• 1 I wish to thank Nicolas Laubry, Maria Letizia Caldelli, Fausto Zevi and Alessandro D’Alessio for or (...)

1In Ostia, shortly before the outbreak of the First World War, the excavation director Dante Vaglieri and his assistant Raffaele Finelli started a new era of scientific work. Dante Vaglieri organised excavations according to the highest standard available at the time, and for the first time, photography also played an important role. In addition, Vaglieri almost immediately published the finds that had come to light from various parts of the city with the help of his excavation assistant Raffaele Finelli. However, excavating a site which had never been forgotten since antiquity can be difficult. Especially the area of the Forum had been ransacked for centuries, mostly unsystematically. Finelli and Vaglieri were aware of the disastrous condition of the Forum after countless undocumented excavations as they found numerous unsorted stone piles, ranging from large chunks to intentionally crushed marble material that had already undergone late antique or medieval 'recycling' processes. Only a few marble paving slabs are seen in situ, rising far above the imperial basalt road. The few, but mostly very high-resolution overview photos form a basis and starting point for a – unfortunately often futile – search for the ancient original state.1

• 2 Photo Ostia A 2421
• 3 Ostia, Archivio dei disegni, “Ostia Colonia Romana. Pianta generale”, Banc. B, Cass. 4, Cart. 1, Nr (...)
• 4 Photo Ostia A 2421.

2For this reason, in 1913, it seems to have become important for Vaglieri to undertake a deep excavation at the Forum down to virgin soil in order to find traces of Ostia’s origins. The planned excavation area was located immediately east of the central intersection of the Cardo and Decumanus, the main axes of the Roman city plan. This part of the Forum had demonstrably never been touched by excavators before as a marble pavement had been almost completely preserved here. For the excavation, the marble slabs of the north-eastern portico of the Forum had to be removed, however without being drawn in a plan. An old photograph taken to depict the eastern side wall of the Capitol is the only coincidental evidence of their former state of preservation.2 The ink plan of the Portico shows only one single row of elongated paving slabs which remained in the north-east of this portico in situ.3 The rest of the ancient marble paving material visible today was relayed and newly arranged after the 1913 excavations and, perhaps intentionally, rotated by 90°.4

• 5 Ceccarelli – Marroni 2011, p. 275; Gering 2023, p. 55. 57-64; Van Haeperen 2019, p. 228.
• 6 Compare GdS 1913.

3Finelli's reports on these excavations below the marble slabs reflect the opinion of the time, which was shared by the director Vaglieri himself, that Ostia’s main temple, which had been visible since Antiquity continuously, was not a 'Capitolium' but the temple of Vulcan. This hypothesis was never questioned as the basis of the epigraphic evidence for Vulcan was overwhelming - in contrast to the (almost) complete lack of evidence for Jupiter or a Capitolium.5 Thus, the north-east portico of the Hadrianic Forum, here abbreviated MFE (“Main Forum East”, fig. 1), was named “area sacra di Vulcano” in the diaries of the excavation.6 The structures uncovered during the 1913 excavations in MFE were carefully described by Finelli. However, finds were listed separately from the structures: thus, their stratigraphical context remained unclear since Finelli did not publish any plans, sections or stratigraphical sketches mentioning absolute heights, neither of the structures nor the finds.

• 7 See below.

4Finelli himself did not yet recognise or discuss the possibly sacred function of the structures he and his workers had uncovered and spoke very generally of “Capanne” (huts) as well as of tabernae (shops) next to them, which had been made of tufa blocks of various dimensions and types. Some of Finelli’s descriptions suggest, apart from the tufa fundaments, the use of earthen architecture, as well.7

• 8 For the definition of TFR-Sanctuary see Gering – Menge – Pedersen 2022, p. 223-230, fig. 2. 26-29; (...)
• 9 Gering – Menge – Pedersen 2022.
• 10 Gering 2023.
• 11 For a first overview of georadar-results in this area compare Gering 2022.

