Pompeii, Its Life and Art

The edge of the compluvium was frequently ornamented with terra cotta waterspouts, representing the heads of animals. In a house near the Porta Marina the projecting foreparts of dogs and lions were used in place of the heads; the remains of a part of the compluvium have been put together again, and are seen in Fig. 119. The lions were placed over the larger spouts at the four corners; the under side of the spouts surmounted by the dogs and lions was ornamented with acanthus leaves in relief. The same illustration presents an example of the antefixes sometimes found.

The tetrastyle atrium differed from the Tuscan in only one respect: there were four columns supporting the roof, one at each corner of the impluvium. In most cases these supports, which interfered with the view of the interior, can hardly have been intended primarily for ornament; they simplified the construction, making the ceiling and roof firm without the use of the heavy and expensive girders.

The Corinthian atrium had a larger compluvium than the other kinds, the roof being supported by a number of columns. There are three examples at Pompeii, the houses of Epidius Rufus with sixteen columns (p. 310), of Castor and Pollux with twelve, and of the Fullonica with six.

The roof of the displuviate atrium sloped from the middle toward the sides, the water being carried off by lead pipes. The aperture for the admission of light and air was relatively much higher above the floor than in the kinds previously described. No example of this type has been found at Pompeii.

The tortoise atrium, atrium testudinatum, was small and without a compluvium. The roof had a pyramidal shape. There were possibly a few examples at Pompeii, as we may infer from the occasional absence of an impluvium; in the only instance, however, in which it is possible to determine the form of the roof (V. v. 1-2), this must have been very different from that referred to by the Roman writer (p. 343).

Vitruvius says further that the atrium should have an oblong shape, the width being three fifths or two thirds of the length, or measured on the side of a square, the hypothenuse of which is taken for the length. The design was obviously to bring the sides nearer together, thus lessening the strain on the two girders which in the commonest form were used to sustain the roof. The height, to the frame of the compluvium, should be three fourths of the width.

In the case of the tetrastyle and Corinthian atriums at Pompeii the height is indicated by that of the columns, but there are rarely adequate data for determining the height of the others with exactness. In regard to length and breadth the proportions harmonize fairly well with those recommended by Vitruvius; but the height, in the cases in which it can be ascertained, is often greater than that contemplated by the rules of the architect.

Looking at the Pompeian atriums in their present condition (Plate VII, Figs. 121, 153) one might easily receive the impression that they were primarily courts rather than rooms. In this respect the restorations of Roman houses in the older books are often at fault, the atrium being generally represented as too low in comparison with the rooms around it.

Such the atrium remained in farmhouses to the latest times. The name meanwhile was transferred to the corresponding apartment of elegant city homes, while in the country it went out of use, being replaced by culina, 'kitchen,' on account of the presence of the hearth. In such a room in his Sabine villa Horace loved to dine, conversing on topics grave or gay with his rustic neighbors, and partaking of the simple fare with relish; while his slaves, freed from the restraints of city life, were permitted to eat at the same time, sitting at a separate table. The remains of an atrium of this kind, with its hearth and niche for the images of the household gods, may be seen in the villa recently excavated near Boscoreale (p. 361).

Without doubt some houses of the ancient type might be found in cities, even in Rome, as late as the end of the Republic. We read of one in Cicero's time in the atrium of which spinning was done. But at Pompeii the hearth had been banished from the atrium at a comparatively early date, in the Tufa Period if not before; and the room was made uncomfortable to sit in, for a considerable part of the year, by the broad opening of the compluvium.

From the architectural point of view, however, the atrium never lost its significance as the central apartment. In all its dimensions, but particularly in height, it presents so great a contrast with the rooms around it as to remind us of the relation of a Roman Catholic church to the chapels at the sides. The impression of spaciousness was perhaps deepened when the atrium was provided with a ceiling. Few traces of such ceilings are found at Pompeii, and in the smaller houses the inside of the roof seems generally to have been visible.

