabra is mentioned. This is the little stream into which the water of the Julia and Tepula falls; which is united at the foot of the hill on which the town of Marino stands, to another stream called the Marrana, and the united water is now generally called by the latter name only[85]. Frontinus mentions several reasons why it was not made use of for supplying the city with water; but it was brought into Rome in the twelfth century as a small mill-stream.

Frontinus also says that the Tepula had its source at ten miles from Rome on the Via Latina, with two more miles to the right on a cross-road[86]. Ten miles on the Via Latina brings us near to Tusculum, at the tenth mile is the Casino de Ciampino, from which starts a cross-road to the right; and at two miles from that point we arrive at the springs called Fontanaccio, before reaching Grotta Ferrata, but close under that village. It follows that the source of the Tepula was at this place, now called Fontanaccio. The spring comes out under a cliff of the rock of lava near the road, and has a modern washing-cistern in front of it; but behind this the ancient work can be seen, with openings into a reservoir in the cliff. This is probably contemporaneous with the time that the conduit of the Tepula was made. The supply of water is small, but of good quality.

As the Aqua Tepula supplied only the Regiones in the northern part of the city, it seems to have passed into the castellum, the remains of which may still be seen in the city wall near the Porta S. Lorenzo, evidently built upon an old agger before the Aurelian wall was erected[87]. It, however, has been a house as well, the lower part only being used as the reservoir for the water, and the upper part, which is large and important, for chambers. The front of this house, or castellum aquæ, still forms part of the city wall. It has been much disfigured, and the old drains walled up during the restorations (!) of 1869.

On the level of the first floor, in which is part of the reservoir, is a row of corbels to carry a wooden gallery or hourd, probably an external passage for the use of the Aquarii, behind which the specus runs at the same level. Immediately above the line of the corbels which carried the floor of the gallery or balcony, is a row of large arches; but these are merely the arches of construction found in most walls of the period. At the south end of this line of corbels, which mark the extent of the castellum in that direction, there is an angle, and the wall recedes a few feet. In this angle is the specus, corresponding in form and dimensions with that of the Aqua Tepula in other parts; it can be seen entering into the reservoir behind the line of corbels. Within the wall are the usual marks of a reservoir of water: the tartar deposit remains visible in the corners of the chambers cut through by the wall of Aurelian, or by the engineers of the Acqua Felice, whose specus runs behind it on the bank within the wall which formed the front of the house. The specus of the Acqua Felice is here at rather a higher level than the Marcian arcade, with the three specus upon it, although at the “Sette Bassi,” five miles from Rome, and near the piscinæ, it is at a lower level. On the bank just above the level of the ground are the water-drains (hidden by the restorers in 1869) for carrying off the superfluous water into a large subterranean drain which runs under the gateway, and which is still in use for purposes of irrigation.

V. Julia.

Frontinus states that the source of the Julia was at twelve miles on the Via Latina, with two more miles added on a cross-road to the right[88]. The twelfth mile is at Frascati; from thence, by a cross-road to the right, we arrive at the bridge of the Squaricarelli, and at the copious springs called the Fontanile, exactly two miles from the starting-point at Frascati. This must, therefore, be the source of the Aqua Julia.

The swampy ground in which the source of the Angelosa or Aqua Julia is found, is full of springs, like the Lucullan fields, from which flow the Appia and the Virgo, and from it run also the streams called di Monte Fiore and la Marrana di Marino.

This ground is on a high level on the Monte Fiore e dell Aglio (the Mons Algidus of the ancients), or “The hill of flowers or of garlic;” and the springs come from the gardens of the modern Villa Aldobrandini, at Frascati. Canina cleared out the ancient specus, which had a curved vault, almost oval (a capanna).

This source of the Julia is on the left-hand side of the road from Grotta Ferrata to Marino, about a mile above the former, and nearly the same distance below the latter. The water gushes out from the foot of the rock and passes under the road to a lavacrum, or washing-place, at a lower level on the opposite side of the road, and then falls into the Aqua Crabra, which passes at the foot of the cliff many feet lower down. A part of the water of the Julia, before it goes into the lavacrum, is carried to the right in a specus to Grotta Ferrata. This is ancient, but has been restored to use, so that it looks modern. Below Grotta Ferrata, the specus has been destroyed in many parts; but remains of it may be seen close to the ancient fortified villages called Pagus Lemonius (on the maps Castellaccio, a small castle). It is ten miles on a branch of the ancient Via Latina, and between the present roads to Frascati and Grotta Ferrata. In a part of the fortifications of this village is a castellum aquæ, or reservoir of the Julia, and near it part of the specus, on an arcade on the brow of the hill. This specus is built of rough stone, and faced with a rude early kind of Opus Reticulatum, more rude indeed than might have been expected at its date (B.C. 34), and not nearly such good work as the Muro Torto, but more like the Emporium. The reservoir is of the same character, but part of the specus in the fortifications is carried on brick arches which agree with that date; the brickwork is not of the time of Nero, to whom it has been attributed. A specus of rough stone follows the line of the hill to the Marrana, after the junction of that stream with the Aqua Crabra, close to the point where the water from that stream enters a tunnel of the Aqua Julia. After it emerges from this tunnel, it arrives at the castellum or piscina in which the water of the Julia was received before it was carried on the Marcian arcade. This piscina is about two miles nearer to Rome, close to the point of junction between the roads to Frascati and Marino, through Grotta Ferrata, and near to the other piscinæ.

