3

Why were peristyles built in Pompeian houses?

3.1 The courtyard architecture of a Mediterranean house

Prior research has clearly indicated that peristyles played an important role in the social life of the house, and that they can be listed among the most important spaces for socioeconomic display.¹ However, the entire premise needs to be re-examined, in particular from the point of view of more extensive assemblages of archaeological evidence, not just on the basis of a few selected examples as has been done previously. The peristyle cannot be an important display space if its functions – whether relating to the architecture of the house or to the human life in the space – do not provide good prospects for display. Consequently, the functions of the Pompeian peristyles will now be examined, starting with the architectural functions, which we will see actually reinforce the peristyle’s potential to function as a display space.

In any geographical region with a similar climate and geographical resources throughout, the architecture will tend to have similar features. In the Mediterranean, houses are often organized around a central courtyard.² This courtyard serves several important architectural functions that are vital for the daily life of the house. They bring air and light into the house, and the courtyard controls movement inside the house. In the Roman house there were several types of courtyards, but the most commonly known are the atrium and the peristyle. Consequently, Pompeian peristyles served very important architectural purposes. Without these courtyards there would be minimal air or light inside the houses, and the movement inside the house would be elementally different. Although these aspects are vital to the function of a house, not every Pompeian house had a peristyle. There were other ways of arranging these functions, and this chapter ends with a discussion of what the choice of building a peristyle in a Pompeian house tells us about the owner’s socioeconomic ranking in Pompeii, and also in the Roman world in general.

Compared with the majority of the rooms, the peristyle garden somewhat breaks with the basic function of a building: it does not provide shelter.³ The ceiling is open, making the peristyle a liminal space between the inside and outside, but its liminality is very controlled. The colonnades, which represent built interior space and the human world, embrace the open-air space, which epitomizes the outside and nature. In the end, the entire peristyle is situated safely inside the architectural structure and is within human control.

The exterior walls of Roman houses did not have many windows,⁴ highlighting the importance of the courtyards. In the model of the traditional Roman atrium house, the atrium has been seen as the most vital space for bringing air and light into the house, but when the house was enlarged, other spaces were built to serve the same functions, the peristyle being one of these.⁵ Besides the atrium and the peristyle, light wells and gardens could also be utilized for the same purpose.⁶

The importance of air conditioning is demonstrated in the Casa di Polibio. Its peristyle garden was the space where most of the interaction between outside air and the microclimate of the house occurred. The trees and the gutter functioned as a sort of air conditioner, and during winters the peristyle garden admitted most of the sunshine into the house. Conversely, the luminosity inside the house was generally poor.⁷ Consequently, this peristyle – like any other peristyle – must have been the focal point of daily life because of its illumination.⁸

Donatella Mazzoleni notes that the north-south orientation of the peristyle of the Casa dei Vettii is optimal for capturing the midday sun and sunset.⁹ This could mean that the peristyle was planned particularly for daylight or early evening activities. However, the arrangement where the rooms are at the north and east sides of the peristyle is not a very common composition in Pompeii. There are only 21 similar peristyles.¹⁰ Altogether, about 37 percent (94) of the peristyles featured rooms only on the north or east sides, or both. Rooms on the opposite sides – south, west, or both – can be found in 25 percent (63) of the peristyles. These peristyles indicate that catching the day and evening sun was not the most important aspect when planning a peristyle, but it was probably taken into consideration, as most peristyles – about 75 percent – had rooms at least on the north or east side, if not on both sides.

A courtyard dominates the movement of the house because the rooms are organized around it, and oftentimes even the rooms that do not immediately open onto the courtyard are inaccessible without first passing through the courtyard. The focus of the design on movement is particularly underlined by the architectural features of the space, namely the division between the portico and the garden. The colonnades are built precisely for movement, but the role of a garden is more nuanced. It was perhaps made for visual display, or may have functioned as a kitchen garden. The garden can be a place of movement, but perhaps for a different type of movement – the pleasure stroll.¹¹ Nevertheless, some peristyles entirely isolate the portico from the garden with a pluteus without openings, such as the small peristyle of the Casa di D. Octavius Quartio.

There are 55 peristyles where rooms can be found on all four sides, which indicates the importance of the role of the peristyle as a space for movement. The high average (8) and median (7) numbers of spaces opening onto a peristyle also highlight this role (Figs. 2.5 and 2.6). Although the peristyle area was adjacent to a significant number of the rooms in a Pompeian house, it also had a vital role in controlling movement throughout the entire house. The majority – 70 percent (176) – of peristyles were connected to a corridor, fauces, and 73 peristyles had at least two fauces leading into the peristyle, indicating that the peristyle was linked to the other parts of the house, not just the rooms around it.

Movement was not only horizontal – 59 peristyles had stairs opening onto the space, and in 24 peristyles the stairs were in the peristyle itself. Therefore, the peristyle was often the space where the upper or lower floors were accessed. These features – such as corridors and stairs – were designed for movement, and they underline the importance of movement through the peristyle.

The role of the peristyle in facilitating movement in Pompeian houses is demonstrated by studies that utilize Space Syntax analysis.¹² The peristyle as a space, however, is problematic for that type of analysis. It presents difficulties when choosing whether the entire peristyle garden should be counted as one space, or whether the colonnade(s) should be counted separately from the garden area. Also, it is problematic whether every colonnade should be treated as an independent unit, or whether all of the porticoes should be thought of as one whole unit.¹³ The methods that enable a detailed computerized use of Space Syntax analysis allow us to bypass the artificial separation of the rooms and spaces, and Michael Anderson has carried out such an analysis on five houses with a peristyle. In all of them, the peristyle – in addition to the atrium – seems to be the most significant space for movement.¹⁴ Although the method still has problems concerning, for example, the missing information on the upper floors,¹⁵ it indicates that the peristyle – at least on the ground floor – was one of the most important spaces for movement in the house.

Even though the peristyle was vital for the architecture of the house, it was not a mandatory feature of the Pompeian house. Most Pompeian houses did not have a peristyle.¹⁶ There were plenty of other architectural options to provide air and light inside the house and control its movement, and if the house size was relatively small it may not have needed a courtyard. Consequently owning a peristyle already placed a person among the wealthier Pompeians. It indicated at least middle class status in the city, likely even upper middle class.

3.2 The multifunctionality of the Roman house and peristyle

Emphasizing the multifunctional nature of built space in the Roman world has recently become very popular.¹⁷ The widespread application of this interpretation has, however, caused its meaning to become almost insignificant. Hypothetically, every space is multifunctional, and therefore this type of broad interpretation does not reveal much about Pompeian or Roman life without a definition of what multifunctionality means in each particular instance. This chapter critically re-evaluates the concept of multifunctionality, and then discusses whether Pompeian peristyles can be seen as multifunctional spaces.

Studies of archaeological finds in Pompeian houses have shaped our opinion of house functions during the last two decades. The schematic view where one room had one function has been displaced, and now the house represents itself as a much more flexible unit of daily life. In particular, Penelope Allison, Lisa Nevett, and Joanne Berry have advocated this type of flexibility and multifunctionality of the house and its rooms.¹⁸ When considering the realities of lived life, this is likely a correct interpretation. However, multifunctionality has become such a popular interpretation that the word itself has possibly lost its meaning, and offers very little new information to us; Laura Nissin, for instance, has questioned some interpretations of multifunctionality in her work on sleeping spaces.¹⁹

I now aim to define what multifunctionality means, and when it can be a practical tool for interpreting the Pompeian house, in particular its peristyle. In contrast to multifunctionality, the concept of a single purpose space is also raised in this chapter. It is merely a hypothetical construction, as no space involving human activity can ever be only for a single purpose, but it can be designed for a single main purpose, which in this case is the opposite of multifunctionality.

After a theoretical discussion, the chapter moves on to investigate what type of activities can be located in the Pompeian peristyles, excluding the architectural functions already discussed, and some activities that can be linked to the social activity and visitors in the peristyle, which will instead be discussed in the next chapter. Here, the focus is mainly on the water supply, on cult activity, and on business activity, as these seem to be the largest groupings of the several activities that can be located in these peristyles. In the end, I return to the topic of multifunctionality by investigating what types of activities occurred simultaneously in the peristyles, and whether the peristyle can therefore be actually defined as a multifunctional space.

Without a more precise definition, multifunctionality becomes a concept that can cover every single space. This is the case with Nevett’s analysis. She has looked for spaces where two or more activities take place, or instances of the same activity occurring in several spaces.²⁰ This barely leaves any space without multifunctionality, which also can be demonstrated with the modern western house, where there is hardly any space that cannot be defined as multifunctional. When considering an even broader span of time, multifunctionality is even more dominant. Pompeii itself is an example of this: its rooms had their domestic or public function in antiquity, but now their main purpose is as a tourist or scholarly attraction. The timeframe is therefore another problem: taking a long enough time span makes all spaces multifunctional, as the functions in a space will inevitably change at some point in history. Therefore, what is a reasonable time period when examining Pompeian house functions? Is it possible to define a space as multifunctional only if several activities take place there simultaneously, or perhaps during the same day, or over the course of a week, or a year, or a generation?

Contrary to multifunctionality, a room could be defined as a single purpose space, but this is only a hypothetical possibility. The storage room perhaps comes closest to this type of space. A stereotypical storage space is very badly suited to other domestic functions. The same scholars who have interpreted Pompeian rooms as multifunctional also often see the spaces in Pompeian houses as storage spaces.²¹ Yet, the same problem emerges again; it is not very clearly defined what a storage space is. If we define it very loosely, a room becomes a storage space as soon as the first item is placed in it. Using this type of loose definition, a space could also be defined as a sort of a living room when a person spends some time in it, and according to these loose definitions every room is multifunctional, which underlines the need for a better definition of multifunctionality.

Given the nature of the archaeological evidence, it might be difficult to determine the exact room functions of the Pompeian house, but occasionally we might be able to figure out what the planned function of the space was. Examining the built structures – and possibly the decoration and small finds if available – can shed light on what functions were planned to take place in the different parts of the house. This helps to form a more rigid definition of multifunctionality: if several planned activities are found in the same space, it is possible to define the space as multifunctional.

Despite the possible anachronism of interpreting Roman house functions emphasizing a single or primary function, it is a possibility that still needs to be considered, even if the current trend in scholarship does not support this model. The rooms might be multifunctional in practice, but this does not mean that they were automatically thought of as such. The modern kitchen or bedroom are examples of this. They often involve plenty of other activities, and in the same way sleeping or cooking can also take place in other rooms. Nonetheless, the dominant function of these spaces transmits an immense amount of information about how we shape our daily lives. In order to define a space as multifunctional, it is therefore not only important to study what occurred in the space but also what was thought should happen in the space: was there a single dominant purpose for the room, or was it planned as a multifunctional space?

Several activities have been suggested to have occurred in peristyles: cooking, dining, loitering or spending (leisure) time, fulling, spinning, tanning, washing, drying, strolling, baking, education, readings, juridical activity, theater plays, etc. However, these are often speculative conclusions, as will be demonstrated in this and the following chapters.

There is no general rule that could be applied to the archaeological evidence to determine the function of a room, but rather every case must be considered individually, in this case one peristyle at a time. The most obvious group of sources to suggest a function are the different types of structures found in the peristyles. On the other hand, occasionally graffiti or small find assemblages can reveal what was going on in the space. However, the documentation of this type of data was not carried out with equal care for every peristyle, and evidence for some activities may well have escaped the modern investigator.

The following detailed discussion focuses on three different areas. First is the role of the peristyle as the water reservoir for a house, concentrating principally on human activity relating to the water supply – the everyday routine of collecting water from the water reservoir. Second is cult activity, again focusing on humans performing these activities. However, the general aspect of the peristyle as a type of sacred space is additionally discussed, as this characteristic has often been associated with peristyles and gardens. Third are commercial activities, covering the peristyle as a space relating to retail and/or small-scale industry. This last area includes several (sub)activities, but can be related to business activities in general.

3.2.1 Water supply as the function of the peristyle

A water supply is and was crucial for the daily life of any house. The open roof of the peristyle permitted the collection of rainwater, a function that is also related to the atrium with its compluvium and impluvium. In a Pompeian house, rainwater was often guided from the roof to a gutter, which then gathered it into a cistern.²² Hypothetically, counting all the cisterns in all of the Pompeian peristyles could reveal if the peristyle in general was primarily planned as a water reservoir for the house. This exercise results in 88 peristyles with a cistern (Table 3.1), but the number is problematic, as the cistern is an underground structure, and the existence or functionality of cisterns can be often verified only if the excavation was continued under the floor level, which is quite rarely the case.

