Terminology

In studies of Early Mycenaean funerary architecture, confusion is sometimes caused (a) by the undifferentiated application of the term ‘tholos’ to all types of circular constructions, irrespective of size and setting, and (b) by the interchangeable use of the terms ‘built chamber tomb’ and ‘built grave’. In this paper, the following definitions are used (Fig. 2):

• tholos tombs – tholoi built as single monuments either above ground or partly underground, usually with a covering mound (although, as we will see, some early examples may have been freestanding);Footnote ⁴

• chamber tombs – earth-cut chamber tombs, usually dug on sloping terrain;

• small tholoi – small tholos-like structures (<4–5 m in diameter), clustered in groups and often covered by a common mound;Footnote ⁵

• built chamber tombs (BCTs) – tombs of apsidal or rectangular form with built walls and side entrances (which differentiate them from plain ‘built graves’)Footnote ⁶ .

Chronology

Pelon has discussed extensively the problems surrounding the dating of collective tombs, especially tholoi (Pelon Reference Pelon1976, 372–91). He has argued that their continuous use over long periods of time renders their contents unreliable for assessing their construction date. Instead, he proposed to employ architectural criteria for tracing stages of evolution at each site. Unfortunately, this method can be applied only on sites with more than one tholos. Moreover, Pelon himself acknowledges that evolutionary patterns are site specific and cannot be applied on regional or supraregional level (Pelon Reference Pelon1976, 380–91). Therefore, if we seek a dating tool that is both omnipresent and quite independent of regional developments, we have to stick with pottery. For all the vicissitudes of preservation, pottery is the only type of evidence that can provide a safe terminus ante quem for the construction of a tomb. Although this can only produce a minimum number of tombs used in each period (for other tombs may have originally contained vases of that period, which were removed in later funerals or cleaning operations), it nevertheless provides a satisfactory level of chronological resolution for the purposes of the present study.

Even more difficult is to determine whether the dromos was constructed at the same time as the tomb. Few dromoi have been excavated in a stratigraphic way that allows for independent dating.Footnote ⁷ Moreover, some tombs evidently underwent changes at a stage of their history, which may have affected the dromos, too.Footnote ⁸ When such information is available, it is included in the discussion. Otherwise, it is assumed that the dromos was an original feature of the tomb.

The study will focus on LH I tombs with lateral entrances, but will differentiate between:

1. a) safely dated examples, i.e. tombs which have yielded complete or almost complete LH I vases, and

2. b) possible examples, i.e. tombs which have yielded only sherd material of LH I date (for such sherds may have derived from the fill rather than from the chamber itself), and tombs which have been dated by their excavators to LH I but are only known through brief preliminary reports without published pottery.

Tombs attributed to ‘MH/LH I’ due to the presence of MH-style vessels among their contentsFootnote ⁹ will be listed with LH I examples. The problem of MH ceramic survivals into the LBA is too complex to discuss here;Footnote ¹⁰ in any case, a distinction between MH/LH I and LH I is not crucial for the present paper, which aims to identify general developments in the period between the end of the MBA and the beginning of LH IIA, not to establish the chronological precedence of one tomb over another.

Detailed dating sources and information about the identification of a tomb as ‘safe’ or ‘possible’ LH I are provided in Appendix I. For the sake of convenience, ‘safe’ and ‘possible’ examples are discussed together, but ‘possible’ LH I tombs are marked with an asterisk (*) throughout the text. All examined tombs are listed per type and region and numbered sequentially in Appendix II: Tables A1a , A1b , A1c , A1d and A2 (with references). Numbers in the tables are indicated in the text by figures in square brackets next to the tomb's name. A map of sites with LH I and pre-LH I tombs with lateral entrance is provided in Fig. 3.

Fig. 3. Map of sites with LH I and pre-LH I tombs with lateral entrance (T = tholos tomb, C = chamber tomb, S = small tholos, B = built chamber tomb).