Frontinus says that the Aqua Crabra flowed past the head of the Julia, but was excluded from it by Agrippa[89]. We find this stream coming from that part of the hill on which the village called Rocca di Papa is situated, passing near the head of the Julia at Fontanile, now Angelosa, and sometimes mixing with the springs there. It has one of its sources in the grounds of the Villa Torlonia, at Frascati. All these three streams now fall into the river Anio.

Near Rocca di Papa an arcade, or bridge of ten arches, of silex or flint, covered with brick, passes over the Aqua Crabra to carry some aqueduct, probably a part of the Julia. This arcade is called Arcioni, [‘the arches.’]

There are some remains of another reservoir within the wall of Rome, on the bank between this and the Porta di S. Lorenzo, which seems to have belonged to the Julia: the specus of the Marcia, with that of the Tepula and remains of that of the Julia over it, are visible on an arcade a little beyond, close to the gate. The lower part of a fine brick wall with bold buttresses, of the usual character of a piscina or castellum aquæ, are visible, and this faces the present Wall of Rome, almost touching it, so that this has been built since the piscina, but that not being exactly in the same line it could not be used. The piscina of the Tepula being on the outer side of the bank or mœnia came into the line of defence of Aurelian, and was used as part of his wall; that of the Julia, being on the inner side of the bank, was too much within the line to be used.

The Castellum Aquæ Juliæ, or chief reservoir of the Julian Aqueduct, is usually, but erroneously, said to have been the one situated on high ground near the church of S. Maria Maggiore, where the picturesque ruins stand, usually called the “Trophies of Marius,” because those trophies were hung under two of the arches there, until they were removed to the Capitol. A lofty arcade carrying a specus at a high level passes across the valley, from the reservoir near the Porta S. Lorenzo to this point. This arcade is of the first century, built of the fine brickwork usual at that period, and agrees with the time of Frontinus; but the only aqueduct that was high enough to have carried water to that spot was the Anio Novus.

This great reservoir was rebuilt by Alexander Severus, and called a Nymphæum, being given as such on one of his coins. This name is rather a vague one; but it is evident, from some excavations made in 1871, that there were extensive thermæ of the Emperors of the third century near this place.

From this lofty reservoir the distributing channels may be seen branching off in different directions, one going to the Thermæ of Titus, towards the reservoir called the Sette Sale, another going in a different direction.

This Nymphæum (?) is on lower ground than the outer wall, though still on ground so high that the only water that could reach it was that of the Claudia and Anio Novus, of which part of the arcade remains between it and the reservoir on the high bank. From that point the ground descends gradually and gently along the line of the agger of Servius Tullius either way, and still more towards the interior of the city. The aqueduct, therefore, could not pass underground and then emerge in this part. The Porta Esquilina of the inner wall was on the same level, or nearly so. The Porta Tiburtina of the inner wall was near the thermæ of Diocletian, with a gradual and gentle descent to it along the line of the agger. The only possible explanation of the text of Frontinus is that the same names of the gates in the inner wall were applied to those in the outer wall on the same roads.

A piece of leaden pipe, with an inscription upon it, was found at the Porta S. Lorenzo, which gives the names of Dolabella and Silanus as Consuls; this fixes the date at A.D. 10, and shews that the Marcian water was conveyed in leaden pipes at that point in the time of Augustus. The stone specus was carried over the gate, as we see by the remains of it in the wall, and the inscription upon it recording repairs by Augustus. The one upon the leaden pipe is given by Gruter[90].

Another leaden pipe was found by Panvinius on the site of the Prætorian Camp, with an inscription, given by Gruter[91], which records that it conveyed the Marcian water to that point.

Three aqueducts are mentioned in other inscriptions found during the excavations near the railway station in 1869[92]; the three aqueducts intended can hardly be other than the Marcia, Tepula, and Julia. The upper one only was excavated; but if the noble Roman princes who conducted this excavation had dug a little deeper, they would probably have found the other two under the one that they had discovered. These three specus must have been carried along and in the agger of Servius Tullius.

To reach that part of it on the side of which these cippi are found, they probably went along the side of the road that passed on the south side of the Prætorian Camp, on which was the gate called Porta Chiusa. Close to this gate remains of a large reservoir were found in the researches of 1869 on the bank, with the Wall of Rome carried right across it; but this part of the wall is medieval, built of old materials, and it is probable that, originally, the front wall of this reservoir was included as part of the wall of the city, in the same manner as that of the Tepula near the Porta S. Lorenzo. This reservoir, with the Porta Chiusa by the side of it, and the road from it to the agger, at the place where the cippi were found, is now in the garden of the Baron Grazioli.

VI. Aqua Virgo (B.C. 21).

This Aqueduct was made by Marcus Vipsanius Agrippa to supply water to his Thermæ, on the south side of the Pantheon, which was the hall of entrance to them.

Frontinus writes:—

“The same (M. Agrippa) when he had been consul for the third time, and when C. Sentius and Quintus Lucretius were consuls, that is, thirteen years after he had brought down the Julia, he brought the ‘Virgo’ also, the water of which was collected in the Lucullan fields[93].

“The Virgo begins on the Via Collatia, at the eighth milestone, in some marshy places: a cemented wall [signino[94] circumjecto] being placed round it, to retain the bubbling waters; the source was increased by many other additional supplies. It comes for a length of 14 miles and 105 paces. Out of this,—by a subterranean stream 12 miles, 865 paces; above ground for 1 mile, 240 paces, of which it runs on a substructure in several places for a distance altogether of 540 paces; on arched work 700 paces. The channels of the additional supplies of the subterranean stream make 1 mile, 405 paces[95].”