Table 3.1 Pompeian peristyles that can be connected to water supply, cult activity, or commercial/industrial activity.
N.	Name	Address	Cistern	Cistern head	Lararium painting	Lararium statue	Altar	Commercial	Industrial	Specific	Multiactivity
1		I,2,6		1
5	Casa della Grata metallica	I,2,28	1	2
6		I,3,3	1	1
7		I,3,8	1
8		I,3,20						1
9	Casa della Rissa nell’Anfiteatro	I,3,23	1	1	1	1					1
10		I,3,24			1
11		I,3,25		1
12		I,3,30			1
13		I,4,2		1
14	Casa del Citarista	I,4,5/25		1
17	Conceria	I,5,2							1	Tannery
18	Casa del Criptoportico	I,6,2			1
19	Fullonica di Stepahus	I,6,7							1	Fullonica
20		I,6,9						1
21	Casa dei Quadretti teatrali	I,6,11	1
22	Casa di Paquius Proculus	I,7,1	1	2
23	Casa del Sacerdos Amandus	I,7,7	1	2
24	Casa dell’Efebo	I,7,11/19			1
25	Casa dell’Efebo	I,7,11/19	1	1		1					1
26	Casa del Pomarius Felix	I,8,2	1	1
27	Casa della statuetta indiana	I,8,5		2	1						1
29	Taberna vasaria	I,8,10			1	1
30		I,8,13		1
33	Casa di Successus	I,9,3	1	2
34	Casa dei Cubicoli floreali	I,9,5	1	1
37	Casa di Cerere	I,9,13–14	1	1
38	Casa del Menandro	I,10,4/14–17	1	2
39	Casa del fabbro	I,10,7		1				1			1
42		I,11,14	1
43	Casa del Primo Piano	I,11,15/9	1		1						1
44	Panifico di Sotericus	I,12,1/2		1
46		I,12,8						1		Garum shop
49	Casa di Suotoria Primigenia	I,13,2			1
50		I,14,11/15		1
52	Casa delle colonne cilindriche	I,16,2-a		1
55	Casa degli archi	I,17,4	1	1	1						1
56	Complesso dei Riti magici	II,1,12					1
59	Casa della Venere in conchiglia	II,3,3		1	1						1
60	Caupona con abitazione	II,8,2–3					1
62	Casa detta di Trebius Valens	III,2,1		1
63	Casa di Pinarius Cerialis	III,4,4	1
64	Casa del Toro	V,1,7		1
65		V,1,15	1	1					1	Bakery
68	Casa del Triclinio	V,2,4			1
70		V,2,15	1		1						1
71		V,2,15	1		1						1
73	Casa delle nozze d’argento	V,2,i	1
78	Casa di M. Lucretius Fronto	V,4,a	1
79		V,4,b	1		1						1
80	Caserma dei Gladiatori	V,5,3	1
81		VI,1,1	1	1							1
83	Casa del Chirurgo	VI,1,10		1
84	Casa di Sallustio	VI,2,4	2	2			1				1
85	Casa di Sallustio	VI,2,4	1	1
86		VI,2,16		1
87	Casa delle Danzatrici	VI,2,22	1	3
88		VI,2,25		1
89	Accademia di Musica	VI,3,7			1
94	Casa del Granduca Michele	VI,5,5		1
95		VI,5,10	1	2
96		VI,5,14		1
97	Casa di Pansa	VI,6,1		2
99	Casa d’Ercole	VI,7,6			1
100		VI,7,7		1
101	Casa di Adone ferito	VI,7,18		2
102	Casa di Inaco e Io	VI,7,19		1
104	Casa del Poeta tragico	VI,8,3/5	1	2
105	Fullonica	VI,8,20		1	1				1	Fullonica	1
106	Casa della Fontana grande	VI,8,22		2
107	Casa della Fontana piccola	VI,8,23/24	1
108	Casa di Meleagro	VI,9,2/13		2
109	Casa del Centauro	VI,9,3/5		1
112	Casa del Centauro	VI,9,3/5	1
113	Casa dei Dioscuri	VI,9,6/7	2	1	1						1
115	Casa dei Dioscuri	VI,9,6/7	1
117		VI,10,6		1
118	Casa del Naviglio	VI,10,11		1
120	Casa del Labirinto	VI,11,8–10		4
121	Casa del Fauno	VI,12,2	1	2
122	Casa del Fauno	VI,12,2	1	4		1					1
123	Casa del Gruppo dei vasi di vetro	VI,13,2	1		1
124	Casa del Forno di ferro	VI,13,6							1
125		VI,13,13	1	2
126	Casa di Sextus Pompeius Axiochus	VI,13,19	1
127		VI,14,12		1
130		VI,14,38	1	2
133	Casa degli scienziati or Gran Lupanare	VI,14,43	1	1	1						1
134	Casa dei Vettii	VI,15,1	1	1
135	Casa dei Vettii	VI,15,1	1	1
137	Casa del Principe di Napoli	VI,15,7/8	1
138		VI,15,23
139	Casa degli Amorini dorati	VI,16,7	1	3		1					1
141		VI,16,26		1					1		1
142		VI,16,36/37	1		1						1
147	Casa di Sirico	VII,1,25/47		1	1						1
148	Casa di Sirico	VII,1,25/47		1
149	Casa di M. Caesius Blandus	VII,1,40	1	2
150	Panificio di Terentius Neo	VII,2,3							1	Bakery
151	Tintoria	VII,2,11–12	1	1					1	Dye shop	1
152	Casa di M. Gavius Rufus	VII,2,16–17	1	3
153	Casa di C. Vibius Italus	VII,2,18		1
156	Casa delle Quadrighe	VII,2,25				1
157		VII,2,51		1
158	Casa del Larario doppio	VII,3,13		1	1						1
159	Casa di M. Spurius Mesor	VII,3,29			1
161	Casa dei Capitelli colorati	VII,4,31/51	1	2
162	Casa dei Capitelli colorati	VII,4,31/51	1	1
163	Casa della Caccia antica	VII,4,48	1	1
164	Casa del Granduca	VII,4,56	1	2
165	Casa dei Capitelli figurati	VII,4,57	1	2					1		1
166	Casa della Parete nera or Casa dei Bronzi	VII,5,59		1
168	Casa di M. Spurius Saturninus e di D. Volcius Modestus	VII,6,3					1
169		VII,6,7		1
170		VII,6,28	1				1				1
171		VII,6,30	1	1
173		VII,7,2					1
174	Casa di Trittolemo	VII,7,5		1
175	Casa di Romolo e Remo	VII,7,10	1
177		VII,7,23	1		1						1
178	Casa delle Nozze di Ercole	VII,9,47		2
179	Casa della Pescatrice	VII,9,63		1
181	Casa e lavanderia	VII,10,5	1	1					1	Washing?	1
182	Albergo e Caupona	VII,11,6–8		1
183		VII,11,9–10		1
188	Casa di Ganimede	VII,13,4/17–18	1	1
189	Casa del Banchiere or Casa della Regina d’Inghilterra	VII,14,5							1	Dye shop?
190		VII,14,9	1	1
191	Casa di A. Octavius Primus	VII,15,12–13	1	1
194	Casa di Ma. Castricius	VII,16,17	1
195	Casa di Championnet II	VIII,2,3–5	1	1
196		VIII,2,13	1	2
197		VIII,2,14-16	1
201	Casa del Cinghiale I	VIII,3,8–9	1	3
205	Casa di Apollo e Coronide	VIII,3,24	1	1
206		VIII,3,27	1	1
207	Casa di Pane	VIII,3,28/31		3
208	Casa dei Postumii	VIII,4,4/49		1
209		VIII,4,12-13						1
210	Casa di Cornelius Rufus	VIII,4,15/30	1	2
213	Casa del Gallo	VIII,5,2/5		1
214		VIII,5,9		1
216	Casa del Medico	VIII,5,24	1
217	Casa della Calce	VIII,5,28	1
219	Casa di Acceptus e Euhodia	VIII,5,39	1
220		VIII,7,6	1	1	1						1
221	Casa dello scultore	VIII,7,24/22	1	2
222		VIII,7,26-27		1
223	Abitazione e botteghe 10, 11 e 13	IX,1,12	1
224	Casa di M. Epidius Rufus	IX,1,20	1	3
225	Casa di M. Epidius Sabinus	IX,1,22/29		1
226	Casa di M. Epidius Sabinus	IX,1,22/29		2
227		IX,2,10		1
228	Casa di T. Dentatius Panthera	IX,2,16			1
229		IX,2,17		1
230		IX,2,18			1
231		IX,2,19–21			1
232		IX,2,26	1	1	1						1
233	Casa del Granduca di Toscana	IX,2,27	1				1				1
234	Fabbrica di prodotti chimici	IX,3,2	1	1	1						1
235	Casa di Marcus Lucretius	IX,3,5/24	1	1
237	Casa di Achille	IX,5,1–3		1
238	Casa dei Pigmei	IX,5,9	1	1
240		IX,5,14–16		1
241	Casa di Giasone	IX,5,18	1	1
242		IX,6,4–7	1	1
243		IX,6,f–g		1
244	Casa della Fortuna	IX,7,20	1
245	Casa del Centenario	IX,8,3/7		3
247		IX,9,1	1		1						1
248	Casa di vinaio	IX,9,6		1		1			1		1
249	Casa dei pittori al lavoro	IX,12,9		1
250	Casa di Polibio	IX,13,1–3	1	1
251	Casa di Obellius Firmus	IX,14,4		2

Wilhelmina Jashemski reports that in the Casa del Gruppo dei vasi di vetro (VI,13,2) the bombings during the Second World War exposed a puteal and made it possible to measure the cistern of the peristyle. Her text does not reveal whether the cistern was visible before the bombings; however, if the puteal was not visible, it would seem safe to assume that the cistern was not visible either.²³ This event suggests that there are unidentified cisterns under other peristyles.

As the case of the Casa del Gruppo dei vasi di vetro reveals, the number of reported cisterns is not reliable enough to estimate the role of the peristyle as a water supply. Additionally, thinking of household members using a peristyle for activities relating to the water supply, the most central feature is the cistern head, as it identifies the spot where the cistern can be accessed.²⁴ A cistern head is often the only visible remains of a cistern, due to the previously mentioned fact that the excavation rarely extends under the floor level. A puteal could also mark the presence of a cistern,²⁵ but it is movable, and the connection between the puteals, cisterns, and peristyle can be ambiguous. There are 116 peristyles with a reported cistern head in Pompeii (Table 3.1). All of the peristyles with a cistern head most likely had a cistern, at least at some point in their history, but it is uncertain whether the cistern was functioning during the last phase.

There are several cases indicating that some peristyles might have lost their function as a water supply over time. The availability of aqueduct water reduced the need to collect rainwater. Gemma Jansen states that compared to Ostia, Pompeii only had a few reservoirs and water storage areas, which indicates that there was enough piped water to satisfy the needs of the populace.²⁶ However, piped water was probably dominantly used for fountains and other decorative purposes, as the pipes rarely lead to kitchens and toilets.²⁷ This conclusion is made on the basis of the Casa del Granduca (VII,4,56) and a few other houses, which raises the question of whether this reflects the overall picture of Pompeii.²⁸

Nevertheless, it seems that collecting water in the peristyle was not always a necessity, and in many cases the water from the gutter was directed out into the street.²⁹ August Mau, for instance, states that the water was not collected in the southern peristyle of the Casa dei Vettii during the last phase, because there is no hole connecting the cistern and the gutter. Instead, the water channel ran into the street.³⁰ The cistern might have been filled without water from the roof, or the northern peristyle of the house may have collected the water into a cistern, as Jashemski suggests.³¹

Many assumptions about water use in Pompeian peristyle gardens remain speculative, as the central systems relating to water supply are often under the 79 CE level and therefore mainly unexcavated. Therefore, conclusions are made on the basis of only a few examples, and it seems that there are cases supporting both assumptions: sometimes rainwater was collected in a cistern, but occasionally it was directed into the street and was not intended for household use.

If we count the number of peristyles with a cistern head and the number of peristyles with a reported cistern (but missing a reported cistern head), the total is 147 peristyles. This could be considered to be the current maximum number of peristyles known to be related to the water supply of the house, and it is more than half of all the peristyles, making collecting water one of the most common human activity in the peristyles. Yet, there are more than 100 peristyles that cannot be connected to this function, and it does not seem possible to define water supply as the dominant activity in Pompeian peristyles.

3.2.2 Cult activity in the peristyles

The Pompeian domestic context distinguishes structures related to cult activity: lararia and altars. A lararium – household shrine – was a sacred place in the Roman house where the household divinities were worshiped.³² There are 29 peristyles with lararium paintings or reliefs in Pompeii. In addition, there are five peristyles with a niche and one peristyle with an aedicule, which were reported to contain lararium statues. Seven peristyles without lararia had a masonry or stone altar. Altogether, cult activity can be linked to 41 peristyles (Table 3.1).

