0% found this document useful (0 votes)
143 viewsPrinciples of Roman Architecture - Pantheon
Uploaded by
Luiz Ricardo Lopes da SilvaCopyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
0% found this document useful (0 votes)
143 viewsPrinciples of Roman Architecture - Pantheon
Uploaded by
Luiz Ricardo Lopes da SilvaCopyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
/ 38
PRINCIPLES OF
ROMAN ARCHITECTURE
MARK WILSON JONES
s
YALE UNIVERSITY PRESS NEW HAVEN AND LONDON IX
THE ENIGMA OF THE PANTHEON:
THE INTERIOR
THe PANTHEON AFTRACTS SUPERLATIVES like no other
oman building, being the most famous, the most
sagnificent, the best preserved, and the one with the
Digest span. Yet the brightest star in the Roman
irmament has its dark side. Its genesis, function and
are all open questions, ones that ancient sources
answer only in riddles. Indeed, the silence that shrouds
the eatly history of the Pantheon seems almost con-
spiratorial: its image is completely absent from coins,
while the few known literary references are at best
cryptic and at worse confusing. Modern understanding
of the building therefore hangs to a large extent on
the analysis of its physical fabric, which happens to be
notoriously problematic.
¥.1 Inerior of the Pantheon by Giovanni Paolo Pannini (ca.
734). showing the atic a8 i appeared before the alterations ofthe
so. Nstional Gallery of Ar, Washington
‘An emblematic obstacle is the prominent inserip-
tion in the frieze over the portico (fig. 10.2)
AcaGRIPPA-LFCOSTeRrivM'HECrT. Ie thus declares the
building to have been founded by Marcus Agrippa and
completed during his third round as consul, in 27 oF
25 nc. But near the end of the nineteenth century
archaeological evidence emerged to prove that the real
patron of the existing structure was Hadrian, Thi
fact, is the third Pantheon on. the site: Agrippa’ first
‘incarnation burnt down in ab 80, to be replaced by a
Domitianic version which burnt in its turn to. make
way for the definitive solution begun (oF at least con-
ceived) around 118, and completed some eight or more
years later. So despite the fict that Hadrian rebuile it
9.2 (above) Denil of the main onder of the interior of the Pan-
theon. Note the exquisite carving of the capital and the strong con-
trast created by the hand of red porphyry in the Bieze
7 ool
ooo
120. The section reconseuces the original frm of the atic fiom start to finish, he apparently put up the inscrip-
tion commemorating the original patron as a show of
respect for tradition,
As for the name, Pantheum in Latin, dv@ewov in
Greck, this is equally curious, for it was not the official
original tide, but rather a. sort of nickname." Its
meaning was already a subject for speculation in a
passage from Dio Cassius’ histories which is another
puzzle in itself. Evidently fooled by the inscription, he
‘gne the building to Agrippa. Dio’ mistake may be
tunderstandable, but it hardly inspires confidence in his
Ako he [Agrippa] completed the building called the Pan-
thoon. It has this name perhaps because it received among
the images which decorate it the statues of many gods,
including Mars and Venus; but in my opinion the name
drives because the vaulted roof resembles the heavens.
Agrippa, for his part, wished to place a statue of Augus-
tus there also and to. bestow upon him the honour of
hhwwing the structure named after him; but when the
emperor would accept neither honour, he placed in the
temple itself a statue of the former Caesar fie. Julius
csr] and in the porch statues of Augustus and himself?
So Dio simultaneously provides two readings, one
eriving from celestial symbolism, the other from
satues of multiple divinities, consistent with the
common perception of the Pantheon as a temple to all
the gods (from the Greek pan: all; and theios of a god).
Buc in many ways it eludes this definition. The inserip~
sion is unlike the usual form for temples, in as much
ss ie omits to mention the god or gods concerned.” A
pasage in the Historia Augusta which cites buildings
that Hadrian restored excludes the Pantheon from the
list of temples.* Then there is the form of the great
jomed interior; it resembles that of no other temple,
but finds parallels in imperial baths and palaces, and
ater mausolea. Few temples were circular, and those
sre relatively small. The question of size is relevant
‘once interiors were intended primarily as homes
cule statues rather than for group worship (whic
don the altar outside), so_large dimensions
\were not inherently necessary. Tradition demanded
single occupancy, that is to say one divinity per room,
csplaining why temples to the Capitoline triad have
hhree rooms and why the Temple of Venus and Rome
hs two (fig 1.8). So the Pantheon, with its single vast
canopy, is unlikely to have been a temple in the strict
sense of the term, although this does not rule out a
‘piritual realm of some kind and temple-like associa-
tions, as indeed the presence of the great pediment
\would seem to imply
AGRIPPA’S PANTHEON: THE ‘SPHINX
OF THE CAMPUS MARTIUS
So what then was the Pantheon? This slippery ques-
tion forces us to try to get to grips with a non-exis-
tent building, namely Agrippa’. It is thought t0 be
significant that the gods mentioned by Dio were asso-
ciated with Augustus’ family, the gens lula” The inten
tion to set up a statue of Augustus inside the building
along with Caesar (his adopted deified father) and
Venus (the mythical Genetrix or founder of the Julian
line) echoes Hellenistic, Panthea, where divine images
__were disposed around that of the reigning sovereign.*
~ Augustus’ refusal of Agrippa’ proposal is characteristic
of his studied respect for traditional Roman values and.