The Virgo has no reservoir, i.e. Piscina ... “The arches of the Virgo have their commencement beneath the Lucilian (or Lucullan?) gardens. They end in the Campus Martius, along the front of the Septa[96].”

The Virgo was the seventh in height as to level (c. 18).

The road now called Via Collatia, or Collatina, where the Aqua Virgo has its origin, is between the present roads to Tivoli and Palestrina; it turns off to the left or north of the highway now called Via Prænestina, just beyond the ruins called Torre de’ Scavi or de’ Schiavi, about three miles from the city. But this is an alteration; the old road from Rome to this point in a foss-way may be traced behind the tower in the meadows, with the subterranean aqueduct running under and in the southern bank of it, as traced by the respirators or ventilating shafts. There are remains of very ancient tombs at intervals all along the line of this old road, which goes straight towards a postern-gate midway between the Porta Maggiore and the Porta di S. Lorenzo. This must be the Via Collatia of Frontinus.

Close to this road, at eight miles from Rome, near the ruins of Collatia, there are several springs in marshy ground, from which the water is collected in a series of reservoirs, just below the level of the ground, the vaults over them being sometimes above ground; from these small reservoirs separate short conduits run to the same point, the great central castellum aquæ, part of which is a cave in a rock, with a larger semicircular basin built out in front of it, over which passes the cross-road from the Via Collatia to Salone.

The line of the specus can be distinctly traced from this reservoir near its source to the city by the respirators over the wells. The smaller castella, or cisterns, at the sources are also still in use.

About four miles from Rome, on the road now called Via Collatia, or Collatina, a portion of the Aqueduct was rebuilt under Benedict XIV. in 1753, as stated on an inscription upon it. It is here carried across a shallow valley, on an agger built of rough concrete faced with brick, for about a quarter of a mile. At the further end is a conduit-head, or a modern castellum aquæ. This is a small oblong building with a semicircular head erected under Pius VI. in 1788, by Joseph Vai; and the conduit was continued by him for the length of 175 feet, also stated in an inscription upon it[97].

From this point there is a branch to the south for a short distance, with two respirators visible in the field; but this, though appearing like an additional source, is only a branch leading to an old fountain of a villa.

The main line comes from the north along the bank of the Via Collatia, as shewn by the respirators, at regular intervals of about a hundred yards. These are mostly round masses of concrete with round heads; but some are dwarf pyramids, one of which, numbered “40,” has been rebuilt in 1866. The surplus water is carried into a brook which runs by it, and receives its chief supply from it.

The specus or conduit of the Aqua Virgo passes along the line of the old road in a direct line west, towards the Porta Maggiore, until within about half a mile from that gate; then it makes a great detour to the north, passing under the modern Via Tiburtina, and eventually enters Rome through the Pincian Hill a little to the north of the Spanish steps, and there is a reservoir for it at the end of a short street called the Via del Bottino. It then goes to the present fountain of Trevi, passing at the back of the houses in the Via del Nazareno, where it may be seen in several of the courtyards. In one of these, on the left hand of the street, is an inscription recording its repair by Claudius[98]; in another, on the right, is an ancient lavatory below the level of the street. A branch passes under the Via dei Condotti carried in leaden pipes, enclosed in a brick specus; this branch is the one that led to the Thermæ of Alexander Severus, which were situated to the north of the Pantheon. The main line supplied those of Agrippa, for which this aqueduct was made. The Aqua Virgo chiefly supplies the fountains and houses in the Campus Martius, or lower city, and the main stream terminated originally in front of the Septa[99], considerably to the south-west of that fountain. In some excavations made in the summer of 1871, a portion of it was found in the Piazza di S. Ignazio. This original termination was at the north end of the Septa, very near the Pantheon.

This aqueduct was restored by Pope Hadrian I., A.D. 772-795, after it had been damaged by the Goths. It was afterwards repaired by several subsequent pontiffs, especially by Boniface IX., A.D. 1389, and by Nicholas V. in the fifteenth century. It now supplies the fountains in the Piazza Colonna, erected in 1574; at the Pantheon, restored in 1711; several others of the seventeenth century; and that of the Trevi, erected by Benedict XIV. in 1730.

Near the source of the Virgo passes an abundant stream, also called Marrana; but this has mineral properties, and was therefore carefully avoided in forming the aqueduct for the Virgo. Pliny says[100] that there was a stream which was abhorred by the Virgo, and for that reason the water was called the Virgin. This stream was called by him Rivus Herculaneus, some say for its salubrious qualities, because Hercules was the god of health; others because it was a strong stream. There is little doubt that this is the stream here described. The same name of Herculanean is given by Frontinus to the stream now called Marrana[101] in two places, because it was also a strong stream.

VII. The Alsietina (A.D. 10), Afterwards made the Aqua Paola.

Frontinus says of this water, “What could have induced Augustus, that most careful of emperors, to bring in the water of the Alsietina (which is called Augusta) I do not well know; for it is not pleasant to the taste, and therefore of no use for the people. It may be, however, that when the work of the Naumachia approached completion, in order not to divert the more wholesome water, he introduced this for the special purpose, and gave the surplus to the adjacent gardens and territories.

“It is the custom, however, in the Transtiberine Regio, when the bridges require mending, and there is no water forthcoming from the city side, to make use of this for supplying the public springs, as a matter of necessity.

“It begins in the Alsietine lake on the Via Claudia, at the fourteenth milestone, about six miles and a-half off on the right hand. Its course is in length 22 miles, 172 paces, and over arched work 358 paces[102].”

“It is the lowest of all as regards the level, supplying only the Transtiberine Regio, and the places adjacent[103].”