a) Numerical Index

1. Portes [B]; 2. Kato Samikon [T, B]; 3. Psari [T]; 4. Peristeria [T]; 5. Kephalovryso-Paliolylos [S]; 6. Volimidia [C]; 7. Routsi [T, B]; 8. Papoulia [B]; 9. Pylos [T]; 10. Costa Navarino [T]; 11. Tragana [T]; 12. Voidhokoilia [T]; 13. Osmanaga [T]; 14. Koukounara [T, S]; 15. Kaminia [S]; 16. Nichoria [T, S, B]; 17. Diodia [T]; 18. Sparta [B]; 19. Epidaurus Limera [C]; 20. Analipsis [S]; 21. Kokla [C]; 22. Argos [B]; 23. Schoinochori [C]; 24. Mycenae [C]; 25. Prosymna [C]; 26. Tiryns [C]; 27. Galatas [T]; 28. Corinth [T]; 29. Eleusis [B]; 30. Thorikos [T]; 31. Marathon [B]; 32. Thebes [C]; 33. Dramesi [B]; 34. Xeropolis [B]; 35. Mitrou [B].

b) Alphabetical Index

Analipsis [S] 20; Argos [B] 22; Corinth [T] 28; Costa Navarino [T] 10; Diodia [T] 17; Dramesi [B] 33; Eleusis [B] 29; Epidaurus Limera [C] 19; Galatas [T] 27; Kaminia [S] 15; Kato Samikon [T, B] 2; Kephalovryso-Paliomylos [S] 5; Kokla [C] 21; Koukounara [T, S] 14; Marathon [B] 31; Mitrou [B] 35; Mycenae [C] 24; Nichoria [T, S, B] 16; Osmanaga [T] 13; Papoulia [B] 8; Peristeria [T] 4; Portes [B] 1; Prosymna [C] 25; Psari [T] 3; Pylos [T] 9; Routsi [T, B] 7; Schoinochori [C] 23; Sparta [B] 18; Thebes [C] 32; Thorikos [T] 30; Tiryns [C] 26; Tragana [T] 11; Voidhokoilia [T] 12; Volimidia [C] 6; Xeropolis [B] 34.

IA. Tholos tombs (Appendix II: Table A1a )

Twenty tholoi in Messenia, Elis, Corinthia, the Argolid and Attica can be safely or possibly dated to LH I (or MH/LH I). Two of them have produced no evidence of a dromos at all (Pylos ‘Vagenas’ [1], Galatas 3* [18]). At Galatas 3* (Fig. 4) Konsolaki-Giannopoulou is clear that a dromos had never existed (Konsolaki-Giannopoulou Reference Konsolaki-Giannopoulou and Konsolaki-Giannopoulou2003, 178; Reference Konsolaki-Giannopoulou, Vasilopoulou and Katsarou-Tzeveleki2009, 510; Reference Konsolaki-Giannpoulou2010, 72). Pelon believes that the ‘Vagenas’ tholos was also devoid of a dromos or, if it had one, it was very short (1976, 397). The two tombs share many construction details. They were built entirely above ground and had comparable diameters (4.9–5.5 m). Although only the lower courses of their walls have survived, they were apparently thin (‘Vagenas’: 60 cm; Galatas 3*: 40 cm).Footnote ¹¹ Konsolaki-Giannopoulou believes that none of them could have supported a full-stone superstructure; instead they may have been roofed with a dome of lighter materials, such as wood and clay (Konsolaki-Giannopoulou Reference Konsolaki-Giannopoulou and Konsolaki-Giannopoulou2003, 178; Reference Konsolaki-Giannopoulou, Vasilopoulou and Katsarou-Tzeveleki2009, 510; Reference Konsolaki-Giannopoulou, Schallin and Tournavitou2015). The entrance of Galatas 3* was very narrow (1–1.1 m) and low (c.1.20 m), flanked by two monolithic antae, without a stomion.

Fig. 4. Tholos 3 at Galatas, Troezen (courtesy of Dr E. Konsolaki-Giannopoulou; source: Konsolaki-Giannopoulou Reference Konsolaki-Giannopoulou and Konsolaki-Giannopoulou2003, 226 fig. 70).

