Proceedings of the First International Congress on Construction History, Madrid, 20th-24th January 2003,
ed. S. Huerta, Madrid: I. Juan de Herrera, SEdHC, ETSAM, A. E. Benvenuto, COAM, F. Dragados, 2003.
Foundations and wall structures in the basement
of the Colosseum in Rome
Heinz Jürgen Beste
Following
a generous
offer made by Rome' s
Archeological
Superintendence,
the
German
Archeological
Institute in Rome commenced
an
investigation of the construction details of the arena
and the basement of the Colosseum in 1996. The
Colosseum forms part of a series of well preserved
major structures
of antiquity
whose state of
preservation
is inversely
proportional
to our
knowledge of the history of the building.'
The ultimate aim of the work we have commenced
is to distinguish the multitude of building phases that
followed each other in the five centuries in which the
world's largest amphitheatre was in use.2 Moreover, it
is proposed to attempt design reconstruction based for
the first time on a thorough cataloguing of building
details and history. The Superintendence also expect
a proposal for using the basement as a museum, as
also for the restoration of the arena floor.
The basement of the amphitheatre, which several
walls subdivide into a series of corridors, measures
76.12 m along the longitudinal axis of the oval and
about 44.07 m along the shorter transverse axis. It is
delimited by the so-called encircling wall, which in
its day carried the podium. Constructively this wall is
the counterpart
of the so-called foundation wall
which supports the outer facade. If one today
approaches the edge of the arena from the stands, one
looks down at the basement, because the wooden
arena floor has long since disappeared.
The
longitudinal
axis of the oval-shaped
basement
is
oriented in the East-West direction, and this is also
true for the greater part of the walls one finds there.
These walls, made of tufa blocks and bricks,
subdivide the basement into twelve corridors of
different widths and lengths and, more precisely: six
elliptical corridors that run parallel to the encircling
wall and nine corridors parallel to the longitudinal
axis (fig. 1).
PHASE 1
At the beginning of our investigations attention was
concentrated
on distinguishing
the individual
building phases and, more particularly, the tufa walls,
because
these divide the basement
into the
aforementioned
corridors. For this reason I shall
henceforth describe these tufa walls as Phase l. As
reconstructed on the basis of our building records,
this phase is characterized by a very light but longspanning construction. The walls, which are about 90
cm wide and 6.30 m high, attain arch spans of up to
4.0 m. These filigree walls made of tufa, which are set
between 2.0 and 4.0 m apart and carried the wooden
arena floor, were presumably at first underdesigned
for withstanding
the vibrations
caused
by
earthquakes, ground subsidences and perhaps the
games
themselves
and therefore
had to be
subsequently stabilized, as is clearly brought out by
our findings (fig. 2).
374
H, J, Beste
---
Strutturaintravarllnonella
fondaziOne del CQIoaeeo
Sb'uttunil in tufo negU ípogei
=~~~":'=~O
SlnItturalnmeltonJ
I¡ I
. .,.',..,."..;."..- s'c-'-
~. .".;';-."'_::- "'
f
""'
~.
-" -"
.:',.;...
""w, ''''. -
... ...,
r:iií
"~.'~"~.a.&~~~,"',.;,
lit .h
~C"''''''''-'''''-'_.'
,_:::~.T'~;:::';'-:'"
¡
" - '"
.j
:;' .,
-.........
~...~~~i¡:',r;;~~~~'"
,
.
,
"
\
¡
::::-:-~..,.
~":'"'::: ':¡;:i:::
--:~..
-'o,
~
o
5 10
20"
:10
040
--_so
&1
__.TO
..
!lO
100
Figure I
Foundation pIatform and basement of Colosseum, general site plan
PHASE 11
For this reason all the openings in the arches were
reduced by inserting brick arches 60 cm deep within
the existing ones. What is particularly noteworthy
about this work is that the soffit was covered with
sesquipedales and bipedales, thus making it possible
to use falsework instead of complete centering. At the
same time as the arches were consolidated, all the
openings and passages were narrowed by providing
each soffit with a strong brick pillar (c.60 x 60 cm).
The only reason why we assume that the walls
numbered 4 and 11 belong to phase II is that they are
made out of brickwork with a basic tufa layer. 1
suppose therefore that in phase 1 there may have been
wooden piles instead of the walls 4 and 11. There are
numerous square structures in the basic tufa layer
which 1 believe functioned as the piJe holes for this
walL For reasons of construction the walls 4 and l 1
could not have been built at a later stage as they
support walls belonging to phase IlI.
