PART FIVE: TECHNOLOGY AND SOCIETY 872 Buckingham Palace and St George’s Chapel, Windsor. Coade stone, under its later owners, continued to be used until, in late Victorian times, the fashion for polychromy ousted it from favour. THE ARCH An arch is a structure composed of wedge-shaped blocks, which are called voussoirs, and designed to span a void. The blocks support each other by mutual pressure and therefore capable of withstanding considerable weight and thrust. A wooden framework, centering, was set up underneath the arch to support it until all voussoirs were in position. Parts of an arch Often the central voussoir at the top of an arch is larger and made into an architectural feature: this is then called a keystone. The point or line where the arch springs up from its supporting wall or column is known as the springing line. The lowest voussoir on each side of the arch is thus known as a springer. The support—wall, pier or column—is the abutment. The span of an arch is the distance between abutments and the height in the centre is the rise. The outer curve of the line of voussoirs is called the extrados, the inner is the intrados; the under surface is the soffit. These parts and the different forms of arch are shown in Figure 18.3. History of building with arches Knowledge of arcuated construction, that is, building structurally with arches rather than in the post and lintel manner (see p. 857), was gained by most of the early cultures but in some of these, notably the Egyptian and the Greek, plentiful supplies of stone and marble were available so trabeated structures were possible. Moreover, the builders held the opinion strongly that the arch was a less stable form of structure. Indeed, the Greeks had a proverb which said ‘the arch never sleeps’, referring to the truism that, if the arch support were to be weakened, the ever-present outward thrust of the curve of the arch would cause collapse. So, in both Egypt and Greece, arcuated structures were confined to small and utilitarian buildings. In Mesopotamia, however, in the area between the rivers Tigris and Euphrates, adequate supplies of neither timber nor stone were available for building and sunbaked bricks made from the local clay constituted the chief building material. Bricks are too small to span a lintel opening, so an arcuated BUILDING AND ARCHITECTURE 873 structural system was developed for both arched openings and vaulted roofing. The true arch, that is, one built with radiating self-supporting voussoirs, was known here as early as 2400 BC. Sun-dried bricks are not as strong as kiln- baked ones so the Mesopotamian structural spans of this time were not large and, due to the softness of the material, remains are far less in quantity than are the stone structures of contemporaneous Egypt. In Italy the Etruscans introduced the concept of arcuated principles at an early date, building both arched openings and vaults. A number of their fine arched city gateways survive; most of these date from a later time, the fourth to third century BC as, for example, that at Volterra. These structures were in stone, as were a number of the corbelled vaults but, as in Mesopotamia, Etruscan arcuated work was on a fairly small scale and it was the Romans, particularly in the days of the Empire, who fully exploited the arcuated possibilities for arcades and vaults. The facing and structure was of brick, forming compartments, which were then filled with their exceptionally hard concrete. The arches so far discussed, as well as those built in the revived classical tradition of the Renaissance and the Baroque eras, were round arches. These might be semicircular, wherein the height of the arch above the springing line is equal to the radius on the line or they might be segmental or stilted; in the Figure 18.3: Arch construction. Drawing by Doreen Yarwood. PART FIVE: TECHNOLOGY AND SOCIETY 874 former case the arch is wider and lower, in the latter it is taller and narrower. A further variant is the horseshoe arch of Islamic origin in which the curve of the arch is carried on below the semicircle. Romanesque architecture of the early Middle Ages (Norman, in England) also employed the round arch. Indeed, much of the Romanesque building, especially in Spain, Italy and southern France where remains of Imperial Rome had survived in quantity, was strongly derivative of Roman classical work. For most of the medieval period the pointed arch was in use and this is characteristic of what is termed Gothic architecture. There were many variations on the pointed form from the early, narrow lancet shape, through the equilateral to the late, wide obtuse and to the four-centred types in the ogee and Tudor forms. The pointed arch is fundamental to Gothic architecture in northern Europe. It was developed solely to provide a flexible form suited to the building of the stone vaults which were made necessary by the hazard of fire which destroyed many Romanesque timber roofs, causing them to have to be rebuilt time and time again. It was the French who introduced the pointed arch into Gothic architecture —aptly, they call it the arc brisé, the broken arch—in the Abbey Church of S. Denis, now in a suburb of Paris, in 1135. The pointed arch was not new. It had long been employed in the Middle East and, in Europe, was used in areas under Arab influence such as Moorish Spain, Sicily and Provence. The new style originated in northern France in the area of the lie de France, because here there was a more stable economy at the time. The reason for the introduction of the pointed arch was that the round arch presents great problems in vaulting a church. This is because nave, choir, transepts and aisles often have different heights and widths. Vaulting is made in bays and the semicircular arch lends itself to a square bay. The bay is decided by the positioning of the supporting piers or columns. In Romanesque building the bay was traversed at roof level by diagonals crossing one another from