The atrium of the Corinthian type most nearly resembled a court, on account of the size of the opening to the sky and the use of many columns. A suggestion of the un-Italic character of this type appears in the name; for one can scarcely suppose that atriums in the strict sense existed at Corinth.

Although the Pompeian atriums show no traces of a hearth, there is possibly a reminiscence of the ancient arrangement in the gartibulum, a table which we frequently find at the rear of the impluvium. Varro says that since his boyhood these tables, on which vessels of bronze were placed, had gone out of use; at Pompeii they remained in fashion much longer. The gartibulum with its bronze vases may symbolize the ancient hearth with the cooking utensils. Possibly, however, it represents the kitchen table near the hearth on which the dishes were washed; that it may have served a similar purpose in later times is evident from the fact that in front of it a marble pedestal was often placed for a statuette which threw a jet of water into a marble basin at the edge of the impluvium. This group of table, fountain figure, and basin appears in many Pompeian atriums. In Plate VII we see the gartibulum and the supports of the marble-basin, but the base of the fountain figure has disappeared.

The strong box of the master of the house, arca, often stood in the atrium, usually against one of the side walls. It was sometimes adorned with reliefs, as the one shown in Fig. 120, which is now in the Naples Museum. It stood on a heavy block of stone, or low foundation of masonry, to which it was attached by an iron rod passing down through the bottom. A wealthy Pompeian sometimes had more than one of these chests.

In three atriums the herm of the proprietor stands at the rear. One, with the portrait of Cornelius Rufus, is shown in Fig. 121.

When there were two atriums in a house, the larger was more elaborately furnished than the other, and was set aside for the public or official life of the proprietor; the smaller one was used for domestic purposes. Typical examples are found in the houses of the Faun and of the Labyrinth. In the former the principal atrium is of the Tuscan type, the other tetrastyle; in the latter the large atrium is tetrastyle, the smaller Tuscan.

III. The Tablinum

The tablinum was a large room at the rear of the atrium, opening into the latter with its whole width; the connection of the two rooms is clearly shown in Plate VII and Fig. 121. According to Vitruvius, when the atrium was 30 to 40 feet in width—as in the larger Pompeian houses—the tablinum should be half as wide; when the atrium was smaller, the width of the tablinum should be two thirds that of the atrium, while the height at the entrance should be nine eighths, and inside four thirds of the width. These proportions will not hold good for Pompeii, where the tablinum is generally narrower and higher (Vitr. VI. iv. 5, 6).

The posts at the entrance were usually treated as pilasters, joined above by a cornice; architecturally the front of this room formed the most impressive feature of the atrium. Between the pilasters hung portières, which might be drawn back and fastened at the sides. In the house of the Silver Wedding the fastenings were found in place,—bronze disks from which a ship's beak projected, attached to the pilasters.

In early times the tablinum ordinarily had an opening at the rear also, but this was not so high as that in front, and could be closed by broad folding doors. In winter the doors were probably kept shut. In summer they were left open and the room, cool and airy, served as a dining room, a use which harmonizes well with a passage of Varro explaining the derivation of the name. "In the olden time," says this writer, "people used to take their meals in the winter by the hearth; in summer they ate out of doors, country folk in the court, city people in the tabulinum, which we understand to have been a summer house built of boards." The derivation of tabulinum, of which tablinum is a shortened form, from tabula, 'a board,' is obvious.

The period to which Varro refers antedates that of the oldest houses at Pompeii. The room which we call tablinum was then a deep recess at the rear of the atrium, open at the front, as now, but enclosed by a wall at the rear; against this wall was a veranda opening into the garden, toward which the board roof sloped. People took their meals in the veranda in summer, and to it the name tablinum was naturally applied. In the recess at the rear of the atrium, corresponding to the later tablinum, was the bed of the master of the house, called lectus adversus because 'facing' one who entered the front door. As late as the reign of Augustus, long after it became the custom to set aside a closed apartment for the family room, a reminiscence of the ancient arrangement still remained in the couch which stood at the rear of the atrium or in the tablinum, which was called lectus adversus, or even lectus genialis.

