IOURNAL OF

ROMAN
ARCHAEOLOGY
voLUME 31 20L8
*
ARTICLES AND ARCHAEOLOGICAL NOTES

AN INTERNATIONAL JOURNAL
 Table of contents of fascicule 1

Articles
P. Visonä Rethinking early Carthaginian coinage 7

A funerary monument on the Capitoline: architecture and painting

P. Tucci 30
in mid-Republican Rome, between Etruria and Greece
Roman wall-painting in soutl-rern Gaul (GalliaNarbonensis and Galliq
A. Dardenay 53
Aquitania)
F. D'Andria The Ploutonion of Hierapolis in light of recent research (201,3-17) 90
The impact of_the-German frontier on the economic development of
M. Redd6 131
the countryside of Roman Gaul
Standardization and mass customization of architectural components:
N. Toma ß7
new perspectives on the Imperial marble construction industry
M.Liuzzo, G. Margani &
Thelndirizzo Romanbaths at Catania 193
R.I.A.Wilson
G. Castiglia
Rural churches and settlements in late-antique and Early Mediaeval 223
Tuscany
Archaeological notes
The Ludovisi "Suicidal Gaul" and his wife:
B. S. Ridgway 248
bronze or marble original, Hellenistic or Roman?
A cork model in Aschaffenburg (Bavaria)
V. Kockel 259
giving new evidence for Pompeii's House of Sallust
A. Laidlaw The Aschaffenburg model: addenda to JRA Suppl. 98 267
Gladiators and circus horses inthe lliad fuieze in
A. Pollard 285
Pompeii's House of D. Octavius Quartio?
A bronze portrait of a slave child from a presumed villa near
D. Ojeda 303
Medellin (Lusitania)
Games, banquets, handouts, and the population of Pompeii
M. Osanna 310
as deduced from a new tomb inscription
A view from the margin? Roman commonwares and patterns of g*g
A. Launaro & N. Leone
distribution and consumption at Interamna Lirenas (Lazio)
New investigations of the 6th-c. A.D. "church wreck"
]. Leidwanger at Marzamemi, Sicily
339

S. Ritter, S. Ben Tahar, I.W.E. Landscape archaeology and urbanism at Meninx: results of
357
Fassbinder & L. Lambers geophysical prospection on ]erba (2015)
The relationship between agricultural production and amphoru
S. Gallimore 3Zg
manufacture on Roman Crete
R. Sabar Josephus' "Cydasa ofthe Tyrians" (Tel Qedesh) 3g7
in eastern Upper Galilee
G. Pavlovski Designing the cqueq Stobi
of the theatre at 406

x- Belt types, identity and social status in late antiquity,

5. I omazlnclc 426
the belt setin Emona's grave 18
R. Gonzälez Villaescusa & The Late Roman fort of Can Blai on Formentera
445
J. H. Fernändez and its role in the defence of the Balearic Sea
Two portraits from Aphrodisias: late-antique re-visualizations
Lenaghan 458
J.
of traditional culture-heroes?
Directing the faithful, structuring the sacred space: funerary
S. Ardeleanu 475
epigraphy in its archaeological contexi in late-antique Tipasa
A cork model in Aschaffenburg (Bavaria) giving
new evidence for Pompeii’s House of Sallust
(a) Valentin Kockel
The Casa di Sallustio at Pompeii is one of the houses that belongs in every book and
study on domestic life and architecture in the Roman era. The fact that it was excavated at
the start of the 19th c., however, means that large parts of its decoration have long since
been lost through weathering and neglect, a situation further compounded by the damage
resulting from Allied bombing in 1943. In order to reconstruct, as accurately as possible, a
picture of the house as it stood in antiquity, it is therefore particularly important to evalu-
ate all the historical visual sources that document the house in a better condition. This
task has already been extensively carried out in the monograph of A. Laidlaw and M. S.
Stella (2014). With a similar goal, I propose to take a fresh look at a cork model dating from
1840 today housed in Aschaffenburg (Bavaria) in order to discuss its context and original
function. Next, A. Laidlaw will compare the model’s meticulously detailed copies of the
structure and decoration, still in 1840 almost perfectly preserved, to the present battered
state of the extant remains, thereby confirming the importance of the Aschaffenburg model
as the primary archaeological source for the house.

The invention of the cork model1

In the 1760s, within the sphere of travellers undertaking the Grand Tour, a new form
of reproduction of ancient architecture came into fashion: scale models of ruins. Crafted
out of wood, cork and plaster, these models achieved a degree of clarity and authentic-
ity that far surpassed the architectural drawings and vedute that were all that had been
available. Cork was popular material for models as its porous structure gave it a surface
texture that bore a resemblance to the weathered limestone of the temples at Paestum and
Tivoli. Individual pieces as well as entire series of buildings from Rome and surroundings
soon became popular purchases for visitors, to be displayed in the palaces and acade-
mies of Europe. For decades, such models were considered to reproduce the exemplary
monumental architecture of antiquity more vividly than any other artistic medium. Of the
monuments outside the area of Rome, however, only the Paestum temples where available
as models. The remaining ruins in the Kingdom of Naples remained under the ‘copyright’
of the Neapolitan authorities, who banned both accurate surveys and the making of mod-
els. This ban extended in particular to the buildings of Pompeii, where visitors were not
even allowed to sketch or draw. An exception was made only for the King of Sweden, Gus-
tav III, who in 1784 commissioned Giovanni Altieri (fl. 1765-1797) to make a model of the
Temple of Isis, which is on display today in the Drottningholm Palace just outside Stock-
holm, although he was expressly forbidden to sell copies of this model.2

1 V. Kockel, “Rom über die Alpen tragen,” in W. Helmberger and V. Kockel (edd.), Rom über
die Alpen tragen: Fürsten sammeln antike Architektur: die Aschaffenburger Korkmodelle: mit einem
Bestandskatalog (Landshut–Ergolding 1993) 11-31; id., “Towns and tombs. Three dimensional
documentation of archaeological sites in the Kingdom of Naples in the late 18th and early
19th century,” in I. Bignamini (ed.), Archives & excavations (London 2004) 143-62; id., “Models
of Pompeii from the eighteenth century to the “Grand Plastico”. The three-dimensional
documentation of ancient ruins,” in M. Osanna, M. T. Caracciolo and L. Gallo (edd.), Pompeii
and Europe 1748-1943 (Milan 2015) 266-75 (also in Italian).
2 V. Kockel, Phelloplastica (Stockholm 1998) 72-89.
© Journal of Roman Archaeology 31 (2018)
260 V. Kockel

