Department of Archaeology and Ancient history
The architecture of function
Understanding House K at San Giovenale
Ghaza Alyasin
Master thesis 45 credits in Classical archaeology and ancient history Spring term 2020 Supervisors: Lars Karlsson and Fredrik Tobin-Dodd
Abstract
Alyasin, G. 2020. Funktionens arkitektur: Att förstå Hus K i San Giovenale.
Alyasin, G. 2020. The architecture of function: Understanding House K at San Giovenale. Between 1961 and 1963 a large terraced Etruscan building on the southside of the Acropolis at the archaeological site of San Giovenale in Italy was excavated. The building, named House K, remained largely unpublished and unstudied throughout the years, leaving our picture of the ancient settlement at San Giovenale incomplete.
This thesis aims to get a better understanding of the Etruscan architecture, settlement, culture, and society at San Giovenale, by doing an architectural analysis of the building of House K, using unpublished field journals, drawings, and photographs. By comparing the architecture of House K and any structural elements immediately connected with the building, to other structures at San Giovenale, as well as other Etruscan sites, this thesis set out to learn: what kind of structure House K is; what function it had; what its relative dating is; what role it played in the settlement at San Giovenale; and what it can tell us regarding the society and culture at San Giovenale.
Using environment-behaviour relations and building archaeology, the study first examined the structural elements of House K, before comparing House K to other structures at San Giovenale built 700–400 BCE. Thereafter, the study looked at parallels outside of San Giovenale, looking at the sites of Acquarossa, Luni sul Mignone, Poggio Civitate, Cerveteri, Pyrgi, Montetosto, Vulci, and Narce. Finally, a discussion was had regarding the character and function of House K, whether it was public, private, sacred, or profane.
In the course of the study, an architectural chronology of San Giovenale was established, which dated the construction date of House K to between c. 565 BCE and the late 6th century BCE. The interpretation of the function of House K that seems most probable is that it was a monumental building complex of administrative, political and/or religious significance, which probably also acted as a residence for the ruling elite at San Giovenale – in other words: a
palazzo. House K being a palazzo would mean that there was a centralised authority at San
Giovenale, suggesting that there was some sort of ruler or leader at the top of the social hierarchy, splitting the social division at San Giovenale further than it was previously believed to have been.
Keywords: San Giovenale, Etruscan architecture, ashlar blocks, headers, monumental
building, monumentality, monumental architecture, Archaic architecture, Etruscan settlement, courtyard complex, palazzo.
Master thesis in Classical archaeology and ancient history 45 hp. Supervisors: Lars Karlsson and Fredrik Tobin-Dodd. Defended and passed 2020-06-15.
© Ghaza Alyasin
Department of Archaeology and Ancient History, Uppsala University, Box 626, 75126 Uppsala, Sweden
Cover photo: The King’s terrace with House K in 1963, on the southside of the Acropolis at San Giovenale. Courtesy of the Swedish Institute in Rome.
Acknowledgements
Firstly, I would like to thank my main supervisor, Lars Karlsson, for his constant enthusiasm, encouragement, support, kindness and endless positivity, as well as for sharing his extended knowledge of San Giovenale and its environs. Having a supervisor that has conducted fieldwork and published material from San Giovenale, the very site this thesis is about, has been very rewarding and helpful.
The same goes for my assistant supervisor, Fredrik Tobin-Dodd, who was the one that gave me this thesis subject. He has not only been an excellent teacher, sharing his extensive knowledge and experience of all things Etruscan, but also an extremely helpful and amazingly supportive supervisor, providing me with invaluable feedback, always making the abstract concrete and clearing things up when I have felt the most overwhelmed, as well as pushing me to keep going when I most wanted to give up and quit.
I also want to thank Yvonne Backe-Forsberg who has been a great help and an incomparable source of knowledge and wisdom, continuously and tirelessly giving advice and telling stories of San Giovenale and its excavations, sharing and inspiring with her great love of the ancient settlement. Likewise, I want to thank Ludmila Werkström for her generous kindness, patience, and support, as well as for her superior scanning skills.
I would also like to express my gratitude to the wonderful, kind and helpful staff of the Swedish Institute in Rome, where I was lucky enough to spend five weeks conducting research and gathering material for this thesis. Without the possibility of being allowed access to the archives of the institute, this study could not have been made. In addition, I am grateful to the institute, as well as to the Soprintendenza Archeologia, Belle Arti e Paesaggio per l'area metropolitana di Roma, la provincia di Viterbo e l'Etruria meridionale, for generously allowing me to use unpublished photographs and drawings in this thesis.
A big thank you also goes to the Department of Archaeology and Ancient history at Uppsala University, for five years that have helped shape me as a classical archaeologist, and as a person.
Lastly, I want to say that writing a master’s thesis during the time of COVID-19 has not been an easy task, for reasons that go without saying. However, thanks to the positive and constant support of my supervisors and the Department of Archaeology and Ancient history at Uppsala University, this thesis still managed to get finished on time, for which I am incredibly grateful. Your hard work during these strange and straining times has not been overlooked or taken for granted. Thank you.
Contents
1. Introduction ...1
1.1. Purpose and research questions ...2
1.2. Material and method ...2
1.3. Theoretical framework ...3
1.4. Previous research ...4
2. The focused picture ...6
2.1. The King’s terrace ...7
2.2. House K ...8 2.3. Walls ... 10 2.3.1. W01 ... 10 2.3.2. W02 ... 11 2.3.3. W03 ... 13 2.3.4. W04 ... 13 2.3.5. W05 ... 14 2.3.6. W06 ... 15 2.3.7. W07 ... 15 2.3.8. W08 ... 16 2.3.9. W09 ... 17 2.3.10. W10 ... 19 2.3.11. W11a–b ... 19 2.3.12. W12 ... 19 2.3.13. W13 ... 20 2.3.14. W14 ... 21 2.3.15. W15 ... 22 2.3.16. W16a–b ... 22 2.4. North room ... 26 2.5. Room A ... 27 2.6. Room B ... 27 2.7. Room C ... 28 2.8. Room D ... 30 2.9. Room E ... 32 2.10. Room F ... 33 2.11. Cuniculus E ... 35 2.12. The courtyard ... 36 2.13. The street ... 38
3.1. Area F East ... 42
3.2. Area B ... 44
3.3. The Borgo ... 46
3.4. The Pietrisco bridge complex ... 51
4. The bigger picture ... 54
4.1. Building technique and material ... 54
4.2. Size and layout ... 57
4.3. Location ... 60
5. The final picture ... 63
5.1. Public or private? ... 63 5.2. Sacred or profane? ... 66 6. Conclusion ... 69 7. Bibliography ... 71 7.1. Unpublished sources ... 71 7.2. Published sources ... 71 8. List of illustrations ... 75
1
1. Introduction
In 1956 the archaeological site of San Giovenale, located approximately 55 kilometres north-west of Rome, began being excavated by the Swedish Institute in Rome, in collaboration with the Soprintendenza alle antichità dell'Etruria meridionale.1 The tuff2 plateau of San Giovenale is divided into two parts: its eastern part, the Borgo, is somewhat lower and much smaller than its western part, the Acropolis. Fieldwork and excavations have been carried out on both parts of the plateau: on the Borgo first 1956–65, and then again in 1999, and on the Acropolis 1957– 65.3
The Acropolis was inhabited at least as early as the Neolithic, up until the Middle Ages.4 In modern times the Acropolis has been heavily cultivated, which has resulted in disturbance of ancient remains and finds. The site has thus been in use for a long time with many different archaeological layers indicating its long history.