the rejection of the trappings of absolute monarchism,
with its attendant political dangers. So while overt
worship of the living Augustus was out of the ques-
ugustus’ Forum, where the image of
smperor_enjoyed an intimate relation with
the gods connected to the Julian family”
9.4 Map of the Campus Martius showing the principal structres
in existence in August’ day
19 Dio’ ext makes it clear that chere were many other
gods on view in the Pantheon apart from the ones he
‘mentions, and i s reasonable to suppose that Romulus
was among their number, as at the Forum of Augus-
“tus. Legend had it that Romulus, tansformed as the
god Quirinas, ascended into. the heavens above the
plus apna, the marshy area around the ste of the P
theon (Big. 9.4). Conceivably, by making a formal lnk
between them, the Pantheon promoted Augustus 25 a
new Romulus, the founder of anew. Rome.” With
time, the original complement of gods in Agrippa’
building was no doube supplemented by later deified
cemperots, for the Pantheon in its later reincarnations
comps Gom the Harionic period: +g the became a celebration of the imperial institution rather.
Partcom, se, Haddin Vila, than it individual dynasties. In addition, Hadrian is
supposed to have held court in his rotunda, which was
a fundamental justification for such a large covered
space; it surpassed Domitian’: Aula Regia in the Palax
tine palace while putting the emperor’ justice on
display in a public place. On such occasions the main
apie must have doubled as a tribune where sa the
emperor."
‘The idea thae che Pantheon originated 28 a dynastic
m finds support in various strands_of evix—
dence. Its very name, 3s opposed to ‘Augusteum’, might
be viewed as a smokescreen intended to disguise such
a very un-Roman phenomenon as a building that
accorded divine honours to a mortal. Ie seems signifi
cant too that the building bestrides the axis from the
Baths of Agrippa to the Mausoleum of Augustus (ig.
9.4), the later being the resting-plice of the Jalian
family, to which Agrippa was related by marriage.”
Then there is the eireular form of the plan, which, a5
mentioned, would be exceptional for a straightforward
cemple
‘That none of this isa foregone conclusion is under
lined by the fact that one scholarly interpretation sees
Agrippa building as the ‘missing’ Temple of Mars in
Campo.” The potential for such divergent views exists
primarily because the shape of Agrippa’ Pantheon
‘continues to be a matter of dispute. It has been so ever
since 1892, when the brickstamps collected by Georges
Chédanne’ showed that the structure above ground
belonged to Hadrian’ reign (fig. 9-5)" Where then
vwas the building announced by Agrippa’ famous
inscription? Traces of its foundations were duly uncov-
cred in the ensuing excavations directed by Luca Bel
tami and Piet-Olinto Armanini. Most authorities
follow Rodolfo Lanciani’s view that the evidence
pointed t0 a T-shaped structure of che type illustrated
in figure 3.3, positioned where stands the existing
portico, with the entrance facing south onto an open 27 tp) General plan of the excavations of the 18908
sircted by Luca Belem and Pier-Olinto Armanan
YS Relramsi and Armananis excavations, details (a)
south-east sector of rotunda with buttress wall; ()
Portico and tanstonal Block, forecourt (fig. 9.7)" Bur the opposing school of
thought, that Agrippa’ building incorporated a rotunda
not unlike the present one, is progressively gaining
This question has important ramifications for the
genesis of the existing building, for ether Hadrian’
replacement was a radical new departure, or it followed
its predecessor’ lead. To my mind there is more t0 be
said for the second option. The hypothesis of a T=
shaped Agrippan building rests on litle besides the
uncovering of a projection on the south side of the
structure under the portico (figs 9.7 and 9.8), but there
's nothing a all co show this projection was not merely
a link to a rotunda, I is true chat this theory has the
attraction of explaining the absence of Pantheon-like
buildings before Hadrian’ time, but this may be an
accident of archacology, and nor is it complete
‘At Stymphalos in Greece is a litle heroon with a
comparable plan which certainly dates from before
Agrippa’s day (ig. 9.6), while a less obvious but more
important precedent was the tholos of the Athenian
agora; although this was intially circular it had gained
a rectangular porch by about the time of Caesar (Gig
9.6)”
‘There is a series of arguments in favour of an
‘Agrippan rotunds,"* but itis enough to mention just
three here. Firstly, the restoration of Agrippa’ intcip-
sion in the pediment of Hadrian’ building (fig. 10.2)
implies quite forcefully that both projects had more in
‘common than just the site." In fact Hadrian’ recon-
seruction work in the Campus Martvs was otherwise
respectfl ofthe Augustan character ofthe area” while
this mausoleum mimicked not just the ypology of
‘Augustus’ but alo its dimensions (Table 4.1). snot ll
this inconsistent with a Pantheon which broke radi-
cally with its forerunner? Secondly, there are traces of
‘wo earlier floors below the existing interior one, the
‘uppermost presumably Domitianic, the lowest Augus-
tan, The thinness of the Domitianic paving (only $ t0
Sem), and the types of marble used (including pavon-
azetio) strongly suggest a covered space of some sort,
cone that was apparently as wide as the existing one.”