“The manner of beginning the Alsietine aqueduct is not described in the commentaries, nor can it now be found with certainty; it begins from the Alsietine lake, and then about the Cariæ receives water from the Sabatine lake also, as the Aquarii regulate. The Alsietine gives 392 quinariæ[104].”

The lake formerly called “Lacus Alsietina,” now called Lago di Martignano, is situated on the hills on the western side of Rome, between the Via Aurelia and the Via Claudia, (not far from the old carriage-road from Florence to Rome). It is about 679 ft. above the level of the sea; and as Rome is only 204 ft. at the Porta Maggiore, the lake is 475 ft. above the level of Rome at that high point. There is another small lake about a quarter of a mile from, and a little above the level of, the Alsietina, called Strachia Capra, which has recently been drained by a tunnel in imitation of the emissario from the lake of Albano. The water in the Lacus Alsietina has also been very much lowered in the same manner[105]. In consequence of this reducing of the level of the water, the specus of the aqueducts are now brought to view, being on the bank and above the present level of the water, so that they can distinctly be seen and entered into[106].

To begin with the highest, the specus of the Aqua Paola begins in the upper lake, and forms a junction with the water from the Alsietina Lake, near the bank of the latter, at the south end of that lake. The specus of the Alsietina of Augustus is a tunnel cut in the rock[107]. The Lacus Sabatina[108] is at the elevation of about 523 ft., and therefore 156 ft. below the Alsietina. Augustus drew an additional supply of water from this lake, and this portion only of the aqueduct of Augustus was restored by Trajan. The water for this was not drawn from the lake itself, but from the springs that supply the lake on the western shore, the side most distant from Rome. Some of the work of the time of Trajan is visible there. The specus of the Alsietina has been traced to the point of junction with the Aqua Paola (this water having been restored to use by Pope Paul III., A.D. 1540), and it still supplies the district of the Trastevere. It does not appear that Trajan used the water of the Alsietina at all, but the engineers of Pope Paul evidently did so; a tunnel, or specus, from that lake having been made in the time of Pope Paul, the entrance to which remains, with the grooves for a flood-gate[109]. After the junction with the Aqua Paola, the old wells can no longer be traced; their place is supplied by the constructions called respirators, each of which is a small square structure surmounted by a pyramid. These are evidently built over the wells, and very rudely constructed of the stone of the country, worked rough. When we arrive at the Osteria Nuova,—which is near the site of the ancient city of Cariæ[110], mentioned by Frontinus, and the old village of S. Maria in Celsano, the point of junction of the Alsietina and Sabatina,—there are evident marks of another junction of aqueducts there. The house itself, now used as an osteria or hostelry (auberge), is made out of an ancient Castellum Aquæ, as is frequently the case with similar houses in the country round Rome, and sometimes even within the walls. At about half a mile from this osteria, in a hollow, is a white house called Casale Bianco; and close to this is a fountain, supplied by a spring which runs into a specus. This water, being one of the sources of the Aqua Paola, passes through the hill at a considerable depth. At about half the distance between this and the osteria is a remarkable passage for the Aquarii into the tunnel by a very steep descent, passing sideways across the specus of the Aqua Paola, and going on to a much greater depth to the Aqua Alsietina[111]. This passage is 150 ft. in length from the surface of the hill, and 70 ft. in perpendicular depth, with ninety steps, of which only a few at the top are visible; the others are covered with earth. The water still runs through the upper specus, and is still in use, but not through the lower one[112].

From the valley beneath the sloping passage for the Aquarii, near the Osteria Nuova, Nibby[113] traced the Alsietina as a tunnel cut in the tufa rock, a part of which he saw near an oil mill. The specus then followed the low ground from the tenements of S. Niccola, Porcareccina, Maglianella, and from the Villa Panfili, to the principal gate of the monastery of S. Cosimato (or SS. Cosmas and Damian in the Trastevere), where it is stated by Cassio[114] that the specus was found in 1720, about thirty feet underground. The Naumachia of Augustus is said to have been near this monastery.

The complaints of Frontinus appear to have been listened to by the Emperor, and great changes and improvements were made in this aqueduct under his direction, in the time of Trajan; this is in fact the same as the Aqua Trajana (see X.), and the water still comes from the Lacus Sabatina, but not from the Alsietina. In going from Rome, the respirators can be followed across Monte Mario, and near the high road that passes over Ponte Molli, to the point of junction about ten miles on the road to the Cariæ [now Osteria Nuova]. Here the respirators cease; but their place is supplied by a line of old wells descending into the subterranean specus, which follows the line of the old road. This is not always the same as the new one. The other branch, which supplied the Naumachia, was the only one made in the time of Augustus. Paul V. repaired this aqueduct along the whole line, restored it to use, and put up in various places inscriptions recording this, in which he calls it the Aqua Alsietina[115]. The last of these is on his fountain at the mouth of the aqueduct, where the water still gushes out in great abundance, as it did in the sixth century, when it was observed by Procopius; this is on the Janiculum, above S. Pietro in Montorio, the highest ground in Rome.

VIII., IX. The Claudia and Anio Novus (A.D. 52).

“Afterwards Caius Cæsar, (Caligula,) who succeeded Tiberius, considered seven aqueducts scarcely sufficient for public purposes and private amusements, and began two new aqueducts in the second year of his rule as Emperor, when Aquilius Julianus and P. Nonius Asprenas were consuls (i.e. A.D. 38), in the year of Rome 789, which work Claudius in a most splendid manner finished and dedicated, on the calends of August, in the year of the city 803 (i.e. A.D. 52), when Sulla and Titianus were consuls[116].