The Padiglione family and the model of the town of Pompeii

This would change at the start of the 19th c. when the authorities in Naples realized
that three-dimensional models, as a comprehensive form of documentation, could be used
to great advantage in their own museum. As early as 1805, they commissioned large-scale
models of the temples at Paestum, which at that time were undergoing extensive resto-
ration. In 1806, now under French rule, the Naples museum set up its own designated
model-building workshop, where models — including reproductions of buildings in Paes-
tum, Capua and Pozzuoli — were produced for public display in a “Galleria dei Modelli”.
The exhibits included models of Herculaneum’s theatre and of Pompeii’s Temple of Isis, its
amphitheatre (a quadrant section?) and theatrum tectum, complementing public buildings
from Rome. The museum workshop was headed by Domenico Padiglione (1756-1832) as
“costruttore dei modelli”, assisted from 1818 by his son Agostino (c.1800-1856) in the posi-
tion of aiutante.3
Not until 1821, it seems, did it occur to the authorities to expand the display in the Real
Museo Borbonico to include not just public buildings but also examples of Pompeii’s resi-
dential architecture. In 1822, the Padiglione father and son produced the first model of a
house, the Casa di Campagna (now the Villa di Diomede), which was favourably received
by the king and subsequently placed on show.4 Padiglione senior then conceived the long-
term plan to create a model of “the whole of the city already excavated”. From the Casa di
Campagna to the Quartiere de’ Soldati (porticus of the theatre) the excavations were to be
reproduced in full in a scale of c.1 : 50. The existence of such a uniform concept is confirmed
by various commercial proposals submitted between 1829 and 1831 to the Rijksmuseum
van Oudheden in Leiden5 in which the art-dealers Raffaele Gargiulo and Giuseppe De
Crescenzo (with whom the museum was in negotiation over the purchase of a large collec-
tion of vases) list all the models they can supply; these included not only numerous works
of monumental architecture, but also sections of the town of Pompeii. Although the model
builder is not named, it emerges from a further note that he is Agostino Padiglione. In the
case of the Pompeii models, emphasis is placed on the fact that the individual elements
can be combined into a model of the whole town. Had it been brought to completion, this
project would have fulfilled the criteria of a comprehensive archaeological documentation
in an entirely new fashion, since the model builders during lengthy visits to Pompeii cop-
ied each wall accurately from all sides. This also meant replicating the wall-paintings and
plaster fragments surviving in situ. To this end, the draughtsmen responsible for docu-
menting the wall-paintings were required to make, in addition to full-scale copies of the
most important motifs, drawings in the same, smaller scale as the model, which could

3 A. Milanese, “Il Museo Reale di Napoli al tempo di Giuseppe Bonaparte e di Gioacchino

Murat,” RivIstNazStorArte 19-20 (1996-97) 345-405; Kockel 2015 (supra n.1).
4 Naples, Museo Archeologico Nazionale di Napoli, Archivio Storico, XIV B 8 (unpaginated,
undated).
5 Housed in the archives at the Rijksmuseum van Oudheden are 4 separate commercial proposals
from art-dealers dating from the years 1828-30; these differ both in scope and in such details
as the size and price of the models they offer for sale. The most detailed list was attached to
a letter of 7.2.1830 written from Naples by the museum’s agent J. E. Humbert, in: Ontvangen
brieven 1830, inv. no. 17.01.02/06|1830. See http://archieven.rmo.nl/uploads/r/null/e/4/
e4c817129a84d764bebcf57461ab07c65512775a842f6780cb62af1b428b0a37/069.pdf, digital pp.
94-96 (viewed 22 August, 2016).
On the art dealers De Crescenzo and Gargiulo and their activities, see n.16 below.
 A cork model in Aschaffenburg: new evidence for the House of Sallust 261

Fig. 1. Plan showing the individual sections of the planned model of Pompeii, 1820-40. The numbering cor-
responds to Gargiulo’s list and spelling (del. Valentin and Nikolaus Kockel):
7. Anfiteatro. 9. Teatro tragico. 10. Teatro Lodeo. 12. Tempio di Iside/Ginnasio. 13. Casa detta di Panza. 14.
Casa detta di Sallustio. 15. Casa detta di Arrio Diomede. 16. Strada di Pompei … in più pezzi. 17. Foro diviso
in più modelli (Tempio di Giove/Tempio di Venere/Basilica/Tempio della Concordia/Calcidico/Tempio di Mer-
curio). 18. Le terme di Pompei.

subsequently be glued onto the cork walls or transferred to their ‘plaster’.6 The model of
the Casa di Sallustio discussed below demonstrates the remarkable quality of this form of
documentation. Precisely on account of their detailed nature, however, these models were
not automatically available for purchase: unlike the other pieces, Gargiulo notes that sale
of the Pompeii series was subject to separate approval.
Which elements of this model of the whole town were actually completed, and which
remained unfinished or never got beyond the planning stage, cannot be precisely deter-
mined. By the time that Domenico retired in 1828, the Casa di Campagna, the entire
theatre district and the amphitheatre were finished, and the forum complex was close to
completion. In 1833, a model of the Casa di Sallustio, made by Agostino Padiglione in
collaboration with the painter and draughtsman Giuseppe Marsigli (fl. 1824-1837), first
appeared at the annual Naples art exhibition.7 It was followed by models of the Casa del
Fauno (1835) and the Casa del Poeta tragico (1837/38). The Street of Tombs and the Casa
di Pansa, on the other hand, receive no further mention and were perhaps never built. The
plan in fig. 1 shows the individual segments listed by Gargiulo in his proposal of Febru-
ary 1830, thereby conveying an idea of the overall concept for the model of the town. In

6 This task was specifically laid down in section 93 of the first Regolamento pel Reale Museo
Borbonico (1828): “Il Disegnatore … dovrà inoltre fare i disegni in piccolo di quelle dipinture che
mai dovessero aver luogo ne’modelli in sughero dei principali edifici di Pompei”: Archivio di
Stato Napoli, Ministero della pubblica istruzione, Fs. 358 (after L. Martorelli, “L’organizzazione
del lavoro dei disegnatori di Pompei,” in G. Pugliese-Carratelli (ed.), Pompei. Pitture e mosaici, X
(Rome 1995) 25).
7 This first model was sent to Spain before 1833: ASN Ministero pubblica istruzione. Real Museo
Borbonico busta 336 fasc. 03.
262 V. Kockel

what scale the cork model of Pompeii was to be built is not entirely clear.8 Only once, in
a written reference to 4 palmi, is a precise scale indicated: on the Neapolitan duodecimal
system, this may be calculated to signify 1 : 48 (1 onca : 4 palmi [à 12 once]). Other refer-
ences are only indirect and neither clear nor uniform: they give merely the approximate
size of the individual parts of the model, and ratios of 1 : 42 or 1 : 48 remain the most likely.
These proportions are supported by the scale of the model in Aschaffenburg (1 : 50). The
complete model of the town would thus have required a great deal of display space, and
it is probably on this account that Gargiulo and Padiglione offered their customers in the
Netherlands a much smaller version, occupying a surface area of just 2 x 2 m (and costing
just one-tenth of the large model); but the art-dealer managed to sell neither the collection
of vases nor the model of Pompeii.9
Although the model of the town was evidently never fully completed, its finished parts
could be seen in the museum for many years. After the demise of the Kingdom of Naples,
a substantially smaller model (scale 1 : 100) was commissioned by the new Soprintendente,
Giuseppe Fiorelli. Built between 1860 and 1863 by Felice Padiglione, Domenico’s second
son, this model continues to captivate visitors in the Museo Archeologico Nazionale.10
Nothing is known, on the other hand, of the whereabouts of the older model. In the case
of the cork model of the Casa di Sallustio, we are thus looking at a copy of one part of this
large but lost model of the town dating to the 1820-30s.

