The Lost Architecture of Ancient Rome

Insights from the Severan Plan and the Regionary Catalogues

By: David West Reynolds

Originally Published in 1997

View PDF

Much of the urban fabric of ancient Rome is lost to us. The famous monuments offer a powerful testament to the grandeur of the imper­ial capital, and Rome’s surviving architectural legacy is rich indeed. However, it was in dwellings, shops, work­rooms, and other minor structures that most of the pop­ulation spent most of their lives. Humble constructions like warehouses and taverns made life in this million-strong city possible, and reflected basic social realities rather than grand political image-building. In any city, such volksar- chitektur is an important expression of cultural identity. Built of less substantial materials than the landmarks, the “non-monumental matrix” of ancient Rome has largely perished, or lies inaccessible under modern occupation, leaving icons like the Colosseum and grand temples to shape our image of the ancient city.

Although we are limited in our ability to explore imperial Rome’s non-monumental matrix through excavation, clues exist which can conjure ghosts of the lost thousands of buildings that once coursed with the city’s life. This article will present insights drawn from two remarkable sources of data: the Severan Marble Plan and the Regionary Catalogues (see box on The Purpose of the Marble Plan and the Regionary Cata­logues).

The Severan Marble Plan

The Severan Marble Plan was a 40-foot-high decorative marble map of the city of Rome, mounted on a wall inside the Templum Pacis, at one end of the sequence of imperial fora in the heart of Rome (Figs. 1, 2). Executed during the reign of Septimius Severus, the Plan is dated to between AD 203 and 211 by the inscriptions and monuments appearing on it. The map is delin­eated at a consistent scale of 1:240, and the extraordinary thing is that it depicted nearly every ground-floor room in the entire city. Thousands upon thousands of them are recorded in this smooth gray marble surface, along with staircases, walls, fountains, altars, and other details (see Fig. 3). He discovered piecemeal beginning in the 17th century, 10 percent of the Plan survives in some 700 fragments, about half of which can be assigned to specific locations in the original wall map (Carettoni et al. 1960:199-200).

The Plan is surprisingly accurate. Its overall survey was extremely close to modern measurements (Carettoni et al. 1960:231), and recent work has shown that for specific buildings the errors it contains are minor (Reynolds 1996:92-106; see Fig. 4). The Severan Plan is an amazingly complex document, and a unique record of the imperial city near its zenith. In this respect  it is a unique resource. It is especially interesting for the fact that it includes not only monuments and great pub­lic spaces, but also the tenements and one-room shops of the poor (Fig. 5). The Plan helps us see Rome from top to bottom, filling in many missing pieces of our image of the city.

 Fragment 11

Fragment 11 illustrates several points about Rome’s urban structure (Fig. 6). First of all, it records specific architectural information concerning structures least likely to survive the ages, such as dwellings, shops, and workshops. For example, in Figure 6 can be seen classic examples of the Roman atrium house (A), well-known from Pompeii and Herculaneum. The Plan shows that even in densely crowded early 3rd century Rome, the atrium house plan endured as a dwelling for the rich. Nearby can be seen examples of apartment dwellings (B and C), in which small multipurpose rooms were grouped around open courtyards for light and air.

Single rooms frequently served as the entire dwelling space for poor families. Many dozens of other examples on the Plan show that this particular form of apartment dwelling was especially characteristic of Rome, although the form is not common at Pompeii or Ostia, our two other principal sites in Italy for studying Roman imperi­al urban structure. The Plan serves as a vital and unique source of evidence, providing data that could not be extrapolated from other sources.

Fragment 11 further illustrates several points about the general nature of Rome’s urban fabric. First of all is the close juxtaposition of the houses of the wealthy and the single-room high-rise apartment dwellings of the poor. As this and many other Plan fragments show, there was no significant economic segregation in Rome. While some neighborhoods acquired reputations as higher or lower class than others, they displayed noth­ing like the strict modern American economic segrega­tion with which we are familiar, in which new dwellings are grouped with others of their own economic status, whether they are seas of identical apartments or subdivi­sions of dozens of similar large houses. A lack of eco­nomic segregation was intensely characteristic of ancient Rome and remained common in Italian cities into the modern era.

Second, the fragment illustrates the intermixing of residential and commercial architecture. This occurred in most street-front buildings, which common­ly had shops on either side of an entrance hall leading into a residence behind the shops (F, Fig. 6). Some one-room shops had back rooms as dwellings for the shop­keeper and his family (D). In the illustrated fragment, we can even see the close juxtaposition of a workshop (E) with the houses of the rich. This is another contrast to the modern American urban form created by urban zoning laws which keep separate architectural classes geographically distinct. The same contrast is illustrated by the close juxtaposition of different levels of commer­cial operations—the workshop and the tabernae—where we typically separate such levels as light industrial and retail.