“To one, which was brought from the springs of Cæruleus and Curtius, the name of Claudia was given. This one is next in order of excellence to the Marcian.

“The other, because two streams of the Anio had begun to flow into the city, so that they should be more easily distinguished by their names, began to be called Anio Novus. It ruins all the others[117]. To the former Anio the cognomen of Vetus was added[118].

“The Anio Novus and Claudia are carried from the piscinæ upon higher[119] arches, so that the Anio is the highest of the two. Their arches come to an end after the Pallantian Gardens, and thence they are carried down in pipes for the use of the city.

“But first of all the Claudia transfers a part of its water on to the arches which are called the Neronian, at the Spes (Specus) Vetus. These, being continued in a direct line along the Mons Cœlius, are terminated close to the temple of Claudius. They disperse the quantity which they had received either about the ‘Mons Cœlius’ itself, or in the Palatine, in the Aventine and the Transtiberine Region[120].”

VIII. The Claudia.

“The Claudia begins on the Via Sublacensis[121], at the thirty-eighth milestone, about 300 yards off the road towards the left. There are two very large and beautiful springs, one called Cæruleus, from its blue appearance, the other Curtius[122]. It receives also the spring which is called Albudinus, of such excellence, that when there is need of adding it to the Marcia, the latter loses none of its quality by the addition. The spring of the Aqua Augusta, because the Marcian seemed to be sufficient for itself, was turned aside into the Claudian; nevertheless it was retained as a protection for the Marcian, but so that the Augustan might be added to the Claudian, if the channel of the Marcian was not capable of receiving it[123].

“The channel of the Claudia is 46 miles, 406 paces in length; of this, 36 miles, 230 paces is by a subterranean course: on work above ground 10 miles, 176 paces, and out of this on arched work, in many places in the upper part, 3 miles, 76 paces; and near the city from the seventh milestone, by a substructure of channels for 609 paces, on arched work 6 miles, 491 paces[124].

“The Claudian was the second in height as to level[125].”

IX. The Anio Novus.

“The Anio Novus, at the forty-second milestone, on the Via Sublacensis, at Simbruinum[126], is taken out of the river, which, since it has about it cultivated land in a rich territory, and so very loose banks, flows muddy and turbid even when uninfluenced by violent rains; and therefore at the very entrance of the channel is placed a cistern for the mud (piscina limaria), so that the water on its way from the river to the specus should settle and become clear. From this cause also when heavy showers come down, the water flows into the city in a muddy state.

“There is joined to it the Rivus Herculaneus[127] which rises on the same road at the thirty-eighth milestone, in the same neighbourhood as the springs of the Claudian, on the other side of the river and the road. This is very pure by nature, but when mixed it loses the advantage of its freshness.

“The channel of the Anio Novus is in length 58 miles, 700 paces. Out of this 49 miles, 300 paces is by a subterranean channel;—on work above ground 9 miles, 400 paces; out of this on substructure or on arched work in several places in the upper part 2 miles, 300 paces, and nearer the city, from the seventh milestone, on a substructure of channels 609 paces, on arched work 6 miles, 491 paces. These are the highest arches, elevated in some places 109 feet[128].

“Nor was it enough for our Emperor [Nerva] to have restored an abundant and pleasant supply of water in the other aqueducts; he thought he saw his way to getting rid of the bad qualities even of the Anio Novus, he therefore ordered the source to be changed from the river, which was now left alone, and taken instead from the lock in which the water was most pure, and which is situated above Nero’s villa on the lake (super villam Neronianam Sublaquensem[129]);” i.e. at Subiaco, now a medieval castle and a modern town.

“But the water of the Anio Novus often spoilt the rest, for since it was the highest as to level, and held the first rank as to abundance, it was most often made use of to help the others when they failed. The stupidity, indeed, of the Aquarii was such that they introduced this water into the channels of several others where there was no need, and spoilt water which was flowing in abundance without it. This was the case especially as regards the Claudia, which came all the way for many miles in its own channel perfectly pure, but when it reached Rome, and was mixed with the Anio, lost all its purity. And thus it happened that most of the streams were not in fact helped at all by the addition of the extra water, through the want of care on the part of those who distributed it[130].”

The Neronian Arches.

“Amongst those abuses which seemed to require reform, may be mentioned what took place regarding the supply to the Cœlian and Aventine hills. These hills, before the Claudian was supplied to them, were accustomed to use the Marcian and Julian; nevertheless, when the Emperor Nero gave them the Claudian raised to a greater height on the series of arches extending to the Temple of Claudius, where it was distributed, the older streams, instead of yielding an increased supply, were lost altogether. He made no new castella [for the Claudian], but used those which were there, and which retained their names, although the water brought to them was different[131].”

The River Anio.

The highest source of the river Anio, or Aniene, is in a gorge in the highest part of Mount Cantaro, about 63 miles from Rome. This spring never fails in the hottest and driest weather, and is always cool; it is situated in the close or (serra) of St. Antonio, and is called La Canala, or the canal at the gate of the small castle or fortified village of Filettino. Pliny says[132] that it rises in the territory of Treba[133]. In going from Filettino towards Trevi the road has an ancient pavement, and on the left-hand side is a magnificent substructure of Cyclopean character called Mura Saracine, against the cliff. In 1857 Signor Gori saw some of the large stones of this fine ancient substructure thrown down to the banks of the stream, to make the foundations of a small bridge across it. The stream receives several accessions from other springs in its course, two from a cave called Pertusu, the small limpid stream called Suria, on the hill of Trevi, and another called Capo d’Acqua. The Anio thus augmented falls in two cataracts, or cascades, near Pertusu, and passes under the Ponte delle Tartare with much violence. This bridge is a natural arch of rock formed by the force of the water piercing it. The river then passes on through a gorge in the mountain-pass in small waterfalls, especially at a place called Pendema, where in 1855 an ancient mosaic pavement and a wall of reticulated masonry were found.