The model of the Casa di Sallustio11 in Aschaffenburg: an archaeological source (col. pls.
1a-b on p. 271-72)
In the course of his many trips to Italy, King Ludwig I of Bavaria made several excur-
sions to Pompeii. He must have taken his decision to create in Bavaria a life-size replica
of a Pompeian house at the latest in 1839, when he paid a renewed visit to the excavations
accompanied by the architect Friedrich von Gärtner.12 A piece of land lying prominently
above the Main river on the edge of Aschaffenburg, a small town in the extreme north-west
of the Bavarian kingdom, was identified as a suitable site, and Ludwig’s Pompejanum (the
name is an allusion to the Pompeianum, one of Cicero’s villas) was built and decorated
by 1850. The king and von Gärtner chose as their model the Casa dei Dioscuri excavated
in 1828/29, reproducing parts of its ground plan in only slightly modified form. In its iso-
lated location, however, and with its animated silhouette, the Pompejanum testifies to the

8 This important question cannot be pursued here in detail. Agostino Padiglione, like other
model-makers, offered his models in various scales so as to cater to different requirements. It is
not clear whether this was also the case with the model of Pompeii as a whole.
9 From the long list of items for sale, Humbert, who negotiated on Leiden’s behalf in Naples,
purchased just two models of antique tombs: Rijksmuseum van Oudheiden, inv. H III NNNN
1 and 2.
10 On this model, see V. Sampaolo, “La realizzazione del plastico di Pompei,” Il Museo. Rivista del
sistema museale italiano 3 (1993) 79-95.
11 Eigentum des Wittelsbacher Ausgleichsfonds in München, Leihgabe an die Bayerische
Schlösserverwaltung, ausgestellt im Pompejanum in Aschaffenburg, Inv.-Nr. Asch. V0059
(WAF). Dimensions of the base plate 120 x 146 cm, overall height c.20 cm. On the history of
the model and its archaeological significance, see in detail V. Kockel, “Das Haus des Sallust in
Pompeji,” in Helmberger and Kockel (supra n.1) 135-48 and 321-22 nr. 53.
12 K. Sinkel, Pompejanum in Aschaffenburg. Villa Ludwigshöhe in der Pfalz (Veröffentlichungen des
Geschichts- und Kunstvereins Aschaffenburg 22, 1984) 66-72; W. Helmberger and R. Wünsche,
Pompejanum Aschaffenburg: Amtlicher Führer (2nd. edn.; München 2006) 29-33.
 A cork model in Aschaffenburg: new evidence for the House of Sallust 263

influence of the contemporary villa architecture of Romanticism, and thus differs clearly
from a Roman domus.
Ludwig, who planned the Pompejanum as a museum that would be open to the pub-
lic, wished it to correspond to a Pompeian house in its interiors too. It was consequently
decorated with mythological scenes and ornamental motifs taken from Pompeian sources,
including copies of original wall-paintings that had been removed from the Casa dei
Dioscuri and taken to the Naples museum. The furnishings were likewise modelled on
those of a Roman household: replicas of bronze utensils were cast, while designs survive
for ‘Pompeian’ pieces of furniture that were never built.13 Ludwig thereby turned into real-
ity an idea that in Naples had also been regularly discussed but never implemented: that
of reconstructing a Pompeian house on its original site, or of building a replica close by.14
The purchase of a model that would illustrate, for visitors, the organization of the
rooms inside a Roman house, also fits this educational concept. The choice lay between
the Casa di Pansa and the Casa di Sallustio,15 and it finally fell upon the House of Sallust
perhaps because this latter contained more wall-paintings and because it was the only one
for which a model already existed. On behalf of the Soprintendenza, the model was exe-
cuted by Agostino Padiglione (architecture) and Giuseppe Abbate (painting).16 It arrived
in Munich in 1841 but, instead of being transported straight to Aschaffenburg, it spent over
70 years in the Glyptothek, before being transferred in 1913 to the collection of the Archi-
tecture Faculty at Munich’s Technische Hochschule, only finally to reach Aschaffenburg in
1969. Finally, in the early 1990s, the model underwent thorough restoration in advance of
the major exhibition Rom über die Alpen tragen (Munich 1993) and the subsequent re-organi-
zation of the complete collection of models at Aschaffenburg, following which it was at last
able to take its place in the Pompejanum for which it had been destined from the outset.17
The model stands on a wooden board measuring 146 x 120 cm. It shows not only the
House of Sallust itself but also a larger area of Regio VI, including the streets on three
sides of the house, the prominent fountain at VI 1, 19 and even some of the walls belong-
ing to neighbouring insulae (fig. 2; col. pls. 1a-b). As such, it clearly demonstrates that it
belongs to the large-scale project of a model of the whole town. It extends southwards
to the area around the Casa di Pansa (fig. 1, no. 14) and northwards probably to another
section around the Herculaneum Gate, even if the corresponding model does not appear
explicitly in Gargiulo’s list.

13 Helmberger and Wünsche ibid. 35-36 with fig.; 103-11 with figs. For documents in Naples, see
A. Milanese, In partenza dal regno: esportazioni e commercio d’arte e d’antichità a Napoli nella prima
metà dell’Ottocento (Florence 2014) 313-16 nos. 2.3, 2.47, 3.21, 3.44.
14 A. Milanese, “‘Exactly uniform and corresponding to the houses of the ancients’. The restoration
of a house in Pompeii (1792-1861),” in Osanna, Caracciolo and Gallo (supra n.1) 257-65.
15 Agostino Padiglione and Guiseppe Abbate were instructed in October 1839 to make a model of
the house “either of Pansa or Sallust”: Naples, Archivio Nazionale 328 I.010, Soprintendenza
Generale degli Scavi, Pompei, “Incarico ad Agostino Padiglione di eseguire, per il re di Baviera,
il modello della casa ‘di Panza o di Atteone’” (1839) (non vidi). In his application, Padiglione
asked for the authorisation to take measurements, confusing Casa di Panza and (Casa) altrimenti
detta di Atteone. In the end he obviously did the right job.
16 On the following, see in detail Kockel (supra n.11). On the commercial activities of Giuseppe
De Crescenzo and Raffaele Gargiulo and their (from today’s perspective) dubious dual role as
art-dealers and employees of the Real Museo Borbonico, see Milanese (supra n.13) 201-55.
17 Sinkel (supra n.12) 82 still considers it lost. The present author was able to ‘rediscover’ the
model in a storage depot in 1986.
264 V. Kockel

Fig. 2. Plan of the section of

Pompeii shown in the model
(del. Helge Svensshon).