The removal of the hearth and the bed from the atrium must have taken place when the small hole in the roof was replaced by the compluvium. A broad opening was made in the rear wall, and the place where the bed had been was turned into a light, airy room; this was now used as a summer room instead of the veranda, the name of which was in consequence transferred to it.

Even in later times, when the houses were extended by the addition, at the rear, of a peristyle with its group of apartments, the tablinum may often have been used as a summer dining room; but the tendency now was to withdraw the family life into the more secluded rooms about the peristyle. The tablinum, lying between the front and the rear of the house, was used as a reception room for guests who were not admitted into the privacy of the home; and here undoubtedly the master of the house received his clients.

In the house of the Vettii the tablinum is omitted on account of the abundance of room; but at the rear of the atrium there are wide openings into the peristyle (Fig. 158).

IV. The Alae

The alae, the 'wings' of the atrium, were two deep recesses in the sides (Fig. 115). They were ordinarily at the rear, but were sometimes placed at the middle, as in the house of Epidius Rufus (Fig. 149). Vitruvius (VI. iv. 4) says that where the atrium is from 30 to 40 feet long, one third of the length should be taken for the breadth of the alae; in the case of larger atriums the breadth of these rooms should be proportionally less, being fixed at one fifth of the length for atriums from 80 to 100 feet long; the height at the entrance should be equal to the breadth.

At Pompeii the alae, as the tablinum, are narrower and higher than required by these proportions. In the Tufa Period the entrances were ornamented with pilasters, and treated like the broad entrance of the tablinum.

With reference to the purpose and uses of these rooms we have no information beyond a remark of Vitruvius in regard to placing the images of ancestors in them. This throws no light upon their origin; for only a few noble families could have possessed a sufficiently large number of ancestral busts or masks to make it necessary to provide a special place for these, while the alae form an essential and characteristic part of the Pompeian house. Now and then an ala was used as a dining room; more frequently, perhaps, one was utilized for a wardrobe, as may be seen from the traces of the woodwork. A careful study of the remains only deepens the impression that at Pompeii the alae served no definite purpose, but were a survival from a previous period, in which they responded to different conditions of life.

An interesting parallel presents itself in the arrangements of a type of peasants' house found in Lower Saxony. The main entrance, as in the early Italic house, leads into a large and high central room; at the sides of this and of the main entrance are the living rooms and stalls. At the back the central room is widened by two recesses corresponding with the alae; the hearth stands against the rear wall. In the side walls, at the rear of each recess, are a window and a door. The two windows admit light to the part of the central room furthest from the entrance; the doors open into the farmyard and the garden.

The Italic house in the beginning was not a city residence shut in by party walls, but the isolated habitation of a countryman. The design of the alae, as of the recesses in the Low Saxon farmhouse, was to furnish light to the atrium, which, as we have seen, was completely covered by a roof, there being only a small hole to let out the smoke. The large windows in the rear of the alae of the house of Sallust may be looked upon as a survival; but in city houses generally light could not be taken in this way from the sides. After the compluvium had come into general use, a conservative tradition still retained the alae whenever possible, though they no longer answered their original purpose.

V. The Rooms about the Atrium. The Andron

In front there were rooms at either side of the entrance, ordinarily fitted up as shops and opening on the street, but sometimes used as dining rooms or sleeping rooms, or for other domestic purposes.

On each side of the atrium were two or three small sleeping rooms; in narrow houses these, as well as one or both of the alae, were occasionally omitted.

