by Unknown
It is also worth scrutinizing once more the relationship between the transitional block and the portico. While Colini and Gismondi’s observations, discussed earlier, concerned the actual portico, certain constructional details fit a hypothetical taller one. The original sequence of nine numbered points embraced by the compromise hypothesis can now be extended with reference to a sectional elevation of the transitional block in both its actual and intended form (Fig. 7.13).51
x. The 10-foot-wide concrete strip foundations under the portico are unusually wide for the columns they carry, and would have been adequate for larger columns (Fig. 7.13, A).52
xi. At high level, the front face of the transitional block presents some unsightly projecting blocks (Fig. 7.13, B and C). These facilitated construction in some way or other, though exactly how rather baffled Colini and Gismondi. With a hypothetical taller portico, all such blocks would have been hidden from view between the suspended ceiling and the roof.
xii. The profile of the transitional block sets back where the cornice demarcates the high-level register, just as occurs on the rotunda (Fig. 7.13, D). This set-back follows the classical principle of recession, in tune with structural logic (walls high up in a building need not be as thick as those below). On the front of the transitional block, moreover, the set-back tracks the upper pediment, which was therefore an integral feature of the composition. This arrangement makes most sense if a roof had been planned to arrive here – that is, that of the hypothetical taller portico.53 (Contrariwise, there is no such set-back at the level of the existing portico roof.)
xiii. The ancient bronze trusses that once spanned the portico displayed oddities of configuration, as is clear from surveys made before 1625, including one by Borromini at the time this singular assembly was taken down (see Fig. 10.1).54 In particular, the tie beams over the central aisle did not reach far enough to be seated over the colonnades, and were instead supported by raking struts (Fig. 7.13, E). As Louise Rice has observed, the design need not have been so complicated had the portico achieved its intended height.55 The other puzzle she highlights is the lack of space in the side aisles for semicircular vaults; such a configuration would have conflicted with the trusses (Fig. 7.13, F), and so the ceilings here must have been either flatter than semicircular, or completely flat.56 It is significant that the original project offers a majesty and simplicity entirely in tune with the rotunda and dome. The ideal portico could have accommodated semicircular barrel vaults for all three aisles, and not just over the central aisle.57 Considering the impact of the one section of barrel vault that remains in place, the brick and concrete portion over the vestibule (Fig. 7.14), the effect of the original tripartite and even loftier arrangement would surely have been astounding. Crucially, the main aisle would have been spanned using simple trusses supported over the columns.
At the same time, the original design was consistent with the following advantages: a total height, measured to the peak of the pediment/roof would have been 100 feet (more or less), an eminently satisfying round dimension that echoed other key dimensions (e.g., the 150-foot diameter of the ring of interior columns, the 75-foot datum for the entablature and middle cornice of the rotunda, the 60-foot height of the columns, and the 50-foot height of their shafts). The relatively steep pitch of the pediment is now explained; this particular rake was a necessary ingredient for sweetly resolving these various conditions and intentions.
These last points, especially xi and xii, suggest that when work resumed on the transitional block, there was possibly still the intention to achieve the taller portico. But other features suit the actual portico, including the inclined line of bipedales just above the roof that Colini and Gismondi observed, and the embedded capital blocks already mentioned. As regards the latter, it is noticeable that these are not neatly encased in the masonry as would befit work made all of a piece; there is a slight gap to the sides that would be consistent with their having been lowered and levered into a seating that was fashioned at a later stage to the initial building of the masonry in this area.58 Following the nonappearance of the desired 50-foot shafts, it seems that there was an uncertain phase when both options – to use 50 or 40 footers – were kept open pending a definitive decision.59 The choice of the latter was probably put off until the last possible moment, inducing a hastiness that contributed to the messiness of the final outcome.
The compromise hypothesis, then, can potentially account for most, if not all, of the design puzzles that the Pantheon presents on its entrance side. It also concurs with the relative phasing of construction. But can we be more precise and pin down the specific dates involved?