The model uses a simple form of construction. The base consists of wooden boards onto
which thick sheets of cork have been glued.18 Cut into this base are a series of slots cor-
responding to the ground plan of the house. The walls, pre-decorated and made entirely
of cork, are inserted into the slots. To prepare them for painting, the cork surfaces were
first smoothed with a thin layer of chalk. The decoration itself was executed in tempera.
Masonry structures were imitated with scored lines and then glazed. While the floors con-
sist mostly of grey sand, those that are well preserved in the original are reproduced as
paintings on paper and glued into place. From the records pertaining to other commis-
sions, we know that Padiglione spent several weeks working on one such model at the
excavation site itself, in order to ensure that he reproduced the state of the ruins as accu-
rately as possible. In the case of the House of Sallust, however, this was unnecessary, since
Agostino was able to consult his own, earlier model, first exhibited in 1833.19
The Aschaffenburg model seems to have been intended right from the start to serve
an educational function too. In the area south of the Vicolo di Mercurio, where, strictly
speaking, the Casa del Forno (VI 3,3) should stand, the model leaves a gap. This was occu-
pied initially by a legend in Italian, which was subsequently replaced by an explanation

18 These technical observations are taken from an unpublished report by the conservator Uta
Ludwig.
��. See n.7.
 A cork model in Aschaffenburg: new evidence for the House of Sallust 265

Fig. 3. Model of the Casa di Sallustio (destroyed; formerly Munich, Deutsches Museum; München Deutsches
Museum, neg. no. *03159).
in German more or less identical in content. The text gives the names of Padiglione and
Abbate, the model’s makers, and of the art dealers Gargiulo and De Crescenzo, and also
provides information about the function of the individual rooms, whose location is indi-
cated by tiny numbered pieces of paper within the model. The actual measurements of
individual rooms can be read from a 1 : 50 scale bar. The model thus serves to illustrate, in
exemplary fashion, the functioning of a Roman house.
The model was damaged during and after the Second World War while still housed in
Munich; subsequently it underwent initial and, in places, imperfect repairs. Not until the
conservation campaign at the start of the 1990s were efforts successfully made to restore
the model in such a way that its rebuilt parts are easy to recognize but nevertheless yield a
homogenous overall picture.20
Further models of the Casa di Sallustio are known, but they offer no additional archaeo-
logical information. One was made soon after 1860, probably at the same time as Felice
Padiglione’s model of the town and likewise at the scale of 1 : 100; today it is housed in
Nîmes.21 A second, larger model is known only from a historical photograph taken at the
Deutsches Museum in Munich (fig. 3).22 It can probably be attributed to one of the Bra-
mante family, who continued to work on the model of the town right up to the end of the
19th c., and from whom it was possible to buy models of individual Pompeian houses
in the large scale of 1 : 50.23 There must have been several of these houses but only a

20 The restored columns of the W wing of the peristyle (31), however, are now much too high.
21 Auguste Pelet, before 1863. Nîmes, Musée archeólogique, inv. no. unknown. See A. Pelet,
Description des monuments grecs et romains exécutés en liège (Nîmes 1876) 139-53.
22 Munich, Deutsches Museum, neg. no. *03159. The model entered the museum either between
1903 and 1909 or between 1920 and 1925; it was destroyed during the Second World War. No
information regarding its purchase is known. I thank Titus Kockel for drawing my attention
to this model and Dirk Bühler for his subsequent research, which unfortunately produced no
further results.
23 Mentioned by A. Mau in J. Overbeck, Pompeji in seinen Gebäuden, Alterthümern und Kunstwerken
(4th edn. with A. Mau; Leipzig 1884) 633, n.22.
266 V. Kockel

model of the Casa di Apollo (Dresden, Skulpturensammlung) survives.24 The model of the
Casa di Sallustio formerly in the Deutsches Museum evidently consisted just of the house
itself; of the surrounding urban context, so clearly conveyed in the Aschaffenburg model, it
included only the Via Consolare in front of the entrance. From the photograph, it is impos-
sible to judge whether Bramante was here working to his own design or whether he was
basing himself on the original version by Agostino Padiglione in Naples (the reproduction
of the wall-paintings, in particular, appears more strident and more extensively recon-
structed than in the Aschaffenburg model).
The Aschaffenburg model of the Casa di Sallustio copies a model from the period
around 1830 that was part of the Padigliones’ ambitious project to reproduce the entire
town at the scale of 1 : 48. Unlike the cork models of buildings in Rome, which in most
cases show famous monuments in a state also known from other sources, it documents
a Pompeian house in a condition that is otherwise unrecorded and which is now greatly
deteriorated. Despite some damage, thanks to its authenticity the Aschaffenburg model
possesses its own value as an archaeological source — a value that, in view of its 1 : 50
scale, also far exceeds that of the 1 : 100 town model in Naples.25 This aspect as a primary
archaeological source is explored in the following section by A. Laidlaw.
valentin.kockel@philhist.uni-augsburg.de D-65183 Wiesbaden

Acknowledgements
I wish to thank A. Milanese and V. Sampaolo in Naples, W. Helmberger in Munich, R. Halbertsma
and B. van der Bercken in Leiden for their continuous assistance, and Karin Williams for her transla-
tion of my German text.