Between 1961 and 1963 a large terraced Etruscan building on the southside of the Acropolis, called Area C, was excavated, first under the Swedish King Gustav VI Adolf in 1961, and later under Margareta Lindgren with the help of Lars Gezelius in 1962–1963.5 During the King’s
1 San Giovenale IV:1, 21.
2 There has traditionally been some confusion in archaeological publications regarding the volcanic rock known in Italian as ‘tufo’: even though it has often been referred to as ‘tufa’, which geologically is a different type of rock, the correct geological term for tufo in English is ‘tuff’ – which is the term that will be used in this thesis (Heiken et al. 2005, 38.). See Hemphill 2000, 19–20 for a description of the geology of San Giovenale. 3 Hanell 1962, 299; Karlsson 2000, 99; San Giovenale IV:1, 21; San Giovenale V:1, 58.
4 Gierow 1986, 27. For a discussion of the Neolithic material see Gierow 1984, 17–36; for the medieval remains, a medieval castle believed to have been built in the 12th or 13th century at the north-eastern most part of the Acropolis, see Thordeman 1962 and San Giovenale VI:4–5, 34–51.
5 Gierow 1986, 29.
Figure 1. The King’s terrace before excavation began
in 1961. Viewed from the west. Courtesy of the Swedish Institute in Rome.
Figure 2. The King’s terrace before excavation began
in 1961. Viewed from the south-west. Courtesy of the Swedish Institute in Rome.
excavation, the opened trench was named Trench V, later dubbed The King’s terrace, and under Lindgren, the building within the terrace was named House K. Lindgren began working on publishing the excavations, but unfortunately never published anything before passing in 2017. The terrace and House K have thus remained largely unpublished and unstudied throughout the years, leaving our understanding and picture of the ancient settlement at San Giovenale incomplete.
1.1. Purpose and research questions
The aim of this thesis is to get a better understanding of the Etruscan architecture, settlement, culture and society at San Giovenale, by studying and doing an architectural analysis of the building of House K. The thesis will focus on House K and the structural elements immediately connected to the building and compare it to other structures both at San Giovenale, as well as at other sites. Additionally, this thesis will finally give researchers access to House K for further studies of Etruscan architecture and culture, until a full and proper publication is made.
The main research questions for this thesis is: what kind of structure is House K and what was its function? Relying solely on architecture, the study will investigate the following questions: how was House K built and why was it built the way it was, how does it differ from and relate to other buildings at San Giovenale, and can the building techniques represented at House K inform us of its function, relative dating and its role in the settlement? What information, if any, can House K give regarding the society and culture at San Giovenale? Furthermore, what can House K tell us of the architectural chronology at San Giovenale?
Finally, it is necessary to have a brief discussion about terminology. This study aims to understand the function of House K – function here being defined as the building’s intended use, why it was built at all, why it was built the way it was and with what function in mind? To understand how the building was used and what activities were performed there, on the other hand, it would be necessary to include finds, which this study will not as it is limited to and focused on architecture.
1.2. Material and method
The material used for this study will be documentation of primary data left principally by Lindgren, but also by Gezelius and the King, including field journals, drawings, photographs as well as additional notes. Lindgren wrote four field journals, and a fifth with simple notes, in which she documents the process of excavation, as well as many finds. The King also wrote a field journal in which he documents the process of the excavation during 1961. Drawings made by Lindgren and Gezelius are important since they sometimes are the only documentation of vital information such as the length, width and height of features. To this will be added mostly black and white photographs of House K and the terrace.
This thesis will focus on architecture for an interpretation of function. The relation of House K to the rest of the settlement at San Giovenale will be given attention to first and foremost, and then other Etruscan sites, including Acquarossa, Luni sul Mignone, Poggio Civitate (Murlo), Cerveteri (Caere), Pyrgi, Montetosto, Vulci and Narce. Using the field journals, drawings and photographs, House K will be presented and described, including measurements of structural elements. The building complex and its features will first be dealt with separately (Chapter 2), before looking at and comparing it to the rest of the settlement at San Giovenale (Chapter 3), after which comparisons with a larger Etruscan area will be included, focusing on architecture (Chapter 4) and function (Chapter 5), to give a better understanding of similarities and differences within the Etruscan cultural sphere.
There will be certain requirements for the comparative material in order to keep this study from exceeding its designated limits: the study will only include structures of Etruscan origin;
3
and since Lindgren interpreted House K to be late Orientalizing/early Archaic6 and gave it an estimated date of c. 625–530 BCE, only structures dated between 700 BCE and 400 BCE will be included.7
This study will be limited to the structure of House K and any architectural structures in the immediate vicinity of House K. There are several other architectural and structural elements within The King’s terrace8 that can affect the interpretation of House K considerably. Nonetheless, they will not be included here as the space constraints of this study do not allow for an in-depth presentation and discussion of the entire terrace. Furthermore, roof-tiles will not be included for several reasons: firstly, most of the roof-tiles were discarded of after the excavations and were not well documented, and secondly, the few roof-tiles that remain could not be examined physically for this study.
For a complete and final publication of The King’s terrace, it is necessary to include all the structural elements present at the terrace and to study the roof-tiles, stratigraphy and finds comprehensively, including any prehistoric material. This thesis will nevertheless create a foundation for a full publication of House K and The King’s terrace and hopefully add more to our understanding of Etruscan architecture and its relation to the Etruscan identity, especially at San Giovenale.
Lastly, it is important to keep in mind that studying an archaeological excavation that one has not participated in oneself is problematic for several reasons. Since House K is more or less completely covered by vegetation today, any study of it depends on documentation, photographs and drawings, which in this case are limited and of varying quality. Even though there are good plan drawings of the building and terrace, sections and measurements of many parts of the structure were never made, some measurements possible to ascertain and others unfortunately not. There is also no proper list of photographs, leaving some of the photographs unidentified. Additionally, the photographs lack orientation and a proper scale. Even so, the photographs that can be identified, mainly by comparing them to drawings, will be used in this study.
1.3. Theoretical framework
Since the main research question of this thesis pertains to the function of House K, elements and aspects from different theoretical frameworks will be used as a starting point, including environment-behaviour relations (EBR) and building archaeology.9
EBR is the relationship between people and their environment and their effects on one another. Built environments can reflect social expressions, meaning the study of architecture could inform us of cultural tendencies.10 Amos Rapoport believes there is a link between culture – that is human behaviour – and the built environment, and that architecture to some degree encloses the human behaviour and thus the culture, and can because of this communicate identity.11 In discussing how built environments are created for the purpose of supporting human behaviours, he argues that the activity of said behaviours will likely shape the architecture and that the form of the architecture, in turn, can reflect these activities.12
Along the same lines as Rapoport, Ian Hodder first argued that houses and buildings are reflections, even expressions, of social forms, and hence social processes.13 He later changed his view and instead suggests that the social units that were supposedly reflected or expressed
6 Orientalizing period: c.720–600/575 BCE, Archaic period: c. 600/575–480 BCE.
7 The date range of 700–400 BCE allows for some leeway in the discussion of the architectural chronology at San Giovenale, and a possible date for House K.
8 See section 2.1. in this thesis for a short list of structures within The King’s terrace. 9 Steadman 1996, 68.
10 Rapoport 1990, 10. 11 Rapoport 2000, 148.
12 Rapoport 1990, 11; Steadman 1996, 68–69. 13 Hodder 1998, 85.
in buildings, were instead instilled in those buildings – meaning that the buildings themselves are the social units, which social groups defined themselves in relation to. For example, the residents of a palace are defined as royals or aristocrats, while the residents of a small farm are defined as farmers.
