Tag Archives: Roman Building Types

Roman Building Types

Roman Building Types | Roman Architecture | Second edition

Roman Building Types | Roman Architecture | Second edition

Roman Building Types | Roman Architecture | Second edition

A purely chronological and geographical approach neglects the development of particular building types. Therefore their historical and geographical development is summarised in this chapter.

Town planning

A surveyor (mentor or agrimensor) used a surveying staff (groma)1 to sight along the line to be laid out (Figure 4.2a). This method was used to survey the two main axes of a city, the documents Maximus (main EW street) and the Cardo Maximus (main NS street), which usually intersected in the middle. From these, the lesser streets, parallel to them, could be surveyed.

The groma was also used to survey the territory around the city, commonly in squares measuring 2400 × 2400 Roman feet, theoretically the equivalent of 100 small holdings. Levels were also used by surveyors, especially when laying out aqueducts (Figure 42b). In the center of the city were the basilica, forum, and main temple, often built as a single complex over three city blocks, the so-called ‘tripartite forum’.2

The temple, surrounded on three sides by colonnades, faced across the forum to a basilica built across one short side.

Early examples include Augusta Bagiennorum (Benevaggiena), 5–4 bc; Lugdunum Convenarum (St.-Bertrand-de-Comminges), Augustan or Tiberian; and Velleia near Modena, Augustan or Tiberian. Other examples are Augusta Raurica (Augst in Switzerland) (Figure 2.1), Virunum (Zollfeld in Austria), Bagacum (Bavay in France), Lutetia (Paris), Conimbriga (Portugal) and Herdonia (Ordona), while a variation of the plan can be seen at Leader (Zadar in Croatia).

Religious buildings

Frontality, a feature of Etruscan temples, was assimilated into Roman temple design (Figure 1.1b). For example, in the fifth-century bc Belvedere temple at Orvieto dominated the rectangular enclosure in front of it and had columns only at the front. Large temples like the late sixth century bc Capitoline temple in Rome had columns along the sides as well (Figure 1.1a).

Etruscan temples usually had overhanging wooden roof beams decorated with brightly painted terracotta revetments, often with terracotta statuary on the roof. Up to the second-century bc, the Romans followed the Etruscan design and building methods.

The Capitolium (c. 150 bc) at Cosa, a Roman colony in Tuscany founded in 273 bc, had three cells at the rear, with the staircase and all of the columns at the front (Figure 1.8). It was usually reconstructed with overhanging wooden roof beams in the Etruscan fashion.

In 146 bc Greece became part of the Roman Empire and, in the words of Horace, ‘conquered Greece took her conqueror captive and introduced her arts into rustic Latium’ (Epist.

Figure 2.1  Augusta Raurica (Augst, Switzerland), restored view of the center of the town, mid-second century ad with later modifications. (After A. Boethius and J. B. Ward Perkins, Etruscan and Roman Architecture [Harmondsworth: Penguin, 1970], figure 131.)


2.1.156–7). However, although the Romans adopted the Classical orders, the temple layout remained traditional. The Greeks preferred a low stylobate, usually with three steps, whereas Roman temples still stood on a high podium. In Greek temples, the columns and steps ran all around the building so that every side had equal value and there was no strong emphasis on the façade.

A good example is the Parthenon in Athens. When first seen through the Propylaea (entrance gateway) the back of the temple faced the viewer; the entrance was at the other end. Also, it is noticeable that the Parthenon was not set axially within its enclosure.

As in many other Greek temples, the first view of it was a three-quarter view. The Romans, on the other hand, almost invariably continued to emphasize the frontal aspect of temples in the old Etruscan manner, even though they were now using the Classical Greek orders. A common layout was to have columns only at the front, an arrangement termed ‘prostyle’.

For example, the Temple of Hercules at Cora (Cori), built around 100 bc, was a prostyle temple with a high podium but used a pure Hellenistic Doric order (Figure 1.18). The late second-century bc Temple of Portunus in the Forum Boarium (Figure 1.16) used the Ionic order but was also a prostyle temple, although it did have half-columns attached to the cella walls all round, a system called ‘pseudo-peripteral’.

The circular temple of Vesta at Tivoli, ancient Tibur (Figure 1.17), used an Italic version of the Corinthian order, but the columns stood on a high podium. In each case the Classical order was Greek, but the layout of the temple was Etruscan or what is often termed ‘Italic’.

The Maison Carrée at Nemausus (Nîmes), one of the best-preserved of all Roman temples, is a fine example of this (Figure 2.2). A material associated with Greek temples, particularly in Athens, was marble. It came into use in Rome in the second-century bc, for example, in the Temple of Hercules Victor in the Forum Boarium dating to about 120 bc (Figure 1.16). Once the Luna (Carrara) quarries were opened by Julius Caesar, the white marble was used on a large scale, especially at the time of Augustus.

The Roman temple was not adopted as a model by the early Christians for their places of worship because, like the Greek temple, it was designed to house only cult images, not congregations. Christian worship required a great deal of interior space to accommodate the

Figure 2.2  Nemausus (Nîmes, France), Maison Carrée, begun c. 19 bc.

the congregation, which led to the Roman basilica being chosen because it could be most easily adapted to Christian needs. This does not mean that there were no congregational buildings before Christianity.

The underground basilica near Porta Maggiore in Rome was built for a neo-Pythagorean sect in the first-century ad and has a nave flanked by two aisles with an apse at the end, an arrangement reminiscent of the Christian church. Temples of Mithras (mithraea) followed a similar pattern. A long nave flanked by banqueting couches terminated in a recess containing a sculpture or painting of Mithras slaying the bull in the cave.

Public buildings

As Rome grew from a small town into the capital of a great empire her institutions became increasingly complex. Law courts, money exchanges, treasuries, record offices, and assembly places had to be built in provincial towns as well as in the capital.

These and other public buildings stood in the forum (marketplace), an open space usually rectangular in shape and often surrounded by colonnades on one story or two, as at Pompeii (Figure 6.2). Official weights and measures were often kept there, as at Pompeii.

Important temples faced it, as well as various public offices and the meeting places of the curia (town council) and securities (councilors). Shops (tabernacle), bars (Thermopolis), and sometimes a macellum (provisions market) were built nearby.


The basilica, an aisled building with clerestory lighting and a high, roofed nave, acted both as money exchange and a law court. It was surrounded by aisles, usually on all four sides and the tribunal for the magistrate’s court faced the entrance.

Its position depended upon whether the long or short side faced onto the forum area. Examples of both types of the basilica are common, the Pompeii one (Figure 2.3) presenting its short side to the Forum, and the two in the Roman Forum their long sides (Figure 3.5).

Vitruvius tells us he designed a basilica of the latter type at Fanum (Vitruvius, de Arch. 5.1.6–10). Trajan’s Basilica Ulpia with its twin apses (Figure 3.2) and positioned across the main axis of the Forum represents a striking variation of the normal basilican plan, and one which was followed by Septimius Severus in his forum at Lepcis Magna (Figure 9.13).

The Basilica of Maxentius in the Roman Forum (Figures 12.17 and 12.18) was unusual in that its layout was based on the frigidarium of a Roman bath. However, Christian churches, such as St. John Lateran (begun ad 312), were based on the conventional Roman type.