In three peristyles, excavations under the garden surface have revealed burned organic materials that have been interpreted as offerings.³³ However, dating these actions to the last phase is uncertain, and it is also difficult to interpret how temporary these activities were in the peristyles,³⁴ whereas a built lararium suggests a continuity of activity. Marble, terracotta, or bronze altars have been found in a few peristyles, but as they are movable it does not necessarily imply that they were used in the peristyles.³⁵ Several niches, aediculae, or pavilions, without lararium paintings or reported lararium statues, are identified as locations of cult activity. For example, there are 41 peristyles with a niche, aedicula, or pavilion listed as lararium without any other evidence for this purpose, and as these features could have other functions, the connection with cult activity is debatable.³⁶

In addition to the lararia and altars, peristyle gardens in general have been connected to religion. In 1832, William Gell interpreted the garden of the peristyle of the Casa del Poeta tragico (VI,8,3/5) as a sort of sanctuary.³⁷ More than a century later, Grimal proposed that in the Roman world a garden was a shrine to Dionysus, and he among others refers to the peristyle of the Casa degli Amorini dorati (VI,16,7/38) to support this theory.³⁸ His theory of the Pompeian garden as a sanctuary to a deity is still maintained in much scholarly literature: Eugene Dwyer has stated that the peristyle was connected to the idea of a sacro-idyllic grove. In his interpretation, the garden with porticoes was associated with the worship of Venus, Apollo, and Diana, and more commonly included theatrical and Dionysiac themes.³⁹ Florian Seiler has also studied the Casa degli Amorini dorati, and suggested that the Rhodian peristyle was an imitation of a temple. Seiler sees a sacred connection between the space and its Dionysiac decoration – however, he also notes the possible decorative function of the sculpture.⁴⁰

Although the sanctity of a garden has been a popular theme in scholarship, it has not been accepted without criticism. Jashemski thinks that Grimal’s assumption is too bold, and in the Pompeian context it is difficult to believe that an ordinary Pompeian viewed the distinction between sacred and secular so clearly. Additionally, Jashemski observed the popularity of Dionysus and his followers as subjects in garden decorations, and noted the broader importance of the god of wine in Campania – a wine producing area – but she continued that the only garden that has unequivocal evidence for the worship of Dionysus is the temple garden of the deity outside the city walls of Pompeii.⁴¹

The dominance of Dionysus-themed sculpture decoration in the Pompeian peristyles is evident. There are 23 peristyles where the sculptural decoration can be associated with the deity.⁴² In terms of popularity, Venus comes second after Dionysus. The goddess can be related to the sculptural decoration of 13 peristyles.⁴³ Hercules, Apollo, and Jupiter are each present in the sculptural decoration of four peristyles.⁴⁴ Several other gods are represented in the peristyle sculpture; however, these gods can only be found in one or two peristyles.⁴⁵ Yet, the major problem of the interpretation lies in the question of whether these statues were considered to be decoration or religious symbols. The presence of a sculpture depicting a divinity does not necessarily make the space a sanctuary.

Several problems relating to the interpretation of the cultic role of the Dionysiac sculptures can be presented. Firstly, the imagery of the Dionysiac garden sculptures is problematic: the god himself is rarely represented, and the stage is rather occupied by his companions. As the divinity was usually regarded as being present through an image – especially through a cult statue – the presence of Dionysus in the peristyle gardens is not directly evident. Dionysus is frequently represented as a herm, either with a double face or with one face.⁴⁶ Although herms had a connection with cult activity, especially in a Greek context, in the Roman world they could be equally associated with education, philosophy, and the gymnasia, as Caitlín Barret has noted.⁴⁷ In Pompeii, herms are rarely found in the lararium context: there is one Dionysiac marble herm which was found among the lararium statues, and one other marble herm – which cannot be identified – from another lararium context.⁴⁸ This rarity, with only two known examples, suggests that these few cases are exceptions, and based only on these examples it is risky to assign a religious role to all herms.

In addition, the traditional place for a cult image in a temple would be on the central axis, and probably near the rear, while herms were placed on the borders of gardens, almost as if they were defining the borders of the space. A good example of this is the Casa di Marcus Lucretius (IX,3,5/24).⁴⁹ In this particular peristyle the traditional place of the cult image in the rear niche is taken by a statue of Silenus – not Dionysus. In the gardens, the god of wine is also represented in reliefs, and Dwyer notes: “The images of Bacchus himself which have been found in Pompeii relate more to the two-dimensional representations of the god in the thiasos than to known cult images.”⁵⁰

Secondly, the lararia contained miniature statues representing the gods that were worshiped there.⁵¹ The difference between the lararium statues and the garden statues in the Pompeian domus creates yet another problem when speculating on the sacred aspect of the garden sculpture: their size – or the scale as formulated by Dwyer – is different.⁵² Lararium statuary has a median height of 0.14 m, while the median height of the human-like garden statues in the peristyles is 0.60 m.⁵³ This represents a significant difference.

Thirdly, the garden sculptures and the lararium statues have a different emphasis in their materials. Most (97) of the lararium statues in the peristyles are made of bronze. There are 14 terracotta statues and nine of marble.⁵⁴ Terracotta is seen as the traditional material for Roman cult statues.⁵⁵ This is also mentioned in the literary evidence. Pliny the Elder notes that terracotta and wood were only replaced as the material for the statues of gods after the conquest of Asia, which introduced more luxurious items and materials.⁵⁶ Tibullus’s grandfather also had a wooden lararium statue, but apparently it had a bronze part – its spear.⁵⁷ The preference for bronze in Pompeii might have a connection with Samnite tradition, as the Samnites used small bronze idoletti which were about 0.15 m high and usually represented a warrior. Edward Salmon regards the massive production of these small bronze statues as a consequence of the adoption of cult images among the Samnites.⁵⁸ The material and size of the idoletti is similar to the Pompeian lararium statues. In addition, there was perhaps a workshop of bronze and terracotta figurines near Pompeii,⁵⁹ which also could explain the preference for these materials for cult statues.

The garden statuary, in contrast, is dominated by marble, whereas terracotta and bronze are rarer.⁶⁰ It seems that marble was regularly used for the garden sculpture, but was very sporadic in the domestic religious context; and if it was used, there is even a difference between the type of marble: four of the nine marble lararium statues are made of alabaster,⁶¹ which is not reported as the material of any garden sculpture. The marble lararium statues almost always depict female goddesses.⁶² Therefore, marble in this cult context might relate to female deities – although, it is not the only material used for representing goddesses.⁶³ This particularity further disconnects the male-dominated Dionysiac sculpture groups – nearly always made of marble – from their sacred character.⁶⁴

The use of marble as a material for statues of female divinities might instead indicate the cult use of some aediculae, for example in the peristyle of house I,2,17.⁶⁵ There are three other similar cases where the statues in the aediculae are made of marble and depict female deities.⁶⁶ These aediculae did not have paintings or other statues to link them with cultic activity, and they are similar to the structures of the fountain niches, which were also decorated with sculpture and are usually seen as decorative features.⁶⁷ Therefore it is not certain that the aediculae had a religious role, yet in the cases of marble female statues the association with cultic activity is more likely than in the others – as the material forms a link with the female lararium statues. Nevertheless, their scale is larger than the normal statues used in domestic cults, and it remains ambiguous to what extent these sculptures had much religious meaning.

Certainly there is a connection between divinities and garden statuary – at least, they often represent deities and their companions – or the visual connection between the peristyle and temple architecture.⁶⁸ Nevertheless, interpreting a peristyle as the shrine of a god that is represented in the decoration is an overextended conclusion – particularly for Dionysus. The sculptures – either on account of their material, size, or location – cannot be connected to cult statues. Nevertheless, linking the peristyle to the cult of Venus and Diana – as is done by Dwyer – might be a little more plausible in gardens where the female goddesses are found in the aediculae; however, the lararium paintings or statuary that would verify the cultic activity in these structures are missing.⁶⁹

The question of differentiating between decorative and religious entities is complicated, and our modern views may well influence the situation. Barret has stated that the ancient audience might not have separated the items and themes so strictly into religious and decorative categories. Rather, these themes were complimentary, not contradictory.⁷⁰

Our possibly anachronistic view is also visible when we compare Dionysiac-themed wall paintings and sculpture. Even if the first mentioned also depicts the god, the rooms with these paintings are not generally considered cult spaces.⁷¹ Indeed, if every mythological picture was a sign of cult activity, there would hardly be any space for other activities in the Pompeian house. This does not mean that these paintings – or sculpture – did not have religious value, but merely that they do not indicate cult activity in the space. Consequently, the number of peristyles that can be connected to cult activity is significantly lower than the number connected to, for example, the water supply, and cult activity cannot be considered to be the main function of the peristyle.

3.2.3 Business activity in the peristyles

Some of the Pompeian peristyles were used for small-scale industrial production or business activity. Even though these peristyles have attracted scholarly attention, there is no clear definition of when a peristyle can be identified as a commercial space, as these types of activities are not always very evident in the archaeological record. There are some structures that clearly indicate this type of function, but oftentimes any conclusion that can be made remains at best speculative. Particularly problematic are the peristyles in houses where these types of activities occurred in other rooms and there is no direct evidence that the activity extended to the peristyle.⁷²

There are several examples of speculative identification of peristyles as utilitarian spaces. For instance, Jashemski thinks that the garden with one portico in house V,3,8 was a produce garden,⁷³ but there are no archaeological remains to support this assumption. Ciarallo and Giordano report that the garden with one portico in the Casa di A. Octavius Primus (VII,15,12–13) was utilitarian,⁷⁴ but they do not specify what type of utilitarian garden it may have been, or provide any support for their interpretation. Colomba Serpe even thinks that the complete western part of the house was given over to some utilitarian purpose, but again no evidence is offered to support the speculation.⁷⁵ Mau goes even further and supposes that the whole house was used as a workshop. All of these interpretations might be based on the several graffiti found inside the house, but Mau does not clarify how the graffiti indicate that the apartment was used for working.⁷⁶ In the peristyle of house IX,6,4–7, the simple decoration, easy accessibility through entrance 7, and the traces of folding screens have tempted some to speculate that the peristyle was used for industrial or commercial purposes during the last period.⁷⁷ This use, however, remains hypothetical, and no other sources – such as archaeological finds or architectural structures – are cited to support the assumption. In general, the speculated utilitarian purposes of the above-mentioned peristyles might be due to their plain decoration, or complete lack of decoration. It is possible that an undecorated garden had a production purpose; however, there is no actual evidence to suggest how these gardens were used.

Structures such as ovens, counters, undecorated masonry basins, and a large number of dolia or amphorae in the peristyle are indicators of using peristyles for business purposes.⁷⁸ Nevertheless, not even these can be thought to be clear data of commercial activity. For example, the pools in the peristyle of the Casa del Banchiere (VII,14,5) are assumed to have been used for collecting water for the dyeing process; however, they have also been interpreted to be fish pools.⁷⁹ Consequently, whether this peristyle was used for industrial purposes remains speculative, although it probably was, as the vicinity of the tannery spaces in the house indicates.⁸⁰ Another example of the complexity of sources is the presence of dolia and amphorae. They obviously suggest a storage purpose, but how many dolia and amphorae are needed for us to conclude that the space had an important commercial role?

As a large open space, the peristyle has excited speculations that its garden section was utilized for the industrial activities of the house. For example, fulling and tanning require drying, and it has led to speculation that the open space in the peristyle was used for these purposes.⁸¹ This is a possibility, but direct evidence of this type of activity has not been reported. Furthermore, it has been proposed that drying might also have occurred in other parts of the house. Vittorio Spinazzalo and Sampaolo, for instance, suggest the upper floors for this purpose.⁸²

Table 3.1 lists the peristyles where commercial activity is very likely. Altogether, there are 17. Nonetheless, even in some of these cases the commercial activities can be questioned. Margareta Staub Gierow notes the possibility of the domestic use of the three looms in the peristyle of the Casa dei Capitelli figurati (VII,4,31/51).⁸³ This is one option, but if they were only used for household needs, it would be likely that this type of arrangement would also appear in other houses, and the case is unique in Pompeii.⁸⁴ Additionally, graffiti indicate the places of the persons working in the peristyle, and if the looms were only for household needs, why was there a need for this level of organization, and why did this house require three looms? Small-scale industrial use seems a more plausible explanation in this peristyle than merely domestic use.

The question of differentiating between domestic and commercial use is also complex when we consider food preparation and its possible retail use. Some peristyles are identified as restaurants or inns, for example houses II,8,2/3 and IX,9,1.⁸⁵ They both had features that are associated with dining, but it is challenging to understand whether the dining was business-related. Following Allison’s notion that it is often impossible to distinguish whether the work was done for consumption outside the household or only for domestic use,⁸⁶ it is also imaginable that these two peristyles had a domestic character.

Scholarship has seen a dichotomy between display and industrial use in the Roman house; it has been difficult to see these two types of activities taking place in the same space. This interpretation rests on the tradition of Moses Finley, who constructed it mainly on the basis of the writings of Cicero. The Roman writer was without a doubt very influential, for example Seneca and Pliny the Elder repeat similarly negative ideas towards trade and commerce – but do these reflect the entirety of Roman society? At best, the writers represent the views of the highest social elite males, but not even this elite group can be thought of as homogeneous. This separation of display and commerce in the Roman house has been questioned several times, and the lack of this type of behavior in Pompeii is demonstrated by Wallace-Hadrill.⁸⁷

This idea of the separation of the commercial and the domestic can be seen to lurk behind some interpretations of Pompeian peristyle gardens. It probably relates to the modern scholar’s quest to define single functions for the rooms of Roman houses. Although this reading has been questioned lately, there are cases where some functions might have been spatial-visually separated in the peristyle. For example, Flohr suggests that in the Fullonica VI,8,20, the fullery part of the peristyle was separated from the other parts of the peristyle by raising the fullery to a higher level and building a wall between it and the garden.⁸⁸ Nevertheless, the wall in its current condition is not high enough to visibly separate these spaces. A similar case occurs in the Fullonica di Stepahus (I,6,7), where the fullery seems to be separated in an area that could be loosely defined as the south portico of the peristyle, yet it is on a higher level. In the Fullonica di Stepahus there are some cuts in the floor that suggest that there might have been a door or a partition wall between the other parts of the peristyle and the fullery part.

Occasionally, the presence of commercial activity in the peristyle has been seen as a downgrading of the space.⁸⁹ This interpretation – again – has echoes of the Finleyan tradition, where leisure is seen as an elite function of the Roman house, and commerce is left to the lower classes. In the Pompeian context, the question is almost impossible to answer.⁹⁰ The data does not clearly indicate whether Pompeians better appreciated the commercial or the decorative use of the peristyle. On the one hand, commercial investments might, at least occasionally, have required more funds than the building of a pleasure garden, but equally the use of domestic space mainly for decorative purposes can be interpreted as a symbol of wealth. The relatively low number of peristyles used for commercial purposes might indicate that Pompeians tried to avoid using the space for commercial purposes, but it can also mean that not many Pompeians could afford these types of arrangements in their houses.

However, to return to the question of the multifunctionality of the peristyle, it has been demonstrated here that there is no single dominant function for the peristyle. Water supply is closest to this, but cult activity and commercial use are clearly rarer in the peristyle when examining all of the peristyles of the city. Additionally, when considering these three categories, they almost never require the entire space of the peristyle. Cult activity concentrates around the lararia and altars, and the water supply near the cistern head. Equally, these two activities are temporary. Even if they meant that a large crowd would have occasionally gathered in the peristyle, this still left plenty of time for other usages. Some industrial activities might have required the majority of the space, as well as daylight time. Nevertheless, even in these cases there seems to be room for different purposes, as the example of the Fullonica VI,8,20 demonstrates. There, fulling occurred in the western part of the peristyle and the eastern part was left for other functions – domestic, display, business, retail, or even all of these.

A quick glance at Table 3.1 might give the impression that the Pompeian peristyles were not designed for multiple activities: there are only 32 peristyle gardens with indications of at least two of the three examined activities: water supply, cult activity, or commercial. However, it is worth considering what is not demonstrated in this table. For example, every peristyle served to guide movement in the house, as established in the previous chapter. Of course, how important a peristyle was for household movement can be occasionally debated, such as for house V,3,8, where only two rooms opened onto the peristyle. However, this is the only occasion where the number of opening rooms is less than three, making it rather the exception than the rule.

Taking into consideration that Table 3.1 includes 174 peristyles, this means that in addition to guiding movement almost 70 percent of the peristyles were planned for some additional activity. Furthermore, this investigation has not considered the peristyle as a display space, which will be discussed in the following chapters. Moreover, several activities – domestic, commercial, or other – are invisible due to the shortcomings of the source documentation, or have not left an archaeological record. The peristyle was the center of light and air in the house, and this undoubtedly meant that daily activities occurred around this area. All of these activities place several different people – household members or guests – in the peristyle at some point during the day, meaning that the peristyle was not meant for any specific group of people, such as men or women, or freemen or slaves, but was used by everybody. It is safe to assume that every Pompeian peristyle was multifunctional, meaning that they were planned for several purposes, and there was no one main human activity planned for the space. Peristyles involved movement, and seemed to have a connecting function in the house, but the several indicators of activities that required time to be spent in the peristyle demonstrate that it was not only planned as a passage space.

3.3 The audience of the socioeconomic display

The display function of the Roman house has been connected to the separation of the public and private segments of the house. Display requires an audience, and naturally the public part of a house offers a better opportunity to reach people than the more restricted private section. However, the terms public and private have turned out to be challenging in the context of Roman and Pompeian houses. This chapter introduces some old interpretations of public and private, relating them to the peristyle, but also presents some criticism of them. In the light of new readings, it is impossible to base the interpretation of the Pompeian peristyle as a display space on the literary sources or routines, such as the salutatio, that have previously formed the center of the debate.⁹¹ Consequently, the later part of the chapter focuses on identifying the location of the audience in the Pompeian house, in particular reflecting on the peristyle as the visitor center of the house.

Over the last few decades, the concepts of public and private in the ancient world have been a frequent topic of research, and the house has had a central role in this discussion.⁹² If we consider the two concepts only through their legal definition of ownership, the strong role of the house in this debate seems strange. Should not every space in a domestic context be private? Nevertheless, these concepts are not merely juridical, as holistic life can rarely be strictly separated into these theoretical divisions. Public and private can also be defined through accessibility, which perhaps could be a possible substitute for these words.⁹³ However, in this case – when discussing public and private in a house – accessibility might be seen only as physical accessibility; but the axis of public and private additionally takes into account visual and other types of perceptible accessibility, which do not always require physical movement. Consequently, despite the problems, the terms public and private are better tools than simple accessibility, at least in this case.

Hypothetically, the Roman house was constantly under observing eyes and open to visitors. Scholars have stated that houses were open to the public, and the salutatio brought in a constant flow of visitors.⁹⁴ The house had an important role in public life, and it was in continuous communication with the surrounding world: for example, it could display its owner’s dignitas, social status, or politics.⁹⁵ Vitruvius’s famous passage is often cited as a source for placing public functions inside the Roman house. It describes how an important man needed spaces in his house where he could receive people, and among these requirements the architect listed peristylia amplissima.⁹⁶ According to these, the peristyle is often interpreted as a place for socializing and display, tilting it towards the public end in the sliding scale of public and private. However, there is no clear consensus on this topic, and several scholars have seen the peristyle as either private or accessed only by invitation.⁹⁷

Be that as it may, these views are mainly based on the literary sources, and are additionally problematic in that they are based on the social life of the city of Rome, and as has been stated several times there is a difference between the huge capital of an enormous empire and a small Campanian city.⁹⁸ A house of high social standing in Rome could have been repeatedly visited due to the practice of the salutatio, although whether these visits extended into the peristyle is doubtful. Some scholars leave the peristyle out of the context of the salutatio, limiting it to the vestibulum, atrium, alae, and tablinum.⁹⁹ One can still propose that in many houses the clientes caught – at least – a glimpse of the peristyle, even if they did not enter it. However, in the Pompeian context, it can be questioned whether such customs and habits were the same as in the capital. Was a Pompeian house also equally open, or was this openness only required by the politics of Rome?¹⁰⁰ Did a relatively small town need a ritual morning meeting, such as the salutatio known from Roman literary sources, to organize its social and political life?

Indeed, the practice of salutatio in Pompeii has been questioned. Jens-Arne Dickmann notes that during the Imperial era the salutatio lost its political significance.¹⁰¹ Also, the salutationes were mainly held by the senatorial class, and consequently they might have been restricted to Rome.¹⁰² Although Simon Speksnijder notes that the practice can also be connected to a few rich men, he emphasizes that the examples are very few. He thinks that it is hypothetically possible that freedmen held salutationes, but there are no clear sources for this type of activity.¹⁰³ There are no known men of senatorial rank from Pompeii,¹⁰⁴ and this forces us to question whether there were salutationes in Pompeii, as there were no people representing the class who usually organized this practice. Perhaps the richest house owners in Pompeii held salutationes? However, as Speksnijder notes, in general examples of salutationes held by individuals not of the senatorial class are rare.

Consequently, even if it is perhaps possible that Pompeian peristyles were visited daily due to salutatio, the uncertainty of it does not create a very sound basis for further interpretations. Against this background, stating that salutationes provided an audience for socioeconomic display in Pompeian houses is not very convincing – although hypothetically possible. The practice of the salutatio in Pompeii remains a dilemma, and the interpretation of the use of the peristyle cannot be built on the basis of this concept. It has been suggested that practices similar to the salutatio took place in Pompeii,¹⁰⁵ but it remains for future research to confirm or deny if this occurred, and therefore our understanding of the public nature of the peristyles cannot be built only on the basis of these interpretations. In fact, the idea of the openness of the Roman house might be simple political rhetoric, as Hales suggests,¹⁰⁶ and in a small town such as Pompeii the house owner might have had even stricter control over his or her house and its visitors. The possible audience for display thus has to be searched for from another direction, meaning that we now turn to investigate the houses of Pompeii.

Who used the peristyles of Pompeii, and how public or private they may have been, have been difficult questions to address. Several peristyles have stimulated different type of speculations on these topics. The southern peristyle of the Casa di Sallustio (VI,2,4) is an excellent example of how a peristyle can create several interpretations. It has been suggested that the house was a hospitium, particularly the area around the southern peristyle.¹⁰⁷ In this case, the peristyle would be very public compared to private houses, but there is no data to support this function. There would be a flow of visitors – perhaps paying customers. Anne Laidlaw and John Burge, on the other hand, think that the peristyle had a more private character during the last phase, when a small closet and a mezzanine were added to room 30, which they think was a space for a doorkeeper.¹⁰⁸ The function of room 30 is unclear, making the private nature of the peristyle speculative.¹⁰⁹ However, the peristyle was located south of the atrium, not opposite the house entrance, and thus it was visible only after entering the house, and at least its location made it more private – offering a little support to Laidlaw and Burge’s interpretation – but there is nothing to support its interpretation as a hospitium.

In addition, the southern peristyle of the Casa di Sallustio has been interpreted by many 19th century scholars as a women’s quarter.¹¹⁰ Similarly, the southern peristyle of the Casa del Citarista (I,4,5/25) and the northern peristyle of the Casa dei Vettii have been seen as used mainly by women.¹¹¹ However, Polly Lohmann has demonstrated that these interpretations – along with some other attempts to identify women’s quarters – have little support in the sources, and that there are no architectural or decorative features to distinguish any women’s quarters in Pompeian houses as such. Lohmann studied the find context of objects relating to women’s life, and concluded that they are found in small rooms and atria, but more rarely in the peristyles. In the atria and the peristyles, the finds were in contexts that suggest they were stored in these spaces.¹¹²

Women and guests are not the only groups that have been identified as users of specific peristyles. It has been suggested that the peristyles of the Casa di Ma. Castricius (VII,16,12–15) and the southern peristyle of the Casa dei Dioscuri (VI,9,6/7) were service quarters, but again there is little to support these interpretations.¹¹³ In general, there is no clear evidence to suggest that the Pompeian peristyles were utilized as women’s or slaves’ quarters, or that the use of the peristyle was restricted to a certain social or gender group. It is a possibility in some cases, but none of the Pompeian peristyles clearly indicates this type of use, and as the several activities in the previous chapter suggested, they were likely used by the entire household as well as by visitors. The interpretations of the peristyle as public or private are not built on very convincing ground, and therefore we must take a new approach to the question. The following section will take a look at the archaeological evidence, aiming to identify groups of visitors to Pompeian houses and their peristyles.

Because of the absence of an office culture in the Roman world, business and public matters were conducted inside the house.¹¹⁴ Allison notes that trade and business were not hidden activities within the Pompeian house, and that they could have been conducted in any area of the house.¹¹⁵ In addition, banquets, dinners, and other types of social gatherings were held inside the house. Eating and drinking were related to amici, friends, visiting the house.¹¹⁶ Pompeian houses were thus frequently observed by visitors, and were an integral part of the social life of inhabitants.

The next step is to connect these visitors to Pompeian houses to their peristyles. First, dining with guests occasionally occurred in the peristyles. Allison’s study maps the finds in Pompeian houses and links dining and the entertainment of guests to the gardens, and cooking to the colonnades.¹¹⁷ There are also permanent features that suggest that dining and cooking occurred in the peristyles. Six peristyles had a masonry cooking bench, and one is reported to have had a kitchen, but no remains of this kitchen are visible today.¹¹⁸

There are 23 Pompeian peristyles with a couch group – an outdoor triclinium, biclinium, or stibadium.¹¹⁹ The triclinia and other couch groups suggest that the space was used for dining, and possibly entertaining guests.¹²⁰ Drinking and eating are significant aspects of social display, as the items used for these purposes can be used for display and are associated with luxury and leisure.¹²¹

The architecture of the peristyles that had a dining group is not particularly luxurious compared to the other peristyles of Pompeii: the average size is about 170 m² and the median is 115 m². More than half had only one portico. The average area of the house with a peristyle and triclinium is about 540 m², placing it clearly below the Pompeian average for a house with a peristyle. However, the median size is equal to that of all the houses with a peristyle in the entire city (Figs. 2.5 and 2.6).¹²² It does not seem that an outdoor dining group was a feature of the peristyles in the houses of the wealthiest owners, and it does not seem to be a particularly luxurious element.