Brickstamps
The practice of imprinting bricks and other Roman building products of fired clay with the identification marks of individual production units (officinae) and their parent brickyards (figlinae) happens to have been particularly prevalent in the years spanning Trajan’s and Hadrian’s rule. Usefully, for study purposes these stamps can be dated either roughly or in some cases to a particular year.60 This assigns any building in which they are found a terminus post quem; the building must have been erected after the bricks were made, although it is hard to say precisely how much later.61
On the basis of the prewar studies of Herbert Bloch and Julien Guey, no less than 115 of the 120 stamps observed in situ in the Pantheon belong to the late Trajanic or early Hadrianic period.62 It is significant that similar stamps are dispersed in different parts of the building.63 Thus, all of the brick and concrete parts must be roughly contemporary, including not only the grottoni but also the upper parts of the transitional block. This evidence confirms that the former was built quickly, and that the interruption of work on the latter was brief.
Establishing exactly when works on site began is controversial. As Lise Hetland shows in Chapter Three, the brickstamps that can be dated precisely, or relatively precisely, are mainly late Trajanic. Bloch argued that the Trajanic shipments were stockpiled, not to be taken up until Hadrian instigated the project after coming to power in the middle of 117. Exposing a certain circularity in Bloch’s position, Hetland argues more straightforwardly that the project was Trajan’s, in line with Wolf-Dieter Heilmeyer’s ideas of the 1970s based on stylistic comparisons.64 And is it not more logical, asks Hetland, that Trajan commissioned a replacement Pantheon sooner rather than later after the fire of 110 that ruined its predecessor? In short, a start date between 112 and 115 is more likely than one around 118.
The key consideration for the end date is that in AD 123, a higher than usual proportion of bipedales were produced bearing brickstamps, often with the names of the then-reigning consuls Apronianus and Paetinus. The absence of such stamps in the superstructure of the Pantheon shows that it must have been completed by this time or soon after, in other words by around 124.
It is revelatory to focus on a single stamp that does not fit the general pattern. This is the sole example from the whole building that is unambiguously Hadrianic, one recorded by Rodolfo Lanciani and datable to AD 123. Bloch was struck by the anomalous character of this find, in effect adding another enigma to the building that Lanciani called the “Sphinx of the Campus Martius.” Bloch knew that the rigors of his discipline were unassailable; no structure can be earlier than the latest stamp present (provided it is not connected with out-of-sequence working or repairs). Having been found close to ground level, did not this one stamp postpone the start of construction to later than 123? Bloch resolved this dilemma by supposing that Lanciani had simply been mistaken.65
Lanciani’s record, however, sounds as if it were accurate: “read by myself on the 25th of April on a piece [scaglia] of brick extracted from a sounding made by the north east corner of the brick front, behind the marble pilaster.”66 Rather than doubt his word, there is a way of reconciling it with the evidence of all the other brickstamps. The key is the find-spot, just behind one of the marble pilasters, that is to say exactly where the columnar system of the portico meets the transitional block
. In all likelihood, Lanciani’s scaglia would have formed part of fill material that accompanied the positioning of the pilasters and the erection of the columns.67 Given that the portico could have been put up relatively rapidly (thanks to the prefabricated bases, shafts, and capitals), this evidence suits the dedication of the new Pantheon sometime between 125 and 127. Since Hadrian returned to Rome in the summer of 125 after his first tour of the empire,68 it is tempting to assume that he would have presided over the ceremonies in person.
The Progress of Works on Site
There are thus two main possibilities for the duration of the project from conception to completion: either a period of seven or so years (ca. 118/119 to ca. 125/126), if we give credence to Bloch, or one roughly five years longer (ca. 113/114 to ca. 125/126), if we give credence to Heilmeyer and Hetland, which on balance I think we must. It may also be noted that the papyrus cited earlier that concerns a 50-foot shaft in transit across the eastern Egyptian desert dates to the third year of Hadrian’s reign, specifically the winter months of 119/120.69 If the shaft were indeed intended for the Pantheon, the timing seems too early for a Hadrianic commission; on the other hand, it fits neatly with a start under Trajan.70
In the normal course of events, as DeLaine demonstrates in Chapter Six, a total construction period of six or seven years would be feasible for the Pantheon. But from what we have seen, events at the site were far from normal. Delays were generated by the improvised erection of the grottoni. Delays are also implicit in the interruption of the transitional block caused by the nonappearance of the intended shafts for the portico. (It remains difficult to say whether these delays ran separately or concurrently.)