At the rear were one or two rooms of the same depth as the tablinum, used in most cases as dining rooms. They frequently had a single broad entrance on the side of the peristyle or the garden (Fig. 134, 22), but were sometimes entered by a door from the atrium or from one of the alae (Figs. 115, 121). The door on the side of the atrium seems generally to have been made when the house was built; if the owner did not wish to use it, it was walled up and treated as a blind door, an ornament of the atrium.

The rooms about the atrium in the pre-Roman period were made high, those in front and at the sides often measuring fifteen feet to the edge of the ceiling, which had the form of a groined vault. The rear rooms were still higher, the crown of the vaults being as far above the floor as the flat ceiling of the tablinum. A corresponding height was given to the doors; those in the house of the Faun measure nearly fourteen feet. The upper part of the doorway was doubtless pierced for the admission of light in the manner indicated by wall paintings, and shown in our restoration of one side of the atrium in the house of Sallust (Figs. 261, 262).

The andron was a passage at the right or the left of the tablinum, connecting the atrium with the peristyle (Figs. 115, 121). The name was used originally to designate an apartment in the Greek house, but was applied by the Romans to a corridor. In modern times the passage has often been erroneously called fauces.

The andron is lacking only in small houses, or in those in which a different connection is made between the front and rear portions by means of a second atrium, or other rooms.

VI. Garden, Peristyle, and Rooms about the Peristyle

A few Pompeian houses, like those of the olden time, are without a peristyle, having a garden at the rear. In such cases there is a colonnade at the back of the house, facing the garden; this is the arrangement in the houses of the Surgeon, of Sallust, and of Epidius Rufus. In the large house of Pansa (Fig. 179), we find both a peristyle and a garden, the latter being at the rear of the peristyle; and in many houses a small garden was placed wherever available space could be found.

The peristyle is a garden enclosed by a colonnade, or having a colonnade on two or three sides. When this was higher on the north side than on the other three, as in the house of the Silver Wedding, the peristyle was called Rhodian. In the Tufa Period the colonnade was frequently in two stories, on all four sides or on the front alone. Fragments of columns belonging to the second story have been found in many houses, but in only one instance, that of the house of the Centenary, are they of such a character as to enable us to make an accurate restoration; here the double series of columns extended only across the front.

A separate entrance, posticum (Fig. 115), usually connected the peristyle with a side street. At the rear there was often a broad, deep recess, exedra, corresponding with the tablinum. The location of the other rooms in this part of the house is determined by so many conditions, and manifests so great a diversity that it may be spoken of more conveniently in connection with their use.

VII. Sleeping Rooms

The small, high rooms about the atrium were in the earlier times used as bedrooms; and such they remained in some houses, as that of the Faun, down to the destruction of the city.

The sleeping rooms about the peristyle were much lower, and the front opened by means of a broad door in its whole, or almost its whole, width upon the colonnade. These rooms could frequently be entered also through a small side door from a dining room, or a narrow recess opening on the peristyle (Fig. 146, x). The design of the arrangement is obvious. In summer the inconvenient large door could be left open day and night, a curtain being stretched across the space; in winter it would be opened only for airing and cleaning, the small door being used at other times.

The place for the bed was sometimes indicated in the plan of the room. In a bedroom of the house of the Centaur, of which an end view is given in Fig. 122, a narrow alcove was made for the bed at the left side; the floor of the alcove is slightly raised, and the ceiling, as often, is in the form of a vault, while the ceiling of the room is higher and only slightly arched. A similar arrangement is found in several other rooms decorated in the first style. In several houses, as in the house of Apollo, there is a sleeping room with alcoves for two beds.

In bedrooms with a mosaic floor the place for the bed is ordinarily white, being separated from the rest of the room by a stripe suggestive of a threshold. A similar division is often indicated in the wall decoration, particularly that of the second style; the part designated for the bed is set off by pilasters on the end walls, and differently treated both in respect to the decorative design and in the arrangement of colors.