A number of tholoi had rudimentary access devices. Peristeria South [10] (Fig. 5) and Veves [9] had short projecting walls (Peristeria South: 2.38–2.7 m; Veves: 2.4 m), which created small ‘vestibules’ in front of their entrances (i.e. corridor-like features open at the outer end and blocked at the inner end). That these features were meant to function as stomia is unlikely; although no blocking walls have been reported, the photos and plan of Peristeria South (Fig. 5 and Korres Reference Korres1976b, 549 fig. 11) show a row of stones at the inner end of the feature (along the line of the chamber wall) which must have belonged to the blockage; at Veves, Choremis describes the corresponding feature as a ‘walled dromos’ (Choremis Reference Choremis1973, 60) implying that the doorway was also placed at its inner end. The discovery of a lintel at Peristeria South (Korres Reference Korres1976c, 138) does not allow us to exclude the possibility of a stomion altogether,Footnote ¹² although one should remember that the practice of providing roofing to a walled ‘vestibule’ is also known from the LH IIB tholos at Marathon (Pelon Reference Pelon1976, 229). Choremis asserts categorically that there was no earthen dromos at Veves (Choremis Reference Choremis1973, 60), neither is a dromos reported from Peristeria South. This suggests that access to both tombs was provided through short and narrow ‘vestibules’ leading to equally narrow (c.1 m) and shallow doorways without distinct door-jambs – an arrangement which differs substantially from the wide and deep facades of later tholoi. Both tombs were built above ground, and had small diameters (5.08–5.1 m) and thin walls (Veves: 60 cm; Peristeria South: est. 50–60 cm). Korres has suggested that the upper part of Peristeria South might have been built of light sandstones and covered with layers of clay (Korres Reference Korres1976b, 508; Reference Korres1977b, 322).

Fig. 5. The South tholos at Peristeria, Messenia (courtesy of the Archaeological Society at Athens; source: Korres Reference Korres1976b, pl. 268α).

A slightly different arrangement is attested in the Costa Navarino tholos* [16] and in Peristeria 3 [11]. These tombs also had projecting walls creating corridor-like features, which however were probably used as stomia. This is certain at Costa Navarino*, where a blocking wall has been found at its outer end (Rambach Reference Rambach2009, 363 fig. 21). As for Peristeria 3, Korres interprets the corresponding feature as a stomion, and the fact that it has a finished facade (built with horizontal slabs which create jambs wider than the walls, Fig. 6 and Korres Reference Korres1977b, 328 fig. 6) seems to support his view. However, a piece of low but well-built masonry found in the inner part of this feature (Korres Reference Korres1976b, pl. 272c; Reference Korres1977b, pl. 174b) raises questions about its function. Korres believes that the masonry was not part of the tomb's blockage but a cross-wall associated with a floor depression running through the ‘stomion’ into the chamber, which must have been dug at a later stage of the tomb's history (but still within LH I) as it left the lower part of the ‘stomion’ walls exposed (Reference Korres1976b, 533–4 and pl. 272c; Reference Korres1977b, 337–9; see also Fig. 6). However, the published plans and photos suggest that the masonry in question reached a considerable height in relation to the ‘stomion’ walls, thus probably obstructing circulation between the ‘stomion’ and the chamber. For that reason, it is not implausible that it actually belonged to the tomb's blockage (cf. Dickinson Reference Dickinson1977, 93); if so, then the corridor-like feature of Peristeria 3 might also have functioned as a ‘vestibule’, albeit of greater size (4.1 × 1.73 m) and of more advanced form than those in Peristeria South and Veves. Whether the tombs were provided with earthen dromoi is unclear. According to Korres (Reference Korres1976b, 532; Reference Korres1977b, 339), the dromos of Peristeria 3 would have lain below the West House, and the fact that a published photo (Fig. 6) shows the depression continuing beyond the facade may support this suggestion, although no traces of dromos walls have been identified. No information about the presence of a dromos is available for Costa Navarino*. The tholos of Peristeria 3 was quite large (6.9 m in diameter) but clearly built above ground with thin walls (c.60 cm, Pelon Reference Pelon1976, 345). The size of the Costa Navarino tomb* is unknown, but its walls were obviously thin (Rambach Reference Rambach2009, 363 fig. 21; Reference Rambach2011).

Fig. 6. The facade of tholos 3 at Peristeria, Messenia (beneath a later wall) (courtesy of the Archaeological Society at Athens; source: Korres Reference Korres1976b, pl. 273a).

So far, we have examined tholoi where the presence of a dromos is either safely excluded or reasonably doubted, and the system of access is unsophisticated and non-standardised. With the possible exception of Costa Navarino*, which is not yet documented in detail, these tombs were built above ground and were either freestanding or covered with light materials (e.g. clay).Footnote ¹³ This goes against Zavadil's idea that the lack of a dromos from some Messenian tholoi is due to the erosion of the surrounding tumulus (Zavadil Reference Zavadil2013, 59); in all probability, the examined tombs had never had a covering mound of the type known from later tholoi, not least because their thin walls (<60 cm) and primitive structure would have been unable to bear its weight.

