Brunelleschis Dome (11 page)

Leonardo da Vinci’s drawing of the
castello.

The
castello
would go into operation as soon as the ox-hoist delivered the stone to the working level. Standing on a small platform at the top of the crane — one of the giddiest and most dangerous of all of the stations on the cupola — the operator turned the horizontal wooden screw that moved the load laterally through the air beneath the crossbeam. At the same time, the counterweight at the other end of the beam was adjusted in order to keep the crane in equilibrium. A horizontal arm projecting from the mast prevented the load from swinging at the end of the rope — a danger in the high winds that swirled round the cupola. Then, once the stone was hovering above its final destination, the turnbuckle was adjusted and the load descended into place.

The success of the
castello
is remarkable given the lack of understanding of the strength of materials at the time. Other than through precedent, Filippo had no way of knowing the robustness of his crane’s long horizontal beam when placed under the stress of a heavy load. Not until the studies of the French engineer Claude-Louis Navier in 1813 was the bending strength of beams mathematically determined. In 1420 calculations were based on ancient theories about the various “humors” of trees in the same way that the medicine of the day — equally suspect — was concerned with the interaction of humors in the body. Elm, for example, the wood used in the crossbeam, was said to be “dry;” therefore it did not “agree” with the plane tree or the alder, which were “moist” and so ought never to be used in the same structure as elm — a most dubious set of assumptions on which to rest a sandstone beam weighing over 1,000 pounds. The sight of one of these heavy stones dangling at the end of a crossbeam must have been, initially at least, an unnerving sight.

But the crossbeam held, and the
castello
, like the ox-hoist, needed only minor repairs in the decade that followed. Indeed, in one respect the
castello
was to prove
too
durable. Like the ox-hoist, it remained on the cathedral’s building site through the 1460s, long after Filippo’s death, and was present during the last act of the dome’s construction: the placement of the eight-foot-high bronze sphere that sits atop the lantern. The commission for this bronze ball went to the sculptor Andrea del Verrocchio, in whose workshop there was at this time a young apprentice named Leonardo da Vinci. Fascinated by Filippo’s machines, which Verrocchio used to hoist the ball, Leonardo made a series of sketches of them and, as a result, is often given credit for their invention. How Filippo would have reacted to this misattribution — Filippo, who was so proud of his inventions and so fearful of plagiarism — scarcely bears contemplation.

N
O SOONER WAS
the ox-hoist finished than plans began moving ahead for the first sandstone chain, and in early June a design for the chain had finally been settled upon: a carpenter working for Filippo, a man named Jacopo di Niccolò, was paid for a wooden model demonstrating how the beams would be linked. This chain was complex in design, consisting of two concentric rings of stone laid horizontally around the octagonal circumference of the dome. These long beams rested on, and interlocked with, shorter beams laid transversely, like railway ties, at intervals of every three feet. Before June was out, some eighty-six cartloads of sandstone had arrived in the Piazza del Duomo from the Apennines.

Seen from either the dome or the campanile of Santa Maria del Fiore, the hills surrounding Florence have the sensuous contours of a supine body. Dozens of quarries were worked on their slopes during the Quattrocento, including several near the village of Settignano, the childhood home of Michelangelo, whose wet nurse was the wife of a stone cutter and, according to the sculptor, the source of his genius with hammer and chisel. The hills are formed from
macigno
, a quartz-bearing sandstone so hard that during the Middle Ages it was favored for grindstones. In Florence it was also used in the construction of buildings. So abundant were its seams that all one had to do to build a house in Florence, it was said, was to dig a hole and then pile up the stones. Several quarries were actually found within the city gates, between the convent of Santa Felìcita and the Porta San Piero in Gattolino (now the Porta Romana); one of them was even owned by the convent’s nuns. The Arno also provided the city with supplies of stone: a limestone known as
lapidum Arnigerum
was quarried along its south bank. But the stone for the circumferential chains would come from the Cava di Trassinaia, a quarry a few miles north of the city, near the ancient town of Fiesole. Work sheds were erected on the site in March 1421, and soon afterward a team of nineteen stonemasons went to work.

The sandstone chain.

Most stonemasons served their apprenticeships in the quarries, learning from their masters how to recognize the best beds of stone, how to cut with or against the grain, and how to dress them according to the architect’s templates. The extraction and shaping of a stone made for strenuous labor. A saw was first of all used to cleave the stone from the hillside. In the case of a hard stone like
macigno
, mixtures of sand and iron filings would be sprinkled under the teeth of the saw to act as abrasives and compensate for the comparative softness of the metal. Pried free with a crowbar and wooden wedges, the stone was cut roughly to size with a pickax and afterward dressed using a hammer with a lighter blade. It was then sounded — that is, struck lightly with a hammer — as a test of its quality. If there were no flaws, the stone would ring like a bell, whereas a dull thud indicated a crack or some other defect, and it would be discarded. Another test of quality was the smell. Freshly cut from a quarry, limestone and sandstone smell of rotten eggs, and the stronger this sulfurous stench, the better the quality of stone.

The dimensions and shapes of the stones needed for Filippo’s chains were highly particular. The beams for the circumferential rings had to be seven and a half feet long and seventeen inches in section. As each side of the octagon would require ten of these beams to form the two concentric rings, a total of eighty were needed to encircle the dome. Each stone was to have a series of notches cut into its underside so it could interlock with the shorter beams laid transversely beneath. Even more of these shorter beams were called for, ninety-six in all. These can now be seen protruding several feet from the base of the dome, like sets of teeth, just above the large round windows in the drum.

Templates, either drawn on parchment or carved from wood, were used as guides for dressing the stone. But because of the complexity of Filippo’s design, the stonemasons had difficulty understanding how exactly the stones were to be cut and then fitted together. The enterprising
capomaestro
therefore made other, less conventional models for them to follow. A number of these were made from wax and clay, and some he even carved from
rape grandi
, large turnips that the Florentines ate in winter.

To function effectively, the circumferential stones, which met at 45-degree angles, needed to be linked tightly at their ends. This was achieved by iron clamps that coupled them together. Filippo traveled to Pistoia to oversee the casting of these clamps, which were so specialized that the ironmongers, like the stonemasons, could barely understand what was required of them. Once forged, these clamps were glazed with lead to prevent the iron from rusting and therefore causing the surrounding masonry to crack. Many thousands of pounds of lead were used for rustproofing both these clamps and the iron bars installed elsewhere in the cathedral. Plumbers (whose name comes from the Latin
plumbum
, “lead”) were employed at most cathedrals in the Middle Ages to rustproof iron or make lead tiles for the steeples. Such a recourse, naturally, meant one more danger for the workers at Santa Maria del Fiore, for it had been known at least since the time of the Romans, when the architect Faventinus observed the “deformity” and “dreadful anaemic pallor” of plumbers, that lead was a poisonous metal.
¹

The sandstone chain was only the first of four to be laid, part of a system of four bands that would encircle the dome at regular intervals of 35 feet. In the spring of 1425 Filippo executed a model for the second chain, which was even more complex than the first because the transverse beams were radially disposed, like a set of spokes, toward the vertical center or “hub” of the dome. Also, they were inclined at an angle rather than laid horizontally, a process that would require the expertise of the new
castello
as well as an extremely precise system of measurement.

The Opera’s documents record that the sandstone beams were to be superimposed by continuous iron chains. Iron has a far higher tensile strength than sandstone, meaning that the iron chains encircling the dome would actually have provided most of the resistance to the horizontal thrust. However, these chains, so essential to the dome’s success, are also one of its secrets: it is impossible to know their composition, for the simple reason that all of them are embedded in the masonry and therefore hidden from view. There is no reason to assume that they were not installed, but a magnetic survey conducted in the 1970s failed to detect any evidence of them.

The sandstone chains are not the only circumferential ties in the dome. They were supplemented by a fifth chain, made of wood and installed in 1424, which encircles the dome 25 feet above the first stone chain. Four of these wooden chains were originally planned, but in an example of how “in building only practical experience will teach that which is to be followed,” only the first was ever executed.

The wooden chain created problems for Filippo from the start. The program of 1420 had specified that it should be made from beams of oak 20 feet in length and a foot wide. But a year later, when sufficient quantities of oak proved hard to find, chestnut was chosen instead. In all, twenty-four beams would be needed, three on each side of the octagon, and they would be spliced together with clamps made from oak. Although the chestnut beams were ordered in September 1421, as the first sandstone chain was being laid, they would not arrive until more than two years later — no doubt a discouraging sign to anyone still dreaming of erecting the structure with a large wooden centering. There was, first of all, the problem of finding chestnut trees of an adequate diameter. Then, once the timber was found, it had to be felled in accordance with various rules and traditions, such as waiting for the wane of the moon, since wood cut at this time was thought less apt to breed worms. And once felled, it had to be properly seasoned, a time-consuming operation. First of all, its sap was driven off by soaking the wood in water for up to a month. Alternatively, the timber would be buried for several weeks in ox dung, in much the same way that animal hides were tanned with manure. The wood was then placed on a bed of ashes or bracken and exposed to the air, but protected from the rain and sun, for anything up to several years. Given these various procedures, it is little wonder that Filippo faced such a wait.

Like the four stone and iron chains, the wooden chain was no doubt part of Filippo’s system of invisible buttressing, a means of containing the hoop stress of the dome, for wood, like iron, has a greater tensile strength than sandstone. It may even have been intended to protect against a particularly violent kind of stress. A similar series of wooden ties were incorporated into the base of the dome of Santa Sophia in Constantinople, at the point where the greatest tension would develop, while more timber bonds were introduced into the brickwork following the earthquake of
A.D.
557.
²
Likewise, a double ring of poplar beams embraces the dome of the tomb of Öljeitü at Sultaniya, put there to counteract damage caused by the earthquakes on the Plateau of Persia.

Did Filippo have a similar form of protection in mind when he designed his wooden chain? Manetti alludes to “hidden devices” that were placed inside the dome to protect it from both the wind and earthquakes. Wind loading (the force exerted on the dome by the wind) was not of particular consequence, because of the sheer size of the structure.
³
Earthquakes, on the other hand, were a factor. Quakes would strike the city in 1510, in 1675, and again in 1895. The shocks from the first of these were so severe that many people spent the ensuing nights camped in the open air of the piazzas rather than returning to their homes. None of these earthquakes, however, caused damage to the cupola.