The word basilica is a Greek adjective meaning ‘kingly’ and the noun it originally qualified is likely to have been stoa (Strabo, Geo. 5.3.8). Although several Roman basilicas in some ways resembled stores, they were quite different internally.

They had the advantage of being a covered hall where magistrates could conduct their cases uninterrupted, unlike stoas, which are porticoes open to the weather and the noise of the street. Despite the Greek name, the earliest known basilicas are found in Rome.

However, it is often assumed that the Romans encountered them in a Greek context and adopted them because they had no other building so well suited to their legislative and commercial needs. There is no reason why the Greeks should not have developed such a building. Buildings of the same general type were known in Delos and Alexandria, and Vitruvius’ Egyptian oecus has a similar layout (De Arch. 6.3.9).

Figure 2.3  Pompeii, basilica, 100–90 bc.

Commercial buildings

The oldest known warehouse was the early second century bc Porticus Aemilia (Figure 1.13), a vast building composed of row after row of barrel vaults pierced with arched openings to create a roofed enclosure measuring 48.7 × 87 meters.3

Numerous warehouses have been found at Ostia, of which the best known is the Horrea Epagathiana (c. ad 145–150) with its fine dressed-brick doorway which opens into a courtyard with storage rooms on all four sides (Figure 4.11). Later warehouses at Ostia were more compact and had rows of rooms back to back (Figure 6.22c).

The simple one-roomed shop (taberna) had a wide opening to the street, almost as wide as the shop itself, which could be closed using removable slats and a wooden door. It was lit by a window above the door. Inside there was often a mezzanine floor for storage or sleeping, accessible using a staircase.

A wide variety of shops survive in Pompeii and Ostia, such as bakeries, which had counters near the street where the bread was sold, while at the back of the shop there was storage space for the grain, mills for grinding it, tables for kneading the dough and ovens for baking the bread.

Fish shops found at Ostia had marble slabs for preparing the fish, water tanks, and sometimes facilities for cooking them. Fulleries (fullonica) had presses and complex systems of washing tanks. Shops that sold food and drink were called Thermopolis.

They had solid masonry counters with wide-mouthed jars (dolia) sunk into them, where the foodstuffs were kept (Figure 2.4). The counters were usually L-shaped for sales both on the street and in the shop. Although Roman writers, such as Juvenal (Sat. 3.8.171–8) or Horace (Sat. 2.4.62), stressed the squalor of such places and delighted in a gluttonous emperor like Vitellius frequenting Thermopolis (Suet., Vit. 13.3), the importance of such outlets in the story of the Roman city has recently been pointed out.4

Figure 2.4  Pompeii, Pompeii, bar where hot food and drinks were sold (Thermopolis).

Shops were often built in rows facing the Forum or could occupy ground floor rooms in a house or insula, like the one at Terracina (Terracina) (c. 100 bc), which had a row of shops at street level and apartments above. The House of the Lararium at Ostia is a well-designed complex combining shops on the ground floor and apartments above.

Early shopping centers built around 100 bc can be seen at Tivoli, and at Ferentinum (Figure 1.15), which has a row of five shops opening off a barrel-vaulted concourse. In the early empire, when the Forum became a focus of civic pride, there was a tendency to remove shops, especially those which sold fish or animal products, to a macellum, an enclosure containing counters for the sale of fish or meat, with a circular pavilion in the middle equipped with a pool or fountain.

The first macellum seems to have been built by M. Aemilius Lepidus and M. Fulvius Nobilior in 179 bc (Livy, 40.51.5) and occupied the area where the Templum Pacis and the Forum of Nerva were later built. Good examples of macella can be seen at Pompeii, Puteoli (Pozzuoli), Lepcis Magna, Hippo Regius (Annaba), Cuicul (Djemila) and Thamugadi (Timgad).

The most important planned shopping center in the Roman world was the Markets of Trajan, a multi-story complex built behind Trajan’s Forum (Figures 8.3–8.6) which contained about 150 shops, as well as a substantial warehouse and administrative center.

Domestic buildings

The Romans at first borrowed heavily from both Etruscans and Greeks when it came to domestic buildings. Etruscan houses probably resembled rock-cut tombs such as the Tomb of the Shields and Thrones at Caere (Cerveteri) or the Tomb of the Volumnii at Perugia (Figure 1.2a). Excavations along the Sacred Way in Rome have shown that large aristocratic houses of cappellaccio stone were built on the north slopes of the Palatine at the end of the sixth-century bc.

At Pompeii, the focal point of the house was a large room (atrium) with an inward-sloping, rectangular opening in the middle of its roof (compluvium). A shallow rectangular basin (impluvium) was set into the floor immediately below the compluvium to catch the rainwater from the roof, which then ran into a vaulted underground water cistern (Figure 6.7).

At the edge of the impluvium was a cistern head from which the water could be drawn. However, excavations in the oldest atrium houses at Pompeii, such as the House of the Surgeon, have shown that the impluvium is a secondary feature, dating to the second-century bc. This fact has given rise to considerable controversy about the origin of the atrium.

Whether this means that the original atrium was testudinate, completely roofed over like the Tomb of the Volumnii at Perusia (Perugia), and had a hearth in the middle of the floor, or was an open courtyard is a matter of debate. The opening or compluvium may have been a practical step to allow the smoke from the hearth to escape, which is, incidentally, an argument for connecting the word ‘atrium’ with the Latin ater (black). Seneca (Epist. 44.5) remarked that an atrium ‘crammed with smoke-blackened images’ was a sign of the old nobility.

There were three other types of atrium apart from the testudinate and complicate types: the diluvial atrium, where the roof around the opening in the roof sloped out rather than in; the tetrastyle, with a column under each of the four corners of the compluvium; and the Corinthian, with a row of columns which gave it more the character of a small peristyle than an atrium (Vitruvius, de Arch. 6.3.1–2).

The main external feature of an atrium house was the front door, which was often of great height and singled out for special decoration. The passageway (fauces) into the atrium did not normally give access to the rooms on each side. These were usually either service rooms opening into the atrium or shops (tabernae) opening onto the street (Figure 2.5). As the principal rooms drew most of their light from the atrium (Figures 1.2b and 6.7), windows

Figure 2.5  Pompeii, House of the Painted Capitals, second-century bc: plan.

seldom found on the street facade, and those which did exist were usually high up or barred. There were two or three bedrooms on each side of the atrium and at the far end was a pair of recesses or wings (alae) which gave extra prominence to the three rooms opposite the front door.

Usually, the outer ones were dining rooms (triclinium), while the central one was the principal room of the house, the tablinum, and the main reception room, which was also used for records and personal documents.

Roman kitchens were small, rather dingy rooms, containing a large masonry bench on top of which the cooking was done, and underneath which the cooking pots were stored. There was usually a water basin in a corner and the overflow often flushed the lavatory, which was in an adjoining room.

Until the second century, bc Roman houses must have greatly resembled Etruscan ones. However, they developed a great deal during the second-century bc, a time when the Romans were very much involved in Greece and many Italian merchants set up businesses there, particularly on the island of Delos. It was then that peristyles of the kind found in houses at Rhodes and Delos were introduced to Italy.