7The following analysis and interpretation concentrate on the area underneath the portico MFE, which was so far not included in the former articles concerning the so-called TFR-Sanctuary.8 The first substantial publication about this newly found sacred area focussed on the remains of a “recently discovered Republican Altar”.9 Another already published article serves as an overview of the grid-systems of Ostia’s castrum and the evolution of Ostia’s Forum area including the TFR-Sanctuary and the adjoining Jupiter-temples underneath the later Capitolium.10 However, in these articles there was not enough space to concentrate on the main feature of the TFR-Sanctuary itself – the area of massive tufa and concrete fundaments underneath the later Hadrianic Portico MFE which were never fully excavated nor sufficiently documented and published so far.11

• 12 Mar 1991. For an excellent overview of the status quo of the research see Zevi 2012.

8In 1913, the area of the TFR-Sanctuary was first excavated below the portico level. The excavators at that time documented their finds and findings in an unusually detailed manner (Giornale dello Scavo or degli scavi, in short: GdS 1913). These excavation reports were, compared with all the earlier and later excavation diaries from 1907 to 1924, the most detailed record of an excavation in Ostia from Finelli's pen overall. However, Vaglieri’s sudden death in 1913 and the outbreak of the First World War prevented among other facts the final interpretation and publication of these excavations. Thus, these unfinished excavations and their preliminary documentation in the state of 1913 have hardly been consulted or received by researchers up to now including the most relevant studies on this topic.12

• 13 Calza et al. 1953, p. 74.
• 14 Ibid.

9Even though the extraordinary importance of the finds and built structures was recognized by Finelli and Vaglieri himself, only a small paragraph was summarised in the fundamental monograph on the excavations of Ostia (Scavi di Ostia) by his successor Guida Calza and Giovanni Becatti from 1953.13 The only plan of the area MFE published at that time simplifies important details like the different sizes and building techniques of the walls.14 The explanatory text is limited to the statement that there were five rooms (a-d), all of them tabernae (shops) bordering the Cardo maximus. However, just by looking at the simplified plan from 1953, it is already obvious that the so-called room c and d could never have been shops either because of their size or their missing entrance. Additionally, two of these tabernae walls were built from much smaller and more irregularly shaped stone blocks clearly forming an earlier structure from a deeper level than the large stone-blocks of the above-mentioned Taberna-walls (see below).

• 15 Gering 2011, p. 476-486.

10In 2010-2012, the Ostia Forum Project (OFP) started a surface excavation to analyse and document the visible structures in the area of the northeast portico (MFE) and the adjoining group of three rooms MFR. First, the marble floor of MFR was excavated, so that a complete survey with a Laser scanner and stone-by-stone recording of the various construction phases of the marble pavement became possible.15 This also included its late antique renovation, decay and modern conservation history. In 2011, the first complete georadar survey of the subsoil of the MFE and MFR areas, which were not covered by marble slabs, was undertaken, in order to map the areas of the backfill of Finelli’s old excavations for future campaigns of minimal-invasive archaeology. In 2012, the surface excavation was extended to the southern area of the portico MFE (fig. 2).

• 16 Many thanks to Gabor Bertok and his team. For an overview about the methods and results see Gering (...)

11In 2016-2018, when the area of the Portico MFE once again became the focus of OFP interest, several campaigns of georadar were conducted with new cooperation partners from the Peter Pazmany University. This time, methods with a higher resolution were employed.16

• 17 In this article, our complete interpretation of the 1913 excavations (GdS) cannot be presented due (...)

12Thus, the new observations, which are proposed here, are based on: a.) georadar images and profiles of the whole Forum area, b.) our excavations in the area immediately to the northeast in 2019 (TFR2) and 2023 (TFR1), c.) some small sondages in Finelli’s backfill in MFE in 2012 und 2021 and d.) archive-studies from 2020 to 2022 in order to combine all topographic informations from old photos, plans and the unpublished excavation-diaries (GdS) in the context of new georeferenced 3D models and plans. Based on this, it was possible to determine the height, location and stratigraphic relation to the level raises of most of the structures found by Finelli in 1913. Apart from two rows of shops underneath the north part of the Portico MFE, which will be treated separately, this allows a new interpretation of the structures Finelli did not fully excavate in 1913 in the south of this portico. There are arguments to understand these structures in their original context, a sacred area, and interpret them due to their massive fundaments as temple podia, built on top of each other, and a corresponding structure perhaps as an altar, in line with the podia.17

• 18 The (later) main river harbour is situated west of the castrum. For a possible first river harbour (...)

