The streets of Pompeii vary greatly in width. The widest is Mercury Street, the continuation of which near the Forum has a breadth of nearly 32 feet. Next come Abbondanza and Nola streets, the greatest width of which is about 28 feet; the other streets and thoroughfares vary from 10 to 20 feet. With unimportant exceptions, broad and narrow streets alike are paved with polygonal blocks of basalt, which in laying were fitted to one another with great care; on both sides are raised sidewalks, with basalt or tufa curbing. The sidewalks in some places are paved with small stones, elsewhere are laid with concrete, or left with a surface of beaten earth. As there is no uniformity, the sidewalk varying in front of adjoining houses, it is clear that the choice of materials was left to individual owners of abutting property. The limits of ownership are often designated by boundary stones, laid in the surface of the walk.

Broad ruts, worn by wheels, are seen in the pavement, shallower in places where the basalt flags, cut from the lowest stratum of the stream of lava, are particularly hard; deeper wherever there are blocks quarried nearer the surface. Only the principal streets were wide enough to allow wagons to meet and pass; elsewhere drivers must have waited at a corner for a coming team to go by. It seems likely that driving on the streets of the city was forbidden, wheeled vehicles being used only for traffic; people who wished to ride availed themselves of litters.

At various places along the thoroughfares, but particularly at the corners, large oblong stepping stones with rounded corners were set in the pavement at convenient distances for those wishing to cross, the surface being on a level with the sidewalk. The number varied according to the width of the pavement; in the broadest streets as many as five were used. They were arranged always in such a way as to leave places for the wagon wheels. It is not difficult to understand how Pompeian drivers guided their teams past them; draft animals were attached to the wagon by means of a yoke fastened to the end of the pole, and, as there were no tugs or whippletrees, they had a greater freedom of movement than is allowed to modern teams.

It is not to be supposed that so complete a system of paving existed from the beginning of the city. Some light is thrown on the period of its laying by two inscriptions,—one, EX · K · QUI, cut in the edge of the sidewalk west of Insula IX. iv.; the other, K · Q, in the pavement between the second and fourth Insulae of Region VII. Both are evidently dates, and in full would read ex Kalendis Quinctilibus, 'from the first day of July,' and Kalendis Quinctilibus, 'July 1.' Apparently they relate to the laying of the pavement; this was in place, even in the unimportant side street of Region VII, when the inscriptions were cut, and so must go back to the time before the name of the month Quinctilis was changed to Iulius, our July. Pompeii was paved, therefore, before 44 B.C.

The stepping stones were particularly useful when there was a heavy rain; for the water then flowed in torrents down the streets, as it does to-day in Catania, where the inhabitants have light bridges which they throw over the crossings after a storm. There were covered conduits to carry off the surface drainage of the Forum, one of which runs under the Strada delle Scuole to the south, the other under the Via Marina to the west. Elsewhere the water rushed down the streets till it came near the city walls, where it was collected and carried off by large storm sewers. These are still in successful operation, as are also the conduits at the Forum. One is at the west end of the Vico dei Soprastanti, another at the west end of Nola Street; and a third leads from Abbondanza Street, where it is crossed by Stabian Street, toward the south.

There were other sewers in the city, but they were of small dimensions and have not been fully investigated. They seem generally to have been under sidewalks. They were not designed to receive surface water, but the drainage of houses. They cannot have served this purpose fully, however, for most of the closets were connected, not with the sewers, but with cesspools.

After the lapse of more than eighteen centuries, the visitor at Pompeii will distinguish at a glance the business streets from those less frequented. The sides of the former are lined with shops; along the latter are blank walls, broken only by house doors, with now and then a small window high above the pavement. The greatest volume of business was transacted on the two main thoroughfares, Stabian and Nola streets; next in importance were Abbondanza Street, leading from the Forum toward the Sarno Gate, and the continuation of Augustales Street from the north end of the Forum toward the east. First in the list of quiet thoroughfares is the broad Mercury Street, along which were many homes of wealth; the north end of it is closed by the city wall.