PHASE III
This phase consisted of stabiJization involving all the
walls. It repeated the procedure of phase II by adding
masonry in all the passages, which were thus further
reduced in size. The tufa walls in corridors e, E north,
E south, G north and G south were each reinforced by
the addition of brick walls rising to a height of c.3 m;
they were joined to each other by means of transverse
arches and thus gave rigidity to the original walls
(fig. 3).
Foundations
Figure 2
Reconstructed
scheme of the basement,
and wall structures in the basement of the Colosseum
375
Phase ]
PHASE IV
Phase IV constituted a very substantial alteration of
the existing subdivision of the basement because it
involved the erection of a brick wall (c.2.3 m high) in
corridors F north and F south that closed off the
greater part of the passages that served them. This
phase is distinguished from all the others by the fact
that it does not consist of bricks of uniform size and
rests on a pediment made of material that must ha ve
fallen down from the top of the walls.
Theoretical
treatises about technical details or
traditions regarding the design and dimensioning of
foundation are not to be found in the literature of
antiquity. Vitruvius, who provides us with quite a few
constructional details, dedicates a brief subchapter to
the theme, but never really comes to grips with the
problem.3
The state of research of modern times reflects the
lack of interest of the classical authors. Every
description of the building speaks of the foundation
and the manuals do of course provide an summary
overview of the various solutions, but fight shy of a
true analysis of the problem.4 Though the foundations
are literally fundamental for every building structure,
only a few papers have hitherto been dedicated to this
theme.' And this notwithstanding the fact that some
classical authors speak of building failures during
construction due to inadequate foundations.6
The fact that the 1'oundation is the sole part of a
building that cannot be designed by means 01'
permanent correcting trials during construction shows
us that the problem must be considered to have been
solved in the case of the monuments that are still
standing today. Although considerable interest has
always been shown in the structural concept and the
376
H. J. Beste
Figure 3
Reconstructed scheme of the basement, Phase III
static safety of the Colosseum, the state of our
knowledge about the foundations of the building is
very scant. 7 Presumably
the surviving
ancient
structures and especially the Colosseum suggest a
stability and durability such as to make us tacitly
assume that they are adequately founded.
In his book entitled Ingegneria romana( 1927),
Giuseppe Cozza was the first author to consider the
problem
of the foundation
of the Colosseum.
According to his concept, which he developed in the
light of the structure of the eastern gallery (38), the 80
walls that carry the roughly 50 m high superstructure
and cavea of the Colosseum do not stand on a
foundation platform, but rather on strip foundations
made of opus quadratum, each roughly 3 m wide and
reaching down to a depth of about 6 m. In order to
assure a solid support and thus enhance their carrying
capacity, these eighty strip foundations -in
the
opinion of the authorare joined by transverse
arches, while the space between them has been
backfilIed with soil.
For a long time the theory formulated by Cozza
about the structure of the Colosseum's
foundations
was neither doubted nor checked, and it was on]y in
1977 that a trial boring executed in the area of the
outer circumference
showed that the foundation
consisted
of opus caementicium,
and not opus
quadratum as had previously been assumed.x
The reconstruction of the arena floor commenced
in the year 2000 brought with it the possibility of
sinking six sampling cores in the basement to cIarify
the foundations on which it rested. Three of these
were situated in walls 7, 8 and 13 (S 1-3) and a fourth
in corridor H (SA). A further trial boring was carried
out in niche 39 (SB) and yet another (SC) between
niches 15 and 16. The latter, however, was horizontal
rather than vertical (fig. 4).
The results of borings SB and SC, considered
377
Foundations and walJ structures in the basement of the CoJosseum
ea +9001
+8¡.oQ..sin1
a 80&110 di fondazione
C8. +22,00
b muro di contenlmento
¡pogel
intemo
(US 635)
cantenimento
eslemo
e fondazione
f
muradi
~
20
30
~
ea.