24 K. Knoll, “Das Haus des Apollo in Pompeji. Ein Korkmodell in der Skulpturensammlung,”
Dresdner Kunstblätter 46 (2002) 15-23; Kockel 2015 (supra n.1) fig. on p. 266; 274.
25 The documentary value of the model has so far been taken into consideration only by J.-A.
Dickmann, Domus frequentata. Anspruchsvolles Wohnen im pompejanischen Stadthaus (München
1999) 204 fig. 56; 359 fig. 104.
 (b) Anne Laidlaw
The Aschaffenburg model: addenda to JRA Suppl. 98
V. Kockel has discussed the model as one of the few surviving examples of a form of
three-dimensional archaeological recording that was developed by the Padiglione fam-
ily and other model-makers for the King of the Two Sicilies. Here I provide comparisons
with the extant remains of the house, to illustrate how much more we can learn from the
model of specific details of the structure and decoration that have been lost since 1840,
when the building was still in a remarkably better state of preservation. Aside from the
inevitable gradual deterioration of wall-paintings and pavements, which remained almost
completely open to the elements after the original excavation was completed in 1809, a
direct hit by a bomb on September 23, 1943, left the SE corner a mound of overgrown ruins.
In 1970-72, when the Soprintendenza completely roofed the main house block, cleared
the bomb rubble, and added low modern walls along the lines of the destroyed rooms
to give tourists some idea of the original plan, I directed 37 soundings below the level of
A.D. 79. Then between 2005 and 2007, as part of the Progetto Regio VI under F. Coarelli and
F. Pesando, M. Stella and I added 17 more soundings, mainly in the area of the peristyle
and on the S side of the house. Our final study of the house provided a detailed analysis of
its original excavation during the Napoleonic Wars, a full description of the extant rooms
and building history, and reports on our excavations.1
As noted by Kockel, the Aschaffenburg model of the house (see Kockel above and pls.
1-3 in color) is a copy of an earlier one, which was based on observations and detailed mea-
surements combined with drawings and colored reproductions of the decorations. Made
on site in Pompeii, these were intended not as theoretically complete restorations, but as a
precise record of the actual state of the house as it existed in 1840. That the Aschaffenburg
model is a relatively accurate reproduction is corroborated by details of the structure and
pavements which were uncovered in our excavations between 2005 and 2007.2 The model
also displays a number of almost complete decorations of walls and floors that are now
washed out or destroyed, in some cases supplying a salutary clarification or corrective to
other early 19th-c. descriptions and copies.
The Aschaffenburg version is a model of the house of the Flavian era as it looked some
32 years after its clearance by the Bourbon and French excavators, but the basic plan and

1 A. Laidlaw and M. S. Stella, The House of Sallust in Pompeii (JRA Suppl. 98, 2014). This was
intended to be the definitive publication, but as soon as he saw the published book V. Kockel
informed me of the remarkably complete model of the house on display in Aschaffenburg. I am
grateful to him and his colleague, W. Helmberger, who generously provided me with copies of
their studies made during the restoration of the model in 1993, and arranged for me to study
it in person. I also wish to thank the Wittelsbacher Ausgleichsfond (WAF), the owner of the
model, for permission to publish here some of the Museum photographs of the Bayerische
Schlösserverwaltung, Munich.
2 E.g., in the photograph of the model seen from the south (color pl. 3), the two perpendicular
divisions in the NW corner by the doorway of Shop 5b were clearly visible on the site before
the rebuilding in 1971. They do not, as Dickmann (cited by Kockel n.25; plan, vol. 2, fig. 5b)
thought, represent a blocked doorway to Triclinium 35, but the remnant of a Republican wall
that was incorporated into the structure of the left (E) wall of the Augustan triclinium when it
was first built (see Laidlaw and Stella 119 with fig. 2.77, and Sounding 39, pp. 209-12 with figs.
5.2 and 5.3a).
© Journal of Roman Archaeology 31 (2018)
268 A. Laidlaw

First Style decoration of the public rooms (atrium, alae, tablinum) date back to its origi-
nal foundation as a single-storeyed noble Samnite domus built c.140 B.C. Some rooms were
modified in the Late Republic, but the major conversion to a commercial property belongs
to the start of the Augustan period, when a new owner added a second storey over the main
block, redesigned the two large oeci next to the tablinum and the shops flanking the fauces,
and built a small apartment with a three-sided peristyle over the foundations of earlier
rooms on the S side of the house.3 Evidence for some of the earlier structures and decora-
tions are still visible in the house today (they were faithfully copied in the model too).

New structural information from the model (fig. 1; color pls. 2 and 3)
The most impor-
tant parts of the model
relate to the SE cor-
ner because that was
destroyed during the
War. The bomb struck
near the NE corner of
the peristyle, prob-
ably right (south) of
the entrance to Kitchen
36, obliterating the
kitchen and Diaeta 33,4
along with all but frag-
ments of a column by
the peristyle’s W walk-
way.5 The destruction
extended up to the
N wall of Andron 20,

Fig. 1. Plan showing area of

bomb damage (A. Laidlaw
and M. Stella).

3 For a detailed analysis of the building phases, see Laidlaw and Stella chapt. 3. The original
foundation date is based on finds of black-gloss pottery from the builders’ trenches around
the atrium (chapt. 4, Sounding 2a, pp. 169-73; workmen’s trenches nos. 5-15, pp. 175-87).
For the date of the Augustan renovation, see Sounding 2a, p. 171 with n.38 (denarius of Marc
Antony, found on the surface of the builder’s level for the thermopolium counter in Shop 3 [this
coin, cat. 68, was identified by the late T. V. Buttrey in his preliminary catalogue at the time of
excavation in 1969, although it is listed as “illegible” 45 years later in the final publication]). For
the Augustan dating of the S. Apartment, see chapt. 5, Sounding 39, p. 211, n.4 (sigillata found
in the foundation for the back portico column re-used as a doorpost for Triclinium 35), and
Sounding 46, p. 217 (laying of the final cocciopesto pavement in the W walkway of the peristyle,
when the level was raised during filling in of earlier foundations). For discussion of Dickmann’s
theories for an earlier dating of the peristyle, see below, p. 270.
4 In 1971, part of the left doorpost of the diaeta, with its wall-painting miraculously intact, was
found blown over, 6 cm below the level of the black-and-white mosaic floor which a restorer
was piecing together using a pre-war photograph (Laidlaw and Stella, color pl. 16 [during
excavation]; PPM IV, 1, figs. 67-68 [after restoration]).
5 The columns, which were composed of triangular segments of brick revetted with plaster, were
blown apart; in the reconstruction of 1971, scattered fragments found in the bomb dump were
used to reconstitute the arrangement, following Mazois’ plan.
 The Aschaffenburg model: addenda to JRA Suppl. 98 269