Likewise, Donald Sanders argues that the practical function of a building is only second to its meaning as a cultural and social unit.14 According to him, a building’s function is divided into two basic components: the primary, which is a pure denoted function, and the secondary, which is a connoted and conceptual function. Social behaviour is the result of reactions to the secondary function and its symbolism. Having certain expected and preconceived ideas about the secondary function of a building, signals and prompts certain responses. In turn, the social behaviour that is generated by these responses shapes cultural norms and values within any given cultural group. In other words, architecture forms culture.
Lastly, architectural archaeology, or building archaeology, is based on the principle that one through direct observations of buildings can reconstruct their history, by using the theoretical principal of archaeological stratigraphy, meaning walls and structures are studied using stratigraphic analysis, or elevation readings.15 Simply put, by reading a structure through its different stratigraphic layers, it is possible to understand and define its chronology and construction, alteration, occupation and even destruction phases. However, the method has been developed for the study of medieval structures and is thus better suited for structures that are much better preserved and with longer lifespans than House K. Despite this, the hope is that the method will allow for a partial understanding of the chronology of House K, and its possible function. By focusing on the structural and architectural elements of House K, the idea is to be able to read the building: how it was built, with what technique and in what order, to be able to answer why it was built the way it was.
Furthermore, by studying the construction and building technique of House K and comparing it with other structures at San Giovenale and other sites, the goal is to establish a relative chronological typology of technique and material for the Etruscan architecture at San Giovenale.
1.4. Previous research
To begin with, one of the only publications that has published anything about the excavation and structure of House K is Etruscan culture: Land and people,16 written by members of the Swedish Institute in Rome. The publication focuses on the archaeological research conducted by Swedish archaeologists at San Giovenale and its environs. The English edition, released 1962, a year after the Swedish edition, includes results from the excavation in 1961 when House K first began being excavated. However, it states that the excavation had not been completed and that there were no definite interpretations of the building complex at the time, but that based on the available evidence it was similar to terraced habitation found on the Borgo, which will be discussed in more detail in Chapter 3.17 A preliminary date for House K was also included in Etruscan culture: based on pottery finds from the unfinished excavation, the building was given an approximate date of 6th century BCE.18
When it comes to publications of the excavations on the Borgo, the most comprehensive work of the architecture and stratigraphy is done by Carl Nylander et al. in the work The Borgo:
excavating an Etruscan quarter: architecture and stratigraphy.19 As for the Acropolis, the most important publications relevant for this study are: San Giovenale: Area F East. Huts and Houses
14 Sanders 1990, 45. 15 Schuller 2002, 7; Dessales 2017, 71. 16 Etruscan culture 1962. 17 Hanell 1962, 302. 18 Hanell 1962, 304. 19 San Giovenale V:1.
5
on the Acropolis20 by Lars Karlsson, in which Karlsson discusses the architecture, stratigraphy and finds of three buildings to the east of House K; San Giovenale: Excavations in Area B,
1957–196021 by Eric Berggren and Kristina Berggren in which they discuss a 7th century BCE building north-east of House K; and San Giovenale: The semi-subterranean building in Area
B22 by Björn Olinder and Ingrid Pohl in which they discuss a cut-out square room north of
House K. Lastly, for the bridge complex over the Pietrisco river, Yvonne Backe-Forsberg’s dissertation Crossing the bridge23 will be relied on.
Additionally, two other important works on San Giovenale and its environs, as well as Swedish research in South Etruria, will be consulted: San Giovenale: Materiali e problemi24 which gives a good overview and summary of almost all excavations and remains – excluding The King’s terrace – at San Giovenale, as well as Architettura etrusca nel Viterbese,25 which is focused on the architecture at San Giovenale and Acquarossa. In Architettura etrusca, Nylander dates House K to the second half of the 7th century BCE.26
Finally, a few short words on the publishing history of The King’s terrace: the terrace itself has not been given different names, but the building has. Even though it was left out of Materiali
e problemi, it was still included in a map of the site in the publication, which was taken from
the first publication about San Giovenale, San Giovenale: General introduction by Bengt E. Thomasson, being referred to as “House V”.27 However, in Etruscan Culture the building was referred to as “House K”,28 which is what Lindgren called it as well, hence it will continue to be referred to as House K is this study.
20 San Giovenale IV:1. 21 San Giovenale II:2. 22 San Giovenale II:4. 23 Backe-Forsberg 2005.
24 San Giovenale. Materiali e problemi 1984. 25 Architettura etrusca 1986.
26 Nylander 1986a, 39.
27 San Giovenale I:1, 8, map 3; San Giovenale. Materiali e problemi 1984, 14, fig. 1. 28 Hanell 1962, 302, figs. 275–276.
2. The focused picture
In this chapter, the focus will lie on The King’s terrace. The architectural and structural features and elements of House K will be presented, followed in the next chapter by other architectural and structural features within the settlement at San Giovenale. The main terms for the building technique described will be headers and stretchers, referring to ashlar blocks stacked on their short side (headers) and blocks stacked on their long side (stretchers) along the lengths of the walls. The study will, for the most part, keep names of features given by Lindgren but will also rename and add names to previously unnamed features for the sake of clarity and simplicity.
Since the excavated area is cut diagonally by the north-south line, from the north-western corner to south-eastern for the structure itself, and south-west for additional walls and structural elements, this thesis will mainly refer to the different sides of the terrace and building according to the four cardinal points. Referring mainly to the cardinal points will help to simplify the understanding and reading of both the thesis and the building. Consequently, this means that what is referred to as north (N) is, in fact, northeast; south (S) is southwest and southeast; west (W) is northwest; and east (E) is southeast. All maps and drawings, however, will show true north.
All measurements have been assessed from drawings and will be given in meters unless otherwise stated. When more than one measurement is given it will be in length x width x height unless otherwise stated.
7
2.1. The King’s terrace
The King’s terrace, which here refers to the entire excavated area in Area C, includes the relatively large building of House K; a big courtyard; a cuniculus, that is a water channel; a street; a cave with a connected second cuniculus; a second, possibly prehistoric, terrace; as well as a possible city or defence wall. The dates of the different features seem to differ, some features possibly being prehistoric, while others might be medieval.
When the excavation first began in 1961, the aim was to closer investigate a series of rectangular tuff blocks that were visible on the S slope of the Acropolis, below a wall of a later date (Figs. 1–2). As the excavation progressed in 1962–1963 and the building was uncovered, it became apparent that underneath the later wall, which was by Lindgren dubbed the “secondary wall”, was a rubble collapse, visibly containing broken tuff blocks (Figs. 5–6).29 The collapse covered the span of Rooms A–C of House K.
29 Lindgren field journal I 1962, 16, 22.
Figure 4. Plan of The King’s terrace with relevant features marked out. (1). North room. (2). The oven. (3).
Room A. (4). Room B. (5). Room C. (6). The cantina/cellar. (7). Room D. (8). Room E. (9). Room F. (10).
Cuniculus E. (11). The courtyard. (12). C1. (13). C2. (14). C3. (15). Drain 1. (16). Drain 2. (17). The street. The
Figure 5. The so-called secondary wall. The upper course of W05 at the bottom of the photo. N is to the right.
Courtesy of the Swedish Institute in Rome.
Figure 6. The rubble collapse on the lower terrace, covering Rooms A–C, after removal of the secondary wall. N
is towards the top. Courtesy of the Swedish Institute in Rome.