However, the peristyle did not take the place of the atrium, which was a venerable room showing the wealth of the owner by the area (money chest) which was kept there; the lararium, where the owner’s household gods were kept; and the imagines, which indicated pedigree.

Instead, the peristyle was built behind the atrium, as in the House of the Faun (Figure 6.5) or the House of the Painted Capitals (Figure 2.5), both of which had two peristyles. Whereas the center of the Greek peristyle was usually paved or covered with mosaic, the Romans preferred to make it a frigidarium (garden), as in the House of the Gilded Cupids (Figure 6.6).

Like many other building types, the Roman house was the product of outside influences, but the result was unmistakably Roman. The atrium became the public part of the house ‘which anybody could enter’ while the bedrooms and bathing rooms were kept more private (Vitruvius, de Arch. 6.5.1–2). The peristyle, a peaceful oasis far from the bustle and din of the atrium, was open to the public, but surely a more select public.

The slaves, on the other hand, were crammed into dark and dingy parts of the house. This can be seen in the House of M. Aemilius Scaurus near the Arch of Titus, built around 58 bc. Although the house itself was considered the height of luxury, the dark basement underneath contained 50 tiny cells with beaten floors and a stone bench for a bed.

In the second-century bc House of the Faun at Pompeii looked more egalitarian (Figure 6.5). There were two atria, Tuscan on the left and tetrastyle on the right. The latter was the slaves’ quarters, but behind the atrium, they rapidly tapered off, with the result that they occupied only a small fraction of the space available to the family.

The same pattern can be seen in the enormous villa at Oplontis, where the slaves were concentrated in a dingy area around the kitchen, which is distinguished by the simple zebra-stripe decoration of the walls.

In the House of the Menander at Pompeii the slaves’ quarters, shown hatched (Figure 6.8), were discreetly tucked away behind the main rooms so that the slaves would always be on hand to appear and disappear. It will be observed that their quarters occupied only a fraction of the total area of the house.

By the first-century ad pressure on land led to many large mansions being divided up into apartments. At the same time, the disastrous fire of ad 64 dramatically changed the appearance of the city of Rome. The old winding alleys lined with dangerous and dilapidated tenement houses were swept away and replaced by broad straight avenues of brick and concrete apartment blocks (insulae).

The use of timber was discouraged, and concrete vaulting became more common. Ostia underwent a transformation similar to that of Rome in the prosperous years following the opening of the Trajanic harbor, so that by the end of the second-century ad the majority of its inhabitants lived in apartments.

The Marble Plan (Figure 12.1) reveals a similar pattern in Rome, confirmed by the Regional catalogs, which show that by the fourth-century ad there were only 1,790 Domus in Rome compared to 46,602 insulae. Ostia is the best place to see insulae, with their plain, austere facades of brick-faced concrete, sometimes relieved by balconies over ground-floor rooms (Figure 6.20).

They rarely exceeded four or five stories, probably to comply with regulations governing building height, 70 Roman feet (20.72 meters) under Augustus and 60 feet (17.76 meters) under Trajan. Ground-floor rooms either faced an inner courtyard or were open to the street as shops, and staircases led directly to the upper rooms (Figure 6.21). A water cistern in the courtyard served all the residents of the block, and there was usually one lavatory on each floor.

Country residences ranged from working farms to large luxury villas used by their owners as an occasional retreat. The Villa Sambuco at San Giovenale is an example of the simplest type of farmhouse with timber and mud brick walls and earth floors.

Intermediate types include the late Republican villa Rustica at Boscoreale, which had a small, but luxurious residential section adjoining a large industrial area used for oil and winemaking, and the San Rocco Villa at Francoise in Campania (50–25 bc) with separate residential and working quarters of roughly equal size.

The Villa of the Mysteries at Pompeii was a suburban villa with a very small industrial section, which had many features of a large atrium/ peristyle house, except that it had more outward-facing porticoes, especially to the west, which commanded splendid views over the sea. Several villas at Tivoli were built on a series of terraces climbing the hill with fine views of Rome.

The Villa of the Papyri at Herculaneum revived as the J. Paul Getty Villa at Malibu was purely a pleasure villa built close to the sea. Another was Tiberius’ Villa Jovis at Capri (Figures 5.2 and 5.3). Seaside villas, illustrated in many Pompeian wall paintings, could be of the peristyle or the porticus type.

The Damacuta villa at Capri, with rooms opening off a long colonnade overlooking the sea, is an example of the latter. Pliny gives a detailed description of his seaside villa at Laurentium, near Ostia which contained a bath and a ‘hippodrome’ with a stadium (curved dining couch) at one end (Pliny the Younger, Epist. 2.17 and 5.6).

The very large Villa San Marco at Castellammare had an atrium, porticus, and peristyle with a large piscina in the middle. The enormous villa at Oplontis, whose full extent is still unknown, had a vast piscina, bathing suit, and seemingly endless porticoes.

Nero’s Golden House, set within a huge park that contained vineyards, woodlands, and a large lake, was essentially a villa, despite its urban surroundings. The surviving Esquiline wing (Figure 5.7) had a combination of features belonging to the terraced, the peristyle, and the porticus villa.

Even more complex was Hadrian’s Villa at Tivoli (Figure 8.13), whose baths, pools, libraries, nymphaea, and pavilions followed several unrelated axes and were scattered over an area that measured 1,000 × 500 meters. During the second century ad, there was a tendency to reject the sprawling landscape villa in favor of a more tightly planned complex following the pattern of the rather tower-like villa of Domitian at Albano.

Examples include the Villa at Sette Bassi and the Villa of the Quintilii on the Via Appia. However, the landscape villa never died out, as can be seen in the early fourth-century ad Villa at Piazza Armerina in Sicily (Figure 12.16). Outside Italy, the atrium is rarely found, but some splendid peristyle houses have been found at Visio (Vaison-la-Romaine) in France, and peristyle villas have been found at Glanum (St. Rémy).

There are some enormous villas in northern Europe, such as the late first-century bc villa at Fishbourne, with rooms ranging around a vast peristyle courtyard (Figure 10.14). Most British villas were later in date, such as the beautiful and extensive villa at Chedworth, which belongs mainly to the third and fourth centuries ad.

A large residential villa at Anthée near Namur was part of a large estate with rows of small working farm buildings nearby. One of the largest villas of northern Europe is at Nennig in Germany (third-century ad) with its tower-like wings enclosing a colonnaded central section.

Recreational buildings

Gymnasia provided facilities for exercise and sometimes for swimming. One of the largest is the Augustan palaestra at Pompeii, which measures 141 × 107 meters. There were also facilities for ball games, running, and wrestling in the Campus Martius (Strabo, Geo. 5.3.8).

The best-known recreational buildings were baths, which offered facilities for exercise as well as for bathing. Women attended the baths in the morning and men at about the eighth hour (2:00 p.m.), which was considered the best time to bathe (Martial, 10.48.1–4), although some were forced by the pressure of business to bathe later (Suet., Aug. 76.2).

The normal bathing procedure was for bathers to change in the apodyterium, oil themselves and exercise in the palaestra (Pliny the Younger, Epist. 9.36.4). They then passed into the sweating rooms (sudatorium), after which they took a hot bath in the caldarium, cooled down in the tepidarium and passed into the frigidarium and notation (swimming pool) for a cold dip.

However, the bath was not so pleasant for those who had to live nearby. Seneca lived over a public bath and talked of the noise made by people exercising and swinging weights, the groans they made, their hissing and gasping, the noise of flesh being slapped, and the shouts of people splashing in the water.

He describes the hair plucker with his shrill, high-pitched voice and the shrieks of clients when their armpits were being plucked. In addition, there were the cries of people selling drinks, sausages, and pastries (Seneca the Younger, Epist. 56.1–2).

Greek baths were very simple, consisting of rows of sitz baths filled and emptied by hand. An important link between Greek and Republican Roman baths has been revealed by excavations in the Stabian baths at Pompeii, where a row of sitz baths dating to the fourth-century bc was found on the north side of the palaestra. The water for these early baths came from a deep well and was raised by a water wheel.

The baths were enlarged in the later second-century bc when a tepidarium (warm room), caldarium (hot room), and apodyterium (changing room) were added along the east side of the palaestra (Figure 2.6). The hot rooms were fitted with hypocausts, or underfloor heating, an invention attributed by Pliny (Nat. Hist. 9.79.168)

Figure 2.6  Pompeii, Stabian Baths, second-century bc: plan.


to Sergius Orata, a Campanian who lived in the late second-century bc. At the same time, the walls of the hot rooms were lined with tegulae mammae which created a hollow space through which the hot air from the hypocaust could rise, providing an additional source of heat.

Hot rooms usually faced south to take advantage of the sun and the furnaces heated metal tanks behind the plunges. The hot water flowed into the plunge while the cold water settled back into the tank to be reheated.

Despite these improvements, heat transfer was still poor and windows remained small to conserve heat. The result was that Republican baths were poorly lit even on bright days, as Seneca observes, comparing the small dark bath in the villa of Scipio Africanus with the luxurious ones of his day (Seneca the Younger, Epist. 86.6–12).

Scipio’s baths would probably have looked rather like the Stabian Baths in Pompeii, which were small and dark compared with the later Imperial baths. The complex (Figure 2.6) consisted of a palaestra, although Vitruvius points out that this was not in the Italian tradition (de Arch. 5.11.1), an apodyterium (changing-room), a frigidarium (cold-room), a tepidarium (warm-room) and a caldarium (hot room).

In older baths, the caldarium had an apse with a stone basin for cold water (schola labri). Vitruvius says (de Arch. 5.10.5) that a laconicum (dry hot-room) should adjoin the tepidarium. It should be circular with a hole in the dome which should be hung a bronze disc suspended on chains.

This could be raised or lowered to regulate the temperature of the room. Such circular rooms are found in the Stabian and Forum baths at Pompeii but are thought to have become cold rooms because there is no trace of underfloor heating in them.

Three remarkable transformations altered the appearance of bath buildings in the first-century ad: the growing number of aqueducts that provided the running water essential for plunge baths and swimming pools; the invention of window glass; and the introduction of hollow wall tubes, which allowed more hot air to circulate than tegulae mammatae (tiles with lugs or nipples).

Hot gases were drawn under the hollow floor, up the tubes lining the walls, and out through chimneys in the vault (Figure 2.7). As the tubes heated up more quickly than the floor, which was usually 0.40–0.50 meters thick, the walls became a significant source of radiant heat.

The importance of these developments cannot be understated as they caused almost a revolution in the appearance of Roman baths in the late first-century ad. Many of these changes were taking place at the time of Nero (ad 54–68) when so much else that was novel was happening in architecture.

Unfortunately, Nero’s baths in Rome (ad 62–64) were later rebuilt, a great loss because they may well have been the first to take advantage of these technical innovations, and may have pioneered the double-circulation system used in the Baths of Titus 15 years later.

Although little survives, the plan of the Baths of Titus is known from Palladio (Figure 8.1). The main block was symmetrical about its shorter axis and the rooms were arranged in the axial sequence later to become standard: caldarium, tepidarium, and frigidarium.

The frigidarium was flanked on each side by a palaestra and the caldarium projected from the block to take full advantage of the sun. The Baths of Trajan (ad 109) covered over six times the area of those of Titus and the bathing block was surrounded by a walled precinct containing libraries, halls, gardens, and a running track.

The addition of a swimming pool (natation) to the sequence of bathing rooms resulted in the two main axes of the building intersecting in the frigidarium. The third-century ad Baths of Caracalla (Figure 12.4) and of Diocletian (Figure 12.7) mark the culmination of the development of the double-circulation type of baths. Both followed much the same layout as the Baths of Trajan, except that the bathing block was completely detached from the surrounding precinct.

Baths of the double circulation type sometimes appeared in large provincial cities, some of the finest examples being the Hadrianic Baths at Lepcis Magna (Figure 9.12), the Antonine Baths at Carthage, and the Imperial Baths at Trier (Figure 12.13).

Greece and Asia Minor followed a different tradition from that of the west, with the main bathing block being planned on rectilinear lines. Usually, a considerable portion of the complex was taken up by a gymnasium, frequently featuring a marmoreal, a room lined on three sides with tiers of columns and niches with

Figure 2.7  Diagram to illustrate the heating system of a Roman bath.

an open columnar screen on the fourth. A good example can be seen in the Harbour Baths at Ephesus. Lucian gives a detailed description of a bath house of the eastern type designed by Hippias (Lucian, Hippias 5–8).


Roman Building Types | Roman Architecture | Second edition | (Part-02)

Roman Building Types | Roman Architecture | Second edition

The great imperial thermae represent only a small proportion of the baths built, many of which could be quite small. The latter were termed balnea or balneae, and there were 170 of them in Rome according to Agrippa’s census (33 bc) rising to 856 in the fifth-century ad according to the Regional catalogues.

Some were a continuation of the Pompeian type, with a single row of rooms along one side of a palaestra, as in the Baths of Neptune at Ostia, or the Baths at Glanum. A variation, with the rooms in an axial sequence but running perpendicular to the palaestra, can be seen at Champlieu and Conimbriga.

Smaller baths were often laid out on the ‘ring’ plan, with the rooms arranged in a closed, but not axial sequence. The endless variety of ground plans inherent in this type can be appreciated by comparing such diverse examples as Madaurus and Thence (both in Tunisia), the Hunting Baths at Lepcis Magna (Figure 9.16) and the south baths at Thamugadi (Timgad).5

Therapeutic and spa baths, like those at Baiae, Badenweiler and Aquae Sulis (Bath) featured large plunge baths fed by natural springs (Figure 10.12).


Roman emperors were careful to provide plenty of entertainment for the Roman populace because they knew that people were held in control by two things: bread and circuses (Fronto, Elements of History, 18). The most important buildings designed for mass entertainment in the Roman world were the theatre, odeum, amphitheatre, stadium and circus (or hippodrome).

The earliest theatrical performances date to the fourth-century bc and took place in front of or close to temples. Audiences in the early days were easily distracted. In the Mother-in-Law, a play by Terence (fl. 170–160 bc), the actor delivering the prologue laments that in the first performance of the play the audience was more interested in the boxers and the rope-dancer (Terence, Hec. 1–5).