Thindly, excavations begun in May 1996 have revealed
two flights of stairs Jeading up from the forecourt to
the portico, one superimposed over the other. Pending
publication of the report (and the possibilty of a secure
ating on the basis of small finds), the levels suggest
that the lower stair should belong to cither Domieian’s
‘or Agrippas Pantheon. Thus at least one of these was
‘orientated north, just like Hadrian’. In shore, there are
hints of continuity everywhere.
If Hadrian’s Pantheon did indeed mimic Dos
aR
cor even Agrippa’ this mises che quandary of che
original rotunda. Perhaps it was substantially open 1
the sky, sive for an ambulatory around the perimeter?
‘The possibilty of a roof spanning the entire space runs
‘up agains che problem of its sheer immensity. A con-
crete vault was out of the question, since the necesary
technology was a yet in its adolescence in the Augus-
tan period; besides, the seriousness of subsequent fies
evokes timber instead. A span somewhere in the gion
of 43 metres would have been far in excess of any-
thing known, s0 i seems strange that Pliny, whose
admiration was attracted by the 100ft (20%m) roof
timbers in Ageippas Dinibitorium nearby, made no
mention of it. Akernatvely, there could have existed
columns analogous to those supporting the tholos a
‘Athens (fig 9.6b) that somehow chided Beltrams’s snd
Armaninis tenches. In all of this the only sure thing
is lide concentric portion of opus riulauum at che
base of the existing rotunda documented in Beleramis
report (ig. 9.83)" fe may be impossible to know if it
was originally part of an open court or a full-blooded
building, bue either way itis 2 remnant of the ste
ture which conditioned both the position of Hadrian’
rotunda and its great size
CELESTIAL AND TERRESTRIAL THEMES
The chief novelty of Hadrian’ Pantheon was the
incomparable dome (fig, 9.10)-At one level its function
-was simply to astound the Roman populace; at another
it spoke of a universal cosmology, representing, as Dio
intwits, the celestial home of the gods While the
specific original intentions may remain clusve it seems
inconceivable that this was a dumb masterpiece of
technical, aesthetic and spatial experience: it com-
‘municated unequivocally at che level of symbol and
rhetoric too. This was certainly a factor that had an
influence, and sometimes a determining influence, on
the form of Roman architecture, and as such a subject
which bears on design, even if it must remain at the
fringes of the present discussion. It is however worth
signalling here the sort of interpretational problems
that come to the fore, by looking jst atthe reasoning
bbchind the orthogonal grid for the paving patter, a
choice seemingly at variance with the circular pln
Might the uncentred expanse of the floor be identified
‘with the earth’ limitless horizon, and the rotunda with
the perfect imperial order that bounds and shapes it
Or could it be seen instead as an almost automatic
response based on precedents in the imperial fora, and
in particular the apses of the Basilica Ulpia? My poine here is ewofold. Firstly as stated repeatedly, architectural
design is an inclusive, multi-faceted activity; space can
concerns, so that
be moulded to answer quite dispara
in the field of interpretation it is, theoretically at least,
possible t0 have one’s cake and eat it too. Secondly,
that any specific readings must always be subordinate
to the coherence of the project in its own right
1s exphined below, the play-off between the che-
querboard floor and its cylindrical container is actually
2 vital part of the formal resolution of the whole
Number symbolism doubtless made some contribu-
jon fo the message OF the rotunda, The articulation of
the groundplan according to a sixteen-part geometry
recalls, as does Vitruvius’ radial city plan, the sixteen=
part Etruscan sky, placing the Pantheon implicitly at
the centre of a celestial scheme.” On the other hand,
the successive bisection of the cardinal axes (fig. 9.11)
is an almost instinctive, commonsense way of creating
radial rhythms. Meanwhile, the coffering of th
is divided into nwenty-eight parts, the same number as
hat of the columns and pilasters belonging to the main
of
order.® Twenty-eight was considered ‘perfect,
2 very limited set of numbers that equal the sum of
their factors." Being ako a rough approximation to
the number of days in-the lunar cycle, ewenty-cight
nerhaps invoked a cosmic iconography to0.”” And if
the ewenty-cight vertical divisions represent the moon,
che oculus the sun, and the five horizontal rows of eof-
fering the remaining five planets, then all the seven ele-
aiens of the solar system hen known are present. Is
«ako relevant that the Pantheon has seven exedrae
including the apse on the main axis), one for each of
the asociated divinities?" Did the sunbeam ffom the
dculus mark astronomical cycles and events
The oculus undoubtedly contributed in a more
general sense to the mystical quality of the main
space (fig. 9.9). Since it provided all the illumination,
windows were unnecessary at a lower level, and their
absence was in fact vitally important, Being denied
visual contact with surrounding buildings puts the
visitor ina realm removed from everyday reality
luckily nothing came of a sixteenth-century project to
cap the oculus by a lantern and to introduce windows
2t a lower level). At the same time, the sin
of light provides a wondrous sense of drama. As the
sun moves across the sky, so the pool of sunlight strokes
the cupola, the walls and the floor, acting as a magnet
tention (fig, 9.1). Each day
js different, and the effect changes too with the
weather and the seasons. The sun reaches the floor
around midday between May and August, bathing the
for the viewer’ arse
99 Oculus ofthe Pantheon
9.10 View of the cupola ofthe Pantheon seen from below: Note
how the central fl of the coffer appears to be centrally post
Sioned within the overall frame of the iby an eller which is
achieved thanks to the asymetrial profile of collin in section.
whole interior in glorious reflected light and provid-
ing optimum conditions for reading sculptural detail
and gradations of colour. It is a constant surprise how
good the level of illumination is, despite the dicty grey
render that now covers the cupola.