There were many fountains along the streets of Pompeii, most of them at the corners. They were fed by pipes connecting with the water system of the city. The construction is simple. A deep basin was made by placing on their edges four large slabs of basalt, held together at the corners by iron clamps. Above one of the longer sides, usually near the middle, is a short, thick standard, of the same stone, pierced for the lead feed pipe, which threw a jet of water forward into the basin below; on the opposite side is a depression through which the superfluous water ran off into the street. Most of these standards are ornamented with reliefs, roughly carved but effective,—an eagle with a hare in its beak, a calf's head, a bust of Mercury, a head of Medusa, a drunken Silenus (Fig. 103), or some other suitable design, arranged so that the water would spurt from the mouth of the figure or from an amphora.

Occasionally we find a fountain of finer material. That of Concordia Augusta, of limestone, has already been mentioned (p. 117). In the neighborhood of the Porta Marina there is a fountain of white marble with a relief showing a cock that has tipped over a jar, from the mouth of which the water flowed. Both these more costly fountains were probably the gift of private individuals, one presented to the city by Eumachia, the other by the owner of the nearest house, at VII. xv. 1-2. All the fountains bear witness to long use by the depressions worn in the stone by the hands of those leaning forward to drink.

Water towers stand at the sides of the streets, small pillars of masonry preserved ordinarily to the height of 20 feet. Usually on one side there is a deep perpendicular groove (shown in Fig. 103) in which ran the pipe that carried the water to the top of the tower, where it was received by a small open reservoir, presumably of metal, and distributed through numerous small pipes leading to the fountains and to private houses. The sides of the towers are often covered with incrustations of lime deposited from the water, in which the impressions of the lead pipes are still to be seen; in the case of one tower, at the northeast corner of Insula VI. xiii, a number of the pipes have been preserved. A reservoir was placed also on the top of the commemorative arch at the lower end of Mercury Street, on which stood the bronze statue of Nero or Caligula (p. 48); the traces of the pipes leading from it are clearly seen on the surface of the arch. Similar water towers are in use now in Constantinople and Palermo, having been introduced into the latter city, it would seem, by the Saracens, who very likely took their water system from that of the Turkish capital.

Across the street from the Baths near the Forum, on the west, is a deep reservoir, of which we give the plan (Fig. 104). It is built partly below the level of the sidewalk, and measures about 50 feet in length and 13 in width, being covered by a vault. In the south end is a window (c), reached from one of the stairways; when the reservoir was filled to the bottom of the window, it contained not far from ninety-five thousand gallons. There were two outlets. One was at the level of the floor, closed by means of a bronze slide; the grooves in which the slide worked are preserved. This must have been used only when the reservoir was cleaned. The other outlet was placed about three feet above the floor, so that the water could be drawn off without disturbing the bottom. On the flat roof were rooms the arrangement of which cannot be determined.

Similar reservoirs are found in Constantinople, designed to furnish a supply of water in case of siege. Such may have been the purpose of our structure, which seems to have been built in the early years of the Roman colony. The residents, remembering the hardships of the siege of Sulla, may have thought it necessary to make provision against a similar strait in the future.

The source from which the city received its water supply has not been discovered. Evidently it did not draw upon the sources of the Sarno; the water channel constructed by Fontana (p. 25) runs through the city at a height of less than sixty feet above the level of the sea, while the ancient aqueduct that supplied Pompeii had so great a head that in the highest parts of the city, more than 130 feet above the sea, it forced the water to the top of the water towers, at least twenty feet more. Copious springs can never have existed on the sides of Vesuvius; water must have been brought to the city from the more distant mountains bounding the Campanian plain on the east.

We can hardly believe that the construction of a water channel for so great a distance lay within the resources of so small a town. We find, however, the remains of a great aqueduct which, starting near Avellino, a dozen miles east of Nola, skirted the base of Vesuvius on the north and extended westward, furnishing water not only to Naples but also to Puteoli, Baiae, and Misenum. This ancient structure drew from the same springs, and followed substantially the same route, as the new aqueduct which since 1885 has been bringing water to Naples. No inscription in regard to it has been found, and there is no reference to it in ancient books. The remains—of which the longest section, known as Ponti Rossi, 'Red Bridges,' may be seen near Naples—seem to indicate two styles of construction, extensive repairs having been made after the aqueduct had been partly destroyed; but up to the present time it has not been possible to determine the period to which they belong.

The water system of Pompeii goes back to the time before the founding of the Roman colony. This is evident, not only from the arrangements of the older baths, which contemplated a freer use of water than could well have been provided by cisterns, but also from the existence of three marble supports for fountain basins, which, as shown by their style of workmanship, the use of Oscan letters as mason's marks, and their location in pre-Roman buildings—the temple of Apollo, the Forum Triangulare, and the house of the Faun—belonged to the earlier period. If we may ascribe the building of the great aqueduct to the time of peace and prosperity in Campania between the Second Punic War and the Social War, and suppose that Pompeii, joining with other towns in its construction, was supplied by a branch from it, we have a simple and highly probable solution of the problem. Nothing in the character of the masonry requires us to assign the aqueduct to a later date.

The shrines along the streets, with few exceptions, were dedicated to the guardian deities presiding over thoroughfares, particularly the gods of street crossings, Lares Compitales. The worship of these divinities in Rome was reorganized by Augustus and placed in charge of the precinct wardens, vicorum magistri, who were to see that the worship of his guardian spirit, Genius, was associated with that of the Lares at each shrine. The arrangements at the Capital were naturally followed by the colonies and other cities under Roman rule.

At Pompeii the shrines of the street gods differ greatly in size and character. Sometimes there is a small altar against the side of a building, with two large serpents, personifications of the Genius of the place, painted on the wall near it; one of the serpents, with a conspicuous crest, represents a male, the other, a female.

Frequently the place of the altar is taken by a niche, in which the passer-by could deposit his offering. In our illustration (Fig. 105) we see an ancient street altar which was carefully preserved when the Central Baths were built, a niche being made over it in the new wall.

Sometimes a large altar is found, and the Lares, with their offerings, are painted on a wall above it. Such a shrine may be seen at the northwest corner of Stabian and Nola streets, between the fountain and the water tower (Fig. 103). Back of the altar is a wall terminating in a gable (the tiles are modern) on which was a painted altar with four worshippers clad in togas, and a fluteplayer, the inseparable accompaniment of a Roman sacrificial scene; at the sides were the two Lares, represented as youths, in loose tunics confined by a girdle, holding in one hand, high uplifted, a drinking horn (rhyton), from which a jet of wine flows into a small pail (situla) in the other hand. It is remarkable that we do not find in this or similar paintings at Pompeii, any figure representing the Genius of the emperor, while in private houses the Genius of the proprietor often has a place with the Lares, and sometimes the Genius of the emperor also; in theory at least, as already remarked (p. 104), the emperor stood to all men in the relation that the master of a house bore to the household.

There is also a small chapel for the worship of the street gods on the west side of Stabian Street, near Abbondanza Street. As may be seen from the accompanying plan (Fig. 106), at the left as you enter is a bench of masonry (1), at the rear a long altar (2). In the wall at the right is a niche for the bronze or terra cotta figures of the Lares and the Genius, while the surface of the altar is divided into two parts, for the separate worship of the same divinities. A similar chapel is situated on the west side of Mercury Street (VI. viii. 14). Here also we find a bench of masonry, with two niches above it; in the middle was a block of limestone which may have been used as an altar. At the rear is a door leading into a small back room. This chapel was formerly thought to be a barber shop.

It has been customary to assign to the street gods all of the shrines at the side of the street. Occasionally, however, other divinities were thus honored; and the only street altar found with an inscription is consecrated to a different deity. This altar is near Nola Street, on the east side of Insula IX. vii. On the wall above two cornucopias are painted the words Salutei sacrum, 'Sacred to Salus'; the goddess of health was worshipped here.

Near the upper end of the Forum, on the north side of Insula VII. vii, is another altar, above which is a stucco relief representing a sacrifice; at the sides of the relief are pilasters, and over it a gable, in which an eagle is seen. This indicates that the shrine was dedicated to Jupiter.

The largest of the street altars, of tufa, stands free in a vaulted niche on the north side of Insula VIII. ii, but no traces of painting are to be seen near it (Fig. 107).

Various divinities are painted on the outside of houses. The largest picture of this kind is at the corner of Abbondanza Street, on the east side of Insula VIII. iii. It contains figures of the twelve gods, distinguished by their attributes—Vesta, Diana, Apollo, Ceres, Minerva, Jupiter, Juno, Vulcan, Venus Pompeiana, Mars, Neptune, and Mercury. Underneath are the two serpents, facing each other, on either side of a painted altar; near the altar are other figures that cannot be plainly distinguished, probably of men offering sacrifice. This is not a shrine—there is no place for the offerings. The owner of the property (house of the Boar), desired to place his household under the protection of these gods, perhaps also to preserve the corner from defilement. We often find roughly sketched figures of single gods, to the guardian care of whom the master of a house wished to commit his interests—most frequently Mercury, the patron divinity of traders, and Bacchus; but also Jupiter, Minerva, and Hercules.

Sometimes merely a pair of serpents are painted on a wall, in order to give a religious association to the place, as a means of protection. In one case (east side of Insula VII. xi. 12) an explicit warning was painted on the plaster beside them: Otiosis locus hic non est; discede, morator,—'No place for loafers here; move along!'

CHAPTER XXXII
THE DEFENCES OF THE CITY

From the military point of view, Pompeii at the time of the eruption did not possess a system of defences. For many years previously the city wall had been kept in repair only as a convenience in matters of civil administration, and the gates had long since lost all appearance of preparedness to resist attack. The fortifications are not, however, without interest. They form a massive and conspicuous portion of the ruins, and as a survival from an earlier period they have recorded many evidences of the successive changes through which the city passed.

The relation of the wall to the configuration of the height on which Pompeii stood was pointed out in connection with our general survey of the city (p. 31). Along the southwest side, at the time of the eruption, it had almost completely disappeared. Here, where the slope was steepest and the city best defended by nature, the wall had been removed, and its place occupied by houses, at a comparatively early date, probably in the second century B.C.; enough fragments remain, however, to enable us to determine its location with certainty. Elsewhere the greater part of the wall is in a fair state of preservation. The towers did not belong to the original structure, and one of the gates in its present form is of still more recent origin.

The construction of the wall will be readily understood with the help of the accompanying illustrations.

First, two parallel stone walls were built, about 15 feet apart and 28 inches thick; both walls were strengthened on the side toward the city by numerous buttresses, the inner wall being further supported by massive abutments projecting into the space between (Fig. 108). This space was filled with earth.

When the desired height, 26 or 28 feet, was reached, a breastwork of parapets was constructed on the outer wall; the inner wall was carried up about 16 feet above the broad passageway on the top (Fig. 110) as a shield against the weapons of the enemy, preventing the missiles from going over into the town and causing them to fall where the garrison could easily pick them up to hurl back again. Rain water falling on the top flowed toward the outside, and was carried beyond the face of the masonry by stone waterspouts.

For additional strength there was heaped against the inner wall an embankment of earth, which still remains on the north side, between the tenth and twelfth towers. At the right of the Herculaneum Gate the place of the embankment and of the inner wall was taken by a massive stairway (E in Fig. 108) leading to the top. Originally, the stairs extended east about 270 feet, but afterwards they were demolished for the greater part of the distance, and houses were built close to the wall. There is a smaller stairway of the same kind east of the Stabian Gate (Fig. 111).