13Early Ostia was surrounded by water. Based on geological core drillings, Ferréol Salomon brought up the observation, published in several articles, that the bank of the Tiber in the 4^th to 3^rd centuries BC was closer to the northern gate of Ostia’s first fortified settlement (castrum) than today, more or less in the time when the castrum is believed to have been founded.18 The area to the north of the castrum probably had the character of an indentation of the banks of the Tiber, possibly in the manner of a first river harbour separated and protected from the main traffic axis of the river. This indentation could have served perfectly for the construction of the nearby castrum (which could explain why the castrum had been founded exactly here and not somewhere else). And it could have defined the northern Cardo as the most important traffic axis of early Ostia, while the Decumanus and the Via della Foce connecting to the second river harbour in the west gained more importance only later. However, without further geological core drilling, the position of the first Tiber docks north of the castrum can only be determined very approximately by alterations of the orientation in the later urban street-layout.

• 19 “Sezione stratigrafica dell’urbanistica ostiense”, Ostia Archivio disegni, Banc. B Cass. 6, Cart. 2 (...)
• 20 For the southern Cardo gate compare Calza et al. 1953, p. 70, fig. 20.

14Nevertheless, at least in the course of the Cardo towards the south, i.e. towards the city centre, there are reliable data on the earliest settlement’s ground level: in the area of the Insula di Giove e Ganimede, today located immediately outside the castrum, a test sondage was carried out in 1966, which indicates the height of the ‘sabbia marina’ to be at 0,35 m ASL.19 According to the old excavations, the interior floor of the southern gate of the castrum lies at 0,29 m ASL, but unfortunately the northern gate had never been analysed, because it is still covered by the impressive ruin of the Hadrianic Capitolium.20

• 21 The ancient see-level was about 1 m lower.
• 22 Houses could not be detected underneath the portico MFE. They were probably set behind the shops, e (...)
• 23 Measurements are also given in Roman foot (RF) in the following. It varied between 29,57-29,64 cm.

15The georadar maps inside the castrum (fig. 3-4) show the deepest building structures (walls) along the northern Cardo laying on a level of ca. 0,40-0,70 m ASL with rare deeper foundations (fundaments or sewers) still above the recent see-level.21 These structures, partly already excavated by Finelli in 1913, can be interpreted as a first row of shops along the Cardo front (fig. 10: green shops).22 The corresponding street level cannot be determined yet, since the earliest Cardo still had no basalt pavement, but would have been at least 20 cm lower to protect the shop floors from the rain water running along the streets. Only some strong georadar echos could be interpreted as the foundation of division walls of the first planned insulae (fig. 4: at 0,10 - 0,50 cm ASL with 90/100 RF23 distance from the Decumanus).

• 24 G. Bartok commented the georadar echoes underneath the south part of portico MFE as the “most massi (...)

16Based on these data, the following visual is probable: the streets between the small Tiber harbour in the North and the castrum, and all quarters inside the castrum had a first street level around 0,10-0,30 m ASL. However, it is important to note that the highest elevation inside the castrum seems to have been the area immediately north-east of the street-crossing of Cardo and Decumanus, exactly the area of the TFR-Sanctuary, maybe as a consequence of previous Tiber floodings, at the same time serving as protection against such floodings. So far it is not clear, if this TFR-Sanctuary had been founded a.) on a virgin sand-level of a natural dune up to ca. 0,90 m ASL or b.) on an artificial backfill to raise the level (fig. 4-12-13). In any case, this area remained slightly elevated above the surrounding street levels for a long time – and the buildings erected on top of the “most massive structure” discovered so far by georadar (GPR)24 were thus always well visible in the cityscape of Ostia in Republican times.

• 25 Finelli 1913, p. 250.
• 26 Gering – Menge (in this issue), Phase 1-2: the highest recorded surface of such tufa in TFR2 was th (...)
• 27 Finelli 1913, p. 250.
• 28 Lena 2011, p. 22-23.
• 29 Gering – Menge – Pedersen 2022, p. 239-242.
• 30 Gering – Menge – Pedersen 2022, p. 240: Phase 3 (brown).