2.2. House K
House K is L-shaped and divided into an upper terrace and a lower terrace made up of a total of seven rooms (Fig. 4). On the upper level is the North room, as well as what will here be referred to as a “courtyard”, while the lower level consists of six spaces or rooms: A–F. The lower terrace measures c. 18x4 m, while the entire building complex measures roughly 18x13
9
m. Right above and on the E side of Room F is cuniculus E (not to be confused with cuniculus W connected to the cave). The lower terrace has been shaped by cutting the tuff bedrock vertically and creating a wall to the N in the rooms, whilst simultaneously quarrying ashlar blocks for the construction of walls.
The building complex has ten built walls of blocks with at least one full course remaining (W01–07; W09–10; W13), with only W13 being mostly built out of stretchers, as well as one third of W02. In addition to these ten walls, what remains of another five walls, or what will at least here be treated as walls (W11a–b–12; W14–16a–b) in or near the structure will be dealt with, as well as an additional wall (W08) which is also almost entirely made of stretchers and that frames in the “street” going down the complex. The roof was made out of terracotta tiles, which will not be included in this study.
Each wall will first be described below (2.3.), as they often reach across several rooms, before presenting and describing each room and structural feature and their different architectural elements, whilst orientating and putting them in context to each other and the complex as a whole (2.4.–2.13.).30
30 The order of presenting the architectural remains follows the traditional method of Swedish archaeological publications, beginning with the publication Labraunda I:2.
Figure 7. Plan of the King’s terrace with all relevant walls, W01–W16b, marked out. Courtesy of the Swedish
2.3. Walls
Under this heading, all walls, counting transverse and partition walls, will be presented, including where they are located, how they are constructed and their relationships to each other as well as other contexts (Fig. 7). The walls are overall made with cut stone, or ashlar masonry, meaning regularly cut rectangular blocks. All the walls, except W09 in all probability, are built using dry masonry, meaning they were built completely without mortar or any type of bonding. The walls will be presented in order of name and not location.
2.3.1. W01
Wall 01, called “the south wall” and “the outer wall” by Lindgren, is a W–E going wall, 18.50 m long, and frames in the lower end of the terrace (Figs. 8–9, 18, 25, 29, 40–41, 43–44, 46,
48).31 The wall is substantial, measuring up to 2 m at its widest point and averaging 1 m in width.
The character of W01 changes quite drastically two times; on its W and E ends. It also becomes hard to distinguish it from W02 to the W and W07 to the E at these points. To the W, the wall begins with at least two rows of blocks, one inner and one outer row. The inner row consists of headers, with the first being placed within W02. The top course is made up of four headers, and the second course has at least six distinguishable blocks. The top course of the outer row consists firstly of one stretcher measuring 1.25 m in length and being the last block in W02, followed by a header measuring 0.40 m in width, making a clear distinction between W01 and W02 at least on the outer row of W01. After this, the wall gets a somewhat cluttered look. Around 1.75 m from the W edge, the outer row has a cavity measuring 0.75x0.55 m, whilst the inner row stays intact (Fig. 10). The cavity is followed by 2 m of what at first glance seems like a rubble wall. Closer inspection makes it evident, however, that it is a section of four headers that have been broken and damaged. It is also at the end of this section that the inner row merges into the outer row, leaving a single row of two–three courses of headers for almost the entire rest of the wall, which measures about 1 m in width. The upper course has c. 17 headers and the lower course is placed more or less parallel to the upper course, giving it an estimated number of 21 headers.
31 Lindgren field journal I 1962, 12; Lindgren field journal III 1963, 40.
Figure 8. The SW corner of House K, including
Rooms A–C. N is towards the upper right corner. Courtesy of the Swedish Institute in Rome.
Figure 9. The SW corner of House K, including Rooms
A–B. N is towards the top. Courtesy of the Swedish Institute in Rome.
11
When W01 reaches the end of Room C it is cut by a transverse wall (W05) going in from the outer edge of W01 up to the cliff wall (Figs. 40–41). W05 is made out of headers the entire way, meaning the first two blocks of W05 become stretchers in W01.
Just before the end of Room E, c. 13.50 m from the W end of the wall, the wall retains only one course of headers. The cavity only lasts until the wall parallels W15 at the W end of Room F, where the second course remains in situ, and the character of the wall changes completely again.
Where Room F begins, c. 14.50 m from the W edge of W01, and 3.50 m from its E edge, the wall goes from being constructed by one row of ashlar headers to being suddenly extended into two rows, the outer part made up of stretchers and the inner part a mix of
headers and stretchers. W01 at this point becomes 1.25–1.50 m wide. At the end of the wall, the final block being from a W–E perspective a stretcher, becomes a header and the first block of a transverse wall (W07) looking from a N–S perspective (Fig. 49). The inner row of W01 is made up of six headers, with an additional possibly broken block stacked in front of the fourth block counting from the west. The outer row consists of four stretchers, with the first block broken, possibly in half, though it is hard to say as the other part is missing.
2.3.2. W02
Wall 02 is c. 13.30 m long, oriented N–S and was called “transverse wall 2” by the King (Figs.
11–13, 20, 25).32 The wall constitutes the building’s W frame, and lies parallel to W08, with the two walls together framing the street. The wall is located on the slope, which results in different levels and courses throughout the wall. W02 is the only wall of House K that spans
32 Gustav VI Adolf field journal 1961, 20.
Figure 10. The cavity in W01 viewed from the S. Courtesy of
the Swedish Institute in Rome.
both the upper and lower terraces. The elevation differs with 2.79 m between the N and S ends of the upper terrace, meaning there is a steady incline, up until the vertically cut cliff wall which forms the N end of the lower terrace. After this, the bedrock floor level of Room A lowers the elevation by an additional 1 m. Because of this, W02 is tallest after the cut cliff wall, with five surviving courses. However, there seems to be an elevation difference between the W and E sides of the wall as well. Not much of the street is excavated down to bedrock, and against W02 only a total of c. 2 m in length has been excavated to bedrock, meaning there may be more courses and parts of the wall that are not visible or known. To add to this argument is the fact that towards the wall’s S end there is 0.90 m excavated on its W side to the bedrock. Only three courses are here visible and counted for. However, on the wall’s E side, in Room A, five courses are visible, meaning that the bedrock floor is cut differently, 1 m lower towards the room. Section H covers 7.30 m of the wall, and begins 5.35 m from the wall’s N end, and ends c. 0.60 m from its S end.
The first 4.25 m of the wall, from the N, is made out of five stretchers, the longest being 1 m, the shortest 0.45 m, and with a continuous width of 0.50 m. The next 7.40 m is built out of headers in varying numbers of courses. The headers measure 0.75–0.95 m in length and 0.50 m in width on average. Since large parts of the W side of the wall are unexcavated, any discussion about the exact number of courses for most parts of the wall would be speculation. Only at two places can it be said with certainty how many courses the wall has. The first is 5.75 m from the N end, an excavated gap of 1.20 m in the street reveals four courses at this point in the wall, with the uppermost course ending 0.80 m from the N edge of the excavated area, leaving three courses left for the remaining 0.40 m. The height of the wall is here c. 1.45 m for the part with four courses, and c. 1.24 m for the rest. At the wall’s 10.75–11.70 m point from the N end, there is another excavated area measuring c. 1.20 m. Here, the wall has three courses, with a height of 0.65–0.90 m. However, as mentioned above, the inside of the wall is at this point five courses, giving the inside of the wall a different height than the outside. The average height of each block in these two excavated
areas of the wall’s W side measure c. 0.40 m. For the most part, the blocks have very similar dimensions, with a few exceptions, including the top course at the wall’s 5.35 m point, which measure
c. 0.15–0.25 m in height.