In the original structure both outer and inner walls were built of hewn blocks of tufa and limestone; but we find portions of the outer wall, and all the towers, of lava rubble, the surface of which was covered with stucco. The towers were already standing, as shown by inscriptions, at the time of the Social War. We are therefore safe in believing that in the period of peace following the Second Punic War the walls were not kept in repair, some parts of the outer wall being utilized as a quarry for building stone; that with the advent of the Social War they were hastily repaired on the north, east, and south sides, and strengthened by towers, but that no attempt was made to renew the fortifications on the steep southwest side, between the Herculaneum Gate and the Forum Triangulare, where the line of the old wall was covered with buildings.

When the towers were added—probably not long before 90 B.C.—they were not distributed evenly along the wall, but were placed where they seemed to be most needed. The western portion of the ridge between the Herculaneum and Capua Gates was particularly favorable for the approach of an enemy; hence three towers were built near together here, numbered 10, 11, and 12 on Plan I. Another part of the wall especially exposed was on the southeast side, where the height covered by the city slopes gradually down to the plain; and we find five towers within a comparatively short distance, two east of the Amphitheatre, the other three further south. On the north side, between the Capua and Sarno gates, the slope is steeper and two towers were thought to be sufficient.

That there were once two additional towers, besides the ten that have been enumerated, is evident from several Oscan inscriptions, painted in red letters on the street walls of houses. One of them, near the southwest corner of the house of the Faun, reads thus: 'This way leads between Towers 10 and 11, where Titus Fisanius is in command.' The street referred to runs between the tenth and twelfth Insulae of Region VI, direct to the city wall. Two others refer to a 'Tower 12' near the Herculaneum Gate, this part of the fortifications being in charge of Maras Adirius.

In a fourth inscription we read: 'This way leads between the houses of Maras Castricius and of Maras Spurnius, where Vibius Seximbrius is in command.' In 1897, a fifth inscription became visible on the north side of Insula VIII. v-vi, where it had been concealed by a coat of plaster: 'This way leads to the city building (and) to Minerva.' The street referred to is seemingly the blind alley which formerly ran through the insula (Plan I). If this is correct, the sanctuary of Minerva is the Doric temple in the Forum Triangulare; but the 'city building' cannot be identified.

The five inscriptions evidently date from the siege of Sulla; they were intended for the information of the soldiers, belonging to the army of the Allies, who were quartered in the city to assist in its defence. At this time there must have been twelve towers, that near the Herculaneum Gate being reckoned last in the enumeration, as in Plan I; but the location of the two that have disappeared has not been determined. Another suggestive reminder of the same siege is the name L · SVLA, scratched by a soldier in the stucco on the inside of Tower 10, near a loophole.

The towers, which measure approximately 31 by 25 feet, were built in two stories, with strong vaulted ceilings. The floor of the second story was on a level with the top of the wall, and over this story was a terrace with battlements, as shown in Fig. 110; the roof seen on the two towers in Fig. 101 was a later addition, made when the city walls were no longer needed as a means of defence. Stairways on the inside gave ready access to the lower part of the towers, which could be entered from the city by a door (Fig. 109) opening on the embankment. On the outside were loopholes. Below, at the right, was a sally port, placed thus in order that the soldiers when rushing forth might present their shields to the enemy, leaving the right hand free to use with offensive weapons; when returning to the wall they would, if possible, cut their way to the sally port in the next tower to the right, so as to avoid the danger of exposing their right sides to the enemy.

Four of the gates have been excavated, the Porta Marina and the Stabian, Nola and Herculaneum gates; two others, the Vesuvius and Sarno gates, have been partly exposed to view. The remaining two are still completely covered. All bear evidence of extensive repairs, and one of them, the Herculaneum Gate, was entirely rebuilt at a comparatively late period; with this exception, however, they seem to have assumed their present form in the Tufa Period. Three of them still retain traces of decoration of the first style on the inner parts. The different gateways enter the walls at various angles.

The Stabian Gate may be taken as typical. Entering from the outside, at A, one came through a vaulted passage, B, about twelve feet wide, to a broad middle passage, or vantage court, open to the sky, into which missiles and boiling pitch could be hurled from above upon the heads of an enemy attempting to force the gates; then followed a second vaulted passage, a little wider than the other, in which were hung the heavy double doors, opening outward. The projecting posts of the doors are preserved, as are also the stones on which they rested when they were swung back against the wall; the vaulting has been restored. The gateway was paved throughout, with a raised walk on the right side. On one side of the inner entrance is a well (a), the Gorgon's head upon the curb reminding one of the protectress of the gate; on the other, the flight of steps already mentioned (b) leads to the top of the wall. Just beyond the steps are the remains of a small building, perhaps the lodge of the gate keeper (c).

The patron divinity of city gates, Minerva, was probably honored with a small statue in the niche still to be seen in the wall of the vantage court. Two inscriptions commemorate the making of repairs on the thoroughfare passing under the gateway. One of them (at d) is the Oscan inscription recording the work of the aediles Sittius and Pontius, to which reference has already been made (p. 184). The other (at e) is in Latin, and of much later date. It informs us that the duumvirs L. Avianius Flaccus and Q. Spedius Firmus at their own expense paved the road 'from the milestone,' which must have been near the gate, 'to the station of the gig drivers (cisiarios), at the limits of the territory of the Pompeians.' The Roman gigs, cisia, were very light, and adapted for rapid travelling; they were drawn by horses or mules, and were kept for hire at stations along the highways. The site of the station between Pompeii and Stabiae is not known.

The Nola Gate, and the partially excavated Vesuvius and Sarno gates, follow the plan just described in all essential particulars. The inner keystone of the Nola Gate, facing the city, is ornamented with a helmeted head of Minerva, in high relief, which being of tufa has suffered from exposure to the weather. There was once an Oscan inscription near by, which stated that the chief executive officer of the city, Vibius Popidius, let the contract for building this gate, and accepted the structure from the contractor.

The front of the Porta Marina has the appearance of a tower projecting from the wall. The gateway consists simply of two vaulted entrances, of unequal width; one for vehicles, the other, at the left, for pedestrians. Both were closed by doors. In the niche at the right of the wider passage the lower part of a terra cotta statue of Minerva was found. There was no vantage court, no inner passage; but in the early years of the Roman colony the steep lower end of the Via Marina for a distance of 70 feet was covered with a vaulted roof, which still remains. Opening into this corridor on the right is a long narrow room, which formed a part of the foundations of the court of the temple of Venus Pompeiana, and is now used as a Museum.

This gate in its present form could hardly have been intended for defence; it was adapted rather for administrative purposes, and must have been built—probably in the place of an earlier structure—in a period when the possibility of war seemed remote. Such a time, as previously remarked, was the second century B.C., particularly the latter half, after the destruction of Carthage.

A still more peaceful aspect is presented by the Herculaneum Gate. The style of masonry—rubble work with opus mixtum at the corners—points to the end of the Republic, rather than to the Empire, as the period of construction. Here we find three vaulted passages, the middle one for vehicles, those on either side for pedestrians. The vaulting over the middle part of the gate has disappeared; but according to appearances a vantage court was left here, in the middle passage, if not in those at the sides; at the inner end of this court the gates were placed. The greater part of the structure served no purpose of utility; it was obviously designed as a monumental entrance to the city.

PART II
THE HOUSES

CHAPTER XXXIII
THE POMPEIAN HOUSE

Our chief sources of information regarding the domestic architecture of ancient Italy are two,—the treatise of Vitruvius, and the remains found at Pompeii. The Pompeian houses present many variations from the plan described by the Roman architect; yet in essential particulars there is no disagreement, and it is not difficult to form a clear conception of their arrangements.