In the second performance, the play broke up in disorder when a rumour circulated that there was to be a gladiatorial display. Things quietened down a little by the Augustan period when attendance at the theatre was very formal and tightly controlled.

In Italy, the first Roman theatres, such as those at Iaitas, Pompeii and Pietrabbondante, were based on Hellenistic theatres, but the fully developed Roman theatre was somewhat closer in design to the modern theatre.

Like Greek theatres, Roman theatres were composed of three elements: semicircular cavea (auditorium), orchestra and scene building (Figure 10.7), but, unlike those of the Greek theatre, the stage and scene building of the mature Roman theatre were joined to the auditorium and rose to the same height, for example, in the well-preserved theatre at Bostra (Bosra, Syria) (Figure 2.8).

This created a sense of enclosure made more emphatic by the vela (awnings) overhead to shade the spectators. Whereas the Greek orchestra was circular, in Roman times it became semicircular, because it had lost most of its importance by then, with the result that much of its space was taken up by the magistrates’ portable thrones (Visalia).

Above the lateral passages leading into the orchestra were the boxes (tribunal) for the presiding magistrates. Unlike the high Greek stage, the Roman stage (pulpitum) was broad and low, so that the magistrates in the orchestra could see what was going on.

In the front of the stage, there was usually a slot into which the curtain sank at the beginning of the performance. Behind the stage rose the scaenae frons (the front of the scene building), with its three doors for the entry and exit of the performers. It was richly decorated with two or three storeys of columns, niches, statues and honorific

Figure 2.8  Bostra (Bosra, Syria), Roman theatre.

inscriptions. Behind was the postcranium, an area used by the actors and for props. In the case of Orange, ancient Arausio, (Figure 10.6) the back wall (postscaenium) is particularly well preserved.

At the sides of the stage were large rooms (used as foyers) often richly decorated, such as those at Orange (Figure 10.7) which fitted against the sides of the cave uniting it with the stage building.

Most Roman theatres were built against a slope to save expense, but the Romans were perfectly capable of building on a flat site, the classic example being the Theatre of Marcellus in Rome (13–11 bc) with a fully built-up cave resting entirely on alternately annular and radial vaulted substructures (Figure 3.9).

These both supported the cavea and ensured easy circulation throughout the building. The stage of a Roman theatre was enormous, often 60 metres or more in length, a massive space in which performed spectacles such as popular plays, mime and pantomime, often accompanied by musicians and choirs.

The decoration of the sets as well as the costumes of the players were sumptuous, judged by wall paintings and descriptions (Apuleius, The Golden Ass 10.30–32). In the late Roman theatre, the orchestra was sometimes flooded to provide water spectacles.

As early as the first century bc valuable presents were thrown among the people by a magistrate (missile). These were often advertised in advance, as was the provision of awnings (vela) on a hot day. Benefactors donated money not only for the fabric of the building but also in support of the games (Ludi scenic), as numerous inscriptions attest.

The odeum, designed for musical performances and essentially a smaller version of the theatre, was usually roofed, for example, the odeum at Pompeii (c. 70 bc), which is referred to in an inscription as Theatrum tectum (roofed theatre).

Despite its early date, it had an astonishing internal span of 26.25 metres. Some exceptionally large odea were built, like the Odeum of Domitian at Rome (more than 100 metres wide) and the odeum at Carthage (96 metres wide).

Odea is said to have catered for a more refined taste than theatres, although a story is told of an odeum being emptied of its spectators because the fish market had opened (Strabo, Geo. 14.2.21).

Amphitheatres were constructed in much the same way as theatres, but with the important difference that they were not semicircular, but elliptical with an oval arena in the centre, as at Pompeii (Figure 6.9). The amphitheatre at Pompeii, the earliest known, was built against the town walls and upon an earth fill (aggestus), which means that access was limited and most spectators had to climb one of the two external staircases to reach their seats.

Amphitheatres were also equipped with vela to protect spectators from the sun. There are well-preserved remains of the mast holes for the vela in the Colosseum (Figure 7.1) and at Nemausus (Nîmes) (Figures. 10.8 and 10.9). At Capua and the Colosseum (Figure 7.3), there are also bollards at ground level around the building which were used to hold the winches for the ropes to control the vela.

Amphitheatrical spectacles were mainly of two kinds: animal hunts (venationes) and gladiatorial combat. Animal hunts became popular when more than 100 African elephants were captured at Palermo in 251 bc and taken to Rome to provide a public spectacle in the Circus Maximus (Polybius, 1.40).

An animal hunt is depicted on the walls of the frigidarium of the Hunting Baths at Lepcis Magna. As for gladiatorial games, there was a very old tradition of prisoners being forced to fight each other to the death on the tomb of a dead hero, to placate the gods of the underworld.

Such a spectacle is known to have taken place in the Forum Boarium at the funeral of Brutus Pera in 264 bc (Valerius Maximus, 2.4.7). Gladiatorial games were commonly held in the Roman Forum until fire damage caused them to be transferred to the Saepta Julia in 7 bc.

However, the enormous Colosseum (Figures 7.1 and 7.2), built between ad 75 and 80 and measuring 188 × 156 metres, became the venue par excellence. Accommodating about 50,000 spectators the Colosseum exceeded every other amphitheatre in size, although some are quite large, such as Capua, 165 × 135 metres, Italica, 157 × 134 metres, Augustodunum (Autun), 154 × 130 metres, Verona, 152 × 123 metres, Pozzuoli, 149 × 116 metres and Thysdrus (El Djem), 148 × 122 metres.

Later amphitheatres often had provision for storage of props and animals under the arena floor, and in the amphitheatres, at Capua and Puteoli (Pozzuoli) there was a complex of underground cells where the animals were kept until the show began (Figure 7.4).

Counterweights were used to haul the cages up to an upper level where the animals could escape through a series of trapdoors into the arena. The show was a dazzling one and included magnificent processions, exotic animals, executions of prisoners and even mock naval battles.

A spectacle lasting 100 days was put on for the opening of the Colosseum in ad 80 and 9,000 animals were killed. In Trajan’s games of ad 108 10,000 gladiators fought and 11,000 animals were killed.

Gladiators, mostly men but occasionally women, were usually slaves who had been trained at a gladiatorial school. The more agile became net-men (retiarii) armed only with a trident and a net and wearing very few clothes.

They were often pitted against heavily armed gladiators with swords, who depended on brute strength. Often the losers were killed, but not invariably because training a gladiator was a lengthy and expensive business. Some gladiators enjoyed the same kind of adulation that today is accorded to soccer players, as graffiti in Pompeii shows.6 However, spectators could easily become drunk with blood-lust, as St. Augustine warns (Confessions 6.9).

The circus or hippodrome was another well-established Greek building type, used in Roman times for races with four-horse (quadrigae) or two-horse (bigae) chariots, and sometimes for a variety of popular entertainments, including gladiatorial games and venationes.

It was by far the largest of all buildings used for entertainment in the Roman world, the length of full-sized circuses ranging from 400 to 620 metres. The Circus Maximus in Rome was the largest of all and it has been calculated that the arena of the Colosseum would fit into its arena about 12 times.