The combined evidence of the sources, brickstamps, worksite logistics, and the present examination of the fabric thus allows the sequence of operations and chronology of the project to be reconstructed as shown in Plate XIII, which is to say along the following lines:
* * *
110 Previous Pantheon burns Trajan reigns
112–114 Conception of the new Pantheon; scheme design
114–116 Site preparation and foundations
116–119 Progress on brick and concrete superstructure (117) Hadrian’s accession
118–121 Rotunda suffers cracking; progress interrupted; improvisation of the grottoni; nonappearance of 50-foot shafts for the portico (118) Hadrian returns to Rome
120–123 Grottoni completed; work begins on the dome; work on the transitional block interrupted (121) Hadrian leaves Rome
122–124 Completion of the dome; completion of the transitional block; decision to use 40-foot shafts for the portico
124–125 Completion of the portico; installation of statuary and fittings; finishing and inspections (125) Hadrian returns
125–127 Dedication of the Pantheon
128 Hadrian leaves Rome
* * *
Apollodorus and Hadrian
Inception under Trajan as opposed to Hadrian makes it more likely that the Pantheon was designed by the architect-engineer Apollodorus of Damascus, who was Trajan’s preferred designer but apparently at odds with Hadrian. Certainly Apollodorus is the more credible author of the Pantheon than Hadrian himself, who has also been proposed.71 As we have seen in the Introduction, ancient sources credit Apollodorus with Trajan’s Forum and Baths, both quite exceptional projects.72 Attribution of the Pantheon to him, too, makes sense on several levels. There are shared stylistic traits in the marble decoration of the Pantheon and Apollodorus’s Forum of Trajan.73 The open-air half rotundas of Trajan’s Baths offer points of similarity with the Pantheon rotunda in respect of both proportions (see Chapter Five), and the form of coffering (see Figs. 5.2, 5.3). It is especially significant that in elevation, these spaces present a comparable syncopation to that manifest inside the Pantheon. In two out of three of the surviving exedras in the Baths, the niches in the walls are synchronized with the coffers of the vault on the main and diagonal axes, but nowhere else.74 As a result of the present study, it is possible to identify further parallels between both projects as regards staircases. The unusual trapezoidal configuration of the Pantheon stairs is in fact closely anticipated in the staircases behind the half rotundas of the Baths (see Fig. 5.2). Individual flights of stairs also share constructional similarities, for example, in the disposition of bipedales.
It is well, furthermore, to recall discussion about the centering used to build the dome. This would have been a considerable work of engineering in its own right, and Apollodorus was evidently a master architect-engineer with extensive expertise in the erection of giant timber structures, as attested by his authorship of the Poliorcetica. Ancient sources also credit him with a pertinent technological feat, a huge wooden bridge over the Danube, which apparently approached 170 Roman feet or 55 meters in span (though probably less in reality). This sensational structure, which is represented in compact form on Trajan’s Column, was destroyed on Hadrian’s orders out of fear that it would provide a conduit for barbarian invasion. (Some of its stone and concrete piers still survive.)75 The bridge was the subject of another treatise by Apollodorus, a work which, though since lost, was referred to by the sixth-century historian Procopius in such a way as to suggest that it was still well known in his own day.76 Apollodorus, then, was in a singularly good position to have mastered the erection of large timber superstructures, and so too the centering of the Pantheon’s dome. Against the background of the prevalently anonymous history of Roman architectural practice, this completes as good a case for attribution as can ever be made for an architect of the period on the basis of design approach and circumstantial evidence.
It is curious, too, that the persons of Apollodorus and Hadrian come into conflict, according to the testimony of the third-century senator and historian Dio Cassius.77 Apparently, the emperor first banished and later put to death the architect on account of bad feeling that began long before, when Trajan was consulting Apollodorus, who tactlessly put down one of Hadrian’s interruptions with the remark: “be off and draw your pumpkins, you don’t understand any of these matters.” Later, after becoming emperor, Hadrian sent his own design of the Temple of Venus and Rome to Apollodorus, only to receive intolerable criticisms. The divine statues had been made too tall for the height of the cella, so much so that “if the goddesses wish to get up and go out, they will be unable to do so.”