The houses of Greco-Roman antiquity differed from those of modern times in several respects. They took their light and air from the inside, the apartments being grouped about a court or about a large central room which ordinarily had an opening in the ceiling; the distribution of space being thus made on a different principle, the large rooms were often larger, the small rooms smaller and more numerous than in modern dwellings of corresponding size; and in the better houses the decoration of both walls and floors was more permanent than is usual in our day. The ancient houses were relatively low, in most cases, if we except the crowded tenements of imperial Rome, not exceeding two stories. The windows in the outside walls were generally few and small, and the external appearance was not unlike that of Oriental houses of the present time. In the city house the large front entrance was frequently ornamented with carved posts and lintel.

The development of the Italic house can be traced at Pompeii over a period of almost four hundred years. The earlier form consisted of a single series of apartments,—a central room, atrium, with smaller rooms opening into it, and a garden at the rear; an example is the house of the Surgeon (p. 280). A restoration of such a house with its high atrium, wide front door, and garden is shown in Fig. 114.

Later, under Greek influence, a court with a colonnade and surrounding rooms was added. This was called peristylium, 'peristyle'; it is simply the more elaborate inner part of the Greek house, andronitis, joined to the dwelling of Italic origin. We find the union of atrium and peristyle with their respective groups of apartments fully accomplished in the second century B.C., the Tufa Period; the type of dwelling thus developed remained in vogue during Roman times and is often called the Roman house.

The double origin is clearly indicated by the names of the rooms. Those of the front part are designated by Latin words,—atrium, fauces, ala, tablinum; but the apartments at the rear bear Greek names,—peristylium, triclinium, oecus, exedra. In large houses both atrium and peristyle were sometimes duplicated.

The houses of Pompeii impress the visitor as having been designed primarily for summer use. The arrangements contemplate the spending of much time in the open air, and pains was taken to furnish protection from the heat, not from the cold. The greater part of the area is taken up by colonnades, gardens, and courts; from this point of view the atrium may be classed as a court. The living rooms had high ceilings. In summer they were cool and airy, in winter difficult to heat; they were dark and close when the door was shut, cold when it was open.

With a single exception the arrangements for heating so often met with in the remains of houses discovered in northern countries are found at Pompeii only in connection with bath-rooms; the cold was ineffectively combated by means of braziers. We are led to believe that the Pompeians were extremely sensitive to heat, but endured cold with great patience. One who makes himself familiar with the arrangements of Italian houses to-day will receive a similar impression, although the peculiarity is perhaps less obvious than in the case of the ancient dwellings.

In describing the Pompeian houses it is more convenient to designate the principal rooms by the ancient names. In Fig. 115 we present an ideal plan; in it the names are given to the parts of the house, the relative location of which is subject to comparatively little variation. These parts will first be discussed; then those will be taken up which present a greater diversity in their arrangements.

I. Vestibule, Fauces, and Front Door

The vestibulum was the space between the front door and the street. The derivation of the word (ve- + the root of stare, 'to stand aside') suggests the purpose; the vestibule was a place where one could step aside from the bustle and confusion of the street. In many houses there was no vestibule, the front door opening directly on the sidewalk; and where vestibules did exist at Pompeii, they were much more modest than those belonging to the houses of wealthy Romans, to which reference is so frequently made in classical writers. Roman vestibules were often supported by columns of costly marbles, and adorned with statues and other works of art. Only one vestibule at Pompeii was treated as a portico, that of the house of the Vestals near the Herculaneum Gate. This was once as wide as the atrium, the roof being carried by four columns; but before the destruction of the city two partitions were built parallel with the sides dividing it into three parts, a narrow vestibule of the ordinary type, with a shop at the right and at the left.