VIII. Dining Rooms

As long as it was customary to sit at meals any fair-sized apartment could be used as a dining room. When the early Italic house was extended by the addition of a peristyle, and the Greek custom of reclining at table was introduced, it became necessary to provide a special apartment, and the Greek name for such a room with the three couches, triclinium, came into use. For convenience in serving, the length of a dining room, according to Vitruvius, should be twice the width. At Pompeii, however, the dimensions are less generous; with an average width of 12 or 13 feet the length rarely exceeds 20 feet. In many cases one end of the room opened on the peristyle, but could be closed by means of broad doors or shutters.

The plan of a typical dining room is given in Fig. 123. The couch at the right of the table was called the upper couch; that at the left, the lower; and that between, the middle couch. With few exceptions each couch was made to accommodate three persons; the diner rested on his left arm on a cushion at the side nearer the table, and stretched his feet out toward the right. Hence, the first on the upper couch had what was called 'the highest place.' The one next was said to recline 'below' him, because lying on the side toward which the first person extended his feet; the man at the outer end of the lower couch was said to be 'at the foot,' imus. When in the Gospel of John we read of a disciple "lying on Jesus' breast," the meaning is easily explained by reference to Roman usage; John was reclining in the place next below the Master. This arrangement makes clear to us the reason why the couches were so placed that the lower one projected further beyond the table than the upper one; the feet of those on the lower couch were extended toward the end furthest from the table.

To the couches grouped in the manner indicated the same name was applied as to the dining room, triclinium. Of those in the dining rooms only scanty remains are found. In summer the Pompeians, as the Italians of to-day, were fond of dining in the open air. In order to save the trouble of moving heavy furniture couches of masonry were not infrequently constructed in the garden, and have been preserved; such a triclinium is that in the garden of the tannery (p. 398). The arrangement is in most cases precisely that indicated in Fig. 123, the outer end of the lower couch projecting beyond the corresponding end of the upper one. In the middle stands the base of the table, also of masonry; the top is rarely preserved. Near by is a little altar for the offerings made in connection with each meal. The appearance of such a triclinium may be inferred from that of the triclinium funebre shown in Fig. 245, which has a square table and round altar.

In many gardens we find about the triclinium the remains of four or six columns. These supported a frame of timber or lattice-work, upon which vines were trained, making a shady bower, as in the garden of the tavern in the first Region, referred to below (p. 404).

The couches were ordinarily not provided with backs, but the outer ends of the upper and lower couches sometimes had a frame to hold the cushions, as indicated in Fig. 123 and shown more clearly in our restoration, Fig. 188. In the dining rooms small movable altars must have been used for the offerings, such as those of terra cotta or bronze not infrequently met with in the course of excavation. A fixed altar has been found in only one instance, in a small dining room in the eighth Region (VIII. v-vi. 16). Here, as our plan (Fig. 124) shows, the front of the apartment is set off as an anteroom, and in this was placed an altar of tufa.

In accordance with an ancient custom the children, even those of the imperial family, sat on low stools at a table of their own on the open side of the large table. In an open-air triclinium in the ninth Region (IX. v. 11) the children's seat is preserved, a low bench of masonry about forty inches long connected with the projecting arm of the lower couch (Plate VII.).

The inner part of the dining room, designed for the table and couches, was often distinguished from the free space in the same way that the place for the bed was indicated in bedrooms, sometimes by a difference in the design of the mosaic floor, more frequently by the division of the wall decoration and the arrangement of the ceiling. In the third and fourth decorative styles the division is less plainly marked than in the second; but often the side walls back of the couches and the inner end of the room have each a single large panel with a small panel at the right and left, while on each side wall in front are only two panels, of the same size.

In one respect the ordinary dining room was far from convenient; those who had the inner places could not leave the table or return to it in the course of a meal without disturbing one or more of those reclining nearer the outside. Large rooms, in which an open space was left between the couches and the wall, or in which several tables with their sets of couches could be placed, were unknown in pre-Roman Pompeii. In the time of the Empire a few of these large dining rooms were built in older houses. There is one measuring about 25 by 33 feet in the house of Pansa; another, of which the dimensions are 23 by 30 feet, in the house of Castor and Pollux; and a third, 36 feet long, in the house of the Citharist.