The comparison to modern America helps illustrate that there is no “default” urban structure. Cities of different cultures differ in fabric as much as they may differ in architectural styles. There is a tre­mendous range of possibilities—some ancient cities even lacked streets—and each facet of urban fabric may be examined as an expression of culture. In our present example of imperial Rome, it is interesting to consider the reality of close physical mixing of social classes against the literary image of the distinct separation of those classes in many social practices. The contrast to modern America is again striking, where the social ideal of equal citizens meets the structural practices of strict architectural segregation.

Rome’s non-monumental architecture was thor­oughly intermixed, with the urban fabric being funda­mentally of the same composition all over the city (Fig. 7). With its characteristic type of high-rise courtyard apartments, this fabric was significantly different from the urban composition of such preserved sites as Ostia and Pompeii.

Purpose of the Marble Plan  and the Legionary Catalogues

These two records of ancient Rome take very different forms, yet they share a similar enigma: What were they for?

Many suggestions have been advanced over the years, especially for the Severan Marble Plan, which is most frequently supposed to have served as a cadastral (tax) record. The reality was that the Plan was mounted with its uppermost reaches over 40 feet straight up from the nearest viewer’s eyes (see Fig. 1). Inaccessible for close consultation, the Plan was also immutably carved in marble and included few inscriptions—it was no tax record. Yet the Plan’s incredible detail has seemed to argue for some practical justification of all the work involved in collecting its information. The answer is that the Plan was derived from city survey documents that did serve practical purposes, inked on papyrus rolls that more than likely filled the very room in which the Plan was mounted (there appear to have been book niches in the ruins of one of the walls of its building). These plans of the city were maintained as the necessary records, piecemeal and on perishable, correctable materials. The great marble collation of the whole body of papyrus plans was a symbolic creation, emblematic of the city’s records program and possibly of the New Rome that the Emperor Septimius Severus had created by restoring so many monuments devastated in the great fire of AD 192. Severus celebrated the Century Games (Lodi Saeculares) during his reign in AD 204, marking a redefinition of his formerly military image with a revised emphasis on Peace and Concord. The Plan, mounted perhaps meaningfully in the Temple of Peace, may have been rendered in that spirit.

The Legionary Catalogues, 4th century architectural census tables, are similar to the Marble Plan in that they present a great amount of difficult-to-gather information in a form that seems to defy practical usage. Though they have been dismissed as fictionalised “Rome worship,” close analysis of the figures shows that the urban patterns they record are too complex and coherent to be mere inventions. The Elder Pliny (Natural History 3.66) refers to the Loman Census under Vespasian as including not only an assessment of the citizenry, but of many physical details of the city—such as the numbers of city gates and neighborhood shrines—which are among the very features tallied in the Regionary Catalogues. The Legionaries, like the Severan Plan before them, must have been abstracted from city census records for the purpose of celebrating Rome. Both documents made use of accurate, detailed information that had been compiled for official purposes and rendered it into more accessible forms for more general appreciation.


From the Severan Plan, we turn to another unique resource for the investigation of ancient Rome: the Regionar Catalogues (see Jordan 1907). The Legionaries are 4th century architectural census tables for the city of Rome (divided into fourteen wards, or regiones, by Augustus; hence the name given to the tables). Included in these tables are tallies of the num­bers of various kinds of buildings and urban features—bakeries, houses, street fountains, and so on—organized region by region. There are also lists and tallies of the city’s landmarks. There are two versions of the Legion­aries. That known as the Curiosurn is identified as the older and more reliable manuscript, the original dating to shortly after AD 357 (Nordh 1936:8-11). The Legion­ary Catalogues are a special topographic resource which provides vital assistance in the assessment of ancient Rome’s urban structure. Though frequently given passing mention, the Regionaries are almost never engaged in depth, and important aspects of these documents have never been explored.

A crucial step in making the regionary statistical data useful is their conversion into density figures. First, the density figures show that the numbers recorded in the Legionaries are not, as has often been claimed, exag­gerated: they form a pattern which is comparable to density figures that may be derived from excavated sec­tions of Pompeii—where higher, the figures for Rome are not unreasonably higher, and in some cases they are lower (Leynolds 1996:234, 244). Through these figures then, the macrostructure of Rome can be explored. The relative levels of development in different parts of the city can be examined in ways not possible through the Plan or any other means. This macrostrucrural study complements the microstructural data of individual buildings and neighborhoods provided by the Marble Plan, and makes possible a more comprehensive urban analysis of ancient Rome. Here we will focus on two particular topics illuminated by the Regionaries: the high-rise apartments and the neighborhood baths.