pier to pier. As the diagonals were longer than the four ribs connecting the four sides or faces of the bay, it was impossible for all of these ribs to be semicircular in form. Either the vault had to be domical, so giving an undulating ridge rib, or the side arches stilted. The pointed arch, which could be varied infinitely in proportion of width to height, was far more flexible for constructing stone or brick coverings to nave and aisles of different spans and roof levels. THE VAULT In building structures the term ‘Vault’ is used to describe a covered form of roofing based upon the principle of the arch. Traditionally, over the centuries, such vaults were constructed, as were the arches, from wedge-shaped blocks of BUILDING AND ARCHITECTURE 875 stone or brick supported upon a wall, pier or column. During the nineteenth century plaster vaults were often erected reinforced by an iron structure. In modern architecture reinforced concrete, steel and glass fibre are among the materials used for this purpose. As in the building of an arch, a wooden supporting framework—centering—was erected underneath the vault to be constructed and removed when the vault structure had been completed. It was the Romans who first fully exploited the structure of the vault, although knowledge of such building principles had been known to most of the earlier cultures. Using their strong mix of concrete together with brick as a structural and facing material, Roman builders covered very great spans with such vaulting, particularly in their great basilicas and thermal establishments. The vaults were massive but of the simplest forms. Three types were used, all based on the round arch shape. Commonest was the half-cylindrical, known variously as the barrel, wagon or tunnel vault. This was a continuous semicircular or obtuse arch, extrapolated as in a railway tunnel. Where two such barrel vaults of equal diameter met at right angles as in a cruciform building, this created a square plan groined vault, the groin being the angles of intersection. The Romans also built domes which were generally hemispherical as in the Pantheon. In the basilicas and buildings constructed for the bathing establishments, the vaults were very thick. They were supported either on the also massive walling or on heavy columns or piers. The square piers in the Basilica of Maxentius in Romé (AD 308), for instance, measure 4.25m (14 ft) across. The majority of important Roman vaults were coffered, that is, they were decorated by carved octagonal, square or diamond-shaped sunken panels, partly for ornamentation, but also because coffers reduced the weight of the vault and so its lateral pressure upon the walls. Romanesque builders, basing much of their church construction upon the Roman basilican model, continued to erect round-sectioned barrel and groin vaults but, since the knowledge of Roman concrete had been lost they dared to cover only narrow spans. In a few instances, notably the remarkable early example of Durham Cathedral where the choir was vaulted in 1104 (later rebuilt) and the nave in 1135, Romanesque builders constructed rib vaults, but these were rare, being unsuited to vaulting over square bays. With the introduction in France of the pointed arch into vaulting design, builders were able to break away from the constraints imposed by the square bay. They adopted rectangular ones which, with the use of the pointed arches, gave greater flexibility. They also constructed rib vaults in which the stone ribs were built first, supported by centering, then the infilling between the ribs—the web—which consisted of pieces of stone or of bricks, was inserted. This web was of much thinner material than the Roman or Romanesque groined vaulting had been, so rendering the whole roofing of greatly reduced weight. PART FIVE: TECHNOLOGY AND SOCIETY 876 The early rib vaults were quadripartite in design, that is, the bay roofing was divided into four compartments by two diagonal ribs crossing one another in the centre. On the Continent, especially in France, the quadripartite design continued to be built for a long period but in England, often, the quadripartite vaulting bay was divided transversely into two parts so giving six compartments: a sexpartite vault. In most English cathedrals and churches a ridge rib extended longitudinally along the whole interior at the roof apex level and the rib intersections were covered by decoratively carved bosses. In France the ridge rib was rarely used. From the early fourteenth to the mid-sixteenth century vaulting designs gradually became more elaborate and complex. This was especially so in England where the last phase of Gothic architecture, the Perpendicular, lingered late and had no parallel on the Continent. The first development was to insert tiercerons, intermediate ribs extending from the vault springing to the ridge rib. Exeter Cathedral is a fine example of this. Soon the lierne vault evolved (Figure 18.4 shows a modern replica). Lierne ribs, named from the French lier, to tie, might extend in any direction from the structural ribs and might join any other rib. Their use was entirely decorative and later examples are extremely complex as is evidenced in the choir vault of York Minster. The final phase of English Gothic vaulting was the fan vault which evolved from the desire for a structure which would accommodate ribs of different curves as they sprang from the vaulting shaft of the capital. The radiating ribs of the fan are of equal length and the bounding line is in the shape of a semicircle. The whole group of ribs is made into an inverted cone, the radiating ribs so creating a panelled and cusped all-over surface pattern. King’s College Chapel, Cambridge (Figure 18.5), Sherborne Abbey Church and Bath Abbey Church are supreme examples illustrating the use of this type of vault over a long period. An essential concomitant to the building of arches, vaults and domes was the development of the buttress. This