In a number of houses we find a large, fine apartment—designated by the Greek word oecus—which seems often to have been used for a dining room, especially on notable occasions. A particularly elegant form was the Corinthian oecus, which had a row of columns about the sides a short distance from the walls, the room being thus divided into a main part with a vaulted ceiling and a corridor with a flat ceiling. The couches would be placed in the main part; the guests could pass to their places along the corridor, behind the columns. The remains of such an oecus may be seen in the houses of Meleager and of the Labyrinth.

A specially interesting example—unfortunately not yet wholly excavated—is in the house of the Silver Wedding. In this case only the inner part, designed for the couches, is set off by columns. We may assume that there was a vaulted ceiling over the middle, resting on the entablature of the columns; that the ceiling of the corridor between the columns and the wall was flat, and of the same height as the entablature; and that the front part of the room had a flat or slightly arched ceiling of the same height as the crown of the vault over the middle.

In the more pretentious Roman houses there was sometimes a dining room for each season of the year; when Trimalchio in Petronius's novel boasts that he has four dining rooms, we are to understand that he had one each for winter, summer, autumn, and spring. In the case of the Pompeian houses we are warranted in assuming that dining rooms opening toward the south were for winter use, those toward the north for use in summer. Other airy apartments, with a large window in addition to the wide door, may well have been intended for summer triclinia. Further than this it is hardly possible to classify Pompeian dining rooms according to the seasons.

IX. The Kitchen, the Bath, and the Storerooms

In the Pompeian house the kitchen had no fixed location. It was generally a small room, and was placed wherever it would least interfere with the arrangement of the rest of the house.

The most important part of the kitchen was the hearth. This was built of masonry, against one of the walls. It was oblong, and the fire was made on the top. The cooking utensils sometimes rested on rectangular projections of masonry, as in the kitchen of the house of Pansa, sometimes on small iron tripods, as in the house of the Vettii (Fig. 125). The hearth of the latter house was found undisturbed, with a vessel in place ready to be heated. In one house the place of an iron tripod was taken by three pointed ends of amphorae set upright on the hearth. Underneath there was often a hollow place, like that shown in our illustration, in which fuel was kept, as in similar openings under the hearths of Campanian kitchens to-day.

Sometimes we find near the hearth a bake oven, not large enough to have been used for bread, and evidently intended for pastry; bread must ordinarily have been obtained from the bakers. In one of the cellars of the house of the Centenary there is a larger oven, which may have been used to bake coarse bread for the slaves; the heat was utilized in warming a bath above.

Over the hearth was a small window to carry off the smoke. As the kitchen was ordinarily high there may have been a hole in the roof also, but the upper parts have been destroyed, and their arrangement cannot be determined. From the small size of the kitchens and of the hearths in even the largest and finest houses, we may infer that the luxury of the table prevalent in the Early Empire had made only slight progress at Pompeii.

Close by the kitchen, frequently forming a part of it and next to the hearth, was the closet; a separate closet of good size is found in the houses of the Faun and of Castor and Pollux.

In connection with this group of rooms we may mention the storerooms, which are found in various parts of the houses and may be identified by the traces of the shelves that were fastened to the walls.

Comparatively few houses were provided with cellars. In the house of the Centenary, however, there are two. One, entered from the atrium by a stairway, extends under the tablinum and the front colonnade of the peristyle; the other is accessible from a side atrium and is divided into several rooms, in one of which is the oven mentioned above. The cellar belonging to the house of Caecilius Jucundus is under the garden; that of the villa of Diomedes will be described later.

X. The Shrine of the Household Gods

In ancient Italy each household worshipped its guardian spirits and tutelary divinities, which formed a triple group, the Lares, the Penates, and the Genius. In Pompeii the remains associated with domestic worship are numerous and important.