At the two bridges of Communacchio (comune acqua), where an amphitheatre of mountains and the castle of Valle-pietra are situated, another stream coming from the mountains of the Trinitá and Autore, the highest in that district, joins the Anio.

In the basin of Valle-Pietra, two fine cascades fall over the massive rocks and unite in a single stream. From the bridge of Communacchio a road leads to Arcinazzo, situated in a large plain, surrounded by the mountains, with several roads leading into it. One of these from Palestrina, passing by Piglio, is an ancient paved road, and some persons consider that an arcade of Cyclopean masonry, near Guarcino, carried an aqueduct to the thermæ there. On the south-west side of this plain are the ruins of an imperial villa, commonly called of Nero; but two inscriptions found there upon leaden pipes in 1860, shew that it was of Trajan[134]. Great quantities of marble were dug up here in the time of Pius VI., A.D. 1795.

The river passes below the monastery of S. Benedict, called the Sacro Speco, and a little lower down, the monastery of S. Scholastica, both of which stand on the brink of a precipice, over which the Anio makes a large and picturesque waterfall, and passes under the bridge of S. Mauro or Piedilago, where the rocks approach so closely as to leave only a narrow passage, which was formerly closed by a gigantic wall, forming a long lake or lock. The stream now passes through the ruins of the wall, the demolition of which was caused in 1305 by two ignorant monks, who pierced a hole at the foot of the wall, the result being that the whole country was inundated up to the walls of Rome, and serious mischief done, as recorded in the anonymous chronicle of the monastery[135].

On the banks of the lake are the ruins of a villa of early character, called Casa de’ Saraceni and Carceri, with a nymphæum and baths. At Pianigliu, are other ruins on an enormous scale.

Subiaco is more than two thousand feet above the level of Rome, and all the foregoing description applies to the part of the river above Subiaco. In all this upper part of the stream, the river Anio has very much the character of a large mountain-torrent rushing through the rocks with great violence. It usually appears to be a clear stream of beautiful water, excepting in time of floods, when the clayey soil is brought down into the stream and makes it muddy. To guard against this evil, Nero made his great lakes, which do not correspond at all to the usual English idea of a lake, by which is commonly understood a considerable sheet of water, like the Swiss lakes or the lakes of Westmoreland, Scotland, and Wales. The Roman word lacus, and the modern Italian lago, may mean a lake of this description also; but it includes a reservoir of water of any kind. We are expressly told in the Breviarium, at the end of the Catalogue of the Regionaries, that a lacus is a well, puteus, and the numerous lacus of that Catalogue are the same as the castella aquarum of Frontinus, within the walls of Rome. These justly-celebrated lakes of Nero are in fact portions of the river Anio, intercepted in a gorge of the rocks about a mile above Subiaco, and are formed by cutting away some large pieces of the rock on each side in large masses, and with these building a great high and massive wall across the stream, forming an effectual barricade or dam to stop the water and raise it to the level of the top of this great wall. There were three of these walls across the stream, over each of which the river fell in tremendous cascades. The first loch (lacus) commenced at the Mola di Ienne, where the first great wall of enclosure is situated. The second at the cascade of the river, under the great monastery of S. Benedict, and called “the Sacred Specus,” from a cave in which the saint is said to have lived, extends to the bridge called Ponte di S. Mauro, where the specus of Trajan commenced. The third is from the Ponte di S. Mauro, through a gorge, where the wall of enclosure is visible. Here was the Piscina Limaria of Claudius, and the specus of the Anio Novus originally commenced at the Emissarium, restored by Cardinal Barberini. This was made by Claudius, but was abandoned by Trajan, because the earth from the adjoining fields had fallen between the river and the specus.

The Villa Sublacensis of Nero was below the level of the upper lake or loch, as mentioned by Frontinus. This was just above the present bridge called S. Mauro or Piè-di-lago, which seems to be made upon the two ends of the dam. This dam, wall, or barricade, was quite 150 feet above the surface of the water in ordinary times. The bridge is still 144 feet above the water by measurement, and the specus is nearly ten feet higher. This specus or conduit is cut in the rock of the cliff on the side of the valley, at the level at which the water originally stood in this lake or loch; the wall was a few feet higher, in order to force a portion of the water to pass through the specus before the rest fell over the cascade. There are ruins of the piscinæ on the bank where the specus began, and of the villa of Trajan on both sides of the loch, with the well-known brickwork and Opus Reticulatum as facings for the walls. These magnificent cascades still remain, being a natural formation; but as the bed of the river is very deep, they are much concealed by the banks, and the shrubs upon them. The great walls or dams of Nero and Trajan being brought out to the edge, the cascades falling over them must have had a much finer effect, although the natural site is extremely grand and most picturesque; in fact it is celebrated among all the landscape painters of Italy, the scenery about Subiaco being among the finest of its kind of river and mountain scenery that is known. The enormous reservoirs or lakes, or lochs of Caligula and Claudius (commonly called of Nero), cover the space between the natural cascade and the outer wall of rock artificially constructed, over which the water was made to fall. When the ignorant monks in the fourteenth century made an aperture in the great wall or dam, the force of the water soon enlarged it, and washed the whole structure away, leaving the great masses of stone or rock, of which it had been built, scattered in the stream below as if they were natural rocks, where they still remain. The object of the monks was to release their fields adjoining to the monastery above the falls from a temporary flood.