The disparaging reference to pumpkins, or gourds, was most likely an allusion to the scalloped vaults that Hadrian and his architects used to such effect at his villa at Tivoli.78 It is tempting to wonder if the story about the Temple of Venus and Rome was a corruption of a text in which the Pantheon was the real focus of dispute.79 Dio could hardly have endorsed this possibility since, after all, presumably on the basis of the inscription on the portico, he believed the Pantheon to have been built by Agrippa, as is clear from a passage discussed in Chapter Two. Bad feeling between Apollodorus and Hadrian may have been further fueled by the demise of the former’s bridge on the latter’s orders. Be that as it may, the fact is that Dio reports antagonism between the two men, and rivalry that revolved around contrasting approaches to design. If there were even a kernel of truth to this, and if Apollodorus were indeed the designer of the Pantheon as I argue he was, it is easy to imagine the two men taking up opposing positions over this project and how to resolve the misfortune that had befallen it.
Presumably, Apollodorus held out for the taller portico and its majestic 50-foot shafts, while the emperor sought to prevent further embarrassing delays by resorting to compromise. From his knowledge of Athens, Hadrian may have been aware that the Propylaea of the Acropolis had two separate pedimented roofs, and that when seen from a distance, one might look as if it were superimposed on the other. Was it he who imposed the double pediment solution, while commandeering a batch of 40-foot columns from another project under way in the capital?
Leaving aside such conjecture, the building site of the Pantheon was eventful, to say the least. Improvisation at the sou
th end suggests that the dome was thought to be in jeopardy. Then there was the dilemma caused by the nonappearance of the intended column shafts at the north end. The architect, whoever he was, no doubt had to shoulder the consequences and perhaps the blame for them, too, even if unfairly so. Remembering all the while that design represents a team effort, the architect(s) of the Pantheon can stake a claim to one of the most sublime architectural experiences of all time. As the product of a rare genius and extraordinary technical audacity, it must have given its author immense satisfaction, yet by this interpretation, the building of it was harrowing in its uncertainty and immensely frustrating. The awesome magnificence of the interior should have been matched on the exterior, but instead the designer saw his vision spoiled. Much of his efforts must have been directed at artfully minimizing the negative impact of circumstances that could not be avoided. But compromise is part and parcel of an architect’s business. Building the Pantheon was a dream that turned nightmarish, though in the end it sends all who enter into reveries.
1 The diameter of the Temple of Diana, part of a thermal complex, is fractionally greater than 29.5 meters, or 100 Roman feet. The so-called Temple of Apollo, also at Baiae, apparently measures about 35 meters (ca. 120 ft) in diameter, but too little is known about this structure to be sure that it once supported a dome. In Rome, the caldarium of the Baths of Caracalla, originally domed, spanned about 35 meters too.
2 Selected studies of imperial construction include G. Lugli, La tecnica edilizia romana, Rome 1957; William L. MacDonald, The Architecture of the Roman Empire, vol. 1: An Introductory Study, London 1965, 2nd ed. rev. New Haven 1982; M. E. Blake, Roman Construction in Italy from Nerva through the Antonines, Philadelphia 1973; Jean-Pierre Adam, La construction romaine. Matériaux et techniques, Paris 1984, translated as Roman Building: Materials and Techniques, Bloomington 1994; F. C. Giuliani, L’edilizia nell’antichità, Rome 1990; Janet DeLaine, The Baths of Caracalla in Rome: A Study in the Design, Construction, and Economics of Large-Scale Building Projects in Imperial Rome (Journal of Roman Archaeology, Supplement 25), Portsmouth, RI, 1997; Larry F. Ball, The Domus Aurea and the Roman Architecture Revolution, Cambridge 2003; Rabun Taylor, Roman Builders, Cambridge 2003; Lynne Lancaster, Concrete Vaulted Construction in Imperial Rome: Innovation in Context, Cambridge 2005.