~ +9.00
+22,00
g~
h~
I +8,00 slm
~ +15,50
$1m
ea. +15.50. +12,901 +11.35 slm
~
ca.+22.00~
40
Figure 4
Section through the shorter transverse
50
i galleria
probabilmente
funzionale
al cantiere
¡pagel
ea. +16,20
slm
+16.00slm
60
70
axis from the CoJosseum
together with those of the 1977 trial, made it possible
for the first time to make some concrete statements
about the construction process and the size of the
foundations. Accordingly, the walls in the basement
would seem to have a separate foundation, while the
fa"ade and the cavea of the Colosseum rest on a
foundation ring made of opus caementicium that is
about 60,30 m wide and 12-13 m deep. The manner
in which it was constructed
permits its being
subdivided into two sections. The coring results show
that the shallow footing of the ring is situated at a
level of 10.0 m as\. Section 1 of the foundation rises
from this level and reaches up to about 16.0 m as\. It
is now documented that it was constructed with the
help of timber shuttering, which actually be seen in
the basement at a distance of about 0.50 m below the
upper edge of this first section. Section II of the
foundation reaches from about 16.0 m asl to about
22.0 m asl and is lined both on the side of the fa"ade
and on the inside (i.e. in the area of the basement) by
a 2.30 m wide brick wall, the so-called inner and
of view, the fact that different techniques were
employed for the two sections means that the lower
section must have been built below ground level and
with the help of formwork.
To permit the construction of the two encircling
walls, on the other hand, the surrounding ground must
either ha ve been excavated down to level 16.0 m asl
or must not ha ve risen above that level in the area in
the area in which the Colosseum was erected, our
state of knowledge suggesting the latter assumption
to be more probable.
This assumption
is al so
confirmed by ongoing excavations, which show that
the level of the buildings destroyed by the fire of 64
A.D. is of the order of 15.50 m asl and therefore lies
only some 0.50 m above the floor level of the
Colosseum basement.
The result of the horizontal boring SC through the
inner encircling wall made it clear once more that the
brick wall, quite apart from its architectural
articulation
into 40 niches, performs
first and
foremost a structural function. The wall sections
outer encircling
between the niches, 1.80 m wide by 2.30 m deep, act
walls.9 From the construction
point
378
H. J. Beste
as buttresses or counterforts and therefore enable the
brick wall, which within the niches has a thickness of
no more than 0.72 m, to resist the pressure of the
foundation ring between 16.0 m and 22.0 m as!.
Whether the outer encircling wall is constructed in the
same manner has to remain an open question for the
moment, because so far our knowledge is limited to
its uppermost layer.
Trial borings SI and S3 in the basement of the
Colosseum showed that walls 7 and 8 stand on strip
foundations that are dug into the graund of the
basement and have a width of about 3.25 m. The
material contents of the two cores enab]e us to
distinguish two sections of these strip foundations.
Section 1, which has a depth of about 3.25 m and
extends between 12.75 and 15.50 m asl, is made of
opus caementicium and is topped by a layer having a
thickness of the order of 25-50 cm, the so-called
]evelling course (16.0 m asl), likewise made of opus
caementicium.
The contents of core S3, which
reached down to 11.20 m asl for an overall length of
4.80 m, consists exclusively of opus caementicium,
but in this case the levelling course cannot be
demonstrated. The results of the two borings are very
clear and show that the 6.0 m high walls in the
basement rest on foundations having a depth that
varies between 3.25 and 4.80 m and are therefore
adequately founded. The borehole in corridor H (SA)
showed that only the levelling course was poured
there and that it has a thickness of about 50 cm
(fig. 5).
The archaeological investigations undertaken at the
same time made it clear that the structure of the
foundations in the basement is considerably more
complex than is suggested by the results ofthe coring.
Four of these (S 1, S6, S 19 and S 27) are of particular
interest: they braught to light wooden residues along
the strip foundations and thus made it possible to
conclude that these must have been constructed
within 4-5 cm thick timber formwork, which -given
the high water table in this area, has been well
preserved to this day.
Though the high water level prevented
the
boreholes fram reaching a really substantial depth,
they showed that the strip foundations of waJ] 1, 5, 6
and 14 consist of three sections, which contrasts with
those of wall 7 and 8, which consist of only two
sections. The layer of opus caementicium (US 635) is
followed by another layer of opus caementicium,
lit-I~ ~ ~ ; 'i...~ j
m14
11113 002
m11 m10
!!~! !!!tli
<
. . . 8 8 . 8.