The evidence, however, consists mostly of peristyles with a masonry triclinium – a wooden dining group is reported in only three of the peristyles.¹²³ Among these are the two largest peristyles with outdoor triclinia: the Casa del Menandro (I,10,4/14–17) and the Casa di Paquius Proculus (I,7,1). Both had four porticoes and extensive decoration, such as fountains, pools, and wall paintings.¹²⁴ It is likely that other large or luxurious peristyles had wooden dining groups, but that they have not been documented.¹²⁵ In addition, the wooden triclinium seems to be smaller (1.74 x 2.69 m) than the masonry triclinia (ranging from 2.50 x 3.00 m to almost 5.00 x 5.00 m).¹²⁶ Zanker states that pieces of furniture, including dining couches, were light and portable.¹²⁷ It is possible that a dining group could have been relocated in the peristyle when required, and thus as a movable item it was not necessarily in the peristyle when the eruption occurred.

The peristyle area – the rooms around the peristyle – was also considered suitable for dining and the entertainment of guests.¹²⁸ Based on finds, Allison particularly connects her room Type 7 – traditionally called tablina – and Type 11 with dining, although this was not necessarily their only function. Both of these room types are open to the garden and/or the peristyle area.¹²⁹ In 107 peristyles the so-called tablinum opens onto the portico. Without a careful investigation of the finds, such as those made by Allison, there can be no certainty that in all the cases the rooms were actually used for dining and entertaining guests, but according to Allison’s results many of the tablina were used thusly. Her results were confirmed by Ambra Spinelli with a much larger sample, in particular when examining the floor decoration.¹³⁰ Yet if these activities cannot be as certainly connected to every single tablinum, it is very likely that at least one room opening onto the peristyle was used for dining and entertaining guests. This room likely had a fine view – through a large door or window or several windows – to the peristyle.

Even though there is no certainty about how most of the rooms around the peristyle were used, their large number indicates that plenty of human activity occurred around the peristyle area. It was literally impossible to enter many of the rooms without walking through the peristyle. This movement in and through the peristyle guaranteed an audience for display.¹³¹ A large part of the movement might be persons living in the house, but there must have been visitors too – even if entering the peristyle required an invitation. When guests arrived in these rooms, they had to pass through the peristyle, and in addition the peristyle created a pleasant background for these rooms.¹³² It has been said that peristyles were planned so that visitors could see the grandeur of the house when they were going through the peristyle to arrive at the reception rooms that opened off the colonnade.¹³³

The architectural shape of the peristyle is reminiscent of public architecture, particularly gymnasia and palaestra.¹³⁴ Although the public colonnades can be connected with educational activity, there seem to be only a few Pompeian peristyles – and only one where the activity can be confirmed by several graffiti – where this type of activity occurred.¹³⁵ Other activities relating to public architecture have been linked to peristyles, such as theater performances, public readings, or even trials, but the Pompeian evidence does not offer any certainty about these activities.¹³⁶ It has been suggested that the Casa del Criptoportico (I,6,2) was a public gathering area, and Alessandro Gallo speculates that the Casa di M. Epidius Rufus (IX,1,20) was a place of cult meetings of Dionysus, which would also mean a large number of visitors. Nevertheless, he also mentions that there is no definitive proof – such as graffiti – of this function.¹³⁷ If these houses had a public or a semi-public role, it would probably involve their peristyles, but there are no indications to support this type of function for these houses.

The graffiti in Pompeian houses are connected with guests and clients, and may even have been made by them.¹³⁸ Identifying the scribblers, however, is difficult if not impossible.¹³⁹ Graffiti are reported from 94 peristyles. It should be noted that the recording of the graffiti has not been carried out evenly throughout the city,¹⁴⁰ which probably means that more peristyles had graffiti than is documented. Lohmann states that peristyles and gardens seem to be the most popular location to write graffiti – and in particular columns were popular for graffiti. The texts in the peristyles mostly featured names and announcements. According to Lohmann, these were not personal messages – such as greetings, which are rare in peristyles – but a kind of secret confirmation of the presence of different persons, and they were probably not even meant to be read.¹⁴¹ Although the significance of graffiti often remains unknown to us, and they rarely reveal what type of activity occurred in a space,¹⁴² they at least demonstrate that someone spent time in the peristyle – enough to scribble something.¹⁴³ As Allison notes, the graffiti of the Casa del Menandro indicate that a number of people came and went quite freely in the house,¹⁴⁴ and one may conclude from this that other peristyles with graffiti were also used quite freely.

Another type of text found more rarely in the peristyles is electoral notices – only four peristyles have been reported with such.¹⁴⁵ In general, the role of electoral notices was highly public. They likely suggest that these peristyles were at least on some level open to the public. Yet, as the electoral notices are rarely found inside houses, their function in a domestic context is not clear, and we must be careful with this interpretation.¹⁴⁶

There are several activities that took place in the peristyles – meaning that many people also visited the space. The peristyle area – including the rooms opening onto the peristyle – was considered to be the living center of the house.¹⁴⁷ Allison also connects utilitarian household activities with the peristyle.¹⁴⁸ In these cases the display function had a lesser role, as the household members were probably already familiar with the peristyle, but the display features were also constant reminders of the social hierarchy of the house. Working and busy household members were also a part of the display of wealth for outsiders, as a large household was an indication of economic success and high social status.¹⁴⁹

The act of display does not necessarily require that people entered the peristyle. The space can be seen from other parts of the house, or even from outside the house. The peristyle is frequently compared to the atrium, which is understood as being more public, mainly because of its location.¹⁵⁰ Sometimes one peristyle in a house with several courtyards and atria is interpreted as more private than others.¹⁵¹ The difference between the interpretations is often caused by the different definitions of public and private, but the physical location of the peristyle also plays an important role in these assessments. In the traditional model of the Roman/Pompeian house, the peristyle is farther away from the door than, for example, an atrium. This obviously makes it more suitable as a private space: it is not so easily visible from outside the walls of the house. Therefore, the location and visibility of the peristyle in the house plan is important, if we consider the audience.

However, Anderson has conducted a GIS-based computer analysis of visibility inside six Pompeian houses. Five of the houses had at least one peristyle. In all cases the peristyle – or a part of it – is the most visible area in the house.¹⁵² Although Anderson’s sample is small, it implies that the best location for display was in the peristyle. There are no houses where the peristyle took the traditional place of the atrium in his sample, but in all cases the atria were the second most visible location in the house – after the peristyles. Consequently, eliminating an atrium from the house or replacing it with a peristyle would highlight the role of the peristyle in terms of visibility.

There are two houses with two peristyles – the Casa dell’Efebo (I,7,11/19) and the Casa di Sirico (VII,1,25/47) – in Anderson’s sample. In both cases, one peristyle is more visible compared to the other: in the Casa dell’Efebo it is the northern, and in the Casa di Sirico the southern.¹⁵³ The increased visibility could indicate that the other peristyle was more public, but there are also other factors, such as household activities, which could have influenced the public/private nature of the peristyles, and assumptions cannot be made solely on the basis of visibility.

Anderson’s main purpose was to study the storage locations of the construction materials needed for the rebuilding of Pompeian houses. The location of the materials suggests that the visual axis from the main entrance was important for Pompeians, as the materials are not visible when looking from the main entrance. Anderson concludes that this means that the house doors were open to the public.¹⁵⁴ The nature of this undisturbed view from the main door, however, has been questioned. It could have been blocked, for example, by doors, furniture, partition walls, curtains, or individuals living in or visiting the house.¹⁵⁵ However, doors and curtains can be opened, and furniture and people can be moved, and therefore it is at least theoretically possible that Pompeian houses offered a visual axis from the street into the house.

A vast majority of the peristyles (210) are on the main entrance axis. Most of these (113) are located after a room that is located after an atrium – frequently this room is called a tablinum. This type of room arrangement is usually considered typical of Roman and Pompeian houses.¹⁵⁶ The peristyle is visible from the entrance, but it is viewed through three spaces – the fauces, the atrium, and the tablinum (or some other room). Consequently, only a small part of the peristyle is visible from the entrance, and the number of possible visual obstacles is also higher, since there are several rooms between the viewer and the peristyle.

Thirty-two peristyles are located after two rooms – most often fauces and an atrium. The peristyle is closer to the entrance in these than in the houses with a tablinum, but the view is still restricted. Slightly more often, in 37 houses, the peristyle is in the traditional place of an atrium – after the fauces.¹⁵⁷ In these houses, hypothetically the entire rear side of the peristyle and most of the middle part would have been visible from the street.

There were 13 peristyles that were entered directly after the main entrance. Essentially, merely opening the door exposed a major part of the peristyle to the eyes of passers-by. The main entrance was not the only way to connect the peristyle to the public sphere. For example, the Casa di Cornelius Rufus (VIII,4,15/30) had four columns on the south side of the peristyle, and the intercolumniations were likely open, providing direct visual access into the Via del Tempio d’Iside and vice versa.

The architecture and archaeological material places several persons, of different statuses, genders, and ages, in the peristyle. They might have been there to work, to complete their everyday tasks relating to the life of the household, or simply to loiter; or perhaps they were guests spending time in the space. A relatively large number of people walked through the peristyle, and an even larger number gazed at the space everyday – provided that the owner did not have any particular reason to keep the space hidden from visitors and an outsider’s gaze. The multifunctionality of the space strengthens the likelihood of socioeconomic display in the peristyles, as many functions guaranteed a number of people using the space and would provide a large and versatile audience for the display.

If there was an audience for display anywhere in a Pompeian house, it was in the peristyle, making it an excellent space to demonstrate the house owner’s social and economic status. Consequently, locating artwork in the peristyle – which is discussed in more detail in the following chapters – was not a bad idea if it was meant to be marveled at. Furthermore, the size of peristyles supports their role as one of the principal display areas of the house. Besides some large gardens area, peristyles seem to be the largest spaces inside Pompeian houses (compare with the atria, see Table 3.2).¹⁵⁸ Also, the larger the size of the space was, the larger the group of visitors it could host.¹⁵⁹ In conclusion, peristyles played a crucial role in the social interactions that took place inside Pompeian houses, and were one of the most favorable places for display in the house. But this does not necessarily mean that peristyles were always used in this manner. The next chapter will examine the different forms of socioeconomic display found in peristyles.