The specus is nearly six feet high, and only sixteen inches wide; the men must have cut it standing sideways. There are apertures into it at intervals now open, and there probably always were such openings for the use of the aquarii to keep the course clear. The specus which Frontinus calls subterraneus, although that is literally true, does not mean exactly what we now call a tunnel; but this specus is cut in the rock of the cliff, with a few feet of stone only as an outer wall to it, and in this manner it is continued along the edge of the valley of the river Anio for many miles, always on the left side of the river in going towards Rome. An old road runs by the side of it, not now used for carriages, but remaining as a cart-road only; this must be the Via Sublacensis of Frontinus. Another road runs along the right-hand side of the valley, and is the one now in use: this is the Via Sublacensis Neroniana of the time of the Empire. The valley varies in width very much, in some parts it is three or four miles wide; this is the case where the springs of the Marcia and the Cerulean Lake gush out from the rock under the diverticulum of the Via Valeria or present carriage-road, on the right-hand side of the valley.

The lowest of the three lakes of Claudius above Subiaco was circular, the rock being cut away to a half circle on each side of the stream; into this great basin the grand cascade fell from the second lake. The lowest lake or loch was comparatively not very deep. It seems most probable that the lowest reservoir was intended to serve for the Aqua Claudia. The specus has not at present been traced quite so far; but it is found a little lower down, above the modern paper-mill. This is more than a hundred feet below the level of the Anio Novus. It may be that this was one of the springs that fell into the Anio, and was intercepted for the aqueduct. It is probable that the same was the case with the Anio Vetus, as we know it was with the Marcia; but as the water from the springs sometimes ran short in dry seasons, the Anio Novus was taken from the river itself, a part of which was turned into it from the great lake or loch. For this reason that water was always more abundant than all the rest, and was used to supply the deficiency in case of need, as Frontinus tells us. This specus can be entered and examined; it is here a tunnel made in a rock of soft stone, with fissures filled with clay. The specus is lined with brick, and covered with large flat tiles, placed at an angle, so as to form a roof sloping down to the two sides from the ridge in the middle. There are inscriptions recording repairs by Cardinal Barberini, nephew of Urban VIII. This specus is that of the Anio Novus, constructed by Claudius; that of Trajan is about half-a-mile above the town of Subiaco, and on the right-hand side of the river Anio, not on the left, as the Anio Novus is.

The thirty-eighth milestone on the Via Sublacensis was found by Fabretti in situ, and the thirty-eighth milestone in the Via Valeria, now at Arsoli, is said by Gruter[136] to have been formerly at La Sonnoletta, or ad fontem Somnulæ: so that the two sources of the Claudian water, called Cæruleus and Curtius, were at the lake now called S. Lucia, in the territory of Arsoli; and the source called Albudinus is the first of the four springs now called Acque Serene, while the other springs of the same name formed the Aqua Marcia.

The Piscina Limaria, referred to by Frontinus as erected by Claudius for filtering the stream, at the entrance of the Anio Novus, at the forty-second mile[137] on the Via Sublacensis, is visible at the Parata della Cartiera of Subiaco. This piscina is not covered over, but open at the top, and is excavated in the rock of the bed of the river Anio, with a great declivity and with four cascades to throw down the sand, wood, and weeds brought down the river, and so to purify the water; but as it was still liable to become muddy in the time of floods, Trajan, according to Frontinus, excavated another specus in the rock of the mountain near the Ponte di S. Mauro, where the great wall of the lake was situated[138].

The specus of the Claudia seems to have been carried on the right-hand side of the valley as far as Vicovaro, about half-way to Tivoli, and then across the valley on an arcade, and carried on under the Anio Novus, and above the Marcia and Anio Vetus. All of these seem to be at different elevations on the side of the hill to the left of the valley, until they arrive at another valley crossing this, called the Valley of the Arches, about two miles from Tivoli, where they were carried across upon arcades. On these ancient arcades, or out of the materials of them, modern bridges have been made, both at Vicovaro and in the Valley of the Arches. The ruins of these splendid arcades are among the finest and most picturesque ruins of the Roman empire.

After passing the bridges across the valley of the Anio, the Anio Novus, Marcia, and Anio Vetus are continued along the side of the hill in what may still be called the cliff of the valley of the Anio, to near the cascades at Tivoli, at different levels, the Anio Novus considerably higher than the others. To avoid the cascades here the aqueducts wind round the end of the hill. In going out of Tivoli to the promenade of Carciano, on the side towards Rome, there is a large college of the Jesuits on the left hand, the specus of the Anio Novus passing under this, and through a wine-cellar. About a quarter of a mile out of the town, two specus are visible on the side of the hill above the road or promenade, and these may be traced at short intervals for miles, with openings into them in several places. The lower one is the Marcian, passing in a more direct line, and further from Tivoli, and is generally cut in the rock as a tunnel; the upper one is the Anio Novus, and is in some places faced with brick or with Opus Reticulatum, where it has to cross an opening. The Aqua Marcia here passes at a lower level below the road, with reservoirs at intervals, as already described.