17Finelli recorded around June 14^th-19^th 1913 some structures of different time periods in the south end of the portico: “Lato est dell’area del tempio di Vulcano. Con altri due operai fu anche incomiciato [cominciato] lo studio del sottosuolo dell’area sacra del tempio di Vulcano. Le ruine [rovine] rinvenute sotto questo lato dell’area fanno parte di più tempi e di più cose; vi sono dei muri a soli blocchi parallelolipidi [sic!] di tufo nerastro verdognolo posati sulla sabbia avendo uno solo come fondazione. Larg m. 0,43 × 0,26 alt × lung 0,58-0,69. I quali fanno parte delle prime costruzioni, come capanne, ad uno solo piano.”25 The lowest constructions in Ostia usually consist of the greenish Cappellaccio tufa (or “tufo granulare/ pisolitico”) with similar measurements (ca. 0,45 × 0,30 × 0,60 m or 1 ½ × 1 × 2 RF) like from the earliest phases of our 2019 and 2023 excavations:26 in TFR1-2, these blocks were layed on a level of 0,30 m ASL and never rose above the walking-level of 1,54 m ASL. Finelli continues: “Sopra questi muri ve ne sono altri a blocchi bugnati più grandi, ma di tufo più solito e rossastro, ma con fondazione a calce e tufi informi.”27 Finelli connected these walls, which were made with a more reddish tufa, with fundaments made of mortar (‘calce’) with irregular tufa cut-offs, which describes opus incertum. We can identify Finelli’s “tufo rossastro” most probably with “Tufo dell’Aniene”,28 the ancient lapis ruber. Thanks to our excavations in the neighbouring rooms TFR1-2 we know that the fundament-level of the constructions with reddish tufa was at 1,50-55 m ASL and its corresponding floor level in opus signinum was at 1,54-1,62 m ASL.29 In TFR2 we found an altar-construction with a similar reddish tufa, exactly on the same level of 1,54 m ASL with a foundation deposit of Phase 3, dating in the 1^st half of 2^nd century BC.30

• 31 Finelli 1913, p. 250.
• 32 Ibid.

18Furthermore, Finelli recorded a third building technique at the same level, that means more or less on 1,54-1,60 m ASL, which elevates only in some small areas higher according to georadar, up to the maximum of 1,80 m ASL: “muri a cortina a mattoni e altri ad opera reticolata, ma tutti con le fondazioni alla medesima altezza di quella con i blocchi bugnati”.31 Thus, Finelli summarizes three construction types in this position, one lower construction underneath 1,30-1,54 m ASL, which was levelled at this height, and two above this, while every construction used its predecessor as fundament. Finelli himself clearly recognized the importance of his findings as he wrote: “Questi muri veranno [sic! erano] studiati lentamente e ad ognuno di essi si cercherà di assegnarli l’epoca che gli appartiene”.32

• 33 For a detailed study on Finelli’s descriptions and finds-lists see Gering in preparation and Menge (...)

19Unfortunately, the description of the architecture underneath the south end of the portico MFE finishes with the above cited passage without mentioning the complete extension or discussing the function of the lowest massive tufa structure itself, which consisted most probably of the same Cappellaccio-Tuff Finelli had recorded as the earliest construction (fig. 5). Therefore, we concentrate on supplementing informations which old photos, plans and georadar imagery can offer in the following, while the find-lists and descriptions of other structures given by Finelli will be reassessed in a separate study in detail.33

• 34 For the most important epigraphic evidence concerning an Aedes Volcani see Zevi 2018; Van Haeperen (...)

23The size and layout of two massive structures or platforms will be described here, namely a lower platform in tufa, covered with opus signinum, and an upper one in opus incertum with repairs in opus reticulatum. A third ‘podium’ in opus testaceum high above these two older platforms will be discussed elsewhere because of the complex interrelation of the architectural and epigraphic remains.34

• 35 Compare Gering 2022, p. 946-950 with fig. 32.5-6.

25Without prior knowledge of which structure was photographed by Finelli, the photo itself does not seem to be very meaningful at first glance. However, it is fundamental that the Geophysics of 2016 revealed the most substantial echo in the entire underground of the Forum area which explicitly refers to the position of the lowest tufa platform with its internal fill, extending probably as a massive construction at least from 1,80 m ASL down to 0,70 / 0,50 m ASL (fig. 4: echo “T1a”).35

• 36 The oldest opus signinum floor of the area at 1,54 m ASL did not cause any echo in the georadar. So (...)

26In addition, the following observations were based on single georadar sections (fig. 4: upper graphic) and the complete 3D georadar model of all echoes in the entire subsurface beneath MFE (fig. 4: graphic below): georadar profiles are always read from top to bottom, so the upper surface of the coloured georadar echoes is the most conclusive feature. From the height of 1,80 m ASL, we have thus mapped a surprisingly massive platform that exhibits a slope to the north.36 The 3D georadar image represents a sum of all the profiles of the survey, from which the following is deduced:

• 37 Gering 2023, p. 57-64, fig. 9.

35d.) Above that, a fourth layer can be identified clearly: the photo shows a step made of concrete and some excavated parts of the adjoining floor, maybe in opus signinum, as the photo shows a similar texture as the cocciopesto-pavements of TFR1 in the east or the altar-terrace and the Pronaos of the nearby Jupiter temple in the west at 1,14 m and 1,54 m ASL.37 The pavement C (fig. 8: C) starts directly south of the plinth c1 and c2 at a similar height of 1,50-1,60 m ASL (fig. 5), which is, as the geophysics show, lower than the southern half of the pavement C at 1,75-1,80 m ASL (fig. 5. 14). It is not clear if this opus signinum pavement, which as construction technique cannot date prior to the first half of the 2^nd century BC, is contemporary with the first construction of tufa below, or a later repair. However, due to the well-preserved uppermost step (fig. 5) and the fact that it surrounded the plinth c2 (fig. 6), it is evident that it had at least for a certain period the function of a water-proof pavement of the structure T1a. During the excavations which continued until July 1913, further building material was piled on top of this pavement, probably tufa blocks, some with obvious worked surfaces, which may have belonged to the structure. Unfortunately, this material was not documented (fig. 5).

• 38 The original base, probably in tufa, is unfortunately not preserved, only a drum from the shaft of (...)

36e.) Only the above mentioned photo (fig. 5) shows plinth c1 with its ‘foundation’ of smaller and irregularly shaped tufa blocks (c1a-c). The plinths c1 as well as c2 were apparently extremely precisely smoothed on their upper sides. It is noteworthy that the north side of c1 maybe shows a simple profile at its lower end, about 20 cm high, which starts at 1,10 m ASL (fig. 5: c1, the white line shows the profile). The north front of the block c1 and its west side towards the stairs appears to have been relatively carefully smoothed; whether there was formerly a covering layer of stucco cannot be verified from the photo. It is significant, however, that at least two sides of the plinths were obviously designed to be seen, which supports the hypothesis of the open staircase between two pillars or columns in this position. The flat and polished top surface of the plinths c1 and c2 indicates the direct support of a pilaster or column base.38

• 39 Fig. 5: this first step could be reconstructed between the wall edging a2b? and a2b.

37To summarize, the georadar imagery corresponds exactly to the photo and supplements the descriptions of Finelli; in some details, both visual sources of information complement each other. Though the later constructions in this position, which Finelli described (see above), covered or destroyed parts of the older construction, one can still get the impression (which is supported by the georadar profile) that there had been maybe three to four steps leading from the ground level at 0,90-1 m ASL to the level of the Pronaos at 1,50-1,60 m ASL, with yet another step making the actual Cella accessible at 1,80 m ASL. The first step was either already beginning at a level of 0,90 m ASL (in this case the lowest step would be missing39), or at the level of the possibly profiled plinth at 1,10 m ASL.

• 40 Unfortunately, the old drawing of the blocks is not scaled (fig. 6: sketch 1913) and the position o (...)

39To conclude: with the 3D georadar imagery, the gap between the south end of Finelli’s unfinished excavations of 1913 (fig. 3-8-11: south end of plan) and the southern end of the portico MFE is closed now. The main result is the discovery of the “most massive structure” (see above), obviously a tufa platform consisting of a dense fill and covered by cocciopesto underneath the Portico MFE. We know the basic outline in 3D and the dimensions of approximately 5,90 m × 8,90 m. Furthermore, due to the first groundplan (fig. 6: sketch 1913, detail) we can reconstruct an interaxial span of about 4,30 - 4,45 m between the geometric centres of the plinth blocks c1 and c2 (about 15 RF with a possible tolerance40). It does not appear to be a coincidence that the uneven metric measurements of the platform at least correspond to “even” measurements such as 30 RF in length and 20 RF in width.

