Concrete is made by mixing broken stone or gravel and sand (aggregate), a bonding agent, and water, and allowing the mixture to harden through chemical process into a solid mass. So-called cementitious materials had been used in ancient Egypt about 3,000 years earlier and later by the Chinese, Minoans, and Mycenaeans, but this synthetic stone—a new building material—was developed and exploited by the Romans from about the third centuryb.c.
Ambrose advised his protégé Augustine: “When in Rome, live as the Romans do; when elsewhere, live as they do elsewhere.” Throughout the Roman Empire, the architecture they built was a weighty presence imposed upon the subject peoples—a consistency probably more to serve the colonizers, isolated from the familiar things of home, rather than for the colonized. Throughout history architecture has provided a social anchor for migrant peoples. The Roman way was to come, to see, to build, and there was, especially in the days of imperial expansion, a need to build quickly and in a familiar way. That was made possible by the use of concrete.
The Romans used concrete (opus caementicium) for all parts of their structures: foundations, walls, and roofs. It was made by combining pozzolana (a volcanic earth found in many places in southern Europe) with lime, broken stones, bricks, tufa, and sometimes pumice. Such a mixture could set even underwater. Lime was obtained by crushing limestone or seashells, or sometimes replaced by gypsum as a binding agent. The Romans placed a very dry mix of pozzolana and wet lime, layer for layer, over rock fragments, and carefully tamped it into place. Its structural strength depended upon what is now called the water-cement ratio: the higher the proportion of bonding agent to water, the stronger the concrete. The combination of a dry mix and thorough consolidation made the material extremely durable.
At first, concrete was limited to places where it would not be seen. For foundations, it was placed between wooden form boards that were stripped once the mixture had hardened. For building above the ground, its brutal appearance, once the formwork was removed, presented an esthetic problem. Because the many advantages—strength, versatility, economy, availability, and speed of erection—more than offset that single disadvantage, the Romans simply used more presentable materials to face the concrete, usually as a “lost” formwork. For example, as late as 20 b.c. the architectural theorist Vitruvius recommended building two face walls of squared stone (opus quadratum), 2 feet (0.6 meter) thick, tying them together with iron cramps, and filling the cavity with tamped concrete. But that was for prestigious buildings, and a number of alternative wall constructions had already been developed.
From around 200 b.c., slabs of volcanic tufa were used as permanent facings; far more common was the technique known as opus incertum, which employed small, random pieces of tufa, carefully packed together. Over time the shapes were made increasingly regular, and by about 50 b.c. 4-inch-square (10-centimeter) pyramidal tufa blocks were employed. Set diagonally, their sharp apexes penetrated about 10 inches (25 centimeters) into the concrete infill, providing an excellent bond. Because of its netlike appearance, the method was called opus reticulatum. By then, fired clay bricks were also being used for facing. Over the next two centuries the predominant technique was opus testaceum, flat slice-of-pie-shaped bricks, tied at intervals with bonding courses through the wall. The late empire saw a further variation, called opus mixtum, consisting of alternate courses of brickwork and small squared stones. It is stressed that these systems provided only a presentable surface: the real work of the wall was done by the immensely strong concrete mass, which normally supported innovative superstructures, also made of concrete.
Unfinished concrete was not only unattractive to the eye but it also presented an architecture that was, to the Roman mind, inappropriate in appearance. They therefore covered it, whether brick faced or not, with a variety of decorative surfaces: stucco (a mixture of marble dust and lime) perhaps 3 inches (7.5 centimeters) thick in up to five layers, and molded, patterned, painted, and sometimes veneered with mosaics of marble and even glass tesserae. The most important buildings had marble veneers, held in place by bronze pins and nonstructural architectural orders applied as pilasters or half-columns that masked the concrete structure and reduced the visual scale.
Roman public architecture existed to move in and through, rather than around. Such urban buildings as the thermae (bathhouses) and basilicas (law courts) demanded interior spaces uncluttered by columns that could accommodate huge gatherings of people. In order to achieve vast interior spaces, the Romans exploited the semicircular arch, a technology inherited from their Etruscan forebears. The arch and its three-dimensional extensions, the vault (a prismatically extended arch) and the dome (a rotated arch), could span large distances without intermediate supports. With characteristic directness, Roman engineers found expedient solutions. A small rectangular room could be covered by a semicircular barrel vault carried on continuous parallel side walls. A square room could be roofed by a cross vault (two barrel vaults placed at right angles), supported by piers at the corners and allowing the space to be lit from all sides. Larger rectangular spaces could be enclosed by a procession of such vaults built side by side. A polygonal space received a hemispherical dome and an apse a half-dome, carried on drums above the base walls. By using concrete for these roof structures, the Romans enclosed volumes that would not be equaled for over 1,000 years.
The simplest barrel vaults consisted of a series of parallel brick arches cross-tied as in opus testaceum and filled between with concrete; that is, the concrete was packed into brick compartments. The whole structure was supported by wooden centering until the mortar had set. Other vaults and domes were directly formed in mass concrete. The technique had two main advantages: once the centering was designed and placed, it employed unskilled labor, and it enabled complex plan forms to be roofed without the cost of dressed stone construction. Often, the weight was reduced by using hollow clay boxes or even wine jars, especially in the groins of cross vaults; alternatively, vaults were lightened by forming recesses or coffers in their undersides. Domes were generally much thicker toward their base and therefore appeared externally as inverted saucers, while inside they were hemispherical. As they rose, lighter materials, such as pumice—a stone that floats—were used for aggregate.
In order to underline the achievement of the Roman engineers and architects, it is helpful to consider the size of some concrete structures; a couple of examples may suffice. The concrete barrel vaults that spanned the 76-foot-wide (23-meter) side aisles of the fourth-century-a.d. Basilica of Maxentius in Rome were 8 feet (2.45 meters) thick. The 142-foot (43-meter) concrete dome of the Pantheon, also in Rome, is 4 feet thick (1.3 meters) at its apex and 20 feet at its base. The massive loads of these roofs were carried to the ground through huge piers or thick walls, and their horizontal thrusts resisted by buttressing elements integrated with the architectural design.