Table 3.2 Selected ground area sizes for atria in Pompeii, and all the atria area sizes in Herculaneum.160 The atria of Pompeii are concentrated at the larger end of the continuum, which is highlighted by the comparison with all the atria of Herculaneum. It is likely that if all the atria of Pompeii were examined, the average and median would be lower. The Pompeian sizes without a cited reference were measured from the PBMP map, and the calculations of the Herculaneum atria were made on the basis of plans published by Maiuri (1958) and Jashemski (1993), and on the measurements of De Kind (1998, 234‒235, 245, 271, 273, 288 293, 299, 305, 307, Plan VI) and Maiuri (1958, 227, 277, 305, 337).
House number	House name	Atrium area m²
I,2,6		70
I,2,24		41
I,4,5/25	Casa del Citarista	90
I,5,2	Conceria	31
I,9,13–14	Casa di Cerere	70161
I,10,4/14–17	Casa del Menandro	36162
I,11,14		28
I,12,11	Casa dei Pittori	35
I,16,5		34
II,2,2	Casa di D. Octavius Quartio	112
II,3,3	Casa della Venere in conchiglie	88
V,2,i	Casa delle nozze d’argento	198163
V,4,b		21
VI,2,4	Casa di Sallustio	135
VI,5,5	Casa del Granduca Michele	72
VI,6,1	Casa di Pansa	135
VI,7,6	Casa d’Ercole	77
VI,8,23/24	Casa della Fontana piccola	99164
VI,9,2/13	Casa di Meleagro	87
VI,9,3/5	Casa del Centauro	85
VI,9,6/7	Casa dei Dioscuri	84165
VI,10,2	Casa dei cinque scheletri	60166
VI,12,2	Casa del Fauno	170
VI,14,12		78
VI,14,38		75167
VI,14,40		76168
VI,15,1	Casa dei Vettii	92169
VI,15,7/8	Casa del Principe di Napoli	39170
VI,16,7	Casa degli Amorini dorati	48171
VI,17,32–36		55
VI,17,32–36		80
VII,1,25/47	Casa di Sirico	42
VII,2,16-17	Casa di M. Gavius Rufus	43
VII,4,31/51	Casa dei Capitelli colorati	135
VII,4,48	Casa della Caccia antica	70172
VII,4,56	Casa del Granduca	53173
VII,4,57	Casa dei Capitelli figurati	128174
VII,4,59	Casa della Parete nera	81175
VII,4,62	Casa delle Forme di Creta	78176
VII,7,5	Casa di Trittolemo	76
VII,7,23		68
VII,10,5		40
VII,11,9–10		31
VII,12,1–4		99
VII,13,4/17–18	Casa di Ganimede	72
VII,14,5	Casa del Banchiere	98
VIII,2,14–16		135177
VIII,2,26		125
VIII,5,2/5	Casa del Gallo	170
VIII,5,24	Casa del Medico	37
VIII,5,28	Casa della Calce	93
VIII,7,26–27		74
IX,1,12		55178
IX,1,20	Casa di M. Epidius Rufus	205
IX,1,22/29	Casa di M. Epidius Sabinus	54
IX,2,16	Casa di T. Dentatius Panthera	65179
IX,3,2	Fabbrica di prodotti chimici	36180
IX,5,9	Casa dei Pigmei	51
IX,5,14–16		91
IX,7,20	Casa della Fortuna	42
IX,8,3/7	Casa del Centenario	34
IX,9,1		49181
IX,14,4	Casa di Obellius Firmus	238182
Average		80
Median		74
Herculaneum
II,1	Casa di Aristide	60
III,3	Casa dello Scheletro	55
III,11	Casa del Tramezzo di legno	80
III,16	Casa dell’Erma di Bronzo	40
III,17	Casa dell’Ara Laterizia	25
III,19–18/1–2	Casa detta dell’albergo	55
IV,1–2	Casa dell’atrio a mosaico	75
IV,3–4	Casa dell’alcova	30
IV,5–7	Casa della Fullonica	35
IV,5–7	Casa della Fullonica	45
IV,12–13&15–16	Grande Taberna	40
IV,17–18	Taberna	40
IV,19–20	Casa della Stoffa	25
IV,21	Casa dei Cervi	25
V,1	Casa Sannitica	65
V,5	Casa del Mobilio carbonizzato	45
V,6/7	Casa del Mosaico di Nettuno e di Anfitrite	60
V,11	Casa dell’Apollo Citaredo	60
V,15–16	Casa del Bicentenario	95
V,31	Casa del Sacello di legno	60
VI,13/11	Casa del Salone nero	80
VI,17/26	Casa del Colonnato tuscanico	70
VI,29	Casa dei Due atri	55
VI,29	Casa dei Due atri	35
Ins. Occ. I,1	Casa della Gemma	85

Notes

1. See Leach 1997, 52, Sampaolo 1997, 428; 1998 974, Zanker 1998, 12–13, Dickmann 1998, 452, Bragantini 1999, 142, Farrar 1998, 19, Wallace-Hadrill 2008, 193, Ciarallo and Giordano 2012, 389, n. 41, 537 n. 280, 556 n. 306, Trentin 2014, 1.

2. Mazzoleni 2004, 31. Carucci 2007, 18.

3. On this basic function of a house, see Mazzoleni 2004, 7.

4. Jashemski 1993, 10. Farrar 1998, 17. On windows, see Spinazzola 1953, 65–80.

5. Farrar 1998, 16–17. See also Boman 2011, 90.

6. Wallace-Hadrill 1994, 84.

7. Longobardi 2001, 67, 69.

8. Bechi (1825, vol. II, tav. 55, 2) mentions that the portico of the peristyle of the Casa del Poeta tragico (VI,8,3/5) was the best illuminated space of the house. The importance of sunlight is also evident in legal texts, which forbid blocking sunshine for domestic spaces with trees (Liberati 2007, 111).

9. Mazzoleni 2004, 21.

10. Nn. 2, 10, 20, 21, 25, 30, 35, 49, 63, 102, 103, 107, 126, 145, 163, 185, 198, 215, 220, 229, 248.

11. See O’Sullivan 2011, 5–7, 17–18, Gleason & Palmer 2017, 371–374, 395–396, Tally-Schumacher & Niemeier 2016, 59, 64–69.

12. Grahame 2000, 73, 172–196. Von Stackelberg 2009, 120, 147–149. Anderson 2011, 81–86 figs. 5.2, 5.3, 5.5–5.7. See also Grahame 1997, 146–163. On the use of Space Syntax analysis in ancient studies, see Laurence 2011, 397–399. On the houses of Volubilis, see Hilder 2015, 175. Von Stackelberg states that the peristyle of the Casa di D. Octavius Quartio (II,2,2) was relatively inaccessible, but this claim is mainly based on Von Stackelberg’s problematic division of the porticoes. See also the following note 130.

13. Grahame (2000, 41) and Hilder (2015, 163) count the peristyle as a single unit. Grahame (2000, 101, 103, 105) thinks that when there is a garden with one portico, then the portico should be counted as its own unit and the garden as another, but he (see Grahame 2000, 130) is not very consistent with this, as in the Casa della Fontana grande (VI,8,22) he instead counts the garden with one portico as a single unit. Von Stackelberg (2009, 114–116) separates the garden space and the portico, and she also counts every colonnade as its own unit (north portico, west portico, etc.). This creates several problems. First, the corners of the peristyle belong to two porticoes. E.g. the northwest corner is part of the north and the west colonnade. Von Stackelberg has resolved this by simply arbitrarily choosing which portico the corner belongs to. E.g. in the Casa del Menandro (I,10,4/14–17) the northwest corner belongs to the west portico (as room 11 is marked as opening onto the west portico), when in the Casa di D. Octavius Quartio (II,2,2) the northwest corner belongs to the north portico (as room 9 is marked as opening onto the north portico). Second, the atriums should be counted in a similar way as the peristyles, as movement in an atrium also occurs in a similar pattern as in a peristyle, and an atrium can be seen to form four passageways, although they do not often contain columns (see Grahame 2000, 40–41 on the problem of the atrium and the impluvium in the Space Syntax analysis). Von Stackelberg has counted the atrium as one unit, which creates a bias that the atrium was much more important for movement than the peristyle. E.g. counting the peristyle of the Casa di D. Octavius Quartio as one unit provides a control value of 4.56, which is much closer to the control value of the atrium – 5.63 – given by Von Stackelberg (2009, 147), than the separate control values of the porticoes. Nevertheless, separating the porticoes offers a more accurate idea of how the movement in the peristyle probably worked. In addition, in the Casa del Menandro and the Casa di D. Octavius Quartio the peristyle had a pluteus or fence. In the peristyles where the garden is separated with this type of physical obstacle, it is justifiable to count the garden as its own unit. If every portico is counted as its own unit, the line between the porticoes should probably be made to run between the garden corner and the peristyle corner, and therefore all of the rooms that open on the north wall should be counted to be open to the north portico, and all the rooms on the west wall to the west portico, and so on. This does not model the movement perfectly, but as there is rarely information on how the portico borders should be divided, it would be the best method. In addition, the movement in the garden spaces is problematic, as their layout is unknown. E.g., the connection between the northern entrance and the eastern entrance of the peristyle garden area in the Casa del Menandro is assumed, but there is no certainty that there was no barrier (trees, bushes etc.) between the entrances that prevented movement between the two.

14. Anderson 2011, 77, 81–82 figs. 5.2, 5.3, 84–86 figs. 5.5–5.7. On the development of computer assisted methods, see e.g. Landeschi & al. 2016.

15. Grahame 2000, 41–42. See also Von Stackelberg 2009, 59.

16. See the list of different type of houses in Schoonhoven 2006 and Flohr 2017. Additionally, Wallace-Hadrill’s (1994, 80–82) house typology differentiates several types of houses that do not include a peristyle.

17. See Petersen 2006, 123, Nevett 2010, 97–99, Bablitz 2015, 76, Simelius 2015, 131, Cribiore 2015, 156, Ellis 2018, 9–10, Berg 2010, 304-305; 2019b, 60-61.

18. Berry 1997, 183–185, 193–195. Allison 2004, 89–90. Nevett 2010, 97–99.

19. Nissin 2016, 27–28, 57-59.

20. Nevett 2010, 101.

21. Berry 1997, 193–195. Allison 2004, 70, 84–90, 99–103; 2006, 303–305, 360–362, 388–389. Nevett 2010, 97–110, 114–118.

22. On the functioning of water collection in the Roman house, see Farrar 1998, 15, 17, Richardson 1955, 42, Sear 2004, 164–165, De Haan 2010,75.

23. Jashemski 1993, 147–148 n. 280.

24. See Allison 2004, 85–86, 90. Allison (2006, 303) states that the main water collection location of the Casa del Menandro (I,10,4/14–17) was the cistern head covered by a puteal in the north portico of the peristyle.

25. The difference between a cistern head and a puteal is not very clear, particularly when a masonry structure surrounds the cistern opening. In this study, a cistern head is a structure that is supposed to be sunk into the ground, at least partly – although sometimes it is above the ground level – whereas a puteal is above the ground level.

26. Jansen 2011, 71.

27. Jansen 2017, 414. See also Sear 2004, 165–166 and Leander Touati 2010, 122.

28. On the problems of investigating the water distribution in Pompeii, see Olsson 2015, 71–74.

29. The Casa di L. Caecilius Iucundus (V,1,26): Karivieri & Forsell 2015: http://www.pompejiprojektet.se/room.php?hid=13&hidnummer=6388183&hrubrik=V%201,26%20Casa%20di%20Caecilius%20Iucundus%20–%20South%20House&rid=83&ridnummer=5741683&rrubrik=Room%20l%20(peristyle)”. Last visited 26.7.2016. The Casa dei Gladiatori (V,5,3): Mau 1901, 290. House VIII,5,15–16: Mau 1883, 198, Jashemski 1993, 216–217 n. 438. The Casa di Acceptus e Euhodia (VIII,5,39): Mau 1884, 128. House IX,6,f–g: Mau 1881, 22. The Casa della Fortuna (IX,7,20): Mau 1882, 220. Jashemski 1993, 240 n. 501. In house IX,9,1, the connection between the cistern and the gutter was blocked, but it is not stated where the water was directed (Mau 1889, 6). In the Casa di vinaio (IX,9,6), the water collected from the peristyle was directed into the cistern, but the channel continued from the cistern into the street (Mau 1889, 15). This indicates that the water was collected and, when the cistern was full, the surplus water flowed into the street.

30. Mau 1896, 31–32.

31. Jashemski 1993, 155 n. 295.

32. Brandt 2010, 57. Bilias & Grigolo 2019, 82–86.

33. Robinson 2002, 94–95, 98; 2007, 155; 2009, 295. Zanier 2009, 230, 282. The houses are I,9,12, Casa degli Epigrammi (V,1,18) and Casa di Sextus Pompeius Axiochus (VI,13,19).

34. E.g. Robinson (2002, 94) dates the offerings in the peristyle of house I,9,12 to the middle of the first century CE. Therefore, they probably cannot be connected to the last phase of the peristyle.

35. E.g. nn. 2, 8, 16, 136, 178, 189, 235.

36. Niche: nn. 2, 23, 33, 36, 41, 60, 61, 62, 64, 86, 128, 130, 148, 157, 169, 171, 176, 188, 211, 243. The niche without lararium paintings is Brandt’s (2010, 61) Type 30. Aediculae and pavilion: 3, 4, 24, 30, 50, 61, 69, 74, 87, 96, 99, 103, 104, 113, 123, 125, 137, 142, 165, 233, 236. The aediculae and pavilions without lararium paintings are Brandt’s (2010, 60–61) Types 10 and 20. On the several functions of niches, see Allison 2004, 84; 2006, 20, 31. On the decorative function of the niches in the Casa di Sextus Pompeius Axiochus (VI,13,19), see Ciarallo & Giordano 2012, 542 n. 286. Boyce (1937, 71 n. 323) indicates that the niche in the peristyle of the Casa di Ganimede (VII,13,4/17–18) originally had an aedicule shape, but he also states that nothing of the shape is visible, which makes the information doubtful. Boyce (1937, 32 n. 72) reports possible evidence on the bottom of the niche in the peristyle of the Casa del Toro (V,1,7). If the niche had statues, the nature of these statues is unknown, and therefore the cult function is doubtful.

37. Gell 1832, I, 159.

38. Grimal 1984, 324–326.

39. Dwyer 1982, 137.

40. Seiler 1992, 133. See also Loccardi 2009, 69.

41. Jashemski 1979, 123–124.

42. Nn. 14, 24, 67, 94, 97, 104, 122, 123, 125, 134, 136, 137, 139, 147, 149, 164, 166, 178, 218, 129, 235, 244, 245. In the peristyle n. 7, the theater mask can possibly be connected to Dionysus (for the link between Dionysus and masks in general, see Hales 2007, 338).

43. Nn. 3, 25, 33, 37, 62, 103, 107, 123, 126, 136, 139, 235, 244. There are sculptures that are identified as Paris in two peristyles (nn. 122, 134). They can perhaps be linked to Venus.