• 41 Finelli 1913, p. 229-232 (Nr. 8146: Antefix of Jupiter Ammon, written comment by Fausto Zevi in the (...)
• 42 Finelli 1913, p. 201-202 (Nr. 7974, Silenus in front of palmette, found in the TFR-Sanctuary nearby (...)
• 43 For the deposit made of rectangular rooftiles compare Gering – Menge in this issue.
• 44 For the tufa see Jensen in preparation. The terracotta fragments will be published in Menge in prep (...)

44The usually massive terracotta roof of a mid-Republican temple would surely have left some traces. However, after centuries of undocumented robbing activities at the Forum, there is not much, but at least some evidence remaining: An antefix depicting Jupiter Ammon with well-preserved colours (even blue) was found by Finelli close to the main entrance to the TFR-Sanctuary41. The use of blue colour indicates a late Republican date for its original context, which probably was not primarily private housing due to the size and iconography. Other antefixes from the area depicting a palmette with a Gorgo seem to be typical for late Republican temple-roofs: however, considering the relative sequence of Finellis three building phases (see above), only the Silenus antefix from the 3^rd or 2^nd century BC could have belonged to an architectural decoration related to the oldest massive podium mentioned here.42 Other elements of terracotta roofs were excavated by the OFP in 2019 as well in 2023 in the area immediately east of temple and altar: the retaining wall for the first solid tile floor was constructed mainly of large rectangular rooftiles. In the fill of the Libation area, other fragments of rooftiles, partly painted in red, were found in room TFR2: their dimensions seem to be substantially larger than rooftiles used for private houses.43 Additionally, some architectural terracotta and lots of profiled tufa blocks were found in TFR1-2, unfortunately often fragmented and in a very bad state of preservation due to the low quality of tufa involved. These tufa fragments are treated in a separate study.44

• 45 For Cosa see Brown 1980. More parallels to the shape of T1a will be cited in the planned monograph (...)
• 46 For the level of the first (‘green’) phase of the Altar-area nearby compare Gering – Menge – Peders (...)
• 47 A secondary function of temple podia as cistern is possible, compare the ‘Capitolium’ of Cosa (see (...)

45To conclude, the massive structure faced with tufa has close similarities to a podium of a typical Italic temple. The podium has proportions of 20 RF width to 30 RF length or 1:1,5. Comparisons reveal structurally very similar podia in other cities, for example the Forum temple in Cosa:45 the proportions and layout follow the common forms of Italic temples in the 3^rd and 2^nd centuries BC. The podium proportions of 1:1,5 are more elongated than the “classic” temples of the 5^th and 4^th century BC which adopted the proportions 5:6, though there are several exceptions from this “Vitruvian” rule. The podium T1a, which may have a parallel in the slightly later Forum temple in Cosa, still seems to date to the 3^rd century BC due to its relative low fundament level of 0,90 m ASL.46 As houses or shops can be excluded, the working hypothesis as a temple-podium seems to be the most adequate interpretation, with a possible secondary function to collect water in a cistern inside the temple-podium: the strong echo of georadar could have been caused either by a massive fill inside the tufa facing or its opposite, an empty space, as well.47 With the restriction that it was not yet possible to see or analyse this first podium in an excavated state, at least as a working hypothesis it will subsequently be referred to as “temple podium T1a” or “temple T1a” in short.

• 48 For Volterra compare the article Gering – Menge in this issue with bibliography.

47The orientation of the temple T1a to the north, which is verified by the position of the columns and the construction of the steps, seems strange at a first glance, but is not completely unusual. There are a number of parallels like the temples in the Acropolis Sanctuary in Volterra, where the main temple, like the analogous temple of Jupiter in the Forum of Ostia, is oriented to the south, but the smaller temple is oriented to the north.48 In Ostia, the position of the first Tiber harbour seems to be a possible reason for the orientation of temple T1a towards north (fig. 7). Considering the natural “dune” or artificial level-raise underneath the massive podium of T1a, this temple pediment and -roof at least may have been visible already from the ships arriving at the Tiber harbour, raising above the city walls and defining the city shape for all visitors approaching at Ostia.