11'19 1118
417
1116
1115 fn4
m3
m2
11'11
~~ :O1;~ ~~~: ~~ ~tt~
. ,t. l. nI1~f;~~'
~,~~ ~~ ~U:I~ ~
~ ~tl1
DD\~[J
O aa
~:tH ~aJ
.
\
~~
IB1cdl ,l'IJ
DI)
O¿7 (t.
o () O o
\) ~ <, [J
~ c>~
I)\»
<lo
I)~
~O
°¡ry <JQ
~
C.~1-e
a -'kIdllondB:zIone
b_dl~
ctonaazione~(US635J
.\11111
,
"l
c.~
~
~ c>
~ c:?
C::J
CJJ()
Qt.+22,OO-+Q.OO/+8.00slm
el. +22,00-+16.00...
(:8.+16,00-.,2,151+11,20"
'"l
'"l
d~(US628)c:a1O-30o::m
(US62'9)
IItlllURldl¡pog.i
'''''-'
..
I
ca.10-25cm
':' )"N
Figure 5
Detail plan and section through the shorter transverse axis of
the basement
though in this case having a different composition.
The upper layer has the same width as the lower one,
but its height varies fram one wall to another. Just as
in the case of walls 7 and 8, it is topped by a levelling
course (US 629). Having a thickness of the order of
10-15 cm. It is not clear why the structure of the
foundations should vary fram one wall to another. All
1 can do is to suggest that the subdivision into several
sections might be due to unevenness of the terrain in
the basement area. Presumably the lower section of
opus caementicium was in each case at first braught up
only to the level of the adjacent graund. The levelling
was then obtained by means of the second layer of
opus caementicium poured on top of the first, so that
the height of the second layer must have depended on
the graund Jevel. The levelling course, which covers
the entire area, i.e. both the corridors and the strip
Foundations
and wall structures in the basement of fhe Colosseum
foundations, then brought the structure up to 16.0 m
asl, the desired floor level for the basement. Since the
foundation section made of opus caementicium of
walls 2 and 13 neither has a so-called cement-gravel
bedding nor is topped by a levelling course, it seems
possible that these two foundations might have served
as level marks. They presumably also served to define
the location and direction of the other foundations,
since these -following
the pouring of the levelling
coursewere no longer exposed to view.
reconstruction in the area around the Colosseum is as
yet extremely scant, so that a more precise concept of
the size and location of the stagnum wilI have to await
the driJling of further trial holes.
NOTES
l.
2.
CONCLUSIONS
The roughly 50 m high superstructure/fa<,;ade and the
cavea of the Colosseum rest on a foundation ring
consisting of some 246,000 m' of opus caementicium.
The enormous mass of the foundation seems to be
adequately
dimensioned,
because
parts of the
southern side of the structure collapsed only in the
12 century,
presumably
as a result of strong
earthquakes. The fact that the thesis according to
which the Colosseum rested on strip foundations of
opus quadratwn, though formulated in 1927 without
a detailed investigation, had to wait 50 years before it
was corrected is indicati ve of the limited interest that
is still being dedicated to structural problems, and this
notwithstanding
the fact that they are altogether
elementary for the building in question.
Furthermore. recent data about the structure of the
foundations and knowledge of the street level of the
buildings that previously stood on the site of the
Colosseum
enable us to make the following
hypothetical
statement about the Colosseum area.
Unlike the western side of the valley, it would seem
that this area was not redeveloped after the fire,
because
otherwise
the second section of the
foundations with the two encircling walls could not
have been constructed in this particular manner. That
the area kept clear of the housing reconstruction
undertaken by Nero, which reached levels up to about
20.0 m asl, might have been constituted by the
stagnum
Ne ron i.l' seems
plausible
from
a
constructional
point of view and would partially
explain the short time it took to put up the Colosseum,
because no excavation work would have had to be
undertaken from this level down to 16.0 m as!.
However, our knowledge of the level relationships
between
pre-Neronian
buildings
and
Nero' s
379
3.
4.
5.
6.
7.
8.
9.
For the history of the Colosseum see Parker 1876; Rea
1996; La Regina 2001. For the gladiatorial combat e
games
see Golvin
and Landes
1990. For the
amphitheatres see Golvin 1988; Wilson Jones 1993.