183 9.11 Plan of the Pastheon with gromezicl interpretation oer-
Iai, 750.
9.12 Sectiom though the runds, with ideal cice veri. Nore
the comenity ofthe for.
The rising curve of the floor is one of the most
subdle features of the whole project. Some think it is
due to settlement,” while Lanciani thought that it was
a legacy of Agrippa’s project, where it supposedly
served to drain rainwater towards the edge of an open-
18g
air space.” But this convexity was most likely an
optical refinement, one that treads that fine line
“between being noticed oF not: itis clearly visible once
you look out for it, but not nearly 30 prominent as
that_of Michelangelo’ pavement on the Capitoine
hill While there are holes for drainage right under
the oculus, the curve may also have helped to encour
age dispersion and hence evaporation; at any rate i is
remarkable how quickly the floor dries off even after
a thunderstorm. Nor is it inconceivable that the con-
vexity of the floor recalled that of the each, just as the
dome recalled the heavens, thus reinforcing the numic
nous symbolism of the Pantheon 3b fmph mundi, a
celebration of the Roman world, with Rome and the
‘emperor at its centre.” The same message was arguably
reinforced by the provenance ofthe marbles decorating
the space ~ Greece, Asia Minor, Egypt and Numidia
(Tunisia) — providing tangible proof of Rome’ ceres-
al domain
GEOMETRY AND STRUCTURE
Tes now time to move on to les clusive issues, beyine
ning with geometry The interior distil into a her
spherical dome and a cylinder of the same height,
which is the same assaying that the whole space is 35
‘wide a iti high, oF that it contains within its embrace
the figure of a sphere (fg. 0.10). Here is an idea of
such portentous simplicity that it has been commented
con repeatedly, coming to the attention of non-
architect as early asthe time of Rabelais; what is more
recent checks have shown that this relationship is quite
accurate (fig. 9.12). It should also be noted that a
1:1 cros-section characterizes other Roman cincular
halls such as che Domitianic rotunda at Albano and the
“Tor de’ Schiavi (cf figs 4.7f and 4.10), while other con-
figurations work on a similar principle, such asthe ree-
tangular chamber inthe Pyramid of Gaius Cesta, where
the height equals the width of the shorter sides.”
‘The geometry of a dome on a drum brings to mind
cone of antiquity’ most celebrated tracts, Archimedes’
(On the Sphere and Cylinder (pp. 41-3). He rated the
series of proof published therein among, his greatest
achievements, for he had the device of a sphere and
‘oflinder set up over his tomb.” One of the theorems
established the equality in surface area of a hemisphere
“and cylinder with the same radius and height — a°3t
she Dantheon, Thus the surfice area of the Pantheon
‘cupola matches that of the wall supporting it.
‘The starting-point for relating geometry to the
fabric ofthe building was the 130% (100 cubits) ring sngarans, Tae
‘oases
——
1
aria esa a
ia
ona
Slee ete i)
v.13. Part levation of rorunds, projected Ast, with key proportions everid, 1400.
defining the axes of the interior columns. As men-
tioned in chapter 5, the importance of this circle is
confirmed by the way a square inscribed in it doubles
up t locate the axes of the portico colonnade (Fg.
9.11), Meanwhile the height of drum and transitional
block is also approximately the same, thus making a
nncarly perfect square in elevation (fig. 10.16). In effect
the volumetric proportions of the whole project can
be reduced to a hemisphere, a cylinder of the same
veight and a double cube. All this is incredibly appo-
site given the architectonic nature of the whole
project, that is to say the union of a cylindrical/
here ma ~ the rooms — sa cole oe — the
Portico. This grand simplicity also gives the lie to th
snnecesarily complex and often irrelevant tracings of
the sype illustrated in the Introduction.
‘The layout of the lateral exedrae also relates to the
basic diagram, for the circles that contain them touch
‘he inscribed square, Then the clear height of the inte-
rior was made equal to the clear span of about 147f,
as measured to the face of the columns/pilasters. But
the diameter of the cupola itself remains nearer 150f
so itis in fact nota perfect hemisphere, while a portion
of the implied sphere is cut off by the floor (fig. 9.12)"
In any event, the 150% dimension returns as the
approximate height measured to the top of the oculus,
thereby creating two sets of equal dimensions: one set
of 147f ones, another set of 1s0ft ones. In this way
the architect sought to legitimize the lack of geomet-
rical purity and deal with the problem that_archi-
tecture hat thicknes, unlike the Hne drawings used
‘By Geometers and mathematicians. As is only to be
‘expected, other modifications were made during the
course of detailed design. The main openings off
the rotunda are not, for example, completely regular;
the entrance passage is smaller than the similarly barrel-
, vaulted exedrae on the diagonal axes because it faced
"a tougher structural task.
‘When it came to elaborating the interior elevation,
geometry ceased to play the driving role, as is so often
the case in Roman architecture (pp. 120-22, 126-7).
Geometry had indirect effects, of course; fixing the
width of the exedrae, for example, conditioned the
rough column size, since there was a play-off between
185 9.14 (abore) Plan of the Pantheon showing the aan
seractural voids: (3) at ate level; (B) at the level of the springing
‘of the cupola, 1:750.
this and the intercolumnation. But the_composition_
was defined arithmetically The elevation was determined
“by the key decision t'split the notional 7sfe cylinder
into five parts, giving two to the attic and three to the
iain order (fig. 9.13).The resulting dimensions partake
of a comprehensive symmetria based on fractions of
soft (eg. 75, 30, 15, 5, 3%F) and/or multiples of sft
(eg. 45,40, 208)" The portal is emblematic of the same
quest, with its 40 X 20f (2:1) opening, its architrave
2/sft wide (an eighth of the opening) and is trabeation
fe tall (a fifth of the opening),
Any solution had necessarily to marry with the
structure, for it was this, after all, that made the
‘magisterial interior possible. Comfortably surpassing its
nearest ancient rival, the 120ft rotunda known as the
Temple of Apollo near Baiae, the huge span demanded
a complex response, and any similarity with simple
conerete shells like the Temple of Mercury at the same
site is purely superficial. There is nothing unusual about
the relative thickness of the drum (p. 82); what is
exceptional is the way the wall was treated. Together
9.43. Sections aross the rorunda, with Features of the consruc-
tion and materials indicated: (1) foundation of concrete with agar
fake of travertine; (2) concrete with aggregate of alcenating hers
fof uavertinefagiments and lamps of rota 3) concrete with aggre-
ate of alternating layers of ea and fragments of es (4) concrete
‘wth caonenae of predominanly broken bricks (s) concrete with
aggregate of alternating lyers of bricks and eu; (6) concrete with
aemenae of akernating layers of light eufa and voleanc sg. SS a
4
Ae Sa
916 Cutaway of the main structural elements of the rotunda
with the entrance passage and the apse, the six exedrae
_Punch into its depth, reducing the structure to eight
‘principal areas of support. In the middle of each of
these Were inserted a series of half-moon chambers
orientated towards the exterior, and a similar arrange-
_meiit was repeated at high level in the haunches of the
-dome (fig. 9.14). Combined together, these structural
voids save considerable amounts of weight without
reducing the buttressing action of the wall, which
depends primarily on the overall depth. The loads
pressing on these voids were transferred via a series of
solid brick relieving vaults which take up the entire
thickness of the structure (figs 0.14-9.16).
The viability of the structure was further improved
by the gradation of types of aggregate, ranging from
travertine, brick, ufa through to pumice; that is to say
=fromr-the strongest at the bortom to the lightest at the
® The presence of the oculus eliminated
ec e dome, a vulnerable area
mpressive forces are weakest. It may be no
coincidence that oculi were used in some of the very
best-preserved Roman cupolas (like those of the
‘Temple of Mercury at Baiae, the rotunda at Albano
and the octagoral halls in the Domus Aurea and the
Baths of Diocletian). This isnot to deny that the oculus
‘was valued foF its cosmological connotations and as a
source of light: illumination, meaning and structure
‘went hand in hand. Nor should proportion be left
‘out of the equation, since its_zoft diameter equals
both the height of the attic and the shafts of the main
“order, while also. dividing sweetly into the 1sof
iameter of the column ring. So the oculus wonder
fully symbolizes the Roman genius for synthesis (ast
discussed at the end of chapter 7). At one and the same
time it responded to the demands of theory, beauty,
content and practicality
PRAISE AND CRITICISM
‘The splendour of the Pantheon has long ensured it a
special place in Rome. When Ammianus Marcellinus
listed the principal buildings of the city in the fourth
century (ef p. 162), he likened the embrace of the Pan
theon to a ‘ity district’. Is impact was second only to
that of the Forum of Tijan, ranking with the Colos-
seum and the Temple of Jupiter on the Capitoline.
While other antiquities were pillaged for building
materials, the Pantheon was transformed into a church,
thus guaranteeing its survival, None the less to
redieval eyes it was the work of the devil ~ a back-
hhanded compliment if ever there was one; who else
17 9.17 Pantheon interior, seen fom the poral
9.18 Pantheon interior as it ought to be, according to Francesco
& Giorgio.
could have raised a dome of such awesome dimen
sions? The Legenda aurea also contained the story that
the cupola rested on a mountain of earth until con-
struction was completed, one that was eagerly removed
by the Roman poor, thanks to the money that had
been mixed in.** In any event, more than ewelve cen-
turies were to pass before a comparable feat was
attempted. The cupolas of first Florence Cathedral and
later St Peter's were undoubtedly planned to match the
span of the Pantheon while exceeding its height.”