• 49 Finelli 1913, p. 262.

48Finelli observed the remains of a collapsed building with a foundation: three rows of tufa blocks apparently were bordered by an unusually high accumulation of clay.49 These three rows of blocks above the fundament layer were obviously designed as a socle high enough to protect the upper walls made of wattle and daub from rain, even if the roof could not have overlapped by much.

• 50 Finelli 1913, p. 262. I would like to thank Letizia Ceccarelli for her comments on earthen architec (...)

50However, it is already noteworthy that the excavation diaries of 1913 contain clear references to the ancient use of clay and timber framing for the masonry. Finelli uses terms that directly address this – “argilla e straminio” refers to a form of wattle and daub made of branches bound into a sturdy self-supporting wooden framework of posts and beams (“tavole”). In many cases where excavation reports still speak of “clay layers”, this possibility of interpretation must be included.50

• 51 Compare here footnote 12. 13.
• 52 Finelli 1913, p. 263.

53As already mentioned, the structures excavated in 1913 were never published due to the early death of Vaglieri in 1913. Therefore, even a small, but rather unique structure which Finelli described “in or above the old shops” along the Cardo was never discussed by research.51 Only the GdS mentions this (almost) square structure and referred to it as “ambiente” or a “square room”: “Nel centro della taberna v’è un altro muro o fondazione normale a quello est-ovest che ci fa vedere in quadrato di un ambiente sopra le taberne vetere o dietro di queste”.52 However, this interpretation as “wall” or “fundament” or “room” has to be reconsidered.

• 53 Gering – Menge – Pedersen 2022, p. 230-251, fig. 8. The only difference between A1a and A1b-c: the (...)

57The temple T1a is oriented to the north and the enclosure A1a would have occupied an ideal canonical position for an altar precisely there: a reconstruction as an enclosure of an open ash altar could thus gain in probability (fig. 6-7-14: A1a), especially, because the newly excavated altar A1b-c (fig. 6) in the southeast consisted of similar enclosures of a fire pit, which rose only about 20 cm above the corresponding ground-level.53

• 54 Altar A1a was excluded from the later precinct marked in blue.

58The altar A1a refers directly to the temple through the same orientation, slightly different to the orientations of later structures (fig. 10-11: the temple T1b with A1b54). Furthermore, we can see a direct reference to the temple T1a through the same system of proportions: the width of the altar A1a concurs with the width and length of the Cella of T1a, which measures 17 × 17 feet on its outside. Its length measures 13 RF which is equivalent to the distance in between the altar and the temple (fig. 6).

• 55 Gering – Menge – Pedersen 2022, p. 254-58, fig. 27-29; Gering 2023, p. 48-54.

59And last, but not least: there is a connection of the first temple and the altar to the first plot size of the entire sanctuary with 90 × 90 or 75 × 90 RF:55 the altar itself has the same distance to the northern boundary of the sanctuary as the Cella of the temple to the southern boundary, 17 RF each (fig. 6 shows with the red and green lines a very regular planning pattern). All distances from the outer boundaries of the sanctuary to the altar as well as to the Cella measure 17 RF: the possible Altar and the Cella obviously were the two most important elements of the planning. Even the west boundary at the Cardo was 17 feet distant from the Cella. The north wall of the Altar is 17 RF + 13 RF = 30 RF distant from the temple podium, which itself measures 20 RF length from north to south. Only the sanctuary boundary to the east did not conform to this regular pattern, but further excavation is needed to determine this.

• 56 Probably, due to their size, from a temple-roof. For the rooftile-pavement Pa3 see the article of G (...)
• 57 For the phases of the TFR-Sanctuary see Gering – Menge in this issue and Gering – Menge – Pedersen (...)
• 58 Finelli 1913, p. 262-263.
• 59 See Gering – Menge – Pedersen 2022, p. 242-243, fig. 26-28; Gering 2023, p. 50, fig. 2.

61In the TFR-Sanctuary, above the walking levels of 1,54-1,75m ASL there is broad evidence for a substantial construction phase of opus incertum and opus signinum-/cocciopesto-floors as well as a floor of reused rooftiles south of the Libation-area (fig. 10).56 The fourth (“blue”) altar-phase (dated to the second half of 2^nd century and first half of 1^st century BC57) was the first phase involving concrete in construction. This phase also was contemporary with the construction of new shop walls immediately north of the altar A1 (fig. 3-11: blue walls Tb4 nord1-4), which also had a concrete fundament, representing the “second phase of construction” after Finelli.58 The massive tufa blocks of these walls were worked with extreme accuracy (fig. 9). Our excavations in TFR2 revealed similar tufa structures to the east (fig. 11: Tb4). Considering additional information of the old plan (fig. 8), we can assume that the size of the sanctuary had been reduced in the above mentioned fourth building phase to a parcel of 60 × 60 RF (or maximum 70 × 70 RF: fig. 3-11), thus excluding the first possible altar A1a in order to gain more commercial space.59 All the concrete building structures had the same orientation and were built on top of a raised level of pavement as prevention against the regular Tiber-floodings.

• 60 The repair phase with opus reticulatum can be dated in the 2^nd half of the 1^st century BC until Tra (...)

66To summarize: based on photos, plan, small test sondages (and verified by georadar imagery) we have sufficient evidence of a rectangular structure measuring 10,40 × 6 m (fig. 10. 11) on top of the podium T1a. It exhibited extremely massive fundaments (wall thickness in the west: 1,50 m, south: almost 3 m, east: incorporated into later Portico back wall, about 1,50 m). Probably as product of another building phase, the north side, however, exhibited only a relatively thin wall (about 60 cm or 2 RF). This thin wall of opus reticulatum contradicts a hypothetical function as a cistern at least in the last phase, because, due to the water pressure, cisterns need four sides of equal thickness.60

• 61 Georadar showed the underground of the Cella-pavement of the temple T2 and T5 (Gering 2023, fig. 7- (...)

67Structure B was built immediately on top of the opus signinum-Pavement C, which covered the podium A, thus probably replacing the old wooden temple with a more solid construction. The later construction probably consisted of stone (travertine?) walls, which needed equally solid fundaments like structure B. Late Republican temple podia such as that of the Tempio sul Decumano at the intersection of the Decumanus with the Via dei Molini, have a similar layout: the centre of the podia was not filled with concrete, usually underneath the floor of the main Cella, where georadar reflects loose earth or rubble instead of solid masonry. Due to this similarity, the rectangular structure B could have served as the core of the podium of an equally small temple, roughly from the same time period, which probably had been covered at its outside with stucco or stone slabs, and an internal filling of rubble or earth, which was removed during the old excavations.61

• 62 For the hypothetical reconstruction of a possible first temple of Jupiter based on georadar see Ger (...)

74Temple T1b could have served as a prominent landmark in the type of the typical Roman podium temple. As the new podium T1b was much higher than T1a (fig. 12-13) we can reconstruct possibly the same height as the much larger Temple of Jupiter nearby (accepting the hypothetical reconstruction of a predecessor underneath the first excavated “Capitolium” built in the 1^st century BC62).

• 63 Zevi 1996, p. 85. For the introduction of Volcanus in Ostia in the early 2^nd centry BC see Zevi 200 (...)

76We may suggest that temple T1b was built in direct connection with the introduction of a new cult to Ostia, namely into the TFR-Sanctuary. So far, it is a working hypothesis based on later inscriptions that the latter may have become partly a sanctuary of Volcanus, the “dio di agorà”.63

• 64 Gering 2023, p. 51-55, especially p. 53; Gering in preparation; Damgaard in preparation; Menge in p (...)

77Thus, the erection of a new Forum or marketplace on a level of 0,90 m ASL in the area of the “Building B” directly south of the TFR-Sanctuary would seem a logical coincidence with the turning of the temple façade from north (T1a) to south (T1b-c-d, fig. 13). Due to the dating evidence from TFR2 (see above), this happened most likely in the fourth (“blue”) phase of this sanctuary, dating from the second half of the 2^nd century BC to the first half of the 1^st century BC. This may suggest the introduction of the Forum of Ostia at least a century earlier than previously assumed, most probably connected to the TFR-Sanctuary, the newly oriented temple and perhaps the (new?) deities venerated there. The discussion about this hypothesis will continue elsewhere on a broader material base.64