Many Pompeians painted representations of the household gods upon an inner wall, often upon a wall of the kitchen, near the hearth. There was usually a painted altar underneath, with a serpent on either side coming to partake of the offerings.

In a large number of houses a small niche was made in the wall, in which were placed little images of the gods, the Lares and the Genius being also painted on the back of the cavity or on the wall at the sides or below. Such a niche may be seen in a corner of the kitchen in the house of Apollo (Fig. 126); the pictures of the gods are almost obliterated, but that of the serpent—in this case there is but one—and of the altar can be clearly seen. In front is a small altar of masonry; the ferns and grasses with which the floor is carpeted make this kitchen in summer an attractive nook. Sometimes the niches were ornamented with diminutive half-columns or pilasters at the sides and a pediment above.

Frequently a more elaborate shrine was provided, a diminutive temple raised on a foundation, placed against a wall of the atrium or of the garden. An example is the one at the rear of the peristyle in the house of the Tragic Poet (Fig. 153).

In rare instances a small, separate chapel was devoted to the domestic worship, as in the house of the Centenary. In a house of the ninth Region (IX. viii. 7) there is such a chapel in the garden, a niche for the images being placed in the wall.

The Lares are the guardian spirits of the household. Originally but one was worshipped in each house; they began to be honored in plurality after the time of Cicero, and at Pompeii we invariably find them in pairs. They are represented as youths clad in a short tunic confined by a girdle (Fig. 127), stepping lightly or dancing, with one hand high uplifted in which a drinking horn, rhyton, is seen; from the end of the horn a jet of wine spurts in a graceful curve, falling into a small pail, situla, or into a libation saucer, patera, held in the other hand.

Simple offerings were made to these beneficent spirits,—fruits, sacrificial cakes, garlands, and incense,—and at every meal a portion was set aside for them in little dishes. When a sacrifice was offered to the Lares, the victim was a pig.

With the worship of the Lares was associated that of the Genius, the tutelary divinity of the master of the house. He is represented as a standing figure, the face being a portrait of the master. The toga is drawn over his head, after the manner of one sacrificing; in the left hand there is usually a cornucopia, sometimes a box of incense, acerra; with the right hand he pours a drink offering from a patera.

Very rarely we find a representation of the Genius of the mistress of the house. In one painting she appears with the attributes of Juno; the Genius of a woman was often called Juno, as in the inscription on the bust stone of Tyche, the slave of Julia Augusta (p. 418). As a man might swear in the name of his Genius, so a woman's oath might be 'By my Juno.'

The Lares and the Genius are often found together both in the hearth paintings, and in the groups of little bronze images frequently placed in the shrines. They are associated also in an inscription on the shrine in the house of Epidius Rufus: Genio M[arci] n[ostri] et Laribus duo Diadumeni liberti,—'To the Genius of our Marcus and the Lares; (dedicated by) his two freedmen with the name of Diadumenus.' Marcus was the first name of the head of the household.

In a few cases the Genius of the emperor seems to have been revered at a house shrine. Horace (Od. IV. v. 34) speaks distinctly of the worship of the tutelary divinity of Augustus in connection with that of the Lares,—et Laribus tuum Miscet numen. On the rear wall of a little chapel in a garden is a painted altar at the right of which stands Jupiter, at the left a Genius, each pouring a libation. We can scarcely believe that the Genius of an ordinary man would thus be placed as it were on an equality with the ruler of heaven; more likely the Genius of an emperor is represented, perhaps that of Claudius. The face is not unlike the face of Claudius, and the painting is on a wall decorated in the third style (Ins. VII. xi. 4).