With the rehabilitation of pagan culture in the
Renaissance the Pantheon came to be seen as not only
superior to all other antiquities, but also to anything
188
Oy
mee
a
ee sseoyy
§Og882 ons
‘more recent, There is a straight line between fifteenth-
century and nineteenth-century eulogies, from Flavio
Biondo’s declaration that the Pantheon ‘surpasses all
other churches of Rome’ to Henry James’ remark that
St Peter’, by comparison, is absurdly vulgar’* Every
architectural treatise gave the Pantheon pride of place
when it came to appraising the ancient legacy. Along
with the Parthenon it is the ancient building chat has
had the greatest influence on the subsequent practice
of architecture.
Paradoxically, there exists a vigorous parallel current
of negative criticism. Hostility is only to be expected
from the nineteenth-century champions of Greek or
Gothic architecture, since attacking things Roman was
an almost inevitable corollary. (For Viollet-le-Duc the
Pantheon typified the ‘dishonesty’ of the Roman
approach to design, using the orders to conceal vaulted
construction.”) Criticism is more intriguing when
it comes from quarters otherwise disposed t0 praise Ta RE”
spb cs eam a aha Go
pe Ea ibe
[tte co th ogee honey obo
tem fe “eter
ites
Par hc ahah 3
Bp nt gr er
Sn pe om oppo
areata
ooo pm
3 [yore
000
OX LE
ooo
09° ‘priate
ok epee
000 Sepa
dees meee nine Re
bye
: wo pois
poses ee i a
cena PE Pi EE.
tee enlegisilg px ae
=
e
cee
and there is plenty of it. Architects as distinguished
4s Antonio da Sangallo, Michelangelo, Palladio. and
Desgodets judged the building a flawed masterpiece
Apart from the portico (the subject of the final
chapter) dissatisfaction focused on the apparent lack of
tunity berween the constituents of the interior eleva~
tion: the main order, the attic and the dome. Critics
pointed unfavourably to the lack of correspondence in
scale and composition, and especially the fact that the
attic pilasters fail to line up with either the columns
below or the coffers above (figs 9.13 and 9.17).As seen
in chapeer 6, this was directly at odds with Renaissance
theory, as summed up by Francesco di Giorgio:
Asa general rule to be observed without exception
cory clement should sit im a steaigh line over it hike.”
Until the publication of Palladio’s Quattro libri it was
rare to illustrate the Pantheon as it actually was. Archi-
tects and artists tended, consciously or unconsciously,
to correct its perceived shortcomings." Francesco di
Giorgio’ sketch showed everything lining up as he
thought it should (fig. 9.18); by reducing the number
‘of verticals and splitting the attic into two zones, he
I fom Gin jn ot pe
pape hab op
he Tri pe thee
Hee eater sp
9.19 Annotated sketch of the Fanthcon interior and part plan of|
the portico by Antonio da Sangalo the Younger, with accom-
patying wrinen criiqu,
9.20 Drawing ofthe interior ofthe Pantheon by Raphael snd an
nonymous ats, showing the sate of che building before che aler-
tions to che atc of the 1750, but omitting ane of che principal
ered
18 also ensured that the arch over the apse did not cut off
any of the attic pilasters. Antonio da Sangallo sketched
an alternative proposal, adding in a note that the lack
of tectonic rigour was ‘a most pernicious thing’ (fig.
9.19). Since the drum was subdivided by four, eight
and sixteen radial divisions (fig. 9.11), he thought the
cupola should follow on the same principle, with
thirty-two or forty-eight coffers, not twenty-eight.
Settio followed with an adjusted elevation showing
the attic pilasters aligned over the columns and pilaster
below. Numerous later buildings, to cite just San
Francesco di Paolo in Naples, were created in this ‘cor=
rected Pantheon’ mould.” A series of perspectives of
the Pantheon interior, among them one attributed to
Raphael, are often held up as faithful depictions." yet
ceven these show the attic pilasters to be taller than they
really are (fig. 9.20). It seems that until recent times
only Bernini is known to have taken a positive view
of the attic, apparently realizing that the grouping of
its pilasters into bunches of four echoes the distribu~
tion of the main columns below.**The floor was much
admired for its great slabs of coloured marble, but
this feature too was sometimes edited. A characteristic
ambivalence may be seen in Pannini’s cight paintings
of the rotunda dating to the early eighteenth century.
In most of them, including the version now in Wash~
ington (fig. 9.1), the floor is represented faithfully; but
‘two other canvases create a radial organization in sym=
pathy with that of the cupola.”
So the unqualified praise that Procopius of Caesarea
heaped on Justinian’s Hagia Sophia: ‘proud in its inex
pressible beauty; superb in its volume and harmony
of proportions, having nothing in excess nor anything
defective’ could hardly be directed at the Pantheon.