On the individual phases and the function of the various
corridors in the basement, see Beste 1998, 106 ff.; 1999,
249 ff.; 2000, 79 n.; 2002, forthcoming; Beste and
Schingo 1998; Rea and Beste 2000. 311 ff.
Vitruvius JII 4,2.
Adam 1994, 115 ff.; Giuliani 1990. 119 ff.; Ginouves
1992.7 ff.; Durm 1909.
Kienast 1991, 123 ff.
Tacitus. Ann., IV, 62-63; Plinius, Epist. 10,48.
Cozzol927, 204 tI, fig. 133; idem, 1977,20 ff., fig. 12;
Lamprecht 1987, 155 n., fig. ISO.
Mocchegiani Carpano 1977, 10 ff., figs. 2, 6; idem,
1985. 122 ff., fig. 1.
Schingo and Rea 1993, 65 ff.
REFERENCE
LIST
Adam,
J.-P. 1994. Roman
Bulding.
Materials
and
Techniques (London).
Beste, H.-J. 1998. «Relazione suBe indagini in corso nei
sotterranei, i cosiddetti ipogei», RÓmMitt 105 (1998)
106-18.
Beste, H.-J. 1999. «Neue Forschungsergebnisse
zu einem
Aufzugssystem
im UntergeschoB
des Kolosseums»,
RÓmMitt 106 (1999) 249-76.
Beste, H.-J. 2000. «The construction
and phases of
development
of the wooden arena flooring of the
Colosseum», JRS 13, 79-92.
Beste, H.-J. 2002. «Osservazioni
tecniche negli ipogei»,
RÓmMitt 109 (2002)
Beste. H.-J. and Schino, G. 1998. «Documentazione
e
ricerche riguardanti l'arena e gli ipogei del Colosseo»
Forma Urbis 12 (1998) 4-9.
Colagrossi, P. 1913. L 'Anfiteatro Flavio nei suoi venti secoli
di storia (Roma).
Cozzo, G.1927.lngegneria
romana (Roma).
Cozzo, G.1971. l/ Colosseo. L'Anjiteatro
Flavio nella
tecniea edilizia, nella storia delle stmtture, nel concetto
esecutivo dei lavori (1971).
Durm. J. 1885. Handbuch der Architektur
Bd. 2 Die
Bauhmst der Riimer (Karlsruhe).
380
H. J. Beste
Ginouves,
R. 1992.
Dictionnaire
Méthodique
de
l"Architecture Grecque et Romaine. VoI.II, 7 ff. (Rorne).
Golvin, J. C. 1988. L'Amphithéatre
Romain. Essai sur la
théorisation de saforme et de sesfonctions
(Paris).
Golvin, J. C. and Landes, C. 1990. Amphitheatres
el
Gladiateurs (Lattes).
Giuliani, C. F. 1990. L 'edilizia nell"antichita 119 ff. (Rorne)
Kienast, H.-J. 1991. «Fundarnentieren
in schwierigern
Geliinde: Fallstudien
aus dern Heraion von Sarnos»,
Diskussionen zur Archaologischen
Bauforschung Vol. 5
(1991) 123-7.
Larnprecht, H.-O. 1987. Opus Caementicium. Bauchtechnik
der RÓmer' (KOln).
La Regina, A. 2001. Sangue e arena Catalog (Rome)
Mocchegiani
Carpano,
C. 1977. «Nuovi
dati sulle
fondazioni dell'anfiteatro
Flavio», Antiqua 2, Heft 7,
1977, IOff.
Mocchegiani Carpano, C. 1985. «Interventi nell' Anfiteatro
Flavio», Roma, Archeologia nel Centro 1, 122-4.
(7lh edn.
Parker, J.H. 1876. The Flavian amphitheatre
London).
Rea, R. 1996. Anfiteatro Flavio Romana (Rome).
Rea, R. and Beste, H.-J. 2000, "Sotterranei del Colosseo.
Ricera prelirninare al progetto di ricostruzione del piano
dell'arena» RÓmMitt 107 (2000) 311-39.
Schino, G. and Rea, R. ]993. «11 progetto di restauro del
Colosseo. I sotterranei: assetto idraulico e interventi
strutturali tra XIX e XX secolo», BullArch 23-24 (1993)
65-10 1.
Wilson Jones, M. 1993. «Designing
arnphitheatres»,
RÓmMitt 100 (1993) 391-442.