44. Hercules: nn. 39, 94, 139, 166. Apollo: nn. 16, 38, 87, 112. Jupiter: nn.24, 123, 139, 187.

45. Diana in nn. 24, 168. Peristyles with a sculpture of other divinities: nn. 24, 94, 113, 133, 139, 166, 235. A terracotta sculpture in peristyle n. 7 might be Vulcan.

46. Hales 2007, 338.

47. Barret 2019, 172–173.

48. Boyce 1937, 23 n. 13, 30–31 n. 67. Also, according to Boyce (1937, 62 n. 251), one bronze bust is a herm. He (1937, 37 n. 108) additionally reports a terracotta head of a Bacchante, but thinks it is a votive. The busts of bronze, terracotta, and unknown perishable material are represented in the lararia (Boyce 1937, 23 n. 13, 28 n. 49, 31 n. 67, 37 n. 108, 41 n. 123, 62 n. 251, 83 n. 408). The herms might have a link with the busts; however, Brandt (2010, 63) does not think that the heads were cultic.

49. Hales 2007, 338. Herms are usually near the borders of gardens, see e.g. the Casa di Marcus Lucretius (Jashemski 1993, 231–232 n. 479, Kuivalainen 2019, 72–78) and the Casa degli Amorini dorati (Jashemski 1993, 159–163 n. 302). The liminal role is suitable for Dionysus as noted by Aston (2011, 127–132), who identifies Dionysus as the divinity of borders.

50. Dwyer 1982, 123.

51. Boyce 1937, 17–18. Dwyer 1982, 121.

52. Boyce 1937, 49 n. 168. Dwyer 1982, 121–123, 135.

53. The lararium statue heights can be found in Boyce 1937, 21–94, 108–198. Seven statues are taller than 0.30 (see Boyce’s lararium numbers: 202, 220, 221, 439, 445, 467). The average height is 0.16 m, which is close to the median, meaning that the large statues are rare exceptions. I have not included the statues in Boyce’s numbers 10 and 259, as their cult function is doubtful. The garden statue heights are from nn. 2, 3, 16, 24, 33, 37, 38, 39, 62, 87, 94, 97, 104, 107, 122, 126, 133, 134, 136, 139, 164, 187, 218, 219, 235, 244, 245. The average is 0.59 m. If all the characters (children, satyrs, fauns, Pan, nymphs, cupids) that are not identified among the lararium statues (Boyce 1937, 106–107) are eliminated from the count, the median rises to 0.67 m and the average to 0.68 m. Boyce (1937, 72 n. 329) reports bronze figurines of a faun and a cupid in a lararium; however, he states that they were decorative in character. He does not provide any explanation for his interpretation, and therefore their nature and role remain uncertain. The incomplete information on these sculpture – e.g., the measurements are missing – limits the possibilities of making further conclusions about them. Boyce (1937, 37 n. 108) also reports a terracotta head of a Bacchante, but he thinks it was a votive.

54. The materials can be found in Boyce 1937, 21–94, 108–109. I have not included the statues in Boyce’s nn. 10 and 259, as their cult function is doubtful. Also, in Boyce’s lararium n. 176 the statue was probably not from a lararium, as Richardson (1955, 46) has demonstrated. Dwyer (1982, 121) notes the dominance of bronze as a material for lararium statues, but he does not offer any statistics for his notion.

55. McDonnell 2006, 73.

56. Plin. HN 34.34.

57. Tib. 10.15–25.

58. Salmon 1967, 131–132, 181. Salmon speculates that the idoletti might come from Etruria, or are local productions.

59. See Pietilä-Castrén 2019, 117.

60. See Section 4.3.3.

61. Boyce 1937, 41 n. 123, 57 n. 220, 84 n. 416. In addition, a lararium statue reported by Boyce (1937, 30 n. 61) is included, although it is reported to be pseudo-alabaster.

62. Boyce 1937, 26 n. 42, 30 n. 61, 41 n. 123, 53 n. 202, 57 n. 220, 75 n. 350, 109 n. 9. The only male character reported is Horus (Boyce 1937, 57 n. 220). The subject of one marble lararium statue is unknown (Boyce 1937, 84 n. 416). In addition, in Trimalchio’s house there was a marble statue of Venus in the house shrine, besides the silver Lares (Pet. Sat. 29).

63. Dwyer (1982, 121) notes the connection of the use of marble and Venus, but also states that Venus was among the patron deities that were made of bronze. For the female goddesses – Venus, Diana, and Isis – represented in other materials than marble, see e.g. Boyce 1937, 23 n. 13, 32 n. 75, 82 n. 406, 83 n. 408, 198 nn. 2, 5, 7, 9.

64. There are only two peristyles where Dionysiac themes are represented as bronze statues (nn. 97 and 245), and possibly one (n. 244) where a cupid might be connected to the thiasos of Dionysus. There is only one known lararium sculpture of Dionysus, and it is a terracotta statue (Boyce 1937, 89 n. 446). In the lararium context, the god is more often present in paintings (Boyce 1937, 104–150).

65. Boyce 1937, 22 n. 3. See also A. De Vos 1990, 38, Ciarallo and Giordano 2012, 367–368 n. 7.

66. Boyce 1937, 22 n. 10, 63 n. 259, 75 n. 350. The measurements of the statue in n. 350 are not provided. Jashemski (1993, 211 n. 421) states that she was unable to locate the statue, and speculates that it might have been given to the Princess of Saxony. There is also an aedicula in the peristyle of the Casa del Poeta tragico (VI,8,3/5), but it did not have a statue inside, and was interpreted as a decorative structure (Boyce 1937, 48–49 n. 168). Dwyer (1982, 118) thinks that there was most likely a marble statue of a satyr inside this aedicula, but it seems that the statue was actually found in the nearby tablinum (Jashemski 1993, 133 n. 248, Carrella 2008, 84). The aedicula in the peristyle of the Casa del Poeta tragico is often identified as a lararium (Type 20) or shrine, see Bechi 1824, vol. II, tav. 55, 8, Bonucci 1827, 118–119, Gell 1832, I, 170, Fiorelli 1862, 126, Parise Badoni 1993, 528, Narciso 1993, 548, Brandt 2010, 102 n. 151 and Ciarallo and Giordano 2012, 517–518 n. 252. The bronze Apollo of the peristyle of the Casa delle Danzatrici (VI,2,22) is speculated to have been in the aedicula (Bonucci 1827, 97–98, Ciarallo & Giordano 2012, 301–302 n. 215), but its finding place is reported to be only the peristyle in general (Fiorelli 1860, III, 54–55), and hence there is no certainty that it was in the aedicula.

67. Fountain niches with statues: n. 24, 164, 235. See Chapters 4.3.2 and 4.3.3.

68. Grimal (1984, 326) connects Dionysiac sculpture with the term “diffuse religion.”

69. See Dwyer 1982, 137.

70. Barret 2019, 31.

71. For example, Casa di Polibio (IX,13,1–3, see Bragantini 2003, 211). On the other functions of mythological representations in general, see e.g. Hakanen 2020, 124.

72. For examples of houses connected to wine production, but it is uncertain whether the peristyle was used for this purpose, see house II,9,6 (Jashemski 1993, 97–98 nn. 154–155, Ciarallo & Giordano 2012, 463–464 n. 157), house VIII,5,15–16: (Sogliano 1882, 279–280, Sampaolo 1998, 572. House VIII,4,12–13: Lugebil 1861, 237).

73. Jashemski 1993, 114 n. 186.

74. Ciarallo & Giordano 2012, 606–607 n. 399.

75. Serpe 2008, 141.

76. Mau 1874, 96–97.

77. Sampaolo 1999, 748.

78. Flohr and Wilson (2017, 14) mention dyeworks, fullonicae, tanneries, and perfume and pottery workshops as business locations that can be identified as the basis of the built infrastructure. On the methods for identifying workshops on the basis of the masonry structures, see Monteix 2017, 216–221 and Flohr 2013, 20–26. On the pottery workshops (only one is identified inside the city walls, and the building does not have a peristyle), see McCallum 2011, 106–108.

79. See Breton 1870, 455, Fiorelli 1875, 301, Jashemski 1993, 198 n. 389, Sampaolo 1997, 676, Ciarallo & Giordano 2012, 602–603 n. 390.

80. On the house and its tannery, see Fiorelli 1875, 301, Jashemski 1993, 198 n. 389, Sampaolo 1997, 676, Ciarallo & Giordano 2012, 602–603 n. 390.

81. Fiorelli 1873, 30; 1875, 185. Jashemski 1993, 35 n. 33, 172 n. 320.

82. Sampaolo 1996, 496. Spinazzola 1953b, 773.

83. Staub Gierow 1994, 81.

84. There are other locations identified according to graffiti as places where large-scale spinning and weaving may have taken place, but many of these have been questioned. See Flohr 2013b, 66 and Jongman 1988, 161–165.

85. Mau 1889, 5–6. A. De Vos 1991, 316. Sampaolo 2003, 115. Ciarallo & Giordano 2012, 458–459 n. 150.

86. Allison 2006, 14.

87. Cic. Off. 1.150–151. Finley 1973, 41–43. On the criticism, see Wallace-Hadrill 1994, 140–141, Flohr 2013, 2–3, 7–11, Jongman 2017, 418–419, Ellis 2018, 97–101, 184–185. On Roman attitudes towards trade, for example in the works of Seneca and Pliny the Elder, see Tchernia 2016, 3–6.

88. Flohr 2011, 94–98. See also Flohr 2013, 253, 255.

89. Allison 2006, 343–345, 348. On the regression of the Casa del fabbro (I,10,7) during the last period, see Ling & Ling 2005, 144–145, 169.

90. On criticism of the interpretation of downgrading, see Berry 1997, 185, 193–194, Nevett 2010, 112.

91. See Wallace-Hadrill 1994, 5, 12. Grahame 2000, 90. Hales, 2003, 2, 42. Anderson 2011, 87. Cfr. Hales 2003, 36–39, Speksnijder 2015, 88, 96–98.

92. Russell 2016, 13–16.

93. On the complexity of the concepts of public and private, see Bartz 2019.

94. Wallace-Hadrill 1994, 5, 12. Grahame 2000, 90. Hales, 2003, 2, 42. Anderson 2011, 87. Cfr. Hales 2003, 36–39, Speksnijder 2015, 88, 96–98. On the importance of openness and visibility for large Pompeian houses, see Viitanen, Nissinen & Korhonen 2012, 75.

95. Carandini 1985, 119. Wiseman 1987, 393. Wallace-Hadrill 1994, 12. Zanker 1998, 10–12. Hales 2003, 38–39. Pesando 1997, 9. Zanker 1998, 10. Nevett 2010, 5, 7–8. Bablitz 2015, 75.

96. Vitr. 6.5.2.

97. On the public and private in the Roman house and the situation of the peristyle in this dichotomy, see Allison 1993, 1; 2004, 154, Pesando 1997, 263, Leach 1997, 59, George 1997, 310–311, Zanker 1998, 10–12, Petersen (2006, 125-126, 175, Wallace-Hadrill 2008, 195–196, Brandt 2010, 80, 93, Goldbeck 2010, 24, Karivieri 2014, 109. Bablitz 2015, 71–76, Speksnijder 2015, 97–98. Dwyer (1982, 120) states that the peristyle is among the areas that were the most accessible to the public. On the public and private scale, it is also important to note that some people consider some actions more private or public than others (e.g. Brandt 2010, 93). See also, the interpretations of peristyles as a location to display status and wealth: Farrar 1998, 19, Trentin 2014, 1, Sampaolo 1997, 428; 1998, 974, Dickmann 1998, 452, Bragantini 1999, 142. On the southern peristyle of the Casa di M. Epidius Sabinus (IX,1,22/29), see Sampaolo 1998, 974, Ciarallo and Giordano 2012, 389, n. 41, 537 n. 280, 556 n. 306.

98. Ciarallo & Mariotti Lippi 1993, 116, Pesando 1997, 6, 9, Allison 2001, 53; 2004, xv, 14, Viitanen & Ynnilä 2014, 142, Speksnijder 2015, 88.

99. Clarke 1991, 12–13. Wallace-Hadrill 1994, 12. Laurence 1997, 13. Leach 1997, 52.

100. On the openness of the Roman house, see Hales 2003, 36–39. Viitanen, Nissinen and Korhonen (2012, 76) state that according to the mapping of the active streets, visibility and openness were important for large houses.

101. Dickmann 1999, 371–372. See also Berry 2016, 141.

102. Goldbeck 2010, 22–23. Viitanen & Ynnilä 2014, 142, 148–149. Viitanen, Nissinen & Korhonen 2012, 75. Speksnijder 2015, 88.

103. Speksnijder 2015, 88–89.

104. Camodeca 2008, 25.

105. Viitanen and Ynnilä (2014, 148–149) and Viitanen, Nissinen and Korhonen (2012, 75, 76) propose, based on graffiti locations, that there were probably similar social practices.