At about three miles below Tivoli, and half-a-mile after passing one of the great reservoirs of the Marcia at the lower level, the Anio Novus is carried on an arcade across a valley and a small stream, at a place called after the arcade Arcinelli. The great Villa of Hadrian is nearly under this at the foot of the hill, perhaps a mile lower down. This was supplied with water by branches from the aqueducts, but without interfering with the main streams, which went on at the high level still at least five hundred feet above the level of Rome. From this high level the aqueducts led gradually down in a serpentine course, crossing several narrow valleys or gullies through the hills, along which some mountain-stream flows far below. At two important points, the aqueducts have to be carried across such valleys or gullies on fine arcades or bridges. The one nearest to Tivoli is called the bridge of S. Antony, from a small chapel made in the Middle Ages, probably in the fourteenth century, out of a portion of the specus of the Marcia, in one of the chambers of a castellum aquæ belonging to it, the rest of which has been destroyed. The specus of the Anio Vetus under it has also been destroyed, but the arcade is preserved for the use of horses and foot-passengers. This bridge is about eight miles from Tivoli. There is no carriage-road to it, but a tolerable path for horses or donkeys; it is one of the finest and most picturesque objects on the whole line of the Aqueducts, and is about 100 ft. above the water in the stream below which is called the fosse of S. Antony. This valley or gorge in the mountain, with the bridge across it, is not easily seen from any distance; but the site may be indicated by a medieval castle with a tall square brick tower, which is a conspicuous object for miles. This is distinctly seen from the bridge, and as it stands at the mouth of the valley, that object must be seen from thence on looking up the valley. This bridge is a really grand work; it is 373 ft. long, 16 ft. wide, and about 104 English feet high. The dome of S. Peter’s at Rome is visible at a long distance from the hill above the medieval castle.

Following the direction of the aqueducts towards Rome for about a mile, along a very rough cross-country path, we arrive at a good carriage-road from Rome to Poli, and about a mile beyond that is another magnificent structure, as fine and more perfect than the last, called Ponte Lupo, one of the greatest works of all the Roman aqueducts. There are not so many open arches as at the bridge of S. Antony, but a greater extent of substructure and wall at each end, and the height above the mountain-stream is even greater. Several of the arches seem to have been filled up for greater strength, at an early period; it is all work of the first century. Here the two specus of the Claudian and the Anio Novus are perfect, one upon the other, and serve as a lofty parapet to the road for horses, which passes across the valley along the side of them, the specus occupying about one-third of the width of the bridge. The two specus seem to have been first brought together at this point, and afterwards continue one upon the other for the rest of their course into Rome, interrupted only by the piscinæ and castella aquarum at intervals.

There are ruins of another arcade across the valley at a lower level by the side of this grand bridge, only a few yards from it, which can only be that of the Marcia. The Anio Vetus appears to be carried across under the Claudia and Anio Novus at a much lower level, and it is probably for this reason that the arches are closed instead of being left open. They pass upon four other bridges over the streams of S. Gregory and of the Inferno, before they arrive at the piscinæ. They all four meet at this point, and then diverge again for miles. This splendid work is about ten miles from Tivoli, half-a-mile from the carriage-road, five or six miles from the Villa of Hadrian, and out of the line of the direct road to Rome, but very near the road from Rome to Poli. After passing these great works, the aqueducts are continued at their respective levels in the direction of La Colonna and of the piscinæ, and are carried in tunnels through the hill, but these tunnels do not appear to be of any great length; they merely pass through part of one side of the hill. They bring us to the piscinæ, which are about six miles from Rome, where the arcades begin, and from this point the Marrana follows the same direction in the bed of the Almo, winding about but never very distant from the arcades, and always receiving the surplus water. This stream is here divided into two branches, one of which goes through Rome and into the Tiber at the Pulchrum Littus; the other passes through the valley of the Caffarella, and falls into the Tiber near the church of S. Paul beyond the walls, as is explained under the head of the Aqua Crabra and the Marrana.

The specus of the Anio Novus is always faced with brick or with Opus Reticulatum; it is carried upon the Claudian arcade, and the one specus always rests upon the other: what applies to one applies to both after the first junction. The aqueduct of Claudius has a stone specus carried on a stone arcade for the last five miles into the city. It was begun by Caligula, and carried on by his successor; this portion was completed in fourteen years, a very short period for so enormous an undertaking. The Anio Novus was the most abundant of all the aqueducts, as stated by Frontinus[139].

The arches of aqueducts, stretching for miles across the open country and entering Rome at the eastern angle, are the first objects to attract the attention of strangers on approaching the city. These are the arches of the Claudian aqueduct, built of large square stones, with the angles chamfered off, and carrying the streams of the Aqua Claudia and of the Anio Novus.

Of the Piscinæ mentioned by Frontinus, the ruins of one remain at the place now called Porta Furba, where the aqueducts just mentioned meet again about two miles from the city. The lofty Claudian arcade passes over the Marcian arcade at this point, at one of the numerous angles, as it had previously done at the Torre Fiscale. From the Porta Furba to Rome, the Aqua Felice is carried against the side of the Claudian, or on the piers of the Marcian, as most convenient, frequently crossing from one to the other on arches over the road. The specus of the Felice is built in the roughest manner of old materials taken from the other two arcades. The Anio Vetus also passes underground at the Porta Furba. Immediately after the entrance of the Claudian specus into Rome at the extreme east end, in the gardens of the Sessorian Palace (now S. Croce), there are remains of at least four reservoirs or castella aquarum before it reaches the Porta Maggiore[140].

Nero was the immediate successor of Claudius, A.D. 54-69, and carried on the conduit into the city, on what are still called the arches of Nero, which are faced entirely with brick, and some of the most beautiful brickwork in the world. So gigantic a work as these aqueducts could not have been completed in the lifetime of a single emperor, however large the number of slaves he may have employed upon it.

The Neronian Arcade.