architectural feature is intended to withstand a lateral force and it is arch building which particularly requires this means to counteract the thrust which it generates. There are two chief forms of buttressing used for this purpose. One is that pertaining to an arcade, for example, arches extending sideways to divide a church nave from its aisles, where each counteracts the lateral thrust from its neighbour. The other type is where a heavy buttress is built up against the outside wall of a building to reinforce the walls at points believed to be most vulnerable. The Romans, who developed building with the arch and vault, constructed strong, heavy walling, reinforced at intervals with wall buttresses to offset the lateral pressure of their barrel and groined vaults. Romanesque builders did the same. Gothic vaulting aroused greater problems. The buildings were higher, the vaults high and wide and, soon, the walls were being built thinner and being pierced by larger windows. Two types of buttress were developed. BUILDING AND ARCHITECTURE 877 The tower or pier buttress consisted of a reinforcing buttress, thicker at the base than at the top and descending in stages with sets-off from a lofty top pinnacle to a projection at ground level of some eight to ten feet (2.5 to 3m). The lateral pressure is felt at the top of the buttress but the sheer weight of the massive buttress converts the thrust downwards to ground level. As the Gothic period advanced, cathedral and abbey churches in Europe were being constructed higher and wider and with ever larger window openings. The usual cruciform church with nave and choir roofs built much higher than the sloping aisle roofs meant that it was impossible to construct a Figure 18.5: Fan vault. King’s College Chapel, Cambridge, 1446–1515. Drawing by Doreen Yarwood. PART FIVE: TECHNOLOGY AND SOCIETY 878 pier buttress between the clerestory windows to offset the thrust of the nave or choir high vault, so the flying buttress was developed. It was discovered empirically that the maximum thrust was to be felt at a point just below the springing line of the vault on the interior wall so it was at this point on the external wall face that a bridging arch was built to slope outwards and slightly Figure 18.5: Fan vault. King’s College Chapel, Cambridge, 1446–1515. Drawing by Doreen Yarwood. BUILDING AND ARCHITECTURE 879 downwards to join a pinnacled wall buttress. This type of construction served a dual purpose; the counterthrust at the given point on the exterior wall conveyed the vault pressure away from the building and down to the ground; also, by means of the heavy pinnacle above, it helped to offset the vault thrust. In very complex church designs, where five aisles had been built, a double flying buttress system had to be constructed with arches spanning each of the two flanking aisles. The French architects developed the flying buttress and it is in France that the eastern arm chevet of a cathedral displays the elegant and dramatic forest of arches and pinnacles which are so impressive. In England, with lower vaults and roofs, flying buttresses were needed less and are most evident in the later buildings of the fifteenth and sixteenth centuries as, for example, at the Henry VII Chapel of Westminster Abbey in London. With the introduction of Renaissance architecture and the return of classicism the domed covering became more common than the vault. In modern times, however, the development of new materials has made possible the erection of thin shell vaults of half-cylindrical or barrel form which may be extended over long spans without requiring piers which obstruct the interior vista. These are particularly suitable for industrial, sports or entertainment purposes. Reinforced concrete, lightweight metals such as aluminium and plastics reinforced with glass fibre are the most usual materials for this. THE DOME A dome is a convex rounded roof covering the whole or a part of a building witha base on the horizontal plane which is circular, elliptical or polygonal. In vertical section the dome may be hemispherical, partly elliptical, saucer-shaped or formed like a bulb (the so-called onion domes to be seen in eastern Europe). The English word dome derives from the Latin domus, a house, and the Italian duomo, cathedral, or House of God. It has been traditional in Italy to build domes over churches and so ‘dome’ had been adopted for this meaning in England. The more usual Italian term is cupola, derived from the Latin cupula, diminutive of cupa, flask, barrel. While the Romans extensively developed the theme of the arch and the vault they largely neglected the possibilities of domical coverings. They were the first to build domes in western Europe but these were most often supported on buildings with circular walling as in the most famous example, the Pantheon in Rome. Begun in AD 120, this has a very large dome—43m (142 ft) in diameter —which is hemispherical in form. Its continuous thrust is contained by the 6m (20 ft) thick cylindrical walling into which it is embedded and further support is provided by stepped rings of masonry encircling it above wall top level. This results in an uninterrupted view of the dome from inside the building, but on the exterior only the upper part is visible. PART FIVE: TECHNOLOGY AND SOCIETY 880 The Romans also covered square interiors with groined vaults, as in their bathing establishments and it is thought that they knew how to build a domical vault over a square space. This is a domical form made of four webs separated by groins which fuse to a point; it is not a true dome and its design lacks flexibility of proportion. Building a dome over circular walling presented fewer structural problems but limited the spatial capabilities of interior architectural design. In Renaissance and, even more so, Baroque architecture, the dome became a fundamental characteristic of design and a variety of different forms of construction were attempted and built. To achieve, in particular, the Baroque spatial effects of lighting and of solid stone structures ornamented by painted and sculptured curved surfaces appearing to float upwards, it was necessary to support the dome upon free-standing piers instead of the more constricting walling. The basis for this type of construction was the Byzantine development of the dome, evolved in Constantinople, capital city of the Byzantine Empire, a construction which, in its turn, had derived from earlier buildings in Mesopotamia, Anatolia, Persia and Syria. The essential problem in building a dome of circular section upon a drum, also of circular section, which can contain windows to light the interior, over a square is to transform the square base into a circle upon which the drum can be supported. This may be achieved two different ways: by the use of the squinch or the pendentive. Both methods derive from Asia; the former can be seen in extant remains as early as the fourth century in Persia, the latter in Syria and Asia Minor. In squinch construction an arch, or series of arches, is built across the upper angles of the square to transform the square base to an octagonal one and so more easily support a circular base. This method was extensively employed in the Byzantine period and later in both east and west, for example, in the cathedral at Le Puy in France. The pendentive is a more ambitious and imaginative, as well as more successful, solution. It was widely used in the best Byzantine building, from S.Sophia in Constantinople (built 532–8 by the architects Anthemios of Tralles and Isodorus of Miletus) onwards, and has been the basis of dome construction in the west from the Italian Renaissance to the present century. A pendentive is a spherical triangle. In using this form of construction the triangular spaces between the square section of the plan of the piers and the circular section of the base of the dome are built as if they are parts of a lower, larger dome, so their section is like that of an arch carried across the diagonal of the square space to be covered. The lower dome possesses a horizontal section which is concentric with the plan of the intended dome. As the lower dome is too large to fill the square space, it is cut off in vertical planes formed by the four sides of the square. When the four remaining parts of the lower dome have been built high enough to form a complete circle within the square section, this circle provides the basis for supporting the actual dome. If this BUILDING AND ARCHITECTURE 881 design is not further developed it is known as a sail vault because it gives the appearance of a sail filled with wind anchored at the four corners. If the dome set upon pendentives is further constructed, the drum of the dome is set upon the circular section of the lower dome and the drum carries the upper dome. Externally most large domes carry a lantern. Internally the drum is articulated with a classical order and windows while the pendentives and dome are decorated by paintings and/or mosaics. Renaissance domes It was accepted by the classical architects that a hemispherical dome provided the most satisfactory exterior silhouette from the aesthetic viewpoint. This, however, created problems. These were partly of a structural nature, because a hemispherical dome exerts a particularly strong thrust. They were also of a visual character because, while externally the hemisphere provides a pleasing shape and an imposing landmark, internally it is too tall and dark and its decorative covering of paint or mosaic cannot readily be seen. The solution, adopted by many architects, was to build two domes, a taller exterior one and a lower interior one. The first great Renaissance dome was the one built by Filippo Brunelleschi to cover the unfinished medieval Cathedral of Santa Maria del Fiore in Florence. To this commission Brunelleschi had to bring his considerable knowledge of mathematics and of the structure of Roman vaults. The practical problem for this early date (1420) was considerable: how to construct a permanent dome to span the 42m (138 ft) diameter space. Brunelleschi dared not build a hemispherical dome on the existing drum, which had no external abutment, so he compromised and proceeded step by step. His dome is a pointed one, constructed on Gothic principles with ribs supporting a lighter infilling, and is taller than a hemisphere to offset the thrust. He then built an inner dome of slightly less height which is decoratively painted on the interior. Michelangelo faced not dissimilar problems with St Peter’s Basilica in Rome. His exterior design was for a hemispherical dome but when this was completed in 1590, after his death, by Giacomo Della Porta, it too had acquired a slightly pointed form. Michelangelo raised his great dome, also of 42m (138 ft) diameter, almost as large as that of the Pantheon, upon a classical drum and supported both drum and dome on four pendentives rising from four massive piers set at the church crossing. Unlike the Pantheon, however, there was no supporting enclosing circular wall, so Michelangelo encircled the dome with iron chains, which have had to be replaced ten times since the original construction in order to contain the thrust. At St Paul’s Cathedral in London (completed 1710) Sir Christopher Wren supported his smaller drum and dome 34m (112 ft) diameter upon eight piers. . abutments and the height in the centre is the rise. The outer curve of the line of voussoirs is called the extrados, the inner is the intrados; the under surface is the soffit. These parts and the. plan of the piers and the circular section of the base of the dome are built as if they are parts of a lower, larger dome, so their section is like that of an arch carried across the diagonal of. vaulting shaft of the capital. The radiating ribs of the fan are of equal length and the bounding line is in the shape of a semicircle. The whole group of ribs is made into an inverted cone, the radiating