eliminating the walls of the Augustan re-arrangement of the original right oecus (nos. 20,
28, 28a; color pl. 4) down to the pavements, along with all but the W wall of the right
ala with a stub of its SW corner. Only the socle and part of the right half of the paint-
ing of Actaeon survived on the S perimeter wall of the peristyle garden. The force of the
explosion also uncovered the threshold of the doorway (presently exit 31) in Courtyard 26,
which had been blocked (as is shown in the model in color pl. 3) when the south apartment
had been converted in the 1st c. A.D. to a single unit accessed only from the atrium through
what had been Cubiculum 29/30.
In the hasty repairs made toward the end of the War, the doorways to the andron and
Closet 28a next to it were rebuilt incorrectly: corroborated by early photographs of the
atrium, the model demonstrates that the doorway to the closet, hacked through the origi-
nal First Style decoration of the ala as part of the Augustan renovation, was both broader
and much lower in height (color pls. 5a-b).6 The latrine in the back garden (no. 27a, vis-
ible in color pl. 4, bottom right) was completely covered by the fallen rubble; when the SE
perimeter walls were rebuilt, a new exit (no. 30) to the Vicolo di Modesto was mistakenly
located on top of it, which served as a back door from the end of the War until 1971, when
it was excavated. At that time, the back entrance was restored to its approximate original
location, but the blocked exit (no. 31) from Courtyard 26 was re-opened.
The purpose of the reconstruction of 1970-72 was simply to build a protective roof
over the main house block and to restore and roof the few remaining rooms around the
peristyle. In the process of ensuring that the upper parts of the original walls could sup-
port their modern roof beams, the workmen destroyed all evidence of the second storey
and replaced any ancient walls that their foreman judged to be unstable. Subsequently,
the cocciopesto pavement in the room over the bakery
oven (no. 7) was completely covered with asphalt; and
the beam-holes for the treads of the exterior stair outside
Cubiculum 14a/41 were filled in with cement.7 Perhaps the
most misguided “repair” was the removal from Shop 4 of
the remains of the stone stair excavated in 1970 along its
back (E) wall — the only visible evidence for access to the
second storey in the front of the house.8 Here the model
supplies the missing element for its destination (plan fig.
2; color pl. 6): the bottom sill of a window in the upper
wall by the SW corner of Cubiculum 16, where the stair
turned right to reach the room above, after that part of
the second storey had been cut off from the upper walk-
way by the fall of the W colonnade (presumably after the Fig. 2. Location of later stair in
earthquake of A.D. 62).9 Still unclear (since omitted in the Shop 4 to rooms above Cubicula 15
and 16 (A. Laidlaw).

6 For an early photograph, see PPM IV.1 p. 97, fig. 15, probably taken in the 1870s.
7 See Laidlaw and Stella 87, fig. 2.45.
8 See Laidlaw and Stella Sounding 26a with fig. 4.32, p. 193. This later stair copied the Augustan
one on the opposite side of the atrium (Stairwell 18a), which entered Cubiculum 14a/41 through
a vertical window in the back wall at the NW corner (p. 77 with fig. 2.32; p. 87 with fig 2.45).
9 For evidence of the collapse of the W walkway and the renovation of Diaeta 34, see Laidlaw and
Stella 111-13, with fig. 2.67 and color pls. 5, 18 and 23.
270 A. Laidlaw

model) is the purpose of the small stair shown by Mazois and by most later plans in the
SW corner of the back portico.10
Although the evidence for the fallen W colonnade is clear, the original arrangement of
the columns on that side of the peristyle is not. Mazois’ plan shows only two columns in
front of Triclinium 35, with a third indicated with a circle of dots in between them, which
he implies once balanced the middle of the three columns on the opposite side of the gar-
den. The model also has only two columns flanking the triclinium (color pl. 7a), but, as
J.-A. Dickmann already observed,11 a close examination of the pavement decoration in the
intercolumniations indicates traces of an earlier arrangement. On the N side by the corner
columns there was a mosaic strip of diamonds. On the E and W sides, by both corners, the
pattern changed to 4 rows of equidistant tesserae which alternated with the strips of dia-
monds. On the W side (color pl. 7b), however, the second (S) column base is set directly on
top of the centre of the diamond pattern, and a truncated mosaic strip of equidistant rows
appears to continue underneath the entrance wall of Diaeta 34; in between these mosaics
are two blank squares, which should indicate the placement of an earlier arrangement of
the columns. Dickmann, correctly I think, interpreted this as evidence that the original
peristyle columns continued up to the perimeter wall, and consequently the two diaetae
must be later. He dates the earlier version of this peristyle to the Late Republic.12 Yet the
excavation of this area in 2006 showed that the triclinium and colonnades belong to the
early Augustan period.13 The question of when the diaetae were added remains open, but I
would hypothesize, on the basis of the preserved decorations on the back (S) wall of Diaeta
34, that it was not long after the original construction of the whole apartment complex.14
Other structural details supplied by the model are the arrangement of the steps next to
the basin in the back garden (color pl. 8), and in the back stair inside Kitchen 36 (obliterated
by the bomb) which led to the upper rooms over the cubicula in the atrium (above; bottom
of color pl. 4); the short privacy walls of the latrines in Shops 5a and 6; and the extra win-
dows that connected Room 8 with the back garden, when the bakery was added (below,
color pl. 19).

10 In Sounding 19 we looked for a foundation for these stairs, but the bomb had destroyed all
traces except for the lower parts of the foundations of the main house block in the SW corner
of the portico. See Laidlaw and Stella 189. However, our hypothesis for a mezzanine over room
28 accessed by these stairs (266 with fig. 6.12) is clearly incorrect: the model shows that this
room was decorated with mirror images of the Third Style painting on all four walls, below the
remains of a First Style cornice next to the Augustan doorway on the side by the portico. This
painting was copied at least twice, and is also shown in the Naples model of the 1870s (Laidlaw
and Stella 73-74 with fig. 2.28; color pls. 11-12). A doorsill over room 28a preserved on the top
edge of the front wall of the peristyle in the model proves that the upper room was actually
accessed from the N walkway.
11 Dickmann (supra n.2) vol. 1, 359-60 with fig. 104 and hypothetical plan vol. 2, fig. 5d.
12 Ibid., vol. 1, 143, n.106.
13 See above nn. 2-3.
14 The paintings on this wall are a mélange belonging to at least three decorations. At the upper
left two separate layers of painting provide evidence for an earlier upper room before the upper
colonnade on that side was destroyed. The small subject painting of Paris and Helen in the
main zone appears to belong to the Third Style (which fits with the early Augustan decoration
of all the smaller rooms around the atrium), although the larger painting of Venus and Mars,
awkwardly attached to the raw plaster of the upper zone, is clearly Fourth Style. For a detailed
description, see Laidlaw and Stella 111-13 with fig. 2.67, p. 152, and color pls. 18 and 22-23.
Col. pl. 1a. The Aschaffenburg model of the Casa di
Sallustio, view from the southwest (Wittelsbacher Aus-
gleichsfonds, Inv. No. Asch. 0059 [WAF], © Bayerische
Schlösserverwaltung, Foto Maria Custodis).
The Aschaffenburg model: addenda to JRA Suppl. 98
271
 272
A. Laidlaw

Col. pl. 1b. The Aschaffenburg model of the Casa di Sal-

lustio, view from the north (Wittelsbacher Ausgleichsfonds,
Inv. No. Asch. 0059 [WAF], © Bayerische Schlösserverwal-
tung, Foto Maria Custodis).
 The Aschaffenburg model: addenda to JRA Suppl. 98
273

Col. pl. 2. House of Sallust, model seen from the Via Consolare (A. Laidlaw).
274 A. Laidlaw

Col. pl. 3. House of Sallust, model seen from the right side facing the South Apartment (A. Laidlaw).