On the contrary, the idea that its composition could
be improved became so deeply rooted that eventually
it was transformed into action. Using the need for
repairs as an excuse, the attic was completely remod-
elled in the middle of the eighteenth century. The
offending pilasters were dismantled, making room for
the heavy pedimented windows and framed panels
designed by Paolo Posi which conform more to Neo-
classical rather than Roman tastes. The result came in
for criticism; some thought that it would have been
better to keep about a quarter of the ancient pilasters,
that is to say just those which did align over the
columns below, and it seems that Piranesi was the
author of a proposal that sought to reintegrate these
pilasters with a second set artfally overlaid.” But Posis
scheme remains, with the exception of a small portion
of the ancient arrangement towards the south-west re-
created by Alberto Terenzio in the 1930s (fig. 9.17).
190
To avoid the uncomfortable conclusion that the
architect of the Pantheon was not very good at his
trade, many turned to historical explanations. As
Giorgio Vasari narrates, the theory went round that
different parts were built at different times and/or by
different architects. This information appears in an
interesting passage citing Andrea Sansovino’s use of the
Pantheon in defence of his own work at Santo Spirito
in Florence, According to Vasari, Sansovino
made a barrel vaule with richly and variously carved com-
partments, a much admired novelty It is true that the
work would be better if the compartments forming the
divisions of the squares and circles had been in line with
the columns, a ching which it would have been easy 0
do. Some of his old friends, however, have informed
ime that he defended this, saying that he had copied the
Rotunda at Rome, where... the ribs are not in line with
the columns. He added that if the builder ofthe Rotunda,
which is the best designed and proportioned cemple in
‘existence, did not take these things into account... it was
‘even less important on a smaller scale. However, many
artiss, and Michelangelo among chem, are of the opinion
that the Rotunda was built by three architects, the fist
carrying it up t© the comice above the columns, the
second doing so from the cornice upwards
this portion differs from the lower part, the vaulting not
corresponding with the lines of the divisions. The third
is believed to have done che beautfil portico. Thus
modern masters ought not to exeuse themselves like
‘Andeea."
because
A century or so later Desgodets elaborated further on
the faults of the pilasters
they are so slight compared with the ensblaure and
pedestal that one can scarcely cll it an order, eis badly
ordered asembly of pars. They are not ft for an atic at
all, which ought not to have the characteristics of an
order, thus preventing it from being called an atc, unless,
one supposed that once there was some such (plin) one
here, to which were later added the pilasters, an atchi-
crave, and a frieze
Carlo Fontana published a reconstruction of the sup-
posed original rotunda before these and other changes
(fig. 9.21). According to his interpretation, the attic
‘was just a minor aspect of a dramatic series of trans-
formations. The existing building was held to be the
‘outcome of three distinct phases: the first, republican
‘one, with a much lower floor level and no columns,
the second, Agrippa’s reworking, featuring the set of
caryatids reported by Pliny:® the third and definitive
reconstruction. 21 Carlo Fontana’ interpretation of the Pantheon interior as
the elt of sacceasve phases of consmuction: supposed fit phase
ror t the adnon of the main onder.
‘Once it was understood that the structure as. a
whole belonged to Hadrian, chronological explana~
tions such as these had to be trimmed back or aban=
doned altogether. So what, then, might explain the
sarious shortcomings of the interior? It should not be
‘gotten that design is a process, and that che outcome
a project may be the result of modifications and
compromise (chapters 3 and 8). But while it may be
timate to speculate about an original or ideal solu-
tion, its credibility rests on identifying specific causes for
its alteration. Could the cause here have a structural
nature? The superstructure of the rotunda displays
several enormous cracks running from its base to well
up into the dome, which evidently appeared immedi-
ately on completion (fig. 9.212). (This cracking pattern
isa common feature of Roman domes, since th
crete had negligible tensile strength to cope with ‘hoop
tension’. The result behaves less like a monolithic shell
than a collection of masonry arches arranged as orange
segments; it is the need to combat the thrust from
these arches that gives rise to the characterstcally
thick wall construction.) The magnitude of the crack~
ing demanded other remedial measures, including the
likely addition of the rows of walls linking the Pan-
theon with the basilica to the south. It seems they were
invented to act as buttessing — certainly itis dificult
to see what was the purpose of the windowless spaces
in between.” In the interior, however, there are no
signs of modifications of comparable significance. In
short, the idea that the present structure results
from change or adaptation appears to be a mirage.
TOWARDS AN ANCIENT AESTHETIC
In the last few years some scholars have started to shrug
off the shackles of Renaisance theory and look at the
articulation of the interior in a more positive light.”
The apparently conflictual disposition of the main
‘order, attic and cupola may now be seen to be an
inspired solution that avoids a static, sterile effect. The
dislocation between the coffering and the rest of the
structure contributes to the indefinable, but none
the less palpable, impression that the dome hovers over
the drum as opposed to weighing it down, By their
‘21a Elevation ofthe rocunda, projected fat, showing the principal cracks inthe structare 9.22. Pair of hemicycles in the Baths of Trajan (a0 409). Note the
Sangement whereby the niches line up wat the coring ove
head only on the orthogonal and diagonal aves At in the Par
thon the centes of niches can align wit the enees of be rater
than coffer,
semall site and distribution the attic pilasters contribute
positively to this sense_of detachment, as do other
details such as the use of red porphyry in both friezes
40 a5 to create strong horizontal bands separating one
level from another.