The passage inside the front door was called fauces, or prothyron. According to Vitruvius the width of it in the case of large atriums should be half, in smaller atriums two thirds, that of the tablinum; at Pompeii the width is generally less than half. In the houses of the Tufa Period the corners of the fauces where it opens into the atrium were ornamented with pilasters connected at the top by an entablature.

The vestibule and fauces were ordinarily of the same width, and were separated by projecting doorposts with a slightly raised threshold (Fig. 116) and heavy double doors. Sometimes, as in the house of Epidius Rufus, there was in addition a small door at the side of the vestibule opening into a narrow passage connecting with the fauces (Fig. 149). In such cases the folding doors, which on account of their size and the method of hanging must always have been hard to open, were generally kept shut. They would be thrown back early in the morning for the reception of clients, and on special occasions; at other times the more convenient small door would be used.

In several instances the volcanic dust so hardened about the lower part of a front door that it has been possible to make a cast by pouring soft plaster of Paris into the cavity left by the crumbling away of the wood; there are several of these casts in the little Museum at Pompeii. With their help, and with the well preserved stone thresholds before us, it is possible to picture to ourselves the appearance of the doorway.

The doorposts were protected by wooden casings, antepagmenta, which were made fast at the bottom by means of holes in the threshold (α, α in Fig. 116).

The folding doors swung on pivots, which were fitted into sockets in the threshold (β, β) and in the lintel. The pivots were of wood, but were provided—at least the lower ones—with a cylindrical cap of iron or bronze, and the socket had a protective lining of the same metal. Both caps and sockets, especially those of bronze, are found in the thresholds in a good state of preservation. It seems strange that ancient builders did not use smaller pivots of solid metal, on which the doors would have turned much more easily; but a conservative tradition in this regard prevailed against innovation.

The fastenings were elaborate. Near the inner edge of each door was a vertical bolt, which shot into a hole in the threshold (γ, γ); there was probably a corresponding bolt at the top, as in the case of large modern doors. Sometimes there was a heavy iron lock, turned with a key, and also an iron bar which was fastened across the crack in such a way as to tie the two folds together. In many houses there are holes in the walls of the fauces, just back of the door, in which at night a strong wooden bar, sera, was placed; hardly less often we find a hole in the floor a few feet back, in which one end of a slanting prop was set, the other end being braced against the middle of the door. These arrangements bring to mind Juvenal's vivid picture of the disturbances and dangers of the streets of Rome at night.

II. The Atrium

An atrium completely covered by a roof was extremely rare. With few exceptions, there was a large rectangular opening over the middle, compluvium, toward which the roof sloped from all sides (Figs. 114, 118). In the floor, directly under the compluvium, was a shallow basin, impluvium, into which the rain water fell (h in Fig. 118). The impluvium had two outlets. One was connected with the cistern; a round cistern mouth, puteal, ornamented with carving, often stood near the edge of the basin, as in the house of the Tragic Poet (Fig. 153). The other outlet led under the floor to the street in front, carrying off the overflow when the cistern was full, and also the water used in cleaning the floor. In the better houses a fountain was often placed in the middle of the impluvium.

Vitruvius (VI. iii. 1 et seq.) mentions five kinds of atriums, the basis of classification being the construction of the roof—Tuscan, tetrastyle, Corinthian, displuviate, and tortoise atriums. The first three are well illustrated at Pompeii.

The Tuscan atrium, supposed by the Romans to have been derived from the Etruscans, was apparently the native Italic form. Two heavy girders were placed across the room, above the ends of the impluvium (Fig. 117, b). On these, two shorter crossbeams were laid (c), over the sides of the impluvium. The corners of the rectangular frame thus made were connected with the walls at the corners of the atrium by four strong slanting beams (Figs. 117, 118, e). On these and on the frame were placed the lower ends of the sloping rafters (Fig. 117, f), carrying the tiles, the arrangement of which can be seen in Figs. 114, 117, and 118. This was the most common arrangement of the roof at Pompeii.