Col. pl. 4. Augustan remodeling of the right (S) oecus (Nos. 20, 28, 28a); stairs in Kitchen 36 (foreground) and
Latrine 27a in back garden (far right), as of 1840 (A. Laidlaw).
 The Aschaffenburg model: addenda to JRA Suppl. 98 275

Col. pls. 5a-5b. Doorways to Andron 20 and Closet

28a seen from the atrium, showing the Augustan
arrangement in the model in 1840 and the incorrect
reconstruction after World War II (a. Inv. No. Asch.
V0059 [WAF], © Bayerische Schlösserverwaltung;
b. J. Felbermeyer 1969).

Col. pl. 6. Shop 4, window (white arrow) for later stair in Shop 4 to access the apartment above
Cubicula 15 and 16 after the earthquake of A.D. 62 (A. Laidlaw).
276 A. Laidlaw

Col. pls. 7a-b. (a) Peristyle 31 from the northeast, showing the arrangement of columns and location of the
mosaics between the intercolumniations; (b) detail of west colonnade with mosaic strip by Diaeta 34 (Inv. No.
Asch. V0059 [WAF], © Bayerische Schlösserverwaltung).

Col. pl. 8. NE corner of the house, showing the Augustan re-

arrangement of Oecus 22 and the stair from the portico to the back
garden, as of 1840 (A. Laidlaw).

Col. pl. 9. Pattern of the pavement in the fauces, as of 1840

(A. Laidlaw).
 277

Col. pl. 10 (left). NE corner of the atrium and left

ala, showing the patterns of the atrium and ala pave-
ments, as of 1840 (A. Laidlaw).
Col. pl. 11. Detail of the pattern of the mosaic in the
NE corner of Triclinium 35, as of 1840 (A. Laidlaw).

Col. pls. 12a-b. First

Style paintings in the
atrium, left ala, and
tablinum seen from the
southwest.
(a) Actual state, 1969
(J. Felbermeyer);
(b) in model, as of 1840
(Inv. No. Asch. V0059
[WAF], © Bayerische
Schlösserverwaltung).
Col. pl. 13 (top). Oecus 22, A. Laidlaw
showing the Augustan shift
of focus from the atrium and
tablinum to the back garden,
seen from the northeast,
1840 (Inv. No. Asch. V0059
[WAF], © Bayerische
Schlösserverwaltung).
Col. pls. 14a-b (middle).
Oecus 22, back (W) wall.
Second Style painted door
on the blocked doorway
from the atrium, (a) as of
1840 (Inv. No. Asch. 0059
[WAF], © Bayerische
Schlösserverwaltung); (b)
actual state as of 1969
(J. Felbermeyer).
Col. pls. 15a-b (bottom).
Lararium on the blocked
doorway between the atrium
and Oecus 22, (a) as of 1840
(Inv. No. Asch. 0059 [WAF],
© Bayerische Schlösserver-
waltung); (b) actual state as
of 1969 (J. Felbermeyer).
 The Aschaffenburg model: addenda to JRA Suppl. 98 279

New evidence for pavements from the model (see fig. 3)

In the house of the Samnite period, the original pavements were made of red or black
cement (cocciopesto or lavapesta), decorated with standard patterns in white tesserae. In the
fauces the design consisted of a rectangular block of diamonds, separated from the inner door-
way to the atrium by a single row of maeanders set in red cocciopesto (color pl. 9) (only traces
of the diamonds were found by the front entrance in 1971).15 This theme was repeated in the
alae, but here a stripe of 6 rows of diamonds bordered the atrium side; a rectangular block
of interlinked maean-
ders extended from this
stripe to the center of the
room, surrounded by
regular rows of white
tesserae which filled in
the sides and back (color
pl. 10). In the two oeci
flanking the tablinum,
these rows were contin-
ued at the same level;16
slightly larger tesse-
rae in the same regular
pattern decorated the
atrium, here set into
black lavapesta.17
When Triclinium 13
was added in the Late
Republican period, the
theme of diamonds
and maeanders was
repeated, this time,
according to the model,
Fig. 3. Plan showing the location of decorated pavements, now destroyed in lavapesta (fig. 4).18
(A. Laidlaw). Here the narrow room

15 See Laidlaw and Stella 56-57 with fig. 2.5; PPM IV.1, 92, fig. 5.
16 When the level of the left oecus (no. 22) was raised and the room re-aligned with a wide doorway
facing the portico and garden, the Samnite pavement was replaced by a white lithostroton
pavement, similar to the one in the tablinum. This was done in two stages: beginning in the Late
Republic, the atrium door remained in use, with a small section of the lower floor left open by it,
presumably for a small wooden stair; then in the Augustan remodeling this door was blocked,
the stairwell filled in, and the new portico doorway cut through what had been the First Style
decoration of the E wall of the oecus. See Laidlaw and Stella 142 and 148-49; Sounding 1, 167-69
with figs. 4.2-5. The model shows the retouching of the First Style paintings with some Second
Style additions (color pls. 4, 8, and 12b). It is possible that the original pavements also had a
centerpiece of maeanders, a variant of the patterns in the alae, but the excavations in 1969-70
only uncovered the rows of tesserae at the back and front of Oecus 22.
17 Only the bedding is preserved now, but the model shows the complete pattern.
18 In 2006, next to the S wall we found a small fragment of the setting-holes for the tesserae forming
the diamond pattern of the stripe that divided the room in half (Laidlaw and Stella 75 with
n.77). In the model, the tesserae are shown as white set into black cement, which presumably is
correct (in 2006 they were covered with modern lime).
280 A. Laidlaw

was divided in half by a

stripe of 5 rows of dia-
monds inside a box framed
by a single line of tesserae,
set off from the front of the
room by a row of slightly
larger white ones. The
placement of three couches
was indicated by a central
rectangle of interlinked
maeanders alternating
with squares (three on
the short sides, four on
the long). This echoed the
Samnite pattern in the alae,
except that the two designs
were not directly joined
together.
In the South Apart-
ment, Triclinium 35 had a
white mosaic pavement,
outlined by a narrow black
rectangle set back slightly
from the walls of the room.
In 1970, only a small sec-
tion of the white tesserae
was preserved in the SW
Fig. 4. Triclinium 13, mosaic pattern showing the location of the couches, corner, but the model
1840 (Inv. Asch. 0059 [WAF], © Bayerische Schlösserverwaltung).
(color pl. 11) shows two
separate patterns, one across the threshold between the red columns, and one along the
edge of the left alcove. Except for one end, the threshold pattern is almost completely lost,
but it apparently consisted of a row of large black circles, each encasing a cross.19 The
alcove had a simple cable, outlined in black. Both mosaics were also set off with a narrow
black outline. Although parts of the overall design are missing, there is no trace of the inset
square pattern shown in Mazois’ plan. As noted above, the intercolumniations between the
columns in the peristyle were decorated with strips of diamonds alternating with simple
equidistant rows of tesserae. Except for the small swatch of white tesserae in the SW corner
of the triclinium, the model provides our only evidence for these mosaics.