On the inside of the wall along the North room, the wall has two courses resting on bedrock, higher in the N of the room. The blocks of the second course become larger the further S they are in the
room, seemingly evening out the bedrock floor and wall. Overall the bedrock seems to be higher to the E of the wall on the upper terrace, as the street seems to be cut lower.
Several walls are connected with W02, all of them in a W–E orientation, protruding from
Figure 12. W02 seen from the W. Courtesy of the Swedish Institute
in Rome.
Figure 13. W02 seen from the NW. Courtesy of the Swedish Institute in
13
W02’s E side. Between 3.25 m and 4 m from the N end is W03, followed by W04 at 6.50–7.50 m (Figs. 20, 33). Then comes the lower terrace and W09 at 9.25–9.50 m, and finally W01 at 11.50 m (Figs. 8–9). At these points the character of W02 changes: at W03 the wall goes from stretchers to headers; at W04 the headers become smaller where W04 is built; by W09 there are remnants of some sort of collapse, and finally, by W01 the wall changes completely. The S end of W02 merges with the W end of W01. Approximately 0.60–1.75 m from the S edge the first header of W01 is placed, becoming a stretcher in W02. This is followed by the final header, which then becomes a stretcher in W01.
The height of the wall at section F is 1.90 m at the cut by the profile, and 2.35 m further to the N not cut by the profile but visible by the section.
2.3.3. W03
Wall 03 is oriented W–E and consists of one full course of headers, with an additional two stretchers on its W end and two protruding parts with stretchers on its E half (Figs. 14, 33). The first course contains ten headers, measuring 4.75 m in length, and 0.75–1 m in width. The fifth block from the W lies in connection with a smaller stretcher to its S. The header measures 0.70x0.40 m, while the smaller stretcher measures 0.60x0.30 m. The first stretcher in the W end of the wall measures 1x0.50 m, while the second stretcher is smaller and broken, measuring c. 0.75x0.45 m. The stretchers that comprise the second course cover 1.25 m of the first course, although the first stretcher block is placed further into W02. Half of it is within W02 and the other half above the first course of W03. Between blocks six–eight the first protrusion occurs, containing two stretchers and an additional smaller block to the W of the first stretcher. The protrusion is 0.80 m long, to the S of W03, the first block measuring 0.95x0.50 m, the second 0.75x0.30 m and the smaller block to the W of the first block c. 0.35x0.30 m. The second protrusion is right at the E end of the wall and was either at one point connected to the N protrusion at the E end of W04, or it ended in a doorway, which is the more likely possibility. The protrusion consists of two stretchers measuring 0.85x0.50 and 0.65x0.40 m, totalling 1 m out to the S of W03. At section B, the height of the wall measures 0.60 m.
2.3.4. W04
Wall 04 is just like W03 oriented W–E and consists of nine headers as well as two additional
blocks (Figs. 15, 33). The full wall measures c. 4.35 m in length, with an average width of c. 1 m. The wall consists of two courses, possibly three at the W end according to section drawings, but most likely two courses at all times according to photos. To the N of the fourth block from the W, there lies another block, seemingly a stretcher broken in two. Together, both pieces of this block measure 0.60x0.30 m. The
last header to the E of the wall is much smaller than the rest of the headers and is topped to its N by an additional smaller block, the header measuring 0.75x0.40 m and the smaller block 0.50x0.50 m. There is 1.30 m between this block and the last stretcher protruding from the most E end of W03. As stated above, this gap probably created a doorway into the North room. At section B, the height of the wall measures 0.90 m. The wall is located c. 1.80 m N of W09, which also delimits where the upper terrace ends and the lower begins.
2.3.5. W05
Wall 05 is the first N–S going transverse wall on the lower terrace, with W01 to the S, and on its N side it is connected to a shorter wall (W13) built up against the cliffside (Figs. 16, 28, 40–41). The wall, which was referred to as “transverse wall 3” by the King, measures 3.25 m in length, 1.45 m in height and consists of four courses of headers measuring c. 0.90 m in length on average with a width of c. 0.50 m.33 The top course has two intact blocks, with a third smaller broken block, all measuring from the N: 0.90x0.45, 0.90x0.45 and 0.50x0.30 m. The second course has four blocks still in situ, with the last one from the N measuring 0.90x0.50 m. The third course has at least six distinguishable blocks, but has presumably at least seven, and the bottom course has five visible, but must have more. The fourth course, and most likely the fifth as well, go all the way out to W01, meaning they are built into W01, with the two last headers of W05 becoming stretchers in W01. The two blocks of the fourth course measure from the N: 0.85x0.45 and 0.85x0.50 m, measured from their widest points. On the W side of the N end of the wall, there are some additional blocks, seemingly broken and fallen out of place. Lindgren also mentioned that part of the bedrock wall above W05 had collapsed, which is partly visible in Fig. 16 by a crack.34 On floor level a block is placed with its
33 Gustav VI Adolf field journal 1961, 20. 34 Lindgren field journal III 1963, 30.
Figure 16. The N end of W05 seen from the
W. Courtesy of the Swedish Institute in Rome.
Figure 15. W04 seen from the S. Courtesy of the Swedish Institute in
15
short end towards the cliff wall, making it a header when connected to W13, but a stretcher when comparing to the blocks of W05.
2.3.6. W06
Wall 06, referred to as “transverse wall 4” by the King, is the second N–S going transverse wall on the lower terrace and lies, just like W05, between W01 to the S and W13 to the N (Figs. 17,
29, 41).35 However, unlike W05, this wall is not built into W01 but is limited by it. The wall consists of two complete courses and a third assumed one, which has not survived in full. The wall is made up of headers, with a width of 1 m at its widest point and 0.75 m at its narrowest point, and a length of approximately 2.50 m between W01 and W13. The height of the wall is
c. 1.45 m. The first two courses from the bottom each have five blocks, while the top course
only has two surviving blocks in situ at its N end, of which the first block from the N has been broken in two, and the second block damaged. The blocks measure from the N: 0.90x0.35 and 0.85x0.45 m. The blocks of the bottom two courses are generally bigger than the two blocks of the third course. The last three blocks to the S of the second course measure from the N: 0.85x0.50, 0.85x0.45 and 0.70x0.35 m.
2.3.7. W07
Wall 07 is oriented N–S and is more complicated than W05 and W06 as it is hard to distinguish where W01 ends and W07 begins (Figs. 18, 46, 49). On its E side, W07 is built against a rock outcrop, creating a perimeter to the E of the lower terrace structure. Including the blocks at the end of W01, W07 measures 2.50 m to the horizontal bedrock bench cut out of the cliff wall to the N. The wall’s width is 0.75 m, except for the block right after the bench which measures 1 m. This block is also placed further into the room, with the rock outcrop less cut at this point, and the block placed on a higher bedrock foundation. The wall consists of one–two courses, the upper row consisting of five headers, and the lower an unknown number, if any at all. It is uncertain how many blocks from the upper course are resting on bedrock and how many are resting on a second course. The blocks of the top, or only, course measure from the N, including
35 Gustav VI Adolf field journal 1961, 20.
Figure 17. W06 seen from the W in Room D. A possible dolium stand in front. Courtesy of the Swedish
the block closest to the bench: 1.05x0.45x0.35, 0.75x0.47, 0.80x0.40, 0.86x0.45 and 0.95x0.60 m.
2.3.8. W08
Wall 08 is long and somewhat hard to define as many blocks have apparently fallen and moved from their original position (Figs. 19–20). The wall was referred to as “transverse wall 1” by the King.36 W08 follows the bedrock edge to the W – which has been cut down to create the street – and turns W at its S end where there is a lot of rubble from the wall and/or a possible second structure. Lindgren states that the S part of the wall was sloppily and carelessly built, and later, she seemed to think the S part of W08 had collapsed or been dislocated.37 The wall is oriented N–S–SW and
is completely parallel to W02 until its S end. The wall measures c. 14.50 m in length, including the curve to the W. In width the wall measures 0.40– 0.60 m throughout. The blocks of the wall differ in size but are mostly stretchers, squares or irregularly shaped. Either the wall has been heavily damaged, or it was much simpler built
than the rest of the walls, judging simply by the top of the wall. The wall does not seem to
36 Gustav VI Adolf field journal 1961, 20. 37 Lindgren field journal III 1963, 20, 44–46.
Figure 18. W07 to the left, and the E end of W01, seen from the N in Room F. Courtesy of the Swedish
Institute in Rome.
Figure 19. W08 to the left, seen from the SE. Stray block of white limestone in the
17
belong to any type of bigger construction, making it likely that its purpose was to frame in the W edge of the street. There are two sections of W08, section G and F. Section G (Fig. 52) is 5.75 m from the wall’s N end and shows W08 built by at least three courses, measuring in width and height from top to bottom: 0.38x0.25, 0.59x0.50 and 0.42x0.25 m, giving the wall a total height of 1 m. At section F (Fig. 32) the wall has only one course, with one block measuring
c. 0.60x0.60x0.20 m. The E side of
the wall, that is the street, was only excavated about 5.75 m along the wall, as can be seen in the plan (Fig.
7). However, there was an additional
part excavated to the N, but only pictures document this excavated part. Whether or not it was excavated to the bedrock is not known, but it does show the wall with at least four
courses, the three lower courses being made up of bigger blocks which are assumedly stretchers. Section G gives the impression that W08 was built to match the height of W02. Since W08 is to larger parts built on the tuff cliff, however, the wall itself did not need to be built so high but instead used the cliff as support.
2.3.9. W09
Wall 09, which Lindgren referred to as the “wall on the cliff”, is built on the W top edge of the cliff wall (Figs. 21–22, 25).38 The cliff wall is at this point uneven and angled down, and W09 levels the cliff wall out. W09 seems bonded to the cliff wall, which it almost certainly must be, as it would otherwise most likely not have held in place. The wall is oriented W–E and begins
c. 0.72 m from the bedrock ground and measures c. 2.60 m in length at its longest part. The
wall is made of irregularly shaped blocks, that will here be divided into five courses. All blocks are counted from the W and measured in length and height. From the bottom, the first course contains three blocks, measuring: 0.70x0.30, 0.10x0.10 and 0.10x0.10 m, with the last two blocks being one on top of the other. The second course consists of two blocks, measuring 0.60x0.32 and 0.40x0.22 m. The third course is preceded by rubble-like smaller irregular stones, which also precede the remaining courses. The course consists of two blocks, measuring 0.50x0.20 and 0.55x0.15 m. The fourth course consists of two flatter blocks, measuring 0.19x0.09 and 0.50x0.11 m. The final course consists of five blocks, including here one rounder boulder-looking block at the W edge of the rubble, right at the corner of the cliff wall and W02. The boulder measures 0.40x0.32 m, followed by the second block 0.18 m to the E of the first. The second–fifth blocks measure: 0.60x0.35, 0.15x0.35, 0.48x0.40 and 0.65x0.35 m. However, it should be said that the boulder is most likely not a part of the wall but rather some other structure that has fallen, probably a wall on the cliff edge. Lindgren seemed to think that W09 is what is left of a bigger wall placed on the top edge of the cliff wall which had collapsed, as can be seen from the fallen blocks in Room A (Fig. 23).39 She went on to state that the wall could not have been so strong or sturdy, as it completely collapsed.
38 Lindgren field journal III 1963, 32. 39 Lindgren field journal III 1963, 24–26.
Figure 20. The street with W08 to the left and W02 on the right.
Figure 21. Top of W09 seen from the W.
Courtesy of the Swedish Institute in Rome.
Figure 22. W09 seen from the S. Courtesy of the
Swedish Institute in Rome.
19
2.3.10. W10
Only one course survives of wall 10, which is N–S going and forms a partition between Rooms A and B, with an additional block stacked at the N end of the wall (Figs. 24–25). W10 is limited by W01 to the S and the bedrock wall to the N, and is situated to the E, or inside, of W02. The length of the wall is 2.10 m. The wall consists of four bigger blocks alternating with three small ones, as well as an additional bigger block in the N end. The first block, furthest to the S, measures c. 0.50x0.35 m; the second block measures 0.15x0.35 m; the third block is c. 0.40x0.35x0.25 m; fourth block measures 0.05x0.30 m and the fifth block measures 0.40x0.35 m. The final smaller block measures 0.30x0.20 m, followed by two blocks stacked one on top of the other. The bottom one measures c. 0.40x0.50x0.30 m and the top one 0.30x0.48x0.35 m.
2.3.11. W11a–b
Wall 11 is here divided into two parts, both oriented N–S, and forming two ends of a line dividing Room B from C, most likely creating a doorway between them (Fig. 25). The N part, which is here called W11a, consists of two blocks and is placed against the cliff wall. The blocks measure, from the bottom up, 1x0.40x0.38 m and 0.95x0.45x0.30 m. The blocks are placed partly over the built wall in the cellar: the E half of the bottom block covers at least 0.20 m of the top block in the cellar.
Wall 11b is placed 1 m S of W11a and is broken into two blocks but will still be counted as one. The visible part of it measures approximately 0.80x0.35 m.
2.3.12. W12
Wall 12 protrudes out from the outside of W01, to the S, in a straight line with both W11a–b to its N, albeit lower in elevation (Figs. 25–26). The wall consists of one course of three blocks, with an additional block in connection to the SE, probably having been in line with the other blocks but disturbed in one way or another at some point. The first block, from the N, measures
c. 0.34x0.50x0.33 m; the second block is c. 0.40x0.50x0.40 m; the third block 0.48x0.48x0.34
m; and the final block measures 0.30x0.50x0.20 m. The first block is set against the bottom-most block of W01, which is not visible from inside the structure in Room C. Lindgren discusses the possibility of W12 being a border for the passageway or street, which is examined below.40
40 Lindgren field journal III 1963, 54–56.
Figure 24. Room A bordered by W01 to the left, followed by W02, then W09 and the
cliff wall to the right, and W10 at the bottom of the picture. Seen from the E. Courtesy of the Swedish Institute in Rome.
2.3.13. W13
Wall 13 is W–E going and begins at the NW edge of W05, and extends E beyond Room D into Room E, behind W06 and ending where Room F beings (Fig. 27–29). The entire wall measures 5.30 m in length, around 0.50 m in width on average, and 1.30–1.45 in height.
In Room D the wall is somewhat uneven, and consists of four courses, the bottom one being made up of only two blocks, one smaller and one larger, the first one placed c. 0.90 m from W05. The first block measures in length and height c. 0.50x0.20 m and the second block measures 0.90x0.65x0.35 m. The second course contains four visible blocks between W05 and W06, measuring in length and height from the W most block to the E most within Room D: 0.30x0.45, 0.65x0.46, 0.60x0.40 and 0.80x0.40 m. The third course is made up of six blocks, the second being very small and stacked upon the third block. There is also a block (W14) sticking out from the wall between the third and fourth blocks, making it impossible to tell if the third block covers the entire span up to the fourth block or if there is another hidden block in between. The blocks measure in length and height as follows: 0.30x0.45, 0.35x0.10, 0.32x0.40, 0.35x0.38, 0.55x0.35 and 0.32x0.35 m. The fifth and final course has five almost completely visible blocks, measuring from the W: 0.90x0.45x0.50, 0.20x0.35x0.40, 0.80x0.50x0.40, 0.30x0.50x0.35 and 0.58x0.50x0.35 m. The fifth block on the top course is partly behind W06. Right at the edge of where W06 begins, there is a rectangular cut in the block, measuring c. 0.35x0.15 m (Fig. 41). The cut is very intentional and the only of its kind on W13. The wall measures 2.30 m within Room D.
The character of the wall changes from Room D to E. The visible part of the wall in Room E is made up of three courses of stretchers. Each course consists of two blocks, with the bottom course being parallel to the third. The first, or W, block of the second course is longer than the second block and spans the length of the entire visible part of the first block and 0.20 m of the second block of the bottom course. The blocks measure in length and height as follows: 0.70x0.50 and 0.80x0.45 m for the bottom course; 0.90x0.40 m and 0.80x0.50 m for the second course; and 0.72x0.42 m and 0.95x0.45 m for the third course. The wall measures 1.70 m in
Figure 25. The W part of the lower terrace, including Rooms A–
B and most of Room C. N is towards the top of the picture. Courtesy of the Swedish Institute in Rome.
Figure 26. W12 seen from the E. Courtesy
21
length within the room.
It is important to keep in mind however that all these measurements are of the visible parts of the blocks, leaving it likely that several of the blocks are much larger in reality than could be ascertained from drawings and photos.
2.3.14. W14
Wall 14 is not really a wall, but rather two blocks, or possibly one that has been broken, protruding from W13 in Room D (Figs. 28, 40–41, 43). The wall rests on earth and seems intentionally placed where it is, meaning it most likely belongs to a later period of the building’s lifespan. The bigger and outer block measures 0.95x0.40x0.56 m, whilst the smaller block measures c. 0.35x0.50 m. The blocks lie 0.60 m from the bedrock ground level and c. 0.60 m to the E of W05 and 1.30 m to the W of W06. The tops of the blocks are almost in line with the tops of both W05 and W06.
Figure 27. W13 in Room E, seen from the S. Courtesy of the Swedish Institute in Rome.
Figure 28. W14 in Room D built against W13. Seen from the SE. Behind W14 to the left is W05.
2.3.15. W15
Wall 15 is oriented N–S, measures c. 2 m in length and has been partly cut out of the bench in Room F (Figs. 29, 41, 44, 46). The bench makes up 1 m of the wall, though the S tip of the bench has been cut down and a smaller block has been placed on top. The wall further consists of one big block measuring c. 1 meter in length. About 0.20 m of the N part of the block is cut down, and a smaller block is placed in the cavity. The cut part of the bench and the big block thus create the lower course of the wall under the two smaller blocks. The blocks measure from the N: 0.40x0.40x0.25, 0.15x0.30 and 1.10x0.40x0.40 m. The height of the wall is approximately 0.45 m at its tallest point.
2.3.16. W16a–b
Walls 16a–b consist of a few blocks oriented SW–SE, which frame the limestone pavement on the outside of the SW part of W01 (Figs. 25, 56).
W16a consists of what looks in photos as one course of two stretchers. However, when looking at section B it seems that there is an additional smaller block under the first W stretcher. The two stretchers visible in the plan, beginning with the W block, measure approximately: 0.50x0.30 and 0.75x0.32 m. The height of the W part is 0.40 m, with the two different possible blocks measuring from top to bottom c. 0.25 and 0.15 m in height. The first block is located 1.25 m S of W01, while the second block is located c. 1 m S of the gap in W01.
W16b is located c. 0.70 m E of W16a and made up of two pieces that are most likely from the same block. The total measurement is 0.75x0.32x0.40 m, with the lengths of the first part
c. 0.25 m and the second 0.50 m. The block is located 0.75 m S of W01.
Figure 29. W15 seen from the E of Room F. Behind W15 is W06, to the left is W01. Courtesy of the Swedish
23 F igur e 3 0 . S ec ti on A -A 1 -A 2 vi e w ed fr om t h e W . C ourt e sy of t h e S w e di sh Ins ti tu te i n Ro m e.
24 F igur e 3 1 . S ec ti on B -B 1 vi e w e d fr o m t h e W . Co ur te sy of t he S w e di sh In st it ut e i n Ro m e.
25 F igur e 3 2 . S ec ti on F -F 1-F 2 s how ing t he l ow e r te rra c e vi e w e d fro m t he S . Cou rt e sy of t he S w e di sh Ins ti tut e in R om e .
2.4. North room
The North room, which Lindgren referred to as the “upper house”, is the only room on the upper terrace.41 The room has W02 as its border to the W, W03 to the N and W04 to the S (Fig.
33). There is no full wall to the E, but
the E most part of W03 comes down to the S, creating at least part of an E border to the room. The room is located 1.80 m N of the lower terrace, counting from the outside of W04 to the outside of W09. The room measures 4x2.50 m within the walls and is placed on the slope, resulting in quite a difference in elevation between its N and S side. Section B shows that there is 1 m difference in elevation between the N and S end of the room. At about 1 m E of W02 for W03 and 0.50 m for W04, both walls parallel to each other, show that the bottom of W03 is the same elevation as the top of W04, meaning
the bedrock elevation lowers by roughly 1 m between the walls. Section B also shows very dramatic changes in the bedrock floor, which were certainly evened out by higher floor levels. About 0.20 m S of W03, the bedrock floor drops to a depth of 0.80 m, from the bottom of W03 that is, into a hole with a diameter of 0.10–0.40 m, not measuring its full depth. After this, the bedrock floor rises 0.40 m and then drops again 0.70 m, with a final rise of 0.20 m where W04 is built, keeping
relatively even after the wall. At section A the bedrock floor is not quite as dramatic, but has instead a more or less steady incline, lowering 1 m between the bottom outer end of W03 to the bottom outer end of W04.
Large parts of the
room remain
unexcavated, meaning there are a lot of unknowns regarding the room in this analysis.
Against W04, a semi-circular shape was excavated, which was lined by what seemed to be remnants of reddish burnt terracotta. Lindgren and Gezelius interpreted this as an oven (Fig.
34).42 The diameter of the oven is 2.20 m, the top radius 0.82 m measuring to W04, and 0.62 m at its bottom. The depth of the oven measures 0.60 m. The oven is placed 0.50 m E from W02, and 1.50 m from the E end of W04, with its N edge 1.35 S of W03. Section B also suggests that what remains of the oven was under the floor level.
41 Lindgren field journal III 1963, 36. 42 Lindgren field journal III 1963, 58.
Figure 33. The North room, before the oven was excavated.
N is towards the top of the picture. Courtesy of the Swedish Institute in Rome.
Figure 34. The oven in the North room, viewed from the SW. Courtesy of the
27
2.5. Room A
Room A is the first room on the lower terrace and is formed by W01 to the S, W02 to its W, W10 to its E and W09 which evens out the bedrock wall to its N (Figs. 24–25). The room measures 2x1.75 m, with a bedrock floor level measuring c. 2 m below the top of W09. Only the S half of the room is excavated to the bedrock floor and reveals a deep cavity right at the SW corner, going under the visible inner part of W02.
Since the floor level differs quite drastically between the outside and inside of W02, with the inside being lower than the outside, it makes it evident that not only was the bedrock cut vertically to create the lower terrace N wall, but also horizontally to create the floor level. According to section F, the bedrock floor level of Room A is 0.70 m lower than the bedrock floor level of the street to its W. However, the street has a deep cut at this point, with blocks placed in the gap, meaning the difference from the top of the blocks and the bedrock floor level in Room A is at section F 1 m.
2.6. Room B
Room B is the second room on the lower terrace, situated between Room A and C (Fig. 25). To the W Room B is divided from Room A by W10, and to the E W11a–b divide the room from Room C, leaving a probable doorway between the rooms. To the S is W01 and to the N is the bedrock wall. The room measures 3.20x2.50 m, making it one of the biggest rooms on the lower terrace, together with Room F. More of the bedrock wall is cut towards the E end of the room, meaning a difference in size, making the E end c. 0.50 m wider than the W end.
The bedrock wall has several changes in it throughout the room. Right from the W end of the room, the bedrock wall slowly changes its angle, about 10 degrees, being cut deeper towards
Figure 35. The lower terrace viewed from the W.
Courtesy of the Swedish Institute in Rome.
Figure 36. The lower terrace viewed from the E.
the N. About 2.25 m from W10, there is a clear vertical break in the bedrock wall, with an 0.05 m depth added to it along the entire height of the wall. However, at the bottom of the wall, along the bedrock floor, the cut stops 0.02 m above the bedrock floor and instead turns horizontal for 0.90 m, leaving the bottom part of the wall more protruding than the rest. The cut does rise somewhat, and the floor inclines, making the distance between them 0.15 m, before the cut turns vertical again 0.10 m W of W11a, and ends in the bedrock floor. Around this point, 0.35 m above the bedrock floor, 0.20 m W of W11a, 0.60 m E of the edge of the deeper cut in the bedrock wall, and 0.75 m below the top of the bedrock wall, is another cut in the bedrock wall. The second cut measures 0.80x0.10 m and has an oblong shape.
Room B is largely unexcavated down to bedrock floor. There are three blocks that have not been moved by excavators, all of them resting on earth, with at least two of them possibly being strays from the rubble collapse. The first block from the N lies right against the bedrock wall, 0.35 m W of W11a, and measures 0.45x0.25 m. The other two blocks lie close to each other, 1 m S of the first block, and measure 0.40x0.45x0.30 and 0.35x0.25 m.
The bedrock floor level of the room is roughly 1.70 m lower measured from the top of the bedrock wall.
2.7. Room C
Room C is the third room on the lower terrace, bordered by Room B to the W and Room D to the E (Figs. 37, 40). The room is made up of W05 as its E limit, W01 to the S, the bedrock wall to its N and W11a–b dividing the room from Room B to the W. The room measures approximately 2.30x2.80 m.
29
At the N end of the room is a cellar, which Lindgren called the “cantina” (Figs. 38–39).43 The cellar measures 2x1 m in plan. At its W end, the cellar has a built wall of at least three courses of blocks built against the bedrock wall below the ground level of the room. The measurements of the blocks at section F, from bottom to top in width and height, follow: 0.50x0.40 m, 0.50x0.32 m and 0.50x0.42 m. About 0.20 m of the top block is under W11a. The cellar itself measures, including the cellar wall: 2.30 m in length at its top and 2.20 m at its bottom, with a depth of 1.25 m. There is also a circular indentation at the bottom of the cellar, right by the N end against the bedrock wall, 0.90 m from the cellar wall to the W, and 0.30 m from the cellar’s E end. The indentation has a diameter of 0.45 m,
and a depth of 0.10 m. Lindgren believed the circular indentation was made for a large vessel. The depth of the cellar, from its top to the bottom of the indentation, is thus 1.35 m. On its E top end, there is a block going out horizontally over the cellar. The block sticks out 0.30x0.45 m above the top of the cellar, with a height of 0.40 m, and is situated 0.45 m S of the bedrock wall, and 0.25 m N of the S end of the top of the cellar. The E top end of the cellar is according to section F, cut to have an edge, 0.10 m wide, before the bedrock floor is cut down 0.10 m more for the continuing floor level, where W05 is placed.
On the bedrock wall, there are three cut cavities. The first one has a height of 0.50 m, with the width at its bottom being 0.45 m. The cavity has a straight horizontal bottom but goes into a semi-dome shape at its top. The depth at the bottom is 0.10 m, while the top is slightly shallower with 0.05 m. The other two cuts measure from the W 0.40x0.23x0.15 and 0.40x0.20 m and are placed 0.30–0.35 m from each other. The first of these is 0.50 m E of W11a, while the second is 0.45 m W of the E edge of the cellar. They are both placed 1.10 m from the bottom
43 Lindgren field journal III 1963, 56.
Figure 39. The cantina/cellar in Room C. N is towards the top of the picture. Courtesy of
the Swedish Institute in Rome.
Figure 38. Plan of the cantina/cellar in Room C. Courtesy of
of the cellar, and roughly 1.60 m from the top of the bedrock wall. Parallel to these, on the S end of the cellar, are an additional two cut cavities in the bedrock floor, matching the ones in the wall. The cavities on the bedrock floor measure from the W: 0.30x0.40x0.12 and 0.25x0.35 m. The cavities in the bedrock wall and floor were in all probability made for wooden boards or planks, likely creating a hatch to the cellar. To the E of the second floor-cavity, c. 0.15 m, and 0.15 m W of the E end of the top of cellar, is a notch with a diameter of 0.05 m, which Lindgren referred to as a mortice or “taphole”.
On the SE side of the room, there are several other cuts on the bedrock floor. The first cut is 0.50 m S of the S cellar edge, and 0.50 m W of W05. The N end of the cut has a semi-circular shape, followed by a straight line measuring 0.90 m, going SE under W05, 1.50 m S of the wall’s N end. By the wall the cut widens, going from 0.05 m to 0.25 m. The cut is followed by a straight N–S line of three indentations, approximately 0.75 m in total length. The first most N indentation, as well as the last most S indentation, are both oblong and measure roughly 0.25x0.05 m. The middle indentation is smaller and more notch-like, measuring 0.10x0.05 m. To its W is a bigger cut, paralleling it, measuring 0.20x0.10 m. The three indentations in the straight line are placed 0.15 m from each other, while the bigger cut is placed 0.05 m from the middle notch. To the S, at the SE corner of the room, the last indentation opens into a deeper and bigger cut area, seemingly going under both W01 and W05, measuring at least 0.90x0.25 m.
2.8. Room D
Room D is the fourth room on the lower terrace, limited by W13 to its N, W05 to the W, W06 to the E and W01 to the S (Figs. 40–43). The room measures 2.20x2.30 m. In the room’s NW half, W14 protrudes from W13, but only goes about halfway towards W01. The difference in elevation between the top of the bedrock wall and the bedrock floor in Room D is at section D 1.80–1.90 m, with the difference between the top of W13 and the top of the bedrock wall being
c. 0.30 m. There are also several cuts above the cliff wall. One cut is just at the edge of the cliff
wall, parallel to the E end of W06. On top of W13 there is a similar cut, but parallel to the W end of W06. The cuts are in all likelihood traces of the upper structure of House K.
On the floor level, there are several cuts in the bedrock floor. At the N end of the room, between W05 and W06, the floor is cut lower against the bedrock wall. The cut is 0.75 m wide with a depth of 0.25 m at section D, showing the floor was cut to fit W13. On the room’s S end,
c. 2 m S of W13 and 0.30 m N of W01, the floor is again cut lower, this time c. 0.20 m, parallel
Figure 40. Room C–D. N is towards the top of the
picture. Courtesy of the Swedish Institute in Rome.
Figure 41. Room D–E. N is towards the top of the