Banks of seats lined its two long sides and the curved end, and in the early third-century ad, it could hold 150,000 spectators.7  Usually 12 teams of chariots competed by running seven laps in an anti-clockwise direction around the arena which was divided by a median strip or spina.

The careers or starting gates were usually set out on a curve to allow each of the teams an equal chance at the start of the race to get through the narrow gap between the end of the spine and the arena wall. The optimum distance from the careers to the end of the spine has been calculated as 140–160 metres.

In more sophisticated circuses, such as the second-century ad circus at Lepcis Magna (Figure 2.9), the spine was set at an oblique angle to allow the teams more space at the crucial beginning of the race. A further refinement, seen in the early fourth-century ad Circus of Maxentius at Rome, was to angle the seats nearest the careers as well, to bring spectators even closer to the action.

Perhaps because it was most in view, the most lavish ornaments were ranged along the spine, and the finishing post was opposite the middle of it. Along its entire length, it had water basins and fountains, statues of victory and emperors, honorific columns and a shrine or altar to Consus (a chthonic deity associated with the circus).

In prominent positions towards its ends were the seven fishes or eggs which marked the number of laps completed. These were first set up by Agrippa in 33 bc in the Circus Maximus (Dio, 49.43.2). At each end were three cones, about 5 metres high, designed to give the charioteer warning of the turn.

Being a charioteer was a dangerous occupation, but successful ones were idolised by their fans, mobbed in the streets and entertained by the wealthy. One of the most popular charioteers of the later first-century ad, Scorpus, won 2,048 races, but died, probably in a crash (naufragium), when he was only 27 (Martial, Epigr. 10.50, 53).

His epitaph recorded that Fate counted up his victories and thought he must have been old. Horses and riders often suffered injury or even death (Sidonius Apollinaris, Poems 23.323–424). Races were between teams (factions) which raced under the colours red, blue, green and white.

People passionately followed their teams. There were frequent attempts to foul the opposition and sometimes curses were laid upon their horses. Cases are known where a fan committed suicide when his team lost.

The stadium, another well-established Greek type, was used for foot racing and other athletic competitions. Stadia are of similar shape to circuses and are often confused with them, but stadia are very much smaller. Their arena is normally close to a stade long (180–200 metres), and only c. 30 metres in width, compared to c. 70 metres for a circus.

Stadia were seldom built in the western provinces, although they are common in the east where the Greek tradition was still strong. Domitian was very keen on athletics and built a stadium in Rome, well-known, not as a ruin, but as the Piazza Navona.

Figure 2.9  Lepcis Magna (Libya), circus or hippodrome, second-century ad: restored plan.

Honorific monuments

There were so many monuments in the Roman Forum that in the year 158 bc the Censors ordered all statues to be removed except those authorised by a decree of the people or the Senate (Pliny, Nat. Hist. 34.14.30).

The oldest honorific monuments were columns carrying a statue, such as the column of L. Minucius (439 bc) and the Maenian column (338 bc) set up to commemorate the victory of C. Maenius at Antium (Pliny, Nat. Hist. 34.11.20).

In 254 bc during the First Punic War a column bearing the beaks of ships (columna rostrata) was erected on the Capitol in honour of the naval victory of M. Aemilius (Livy, 42.20.1).

A similar column of C. Duilius stood in the Forum. The row of equestrian statues which once stood in the forum of Pompeii represents another old tradition. Being granted an equestrian statue was a rare honour in Republican Rome, decreed for outstanding feats, such as the victories of C. Maenius and L. Furius Camillus in 338 bc.

During the imperial period, emperors were honoured by them, for example, the huge Equus Domitian in the Forum and the still-surviving equestrian statue of Marcus Aurelius.

The earliest known arches (fornices) were the three erected by L. Stertinius in 196 bc on which were placed gilt statues (Livy, 33.27.3–4). In 190 bc P. Cornelius Scipio erected an arch on the Capitol with seven gilt statues and two equestrian ones (Livy, 37.3.7).

The Fornix Fabianus was erected by Q. Fabius Maximus in 121 bc and rebuilt by his nephew in 57 bc. Augustus erected two arches in the Roman Forum, a single arch in 29 bc and one with a lintelled passageway on each side in 19 bc (Figure. 3.8).

Some imperial arches had a single vaulted passageway, for example, the Arch of Titus in the Roman Forum (Figure 7.9) and the Arch of Trajan at Benevento (Figure 8.7); others, like the arch at Orange, had three, the central one being wider and higher than the flanking ones (Figure 2.10).

The most elaborate triumphal arches, such as the Arch of Septimius Severus (Figure 12.2) and the Arch of Constantine (Figure 2.11), had column plinths adorned with victories, soldiers

Figure 2.10  Arausio (Orange, France), ‘Arch of Tiberius’, c. ad 19.

Figure 2.11  Rome, Arch of Constantine, ad 312–315.

and prisoners, keystones containing divinities and spandrels with flying victories. The frieze usually contained a triumphal procession and in the attic was the dedicatory inscription. In the most richly decorated arches, the soffits of the vaulted passages were coffered.

The piers flanking the arched openings, the sides of the arch, parts of the attic, and the walls of the passageways sometimes contained sculpted panels with scenes of triumph, imperial providence, sacrifice and apotheosis. Often bronze figures of horsemen, four-horse chariots, divinities, trophies and barbarians stood in the attic.

Four-sided arches were placed over cross-roads and were covered with a cross-vault, as in the ‘Arch of Janus’ in Rome (ad 315), or a cupola, as in the Arch of Marcus Aurelius in Oea (Figure 9.17). The Arch of Augustus at Rimini, built in place of a city gate to commemorate the completion of the Via Flaminia (27 bc), is the oldest surviving triumphal arch.

Arches are also found at the ends of a bridge, as in Saint-Chamas in France, or the middle, as at Alcantara in Spain (Figure 10.3). The double arch at Saintes in France stood at the entrance to a bridge. Double arches sometimes flanked by towers, belong to city gates, such as those at Nemausus (Nîmes), Augustodunum (Autun), Ravenna, Hispellum (Spello), the Porta Palatina at Augusta Taurinorum (Turin) and the Porta Nigra at Augusta Treverorum (Trier) (Figure 12.11).

Utilitarian buildings

Strabo praises the Romans for paving their roads, constructing aqueducts and building sewers (Strabo, Geo. 5.3.8). It was not just Strabo who admired them. Dionysius of Halicarnassus, a Greek historian who lived in Rome at the time of Augustus, thought that they were the three most magnificent indicators of the greatness of the Roman Empire (Ant. Rom. 3.67.5).

They all provided essential services to modern eyes are some of the most impressive Roman achievements. The Romans probably inherited their skills from the Etruscans, builders of the famous Cloaca Maxima (great sewer) under the Roman Forum which partly functions to this day.

Roman roads, which eventually connected the entire Roman Empire, placed great demands on their builders in terms of planning, surveying, engineering and organisation. The route a road such as the Via Appia (Figures 1.4 and 1.10) was to take had first to be surveyed, a vast undertaking in itself. The road bed had then to be laboriously prepared as Statius describes (Silv. 4.3.40–55).