New evidence for wall-paintings from the model

All four styles of Pompeian painting were preserved in the house at the time of the
eruption. The First Style paneling had been maintained (even occasionally repaired) in the

19 See Laidlaw and Stella 114, n.188, noting a watercolor etching by Giacinto Gigante that shows
the pattern of black circles in a frame continuing along the left side of the room. The design used
in the model as a stripe across the threshold is presumably more accurate.
 The Aschaffenburg model: addenda to JRA Suppl. 98 281

main public areas of what had been the entire single-storeyed central block in its original
Samnite form (color pls. 12 a-b). However, tastes changed in the Late Republic, and the
focus of the living spaces began to shift from the huge central atrium to the garden at the
back, beginning with Oecus 22 to the north of the tablinum (color pl. 13). Here, although
the First Style patterns were left on all three walls, when the room was re-oriented to face
the garden the blocked doorway to the atrium was painted in the Second Style (color pls.
14a-b), as was the upper zone of the N wall (visible at the bottom center of color pl. 8, top
of pl. 12b) above a poor copy of the earlier First Style in the rest of the room. On the atrium
side, the door was re-purposed as a niche and painted with a lararium (color pls. 15a-b).
At the beginning of the Augustan period, the entire house was rebuilt with a second
storey over the main block, a cookshop opening onto the front of the atrium, and a small
apartment with a three-sided peristyle added on the right side. Although the noble Sam-
nite house was now converted into a business property, the proprietor took great pains to
preserve the visible First Style paintings, but the NE garden at the back and most of the
cubicula, including no. 28 in the subdivided oecus to the right of the tablinum (see color pl.
4), were painted in the Third Style (e.g., Cubiculum 14a/41,20 E wall, color pl. 16; Cubicu-
lum 29/30, W wall and E upper zone, later made into a mezzanine, color pls. 17a-b). Less
important rooms, such as those in Shops 4 and 5, if they were decorated at all, had simple
outlined rectangles in the main zone over a red or black socle.
Presumably after the earthquake of A.D. 62, only the separate apartment was completely
repainted in the new Fourth Style, major redecorating necessitated by the collapse of the
upper walkway on the W side of its little peristyle (color pls. 7a and 18).
The proprietor, assumed to be an inn-keeper, must have been thriving during the 1st c.
A.D. since extra rooms were inserted into the side portico at the back, and two semi-
independent businesses were added — a bakery on the left and an independent cookshop
with a private exit to the Vicolo di Mercurio on the right, presumably run by freedmen
connected with the household. The bakery had a room directly over the oven (color pl. 19)
which must have originally been part of an earlier arrangement of that space, since the tri-
partite wall-paintings on the model extend onto the E wall of Room 8 over partly-blocked
rectangular windows which open onto the rear side garden. This room, which must post-
date the building of the oven, has been explained as a room used either for drying wood or
possibly for maturing wine, using the smoke from the furnace,21 but the decoration shows
that at some point it had been used as living space.

20 In Gell’s line-drawing of this wall, made in 1817 (Laidlaw and Stella 77, fig. 2.33), he showed
the two panel frames in the main zone in more detail (traces of the garlands they encased are
still extant today), but his version of the central motif between these panels and the entire
upper zone appear to have been embellished, causing modern scholars usually to assign the
decoration of this room to the Fourth Style. Abbate’s version, which is much simpler, fits into
the Augustan Third Style program for all the other cubicula, as well as being clearly designed to
frame the window for the stair to the upper storey by the NW corner, which must be part of the
Augustan renovation. Since Gell drew numerous details of this house which he then combined
into a variety of pastiches from different rooms (e.g., Laidlaw and Stella 76, fig. 2.31, and 106,
fig. 2.61, which are parts of the same drawing), it seems reasonable to accept Abbate’s Third
Style pattern as the correct one.
21 See Laidlaw and Stella 123, n.210.
 282
A. Laidlaw

Col. pl. 16. Augustan Third Style pattern of the painting Col. pls. 17a-b. Augustan Third Style painting of Cubiculum 29/30, right side of atrium, (a) as of 1840.
on the right (E) wall of Cubiculum 14a/41, left side of the (b) Right (W) wall and left (E) wall, upper zone, retained in the later mezzanine (Inv. No. Asch. 0059 [WAF],
atrium, 1840 (Inv. No. Asch. 0059 [WAF], © Bayerische © Bayerische Schlösserverwaltung).
Schlösserverwaltung).
 The Aschaffenburg model: addenda to JRA Suppl. 98 283

Col. pl. 18. Flavian Fourth Style red and black ground redecoration of Peristyle 31 seen from the northwest, as
of 1840 (Inv. No. Asch. 0059 [WAF], © Bayerische Schlösserverwaltung).

Col. pl. 19. Room over the oven in the bakery, showing the earlier Third Style painting on the right (E) wall,
as of 1840 (A. Laidlaw).
284 A. Laidlaw

Conclusions
The Aschaffenburg model shows, in meticulous detail, the basic structure of the Augus-
tan house, with some modifications made in the 1st c. A.D. Abbate’s copies of the First
and Third Style paintings, and the patterns of the pavements, are a remarkably accurate
record of what was still in a surprisingly good state of preservation in the 1840s — wall-
paintings which are now faded or fallen and had been recorded only occasionally in copies
of greater or lesser accuracy, and patterns of floors which are now entirely lost. The model
provides us with a detailed three-dimensional view of the SE corner of the house, the peri-
style and kitchen and the back rooms that were completely destroyed by the bomb. The
careful reproduction of details such as the bottom parts of windows and doorsills related
to the upper storey confirm several hypotheses that we have proposed about these lost
structures and about how they functioned in A.D. 79. The detailed copies of both pave-
ments and paintings also fill in the gaps in patterns and color schemes for our descriptions.
lablaidlaw@gmail.com Johns Hopkins University, Baltimore, MD

Acknowledgements
I am grateful to the Wittelsbacher Ausgleichsfond, the owner of the model, for permission to publish
some of the Museum photographs of the Bayerische Schlösserverwaltung, Munich. I also thank
J. Stephens for her skilled editing of some of my color photographs.
This publication, most specifically the costs involving in obtaining and printing the images of
the Aschaffenburg model, has been supported by a grant from the Samuel H. Kress Foundation,
administered by the Archaeological Institute of America (see p. 2 above). I am especially grateful for
their generosity which enabled us to print the illustrations of the model in full color.