106. Hales 2003, 36–39.

107. Della Corte 1954, 29–30. Sampaolo 1993, 87, 125. Jashemski 1993, 121 n. 203.

108. Laidlaw & Burge 2014, 264.

109. Sampaolo (1993, 124) identifies the room as a closet or storage room.

110. Bonucci 1827, 102. Niccolini & Niccolini 1862, Descrizione general, 23. Fiorelli 1875, 84–85.

111. M. De Vos 1990, 140. Sampaolo 1994, 565. Ciarallo & Giordano 2012, 551 n. 299. On the Casa dei Vettii cfr. Niccolini & Niccolini 1896, Nuovi Scavi, 75.

112. Lohmann 2015b, 91–94, 100, 104. See also Berg 2010, 304-305.

113. Bechi 1829, Relazione degli scavi di Pompei, 24–25. Bragantini 1997, 889.

114. Wallace-Hadrill 1994, 12. Hales 2003, 2. Tuori 2010, 55, 64–65.

115. Allison 2004, 155.

116. Wallace-Hadrill 1994, 12. Pesando 1997, 249. Zanker 1998, 13, 163. Hales 2003, 2. Feasting was regulated by laws that dictated that the maximum number of guests, which was set to three in 182 BCE; however, the limits were constantly raised by subsequent laws (Wallace-Hadrill 2008, 327–328).

117. Allison 2004, 89–90, 126, 131–132, 139; 2006, 303, 305, 388–389. In addition, house IX,9,1: Mau 1889, 6, Sampaolo 2003, 115.

118. The peristyles with a kitchen bench: nn. 51, 60, 159, 191, 242, 247. I have counted the masonry table in the peristyle of the Casa della nave Europa (I,15,3) as a cooking bench, as interpreted by Jashemski (1993, 61 n. 106). Ciarallo and Giordano (2012, 425–426 n. 108) state instead that it is a kitchen table. Also, remains that have been interpreted as a kitchen bench were found in peristyle n. 7 of house I,3,8, but the terracotta sculpture head is said to have been found inside the construction, which makes the link with cooking doubtful. Also, peristyle n. 17 is reported to have had a kitchen.

119. Nn. 4, 5, 8, 17, 22, 24, 28, 38, 39, 49, 50, 60, 62, 70, 72, 79, 81, 84, 118, 203, 219, 220, 239. There are, in addition, five peristyles with a pergola without an outdoor triclinium: nn. 94, 136, 148, 165, 245. These peristyles might have had a wooden dining group under the pergola, as it and an outdoor triclinia are often together in the gardens (Zanker 1998, 175, see also the peristyles: nn. 24, 28, 39, 60, 62, 70, 79, 81, 84, 219). E.g. Soprano (1950, 308), Jashemski (1993, 169, 171 n. 316), Ciarallo and Giordano (2012, 565–567 n. 321) speculate that under the pergola of the southern peristyle of the Casa di Sirico (VII,1,25/47) was a wooden triclinium.

120. On the connection of outdoor triclinia with dining, see Painter 2001, 40.

121. On the connection of drinking and eating vessels with luxury and display, see Allison 2004, 139.

122. Zanker (1998, 174–175) links the outdoor triclinia to the “smaller houses” and small gardens, but he particularly mentions houses without peristyles. He probably does not count the gardens with one portico as peristyles. However, Cassetta and Costantino (2006, 295) state that outdoor triclinia are often found in medium-large houses. In both cases, the ambiguous definitions of small and medium-large make it difficult to know which other houses are considered to be in the same group. In the houses with a peristyle, outdoor triclinia can be considered to be in the houses that are more-or-less of medium size, but the tendency is more towards the lower medium size than the upper medium size. However, when considering all of the houses of Pompeii, they likely belong to the group of medium-large houses.

123. Nn. 22, 38, 39. Allison (2006, 73 n. 263) is skeptical about the existence of a wooden triclinium in n. 38.

124. Nn. 22, 38.

125. On the problems of the documentation of organic finds, see Allison 2004, 127. Mau (1877, 168) thinks that there was a triclinium around the decorative emblem on the portico floor in house VI,13,13. Bragantini (1994, 449) and Carrella (2008, 97) speculate about the possibility that there was an outdoor triclinium around the pool in the peristyle of the Casa degli scienziati (VI,14,43). This remains a possibility for both houses, but there is no archaeological evidence to support these speculations.

126. Nn. 4, 5, 17, 24, 28, 38, 49, 60, 62, 70, 79, 81, 84, 118, 203, 219, 220, 239.

127. Zanker 1998, 11.

128. Leach 1982, 153. Wallace-Hadrill 1994, 47. Zanker 1998, 13, 165. Brandt 2010, 80. Bablitz 2015, 74.

129. Allison 2004, 82, 92, 94, 131–132, 168. Type 7: “Open rooms leading to garden or open–sided rooms opposite main entranceways” and Type 11: “Medium/large open fronted rooms off gardens/terraces with window or wide entranceway giving view of garden or lower floor.”

130. Spinelli 2019, 154–162.

131. On movement in the peristyle, see Section 3.1.

132. See Zanker 1998, 165, 168.

133. Zanker 1998, 13.

134. Hoffmann 1994, 122. Dickmann 1997, 126–127; 1999, 36–37, 158. Wallace-Hadrill 2008, 173, 175. Simelius 2015, 121. See also Wallace-Hadrill 1994, 21, 28, Leach 1997, 59, Mayer 2012, 187.

135. Simelius 2015, 130. On the educational role of public porticoes, see Garcia & Garcia 2004, 57. On public porticoes and education in general, see Frakes 2009, 34–37. On education in the peristyles, see Zaccaria Ruggiu 1995, 343. For house I,2,16, see Pesando 1997, 215–216, Inserra 2008, 20.

136. Simelius 2015, 127–131. On theatrical and musical performances or readings, see Jashemski 1979, 101; 1993, 146 n. 277. On legal performances, see Bablitz 2015, 67–71, 75–76.

137. Spinazzola 1953, 442, 444. Jashemski 1993, 34. Gallo 2013, 175–192.

138. Viitanen, Nissinen & Korhonen 2012, 74–75. Lohmann 2015, 73.

139. Milnor 2014, 137–138, 148.

140. Viitanen, Nissinen & Korhonen 2012, 69. Milnor 2014, 5, 12–20 (also on the difficulty of dating the graffiti). Lohmann 2015, 71.

141. Lohmann 2015, 73, 75.

142. E.g. Milnor (2014, 160–171) links graffiti to displaying names, salutations, confirming the presence of the writer in the space, and practicing writing. See also Lohmann 2015, 74–75. Lohmann proposes several scenarios for why the graffiti in the peristyles were made. Lohmann links the vertical strokes to industrial, commercial, and domestic activities, but states that these graffiti were usually more common in the atria.

143. Viitanen, Nissinen & Korhonen 2012, 68.

144. Allison 2006, 334.

145. Nn. 77, 178, 250, 251. On the east wall of the garden of house VI,13,16 there was an electoral notice (CIL IV 3463, Tiussi 2009, 502–504). The garden did not have columns or piers, and is not counted as a peristyle in this study. In addition, there are texts written in red paint in three peristyles (nn. 153, 168, 242), but these writings are not necessarily electoral notices, and their function remains unclear for this analysis. On the connection between the use of paint and electoral notices, and the color red and display, see Milnor 2014, 5, 235.

146. On the electoral notices inside the Casa di Polibio (IX,13,1–3), see Milnor 2014, 131. Milnor suggests that the house was, at least partly, transformed into a more public role such as business.

147. Allison 2004, 154. Casa dei Postumii (VIII,4,4/49): Dickmann 1998, 452. Casa dei Pigmei (IX,5,9): Bragantini 1999, 486. Casa di Polibio (IX,13,1–3): Bragantini 2003, 183. Casa dei pittori al lavoro (IX,12,9): Varone 2007, 140.

148. Allison 2004, 89–90, 126.

149. Wallace-Hadrill 1994, 72. Speksnijder 2015, 90.

150. See Grahame 2000, 90, Allison 2004, 154, De Haan 2010, 120, Milnor 2014, 111. Falkener (1853, 68–69) thinks that the peristyle of the Casa di Marcus Lucretius (IX,3,5/24) was more private than the atrium.

151. Mazois (1824, II, 72, 82) thinks that the northern peristyle of house VI,17,23–26 was private, and that the peristyle of the Casa di Pansa (VI,6,1) was in the private section of the house.

152. Anderson 2011, 77–78, 81–82, 84–86. The visibility of the peristyles has been noted earlier. E.g., Niccolini and Niccolini (1854, Casa di M. Lucrezio, 17) state that the peristyle of the Casa di Marcus Lucretius (IX,3,5/24) was visible from everywhere in the house. This is not the case, but the peristyle was visible from several spaces in the house, and it is on an elevated level compared to the tablinum and the atrium (Jashemski 1993, 239 n. 479), which probably made it even more visible. Sampaolo (1999, 670) states that the piers and columns of the peristyle of the Casa di Giasone (IX,5,18) were arranged so that the view from the street was not blocked. On the peristyle of the Casa di Sextus Pompeius Axiochus (VI,13,19), see Zanier 2009, 247.

153. Anderson 2011, 84–85.

154. Anderson 2011, 81, 86–87. On the visual axis from the main entrance in Roman and Pompeian houses, see Zanker 1998, 181, Mazzoleni 2004, 31, Von Stackelberg 2009, 112–113, Brandt 2010, 76–77, Berry 2016, 125. On the Casa della Fontana grande (VI,8,22), see Sampaolo 1993, 613.

155. Speksnijder 2015, 88, 97–98. E.g. in the tablinum of the Casa di Obellius Firmus (IX,14,4), an attachment for curtains has been reported (Sampaolo 2003, 362). Allison (2006, 303–304) reports folding doors in the west portico of the Casa del Menandro (I,10,4/14–17). According to her, the doors were probably salvaged from somewhere else, because there was no need for such doors in this portico. It is, however, possible that they were connected to the private baths, and increased the privacy of the bathers. Bechi (1831, Relazione degli Scavi di Pompei, 8) and Gell (1832, I, 160) report that in Herculaneum evidence has been found of curtains between columns. It is not specified whether the evidence is an actual piece of cloth, or nails in the columns that might have held curtains. In the southern peristyle of house VIII,2,14–16 there is a marble slab (0.18 x 0.25 m) in the northeast corner of the open space. It has a 0.10 x 0.14 m cut. Visually, it appears to be a closing or framing apparatus that is typically found nearby doors. There might have been some sort of closing mechanism between this garden and the portico. On the Casa del Tramezzo di Legno in Herculaneum, see Tuori & al. 2017, 45–46, 61, 68.

156. Bechi 1831, Relazione degli Scavi di Pompei, 15. On the Casa del Toro (V,1,7), see Sampaolo 1991, 481. On the Casa del Forno di ferro (VI,13,6), see Lipizer & Loccardi 2009, 105. On the Casa di Sextus Pompeius Axiochus (VI,13,19), see Zanier 2009, 229. Sampaolo (1999, 41) states that house IX,2,17 follows the canonical order of rooms. However, there is a room between the so-called tablinum and the peristyle.

157. Wallace-Hadrill (1994, 84) states that a peristyle after the fauces might have been called an atrium.

158. On the gardens, see Jashemski 1993, 61–98. The atrium seems to be the most important space that competes with the peristyle in size. There are no comprehensive statistics on the atria area size in Pompeii. Speksnijder (2015, 89) states that the atria sizes range between 100–450 m², but he refers to the entire Empire. His sample contains only four atria, and the largest is the Domus of Scaurus in Rome, but its measurements are based on a reconstruction. The only atrium in the sample from Pompeii is the atrium of the Casa del Fauno (VI,12,2), which is about 170 m².

159. De Haan (2010, 122–123) uses this logic with the private baths: as a small bath can host only a few people, and a large several, this means that the first mentioned was more private than the second.

160. See Tuori & al. 2017, 76 Appendix 3.

161. De Vos 1976, 37.

162. See n. 38.

163. Ehrhardt 2004, 35.

164. Fröhlich 1996, 18.

165. Richardson 1955, 69–70.

166. Rossi 2006, 39.

167. Mau 1878, 110.

168. Mau 1877, 209.

169. Mau 1896, 11. Sogliano 1898, 269.

170. Strocka 1984, 19–20.

171. Seiler 1992, 22–25.

172. Sear 2002, 17.

173. Staub Gierow 1994, 22.

174. Staub Gierow 1994, 50.

175. Staub Gierow 2000, 23.

176. Staub Gierow 2000, 89.

177. Mau 1892, 8. Sampaolo 1998, 74.

178. Gallo 2001, 25–26, 49–50.

179. Trendelenburg 1871, 201.

180. See n. 34.

181. Mau 1889, 4.

182. Spinazzola 1953, 337. Sampaolo 2003, 361.