Two furrows, on which the kerb-stones were laid, marked the width of the road, on average 14 Roman feet (about 4.1 metres) wide to accommodate two passing carriages. Then 0.45–0.60 metres of earth was removed to stabilise the roadbed which consisted of layers of larger stones as a foundation (statement) with broken stones or gravel (rules) above.

Above this was finer material on which were laid large polygonal blocks of basalt, to produce a smooth carriageway (summum dorsum). They fit together so tightly that ‘they do not present the appearance of separate stones, but an unbroken surface’ (Procopius, de Bello Gothic 1.14).

In addition, cuttings, viaducts, bridges and tunnels had to be engineered to allow the road to run as straight as possible. The distances had to be measured and milestones placed along the road, and suitable monuments set up to mark the beginning and the end of the road.

Rivers and streams were regarded by the Romans as barriers which required the ritual of suspicion before they could be crossed. Building a bridge was a demanding operation, especially if it crossed a deep or fast-flowing river. The first bridge over the Tiber, the seventh century bc Pons Sublicius was built on wooden piles or public.

The first stone bridges were built in the third-century bc, a time when the arch was coming into common use. If possible, the Romans avoided placing abutments in a fast-running stream because an obstruction anywhere in the stream increases its velocity, the worst place for obstruction being mid-stream, where depth and velocity are greatest.

Turbulence around an object is caused by the stalling of the fluid particles through friction drag, and they’re breaking away from the smooth flow path of the stream in whirls and eddies, the activity known as ‘scour’, which tends to undermine a bridge’s abutments.

The danger to the bridge abutments is both from the turbulence itself and from the particles of all sizes that the turbulence carries with it. It is interesting to note that the shape of the abutments used by the Romans, with the long faces parallel to the flow, the blunt end downstream and a tapering end or cutwater upstream, is approaching the optimum streamlined shape for minimum turbulence.

Because the Romans bridged a fast-flowing stream in one span whenever possible, the central span of a Roman bridge is often wider than those at the sides, clearly seen in the Pont du Gard aqueduct bridge (Figure 10.10). Using the round-headed arch, the central span is, therefore, higher than the side ones.

Unless a rising bridge was required, the springing heights of the arches had to be carefully adjusted to maintain a horizontal road over a river. The Alcantara bridge in Spain shows how expertly the Roman coped with these problems (Figure 10.3).

Aqueducts must rank among the most impressive engineering achievements of the Romans. There were 14 in the city of Rome alone by the end of the Roman Empire. A great deal is known about them thanks to Frontinus, who was appointed water commissioner (curator aquarium) by Nerva in ad 95 and wrote the work de Aquis Urbis Romae, in which he gives full technical details of all nine aqueducts which existed in his day.

As well as being impressive as buildings they were, in the words of Frontinus, more useful than the ‘idle pyramids’ (de Aquis 1.16). In an aqueduct system, the difficulty was to build a water channel on an exact incline to achieve a steady flow from the source of the water to its destination.

For example, the water for the Roman town of Nemausus (Nîmes) was brought 50 kilometres from the hills outside Uzès. Over this entire distance, the incline was maintained at 1 in 3,000, meaning that the waterfalls were less than 17 metres over the entire distance.

Where the channel crossed a river it was carried across by a bridge, such as the 275-metre-long Pont du Gard near Nîmes (Figure 10.10). It may be argued that a lead water siphon would have been an easier solution, but the cost of transporting and joining lead pipes was prohibitive.

There was also the problem of atmospheric pressure which threatened to burst the pipes. Constructing an aqueduct bridge was a more straightforward solution, especially as the quarry for the Pont du Gard was only 600 metres away.8

The 813-metre-long bridge which carries the Segovia aqueduct across a valley near the centre of the town is perhaps the most spectacular of them all (Figure 10.4). When an aqueduct system crossed an important road its passage was sometimes marked by a single or double archway, such as the Porta Praenestina (Porta Maggiore) (Figure 5.4), the ‘Arch of Drusus’ and the Porta Tiburtina at Rome, all reminiscent of triumphal arches.

Not all the water reached its destination. Greedy landowners often piped off considerable amounts for their private use if the aqueduct passed through their land (Frontinus, de Aquis 2.72–76). The water which did get through was fed into castellum aqua (water distributor) at the highest point in the town from which it was piped to the various districts.

The finest examples of such castella are found at Pompeii (Figure 6.12) and Nemausus (Nîmes). Regularly spaced water towers such as those in the streets of Pompeii distributed water to a particular neighbourhood (Figure 6.13). Three pipes brought the water down from these towers: the top pipe fed private houses; the middle public baths and circuses; the lowest public drinking fountains. This ensured that in the event of a shortage the public fountains would run out of water last.

Harbours and lighthouses

The problem with a river harbour is that the silt eventually produces a sandbar at the river mouth. The solution is to build a separate harbour to shelter ships from storms. Around it can be built shipyards, warehouses and a lighthouse, or towers on each side of the harbour mouth from which chains can be drawn across to protect it (Vitruvius, de Arch. 5.12.1–7).

The remarkable harbour at Carthage, built by the Carthaginians and later used by the Romans, consisted of a rectangular commercial harbour linked to a circular military one which could handle 200 ships. It had a circular island in the middle and contained docking bays with slipways as well as warehouse facilities.

Claudius built a large harbour near Ostia which covered 200 hectares (Figure 6.16). However, it was very exposed and a storm destroyed 200 ships in it (Tacitus, Ann. 15.18.3). Finally Trajan built the splendid 715.54-metre-wide hexagonal harbour, which is still visible from planes near Fiumicino airport.

He also built two other important harbours in Italy, one at Centumcellae (Civitavecchia), greatly praised by Pliny the Younger (Epist. 6.31), and the other at Ancona, overlooked by the tall Arch of Trajan. Other important harbours were at Antium (Anzio), built by Nero (Suet. Nero 9), and at Lepcis Magna, built by Septimius Severus (early third-century ad).9

In 37 bc enormous excavations were begun to create Portus Julius, which was to be the base of the western fleet. This involved building a navigable canal between Lake Lucrinus and Avernus. Under Augustus, the harbour of Forum Julii (Fréjus) became an important naval base but declined in the first century ad to the advantage of Misenum with its magnificent natural twin harbours, and Ravenna which was protected by marshes.

These became the primary Roman naval bases (portal praetorian). The most famous ancient lighthouse was arguable at Alexandria, which is normally reconstructed as a high, slender structure, that, according to Arab descriptions, was between 103 and 118 metres in height.

It was built in diminishing stages like the numerous depictions of the lighthouse at Ostia. The best-preserved Roman lighthouse is the 55-metre-high ‘Tower of Hercules’ at A Coruña in Galicia, Spain. There are also the 13-metre-high remains of the Roman lighthouse at Dover, octagonal in plan and originally c. 25 metres high.

Military and defensive architecture

The Etruscans built fine defensive walls and the Romans followed this tradition during their period of early expansion in Italy. The Servian Wall (Figure 1.5), built c. 378 bc and enclosing 426 hectares (1,065 acres), shows how advanced their defensive techniques were at an early date. Soon mighty walls were built to fortify a chain of defensive sites along the coast and in the Apennines.

The techniques varied to suit the materials available. At Ardea, where there was a ready supply of tufa, the walls were similar to the Servian Wall in Rome. In the hills, enormous polygonal stones were used. At first, the stones were unsmoothed or only partly smoothed, as at Circeii or Anagnia (Anagni), but later they fitted together with admirable precision, as at Signium (Segni) or Alatrium (Alatri).

The Romans first began to build forts after they overran the camp of Pyrrhus in 275 bc (Frontinus, Strat. 4.1.14). There were two types of fort, the larger covering between 1.8 and 2.2 hectares (4.5–5.5 acres) and housing about 800 soldiers (ten centuries of about 80 soldiers each), the smaller covering 1.2 hectares (3 acres) and housing half the number.

The forts were usually rectangular and had curved corners like playing cards (Figure 2.12). In the case of Hadrian’s wall in Britain, they lay parallel with the wall, as at Borcovicium (Housesteads),

Figure 2.12  Borcovicium (Housesteads, England) Roman fort, second to fourth century AD: plan.

or across it, as at Cilurnum (Chesters). Forts had four massive gateways, each gateway with two passages, flanked by guard-rooms. They were divided into five areas by intersecting streets: the via principals passed across the fort from a gate on one of the long sides to the one opposite.

The via praetoria ran from a gate on a short side to meet the via principalis and divided the front area of the fort (praetexta) into two halves, both of which contained barracks. The central part of the fort consisted of the headquarters building (Principia), the commandant’s house (praetorium), granaries, and a hospital (valetudinarian).

The Principia, always in the centre of the fort, consisted of an open columnar courtyard and a cross-hall running the full width of the building which could contain the whole body of troops. At one corner of the hall was a tribunal where the commanding officer stood to address his troops. Opening off one long side of the hall or sometimes opposite it was five rooms.

The central one was the company chapel, which often contained a statue of the emperor, altars and the legionary standards. The pair of rooms to one side was used by the adjutant (cornicularius) and his clerks; the other pair by the standard-bearer (signifer) and his accountants.

The via document ran from the back gate, dividing the rear of the fort (retention) into two halves, which also contained barracks. The via Quintana ran parallel to the via principalis and separated the retention from the central buildings of the fort.

Close to the wall was the latrine. The troops were housed in long barrack blocks each divided into about ten rooms, which were further subdivided into two rooms, one for eating and sleeping and the other for storing equipment.

A centurion and his junior officer had quarters at the end of each barrack block, which would have accommodated 80 or more infantry; whereas in cavalry forts such as Cilurnum (Chesters), the accommodation was a little more generous so that there were only about 60 to each block.

These were the usual arrangements of a Roman fort, although the position of some buildings could vary from the fort to fort. The 21-hectare (52 acres) legionary fortress at Lambaesis in North Africa was built in its present form around ad 268 (Figure 9.4), and the double legionary fortress at Vetera (Xanten) in Germany, which covered 56 hectares (139 acres), had similar features but of course on a larger scale.

In imperial times the Romans did not concern themselves with the defence of the capital itself until the invasions of the third century and when Aurelian built the magnificent walls around the city which have largely survived to this day.

The circuit, mostly built between ad 271– 275, was roughly star-shaped, with the principal roads entering the circuit at angles so that the approaches to the gates would be visible from the walls. It enclosed an area of 1,400 hectares (3,500 acres) over three times the extent of the fourth century bc Servian Wall.

The walls of Constantinople, built by Theodosius II in ad 413–440 after the sack of Rome, were the climax of Roman skill in fortification. There were 300 towers and the total width of the defences was 70 metres, comprising two parallel walls as well as an outer moat and other defences.

The vertical distance from the bottom of the outer moat to the top of the highest part of the wall was 35 metres. Its remarkable success is attested by the fact that it defended the city for more than 1,000 years.


The Romans regarded it as essential to bury their dead, only certain criminals were denied this right. One class of Roman tomb was derived from Etruscan tombs such as those at Caere (Cerveteri) and took the form of a circular masonry drum with a mound of earth on top.

The Tomb of Caecilia Metella at Rome and the Tomb of Munatius Plancus at Gaeta were cylindrical (both c. 20 bc), and both had a drum 29.5 metres (100 Roman feet) in diameter. The Tomb of the Plautii near Tivoli (ad 10–14), 60 Roman feet (17.76 metres) in diameter, was a somewhat smaller example of this type.

The Mausoleum of Augustus (begun 28 bc) was much larger than all of these, 89 metres (300 Roman feet) wide, as was the Mausoleum of Hadrian (Figure 8.25) with a cylinder 66.75 metres (225 Roman feet) wide rising from a base 89 metres wide (c. ad 135).

Two circular tombs in Algeria, the ‘Médracen’ (third-first century bc) and the ‘Tomb of the Christian woman’ (probably second or first-century bc) both had a drum c. 200 Roman feet wide with 60 half-columns around the exterior (Figure 9.1).

A similar but much smaller tomb, dating to the second-century ad, near Capua in Italy and called ‘le carceri vecchie’ had 22 half-columns at ten-foot intervals around the exterior.

The second class of tomb, probably derived from Syria, was composed of several superimposed columnar structures capped by a conical or pyramidical roof. Examples of this type are the ‘Conocchia’ at Capua and the Mausoleum of the Julii at Glanum (St. Rémy) (Figure 10.5).

North African examples include the obelisk tombs at Sabratha, Maktar, Cillium and Thugga (Dougga). There were also temple tombs in Tripolitania and the remarkable tower tombs of Palmyra. One of the finest hypogea (underground tombs) was that of the Scipios, built originally for Lucius Cornelius Scipio Barbatus, consul of 298 bc.

His magnificent monolithic tufa sarcophagus, now in the Vatican Museum, is of great interest because of the Doric triglyph frieze capped by dentils with Ionic volutes at the ends of the lid, a type fashionable in South Italy and Sicily at the time.

Chamber tombs are commonly found and can be seen lining the roads out of Rome, Pompeii and Ostia. Many were elaborately decorated outside, such as the Tomb of Annia Regilla in Rome, and inside, such as the tombs of the Valerii and Pancratii on the Via Latina. An unusual tomb is the pyramid of Cestius built in 330 days c. 18–12 bc (Figure 2.13), which is 37 metres (125 Roman feet) high × 29.6 metres (100 Roman feet)

Figure 2.13  Rome, Pyramid of Cestius, c. 18–12 bc.

at the base. In the later Republic and the first two centuries of the Empire, cremation became common and ash chests or urns were placed in the niches of columbaria (dovecotes), such as the Columbarium of Pomponius Hylas in Rome with its mosaic plaque composed of polychrome glass tesserae.

When inhumation became more common in the third century and the tradition of elaborately carved sarcophagi began. Christians rejected cremation and buried their dead in the maze of subterranean burial grounds known as catacombs.

There were four main burial methods: formal, burials in the ground covered with a slab; loculi, a burial slot in the wall of a catacomb; the arcosolium, an arched recess with the body either immured or in a sarcophagus underneath; and chamber–tombs for richer Christians.