In another house (IX. viii. 13) two Genii are painted, and under one of them is scratched in large letters EX SC, undoubtedly for ex senatus consulto,—'in accordance with a decree of the Senate.' We are probably safe in assuming that the decree referred to is that of the reign of Augustus, by which the worship of the Lares was regulated (Dio Cass. LI. xix. 7); if so, the figure is intended to represent the Genius of that emperor.

The face of the Genius in the house of the Vettii (Fig. 127) bears a decided resemblance to that of Nero. Here the shrine was placed in the rear wall of the smaller atrium. It consists of a broad, shallow niche, the front of which is elaborately ornamented to give the appearance of a little temple, while on the back are painted the household divinities. The Genius stands with veiled head between the two Lares, holding in his left hand a box of incense and pouring a libation with the right. In the original painting the features were unusually distinct.

The Penates were the protecting divinities of the provisions or stores, penus, and the storerooms of the house; under this name were included various gods to whom the master and the household offered special worship. At Pompeii the Penates, as the Lares and the Genius, appear in paintings, and are also represented by bronze images placed in the shrines. In the shrine of the house of Lucretius were diminutive bronze figures of the Genius and of Jupiter, Hercules, Fortuna, and another divinity that has not been identified. Statuettes of Apollo, Aesculapius, Hercules, and Mercury were found, together with those of the two Lares, in another house; in a third, Fortuna alone with the Lares.

Jupiter and Fortuna are frequently met with in shrine paintings, as well as Venus Pompeiana (Fig. 4), Hercules, Mars, and Vulcan as a personification of the hearth fire; Vesta, the patron goddess of bakers, usually appears in the hearth paintings of bake shops.

Underneath the representations of the Lares and Penates ordinarily are painted two serpents, one on either side of an altar, which they are approaching in order to partake of the offerings; these consist of fruits, in the midst of which an egg or a pine cone can usually be distinguished. As early as the beginning of the Empire the significance of the serpent in the Roman worship had ceased to be clearly understood; Virgil represents Aeneas as in doubt whether the serpent which came out from the tomb of Anchises was the attendant of his father or the Genius of the place (Aen. V. 95).

In the Pompeian paintings, when a pair of serpents occurs, one may usually be recognized as a male by the prominent crest. They were undoubtedly looked upon as personifications of the Genii of the master and mistress of the house. When a single crested serpent appears, as in the shrine paintings of both the house of the Vettii (Fig. 127) and the house of Apollo (Fig. 126), we are to understand that the head of the household was unmarried.

XI. Second Story Rooms

With few exceptions the houses of pre-Roman Pompeii were built in only one story; where the peristyle was in two stories, there must have been rooms opening upon the upper colonnade. In Roman times, as the population of the city increased and more space was needed, it became a common practice to make the rooms about the atrium lower and build chambers over them. A complete second story was rare; small rooms were added here and there, frequently at different levels and reached by different stairways. Sometimes the second story on the front side projected a few feet over the street; an example may be seen in a house in the seventh Region (casa del Balcone Pensile), the front of which, with the part projecting over the sidewalk, has been carefully rebuilt by replacing the charred remains of the ancient beams with new timbers.

Houses with three stories were quite exceptional, and the rooms of the third floor must have been unimportant. Along the steep slope of the hill, on the west and southwest sides of the city, a number of houses are found that present the appearance of several stories; they are not properly classed with those just mentioned, however, for the reason that the floors are on terraces, the highest at the level of the street, the others lower down and further back, being adjusted to the descent of the ground.

From the time of Plautus, second story rooms were designated as 'dining rooms,' cenacula. Varro says that after it became customary to dine upstairs, all upper rooms were called cenacula. This explanation is not altogether satisfactory, because other literary evidence for the prevalence of such a custom is lacking. Perhaps in early times, when, on account of the introduction of the compluvium and impluvium, the atrium ceased to be convenient and comfortable for the serving of meals, a dining room was frequently constructed on an upper floor, and, being the principal second story apartment, gave its name to the rest. In some places the ancient custom may still have lingered in the time of the Early Empire.