‘When Roman architects superimposed thythmic
clements upon one another, their default mode was
to place like upon like just a: Francesco di Giorgio
and Antonio di Sangallo advised. Yer lapses of align-
‘ment were admisible provided thee is « substantial
change of scale. As discussed in chapter 6, small atic
pilasters that do not respect the rhythm of the princi
pal order characterize surprising number of Hel-
lenistic and Roman fagades. Such an organization was
4 typical feature of numerous city gateways (figs 0:3
and 6.13), while two of the best-known examples, the
Porta dei Leoni and Porta dei Borsari at Verona, boast
~ as at the Pantheon ~ three levels each working on a
different compositional principle (figs 6.12 and 6.14).”
Other manifestations of the tendency to break away
from predictable vertical alignments are the column
displays in the aymphacum at Miletos and the Library
of Celsus at Ephesos, where the aedicules crisscross
‘over one another (fig. 6.9). So it scems that a minor
ity of ancient architects ~ just as Andrea Sansovino was,
in the minority in his own time ~ consciously searched
192
to fice fasade design from the academic stratjacket of
good Vitruvian manners
‘While the foregoing examples serve to put the
Pantheon atic in the general context of ancient prac-
tice, more direct parallels can be found in Rome in
the years immediately before and after its construction.
The exterior of the Praetorium of Hadrian’s Villa has
‘small pilasters at high level positioned with seant regard
for the openings below: The coffers of the barrel vaule
inside the Hadrianeum fail conspicuously to line up
‘with the scansion of the walls that support i” More
significant stil isthe design of three exedrae at Trajan’
Baths, which 2t oof in diameter are not that much
smaller than the Pantheon rotunda. In each case the
coffering aligns only with the niches below on the
‘main axis and on the diagonal (fig. 9.22). Moreover,
the centre of a niche sometimes aligns not with a
coffer but with a rib ~ a relationship analogous to that
‘occurring on the diagonal axes of the Pantheon.
APOLLODORUS ~ ARCHITECT OF THE
PANTHEON?
‘his last comparison raises intriguing questions, since
ancient sources name Apollodorus as the designer of |
‘Trajan’ Baths, while he has also been attributed with
the Pantheon by Wolf-Dieter Heilmeyer. Just sup-
pposing that the relations between Hadrian and Apol-
lodorus were not s0 hostile as Dio makes out (p. 24),
‘who would be better qualified than he to take charge
of the Pantheon project? As Donald Strong observes:
‘the continuity of tradition in che major public buildings
uring the later years of Trajan's and the extly years of,
Hadrian’ reign is reasonably explained by che fact chat
‘Trajan’ architect, Apollodorus, continued to act as general
‘overseer of public works under Hadrian.”
Although the word ‘fact’ seems misplaced, the basic
point is sound, especially if one drops the insistence on
Apollodorus and Apollodoras only, as opposed to 3
like-minded group of colleagues. However Heilmeyer’
atribucion has generally fallen on stony ground, mainly
because of the failings of his arguments in favour of
‘an earlier date. Perhaps the undoubted stylistic affinities
between mouldings in the Pantheon and Trajan’
Forum betray a commen bond not in the figure of the
architect but in the workshops entrusted with their
‘execution. None the less, the Apollodorus connection,
is worth exploring further, Excellence in engineering
‘was certainly at the root of Apollodorus' fame, so in
this respect he is a prime condidate for conceiving the Pantheon’s structure, Next there is the question of the
syncopation just observed in the exedrae of his Trajan’
Baths. Then there is Trajan’s Column, a likely product
of Apollodorus’ hand, which displays a subtle form of
syncopation in the helical stair: the edge of a step aligns
with the main axis, while the cenae of one aligns with
the cross-axis (fig. 8.9). Moreover the choice of four-
teen steps per turn is unusual in as much as this
number, like the twenty-eight sets of coffers in the
Pantheon, is a multiple of seven. The difficulties of
csublishing the authorship of Trajan’s Markets has
rirealy been discussed in chapter 1, but just supposing.
their designer was cither Apollodorus or one of his
collaborators, it is interesting that the hemicycle again
displays a phy upon local asymmetry and axial sym=
metry (fig. 1.18)
Do a
gies like these represent Apol-
5 it were? Or was he just the fore~
most exponent of a broader movement committed to
snimating an architecture that through its very success
rsked becoming heavy and simplistic? The evidence is
meagre, but perhaps just enough t© suggest that those
in Apollodorus’ ‘school’ or ‘circle’ worked in a differ
cnt idiom from their fellow
Villa. Rather than articulate simple plan-forms in
omples. ways, as at the Pantheon, the Tivoli team
pregnated plans with a geometrical sophistication
that carried up into the vaulting (fig 5.15; cf. fig,
5.203). This approach had consequences for the eleva~
‘on, which usually followed the thythms set up by the
vuking in the interest of avoiding visual confusion.”>
So itis possible to discern two antithetical schools of
ought: one committed to simple plans and sophisti-
ued clevations, the other the converse, Whether
the driving forces were respectively Apollodorus and
ladrian himself must remain a matter for speculation,
specially since the contrast responds in part to a sense
f decorum appropriate for official public projects in
Rome on the one hand, and the emperor’ ‘private’