Rome’s high-rises Apartments: The Insulae

The Roman apartment block, or insula (literally, “island”), has been too often characterized on the basis of the famous ruins at Ostia, where well-built brick apartment blocks of standard plan still stand in testa­ment to Roman order and construction technique. While characteristic of Ostia, they are not necessarily characteristic of Rome, in spite of the fact that they are often pressed into service for scholarly studies describ­ ing the capital. The Loman insulae ran frequently to seven stories and higher, and ancient commentators describe them as built by speculators on the cheap.

One of the friends of Aulus Gellius remarked at the extraordinary rental income realized by landlords of city property, and declared that he would sell off his country estates to buy land in the city, if only the city did not burn so often (Aulus Gellius 15.1.3). Rome’s insulae were commonly built with wood and mud con­struction in their upper sections, rather than the sturdy brick and concrete work familiar from foundations and surviving insulae in Ostia (Vitruvius 2.8.20). This made the apartment blocks infamous fire hazards, as well as dangerously unstable (Seneca, On Anger 3.35.4-5). The fear of one’s dwelling collapsing was real in Rome, where the poor often slept “with the beams in ruin above” (Juvenal 3.194-196). It may be easy for us to lose sight of these crumbling tenements as we imagine the marble Rome Augustus claimed to have left, but accord­ing to the evidence they were an overwhelming pres­ence.

The plot of insula densities (Fig. 8) shows that the greatest concentrations of apartment buildings occurred in the city center, in Region VIII (Roman Forum), Region X (Palatine), and especially in Region XI (Circus Maximus). These high densities are particu­larly surprising, because in each of these regions there were substantial areas given over to open space or public buildings.

One of the implications of this level of density is that the monuments preserved today were absolutely surrounded by dwellings in antiquity. The “monumental center” of Rome was thickly hemmed in, apparently in every available corner, by apartment buildings. Frag­ment 29 of the Plan shows insulae with double rows of tabernae directly bordering part of the Forum of Trajan (Fig. 9), and a remarkable letter records that a precari­ous apartment building collapsed right into this forum in the 4th century AD (Symmachus 37). Even in the age of Augustus, residential settlement was thick in this area. Augustus’s forum is asymmetrical in design because he was not able to buy all the land he had desired from its present owners, who were no doubt reluctant to sell the source of very profitable rents in the city center (Suetonius Augustus 56.2). The monumental center of Rome is often discussed for the design and relationships of its monuments, but the extremely dense residential matrix into which it was set is rarely given sufficient consideration.

The Forum of Trajan was extravagantly praised in antiquity for the great impression it made on visitors, and while the porticoes and Basilica Ulpia were indeed of grand proportions, it would seem that the simple rec­tangular design would not have been especially inspiring (Fig. 10). However, the density figures from the Regionary Catalogues reconstitute in the mind the lost masses of insulae filling every available space in the cen­tral zones of the city, indeed towering over the very Forum of Trajan as the anecdote about the collapsing building attests (Fig. 11). In this overcrowded, over­whelming cacophony of insulae, the clean sweep of the vast open space of Trajan’s Forum would have made a strong impression. In reaching this forum a visitor would have to pass through the dense residential areas, and would always have been conscious of the extraordi­nary contrast presented by the grand plaza.

It was in fact this aspect of the Forum, rather than some feature like the basilica’s ornate appoint­ments, that struck the companion of Constantius when that emperor visited Rome for the first time in AD 356. Constantius entered the Forum of Trajan and, awe­struck at it all, vowed that he would copy the equestrian statue that stood in its center. “First, sire,” replied prince Ormisda of Persia, “build a similar stable for your steed, if you can, so that it can range as freely as the one which we see here” (Ammianus Marcellinus 16.10.15-16). This comment, and other reactions to the imperial fora in antiquity, are better understood in light of the insula density information from the Legionary Catalogues. The extraordinary concentration of dwellings in the city center should be taken into account in any assessment of this part of Rome.

The insulae contained apartments both spa­cious and cramped, both pleasant and miserable; condi­tions grew worse as one climbed higher flights of stairs. Most of the city’s flats would have been single-room garrets, and these underline the degree to which poor Romans necessarily conducted their lives outside their homes. One of the most characteristic of Loman buildings is the bath, an amenity typically enjoyed in a public rather than a domestic setting. The grand imperial baths (or thermal) are well known and justly famous from their surviving ruins, but evidence shows that the very small neighborhood bathhouses, or balnea, were a more sig­nificant part of Rome’s urban fabric than we have given them credit for.

The Neighborhood Baths of Rome

The plot for balnea (Fig. 12) shows that they are found throughout the city. This is what we would expect based on the Marble Plan, on which small baths appear in many locations (Staccioli 1961). According to the Legionary figures, only one region, IX (the monu­mental zone Campus Martius), rates as relatively low in density of balnea. In general the density of small baths follows the density of residential matrix as indicated in the plots for insulae and dornes (private houses). This suggests that while baths were essential all over the city, more were required for the heavy population in the city center. The particularly high densities in regions VII and I can be explained by the fact that these lie along the main transport corridors leading into the city, the Via Flaminia to the north, and the Viae Appia and Latina entering from the southeast in Region I (Fig. 13). Travelers arriving in the city must have desired bathing facilities to refresh themselves soon after passing the main gates.

The great imperial thermae are certainly impor­tant, but they are only part of the picture of Roman bathing, and a smaller part than the spectacular ruins of thermae still standing in Rome would suggest. The Plan and the Regionaries help us to address this skewed image and re-evaluate the role of minor baths in the ancient city.

The imperial baths were huge and luxurious complexes, offering spectacular public amenities the like of which the world has never seen again. How could the city market sustain humble minor baths when such com­petition for patrons existed? One might expect that the imperial baths would replace the old smaller private baths, increasingly as more large thermae were built over time. The first public complex was built by Agrippa in 38 BC. This set the foundation for the long tradition of imperial public baths that was to follow, and which indeed was to become one of the most characteristic traits of Loman urbanism, within the imperial city and throughout the Empire. Most aqueducts to Loman cities throughout the Empire were built not for drinking water supply, but to feed baths large and small (Hodge 1989:128). Beginning with Nero, a succession of emper­ors built public bath complexes in Rome throughout the first four centuries AD. Typically, the newest complex was even larger and more splendid than the last.

The baths of Diocletian, built at the beginning of the 4th century, covered over 30 acres (13 hectares), or an area roughly equal to that of the original settle­ment of the Roman city of Timgad in Algeria. By the time of the Legionary Catalogues there were eleven imperial bath complexes. The amenities of these were extraordinarily luxurious. They were constructed on a titanic scale, made of precious materials, filled with superb art, and offered not only hot, cold, and warm water, but also exercise areas, libraries, lecture halls, and an almost endless variety of physical and mental pur­suits, leisures, invigoration, relaxations, and diversions. These facilities allowed the meanest Roman citizen to live like an emperor; the environment they provided was unsurpassed, and admission to them was free.

What becomes very striking, considering the evidence from the Marble Plan and the Legionaries, is that in the 4th century these imperial baths, for all their luxury and amenities and their free admission, had not eroded the role of neighborhood baths at all. The Plan and the Regionaries attest a perhaps surprising reality, and show that the minor baths held an important place in Rome’s urban fabric that was all their own. In fact, by the time of the Regionaries, neighborhood balnea were more common than ever. The number had grown con­sistently over time. In 33 BC Agrippa ordered the census of baths within the city, and recorded the number 170.

In the 1st century AD Pliny regarded the number as “uncountable” (Natural History 36.121). Fragments of the Several Marble Plan attest that small baths seem to have been tucked in everywhere—in and among ware­houses, in thickly populated areas, and wedged into odd corners. The Regionaries support this impression. Only one region is listed with as few as 15 small baths (Region XI, Circus Maximus); the next lowest number is 44 and nine regions are listed with at least 75. In all, the Regionaries record a 4th century total of 856 private baths. These baths were not expensive, but they did charge admission.

The balnea clearly served a different purpose from that of the thermae. The balnea were regarded as a basic aspect of the city’s identity and structure, and this was doubtless due to the social interactions that they fostered, probably among relatively small bodies of clientele. The familiarity of one’s customary local bath, along with the sight of well-known fellow bathers there, must have provided an important sense of community and identity, beyond the family but still intimately small (Nielsen 1990:146). Balnea were integral to a city’s romanitas, and were probably sources of significant social comfort for the inhabitants of the largest city in the ancient world.

The structure of a city offers useful study as a major expression of the building culture. The urban form of ancient Rome is partially preserved in the Marble Plan and the Regionary Catalogues. It is not quite a “virtual Pompeii” that can be reconstituted from these records, but while most of the non-monumental architecture may be long gone or out of reach, the urban fabric of the city is far from lost, and it yet holds many insights for our investigation.

Cite This Article

Reynolds, David West. "The Lost Architecture of Ancient Rome." Expedition Magazine 39, no. 2 (July, 1997): -. Accessed September 23, 2023.

This digitized article is presented here as a historical reference and may not reflect the current views of the Penn Museum.

Report problems and issues to