We now turn to LH I tholoi which have yielded positive evidence for the presence of a dromos and a stomion. As we will see, there is still considerable variation in the form of the dromos and the entrance system as a whole.

Short, narrow and steep dromoi have been found or reasonably assumed in three cases: Routsi 1 [4], Diodia* [15] and Osmanaga [6]. Tomb sizes vary considerably (4.2–6 m) but all three were sunk in the ground up to the level of the lintel, had quite thick wallsFootnote ¹⁴ and clearly defined albeit rather shallow stomia (1.92–2.3 m deep)Footnote ¹⁵ of comparable height (Osmanaga: 2.75 m; Routsi 1: 2.3 m; Diodia*: 1.8–2.3? m)Footnote ¹⁶ . At Routsi 1 and Diodia*, stomia had finished facades with narrow door-jambs built of flat stones,Footnote ¹⁷ and dromoi were short (c.5 and 3.7 m)Footnote ¹⁸ descending sharply towards the entrance (Korres Reference Korres1989a, 28; Fig. 7). No dromos has been excavated at Osmanaga, but Pelon believes that a steep one must have existed because of the difference in height between the entrance and the original ground (Pelon Reference Pelon1976, 198, 278), and Korres seems to confirm its presence (Korres Reference Korres1977a, 232). There is no information about the presence of a mound at Routsi; the tholoi at Osmanaga and Diodia were probably covered by small earthen mounds (Pelon Reference Pelon1976, 298; Chatzi-Spiliopoulou Reference Chatzi-Spiliopoulou2006, 202).

Fig. 7. The tholos at Diodia, Messenia (courtesy of the Ephorate of Antiquities of Messenia and Dr G. Chatzi-Spiliopoulou; source: Chatzi-Spilioupolou Reference Chatzi-Spiliopoulou1995, 181 fig. 2).

A short (c.3.3 m preserved length) but almost flat dromos has been found at Livaditi* [12]. Unlike the previous tombs, this one is said to have been built above ground with a large covering mound, but documentation is very concise. It was rather small in diameter (4.62 m), and had a shallow stomion (1.76 m) built of flat stones with narrow jambs in the facade (Marinatos Reference Marinatos1958, 189 and pl. 147b; Iakovidis Reference Iakovidis2014, 140 plan 30).

Routsi 2 [5], Tragana 2* [13] and the insufficiently known tombs Gouvalari 1 and 2 [7, 8] belong to a type with well-formed stomia and dromoi, which seem to have sloped gently towards the entrance (Iacovidis Reference Iakovidis2014, 109 plan 26, 126 plan 28, 151 plan 33, 154 plan 34). Among them, Tragana 2* was built partly underground and had a covering mound; Gouvalari 1 and 2 may have been covered by a single mound; the setting of Routsi 2 is not clear, but the available plan suggests that it was also built partly underground.Footnote ¹⁹ At Routsi 2 and Gouvalari 2 (5 m in diameter), stomia were built of flat stones and although shallow (1.5–1.6 m) had well-finished facades (Marinatos Reference Marinatos1957a, pl. 52a; Reference Marinatos1959, pl. 149a). The larger tholoi Gouvalari 1 and Tragana 2* (6.25–7.2 m) have poorly preserved but longer stomia (2.5–2.7 m), and that of Tragana 2* seems to have been built of large, dressed blocks (Korres Reference Korres1976a, pl. 180a). The dromos length has been recorded only at Tragana 2* (7.5 m) and can be estimated from the plan for Routsi 2 (7.9 m). All tombs had floor depressions, which started at the inner end of the dromos and continued into the chamber (see section IIA); Korres (Reference Korres, Oliva and Frolikova1982, 91–3) believes that some of these depressions were the result of later rearrangements due to stability problems related to the tholos, but dates them before the end of LH I.

The tholos at Kato Samikon-Kleidi* [17], reportedly built within a tumulus 14.6 m in diameter, was quite large (5.65 m) but neither its stomion nor the dromos has been excavated (Papakonstantinou Reference Papakonstantinou1983).

Pylos IV [2] and Psari* [14] were the two largest tholoi of the period. They were built partly underground, had spacious chambers (>9 m in diameter), well-formed stomia (4.62–4.9 m deep) and high entrances (Pylos IV: 4.55 m, Psari* 2.85 m). Pylos IV was covered by an earthen mound and had a long and wide earth-cut dromos 10.5 × 4.4 m (Blegen et al. Reference Blegen, Rawson, Taylour and Donovan1973, figs. 176, 177, 186, 188); its blocking wall was not placed at the outer end of the stomion but close to its centre. Psari* was covered by a stone-clad mound and had a shorter and narrower dromos (6 × 3 m) with dressed sides and a low blocking wall at its start (Chatzi Reference Chatzi1983, pl. 57b–c; Reference Chatzi1985, pl. 35) (Fig. 8); since this wall was aligned with the masonry of the surrounding peribolos, it is clear that the length of the dromos was determined by the diameter of the mound. Interestingly, the stomion at Psari* was blocked at both ends (Chatzi Reference Chatzi1985, 103). The covering mound at Psari* had two or more concentric lines of stones around the tholos, which may suggest successive stages of construction (Chatzi Reference Chatzi1983; Boyd Reference Boyd2002, 181).

Fig. 8. The dromos and the stomion of the tholos at Psari, Messenia (courtesy of the Ephorate of Antiquities of Messenia and Dr Georgia Chatzi-Spiliopoulou; source: Chatzi Reference Chatzi1985, pl. 35b).

Voidhokoilia [3] is a unique case of a tholos built within a pre-existing tumulus. Although not particularly large (4.93–5.03 m), it had a well-formed and relatively deep stomion (2.37 m), which may suggest a mature stage in LH I, as is also indicated by pottery (Lolos Reference Lolos1987, 179–81). The ‘dromos’ was an artificial break in the mantle of the tumulus (Korres Reference Korres1989b, 418–9; Boyd Reference Boyd2002, 126), its length (7.8 m) being determined by the size of the mound.Footnote ²⁰

Thorikos IV* [20], built next to a MH/LH I tumulus, is a hybrid tomb (Fig. 9) whose elongated shape may have derived from a combination of tholos techniques with a local tradition of built chamber tombs with apsidal ends (although the latter cannot be shown to date as early as LH I, Papadimitriou Reference Papadimitriou2001a, 93–4). The tomb was built partly underground, and had a low stomion (2 m) with a primitive relieving triangle, and a slightly off-axis 7 m-long dromos, i.e. shorter than the chamber (9.3 m long). The outer end of the dromos was blocked by a low wall, perfectly aligned with the stone peribolos of the covering mound but separated from it by means of two side-jambs. The left jamb (as one approaches the dromos) was built at a small distance from the left wall of the dromos, possibly suggesting two phases of construction (Servais and Servais-Soyez Reference Servais and Servais-Soyez1984, 27–9). The experimental character of this tomb is readily apparent. Although its dromos was not the earliest to be built in Attica (see the BCTs of Eleusis in section ID), it was the first to provide sufficient space for any sort of ritual activities (see section IIC); as such it may have functioned as a prototype for later Attic tombs with dromoi.Footnote ²¹

Fig. 9. The ‘oblong tholos’ (tomb IV) at Thorikos, Attica (courtesy of the Belgian School at Athens; source: Servais and Servais-Soyez Reference Servais and Servais-Soyez1984, 22 fig. 8).

The recently excavated tholos at Corinth* [19] was large (6.7 m) with rather thin walls, and is reported to have had the longest LH I dromos known to date (12.7 m), although no details are yet available about it. The tomb presents an interesting peculiarity: its 2.7 m-long ‘stomion’ seems to have been blocked at its inner, not outer, end (Kasimi Reference Kasimi, Kissas and Niemeier2013, 45–7 figs 3–4); the excavator, Mrs Kasimi, has kindly informed me that this was probably the case in the earlier phases of its use, although later in the history of the tomb the blockage must have been placed in the outer end of the ‘stomion’.Footnote ²² Consequently, this corridor-like feature may have been initially used as a built ‘vestibule’, not unlike those seen at Peristeria South and Veves, something supported by the fact that it does not seem to have had a distinct facade with well-formed jambs. If this observation is correct (and if the dromos was an original feature of the construction), the tomb may represent an intermediate stage in the development of tholos entrances; however, one should await the final publication before drawing definite conclusions.

IB. Chamber tombs (Appendix II: Table A1b )

Approximately thirty chamber tombs in Messenia, Laconia, the Argolid and Boeotia can be safely or possibly dated to LH I (or MH/LH I). Among the north-east Peloponnesian and Boeotian examples, Mycenae 518 [34], Prosymna 25 [35] and Prosymna 26 [36] had the longest dromoi (6.6–8.15 m), with vertical or almost vertical sides widening towards the stomion; at Prosymna dromoi presented a sharp inclination (Blegen Reference Blegen1937, figs. 181–2, 201). Tombs were rectangular in shape with side chambers (one in Mycenae 518 and Prosymna 26, three in Prosymna 25)Footnote ²³ , shallow stomia (0.9–1.27 m), and relatively low entrances (1.70–2 m to 2.25 m). Prosymna tomb 52 [37] was smaller, and semicircular in plan, but its dromos was destroyed. Tiryns XIX* [46] was rectangular in plan with a short (3.2 m) and steep dromos. Schoinocnori E* [47] had an elliptical chamber but its dromos was destroyed. Kokla V and VIIB [38–39] are only known through a preliminary report (Demakopoulou Reference Demakopoulou1982); the length of their dromoi can be estimated from the plans at c.3–4 m. Thebes Agia Anna 2* [48] was also large with shallow stomion, and had a wide (2.6 m) but only partially excavated dromos.

It is unclear which of the Epidaurus Limera tombs contained the two LH I vases published by Demakopoulou (Reference Demakopoulou1968, cat. nos. 3 and 64) and discussed by Dickinson (Reference Dickinson1977, 63–4 and note 34). The available plans of the Agia Triada tombs* [42–43] show elliptical main chambers, circular secondary ones (with separate entrance on the left side of the dromos), shallow and low stomia, and very short dromoi (est. <2.2 m) possibly accessed through steps (Christou Reference Christou1956, 208–9 figs. 1–2). The Vamvakia and Palaiokastro tombs* [44–45] are described as being similar to the Agia Triada ones, the former with steps leading to a short dromos, the latter without a secondary chamber. Gallou, who is restudying the material from the site, reports that the Palaiokastro tomb(s)* had carefully built stomion but no dromos (Gallou Reference Gallou, Cavanagh, Gallou and Georgiadis2009, 87).

The Volimidia cemetery included many LH I tombs [21–33, 40–41*] but not all of them have been illustrated or described in detail. According to Iakovidis (Reference Iakovidis1966, 100–1), they were strikingly homogeneous in style, with circular chambers, tholos-imitating roofs, shallow stomia (<80 cm) and short dromoi with vertical walls widening towards the stomion. The published plans, however, suggest greater diversity in shape (Marinatos Reference Marinatos1965, 105–8 figs 2–4; Iakovidis Reference Iakovidis2014, 3–90). At least four early tombs (Koronios 6 [23], Angelopoulos 7 [25], Angelopoulos 9 [27], and Koronios 1* [40]) had secondary chambers accessed through separate entrances, usually on the left side of the dromos (Fig. 10) (Marinatos Reference Marinatos1953, 238, 239, 245 fig. 6; Zavadil Reference Zavadil2013, 100; Iakovidis Reference Iakovidis2014, 5 plan 1, 32 plan 9, 44 plan 14, 45 fig. 7). Most dromoi were less than 6 m long (usually much shorter), rather narrow (1–2 m) and steep.Footnote ²⁴ Only tombs Angelopoulos 4, 7, 8, 9 [24–27] had longer but also steep dromoi (6–7.8 m) (Fig. 10). Entrances were low (1.2–1.8 m), at least in the recorded cases (see also Boyd Reference Boyd2002, 138–47).

Fig. 10. The dromos of chamber tomb Angelopoulos 9 at Volimidia, Messenia (courtesy of the Archaeological Society at Athens; source: Iakovidis Reference Iakovidis2014, 45 fig. 7).

IC. Small tholoi (Appendix II: Table A1c )

Nine small tholoi in Messenia can be safely or possibly dated to LH I (or MH/LH I). They are said to be included in large mounds, whose precise form and structure are unclear (Boyd Reference Boyd2002, 109, 116). The tombs range in size from 2.73 to 5.2 m, although most are less than 3.5 m in diameter. Among safe LH I examples, only Gouvalari Tα10 [52] had a dromos (c.3.6 × 2.1 m) (Korres Reference Korres1975b, 87 fig. 1). Dromoi are also known from Gouvalari Tα4 (est. 2.5 m) and Kaminia 3 (est. 4.7 m), which, however, have yielded only LH II–III pottery (Korres Reference Korres1975a, 432–40, 493–9). Gouvalari Tα1* [56] was clearly devoid of a dromos (Korres Reference Korres1976–8, 340) and this may have been true in other cases, as most tombs at Gouvalari and Kaminia were placed with their entrances towards a steep slope. No dromos was found at Kephalovryso-Paliomylos [54]; a kind of peribolos creating an enclosed space (approximately 2 × 2.5 m) with a side entrance has been excavated in front of the stomion but its chronological relation to the tomb is unclear (Chatzi-Spiliopoulou Reference Chatzi-Spiliopoulou2007, 357–8 figs 4–5).

Kaminia 4 [53] had a small stomion (est. 1.3 m deep) with a blocking wall at its outer end, probably without a dromos (Korres Reference Korres1975a, pl. 325; Reference Korres1980, inset pl. ΣΤʹ). Gouvalari Τβ1 [50] had short walls projecting from the entrance, which probably formed an open vestibule rather than a proper stomion (Korres Reference Korres1975a, pl. 317b); in the nearby Gouvalari Tα2 (of unknown date) a similar feature was clearly meant to be a vestibule, since a blocking wall was found at its inner end (Korres Reference Korres1974, 141–2, pl. 109b).

The small tholoi at Analipsis* [58] (2.48–3.2 m) had small corridor-like features in their entrances (0.9–1.7 m deep) but it is unclear whether they were meant as stomia or ‘dressed dromoi’, i.e. built vestibules (Kalogeropoulos Reference Kalogeropoulos1998, 77–8). In any case, they seem to lack proper facades, thus possibly resembling the tholos at Costa Navarino*.

ID. Built chamber tombs (Appendix II: Table A1d )

Approximately 30 BCTs from various regions can be safely or possibly dated to LH I (or MH/LH I). They include apsidal tombs, rectangular tombs with entrance in a long or a short side, and complex tombs with more than one compartment and entrance in a short or a long side. The BCTs of Portes [60–62, 81–83*] and Marathon* [84] were included in tumuli; Akones I* [80] and III [59] are said to have been covered by a large mound, whose precise nature remains obscure. Other examples, like the tombs at Argos [64, 65] and Eleusis [66–75, 85–87*] were covered by individual mounds.

BCTs are generally small in size, ranging between 1.55 and 4 m in length, but Mitrou 73* [88], Eleusis Hπ3 [68] and Eleusis E.III.7 [76] were c.5 m, and Portes Γ1 [60] 8 m long. Their height does not exceed 1.75 m, at least where measurements are available. Dromoi are attested in seven cases but most are very small (1.6–2.3 long, 0.9–1.4 m wide) and steep (Fig. 11).Footnote ²⁵ Mitrou 73* had a larger slab-lined dromos (3 × 2 m) featuring a porch with two support bases and a thin wall of clay at its outer end (Moortel Reference Moortel, Dakouri-Hild and Boydforthcoming). Eleusis Iπ1 [73] had the inner part of its 1.85 m-long dromos dressed with horizontal slabs (Mylonas Reference Mylonas1975, vol. 2, 104). A small but well-built wall projecting from the stomion of Portes Γ1 (Moschos Reference Moschos2000, 16 fig. 7) may have belonged to a ‘vestibule’ but has not been discussed by the excavators. A few more Eleusinian tombs may originally have had short dromoi (Papadimitriou Reference Papadimitriou2001a, 216).

Fig. 11. The built chamber tomb 164 at Argos, Argolid: plan and section A–A′ (source: Papadimitriou Reference Papadimitriou2001b, 45–6, figs. 3–4).

Many BCTs had shallow (<1 m) but quite distinct stomia: at Eleusis they were formed by upright slabs or rectangular blocks flanking the entrance (e.g. Zπ6 [66], Ηπ3 [68], Ηπ20 [69], Ιπ1 [73]); at Akones I* [80] by two small antae; Portes A2* and A3* [83–83] had ‘an entrance with pillars’ in one short side (Moschos Reference Moschos2000, 13); Argos 164 [64] and Portes Γ1 [60] had built antae and thresholds (Papadimitriou Reference Papadimitriou2001b, pl. 4d; Moschos Reference Moschos2000, 17 fig. 8).

Among BCTs with dromoi, Eleusis Iπ1, Argos 164 and Mitrou 73* evidently underwent some kind of refurbishment at a later stage of their history. Eleusis Iπ1 was enlarged during LH I and it is possible that the partially dressed dromos was added by that time (Mylonas Reference Mylonas1975, vol. 2, 102–4; Papadimitriou Reference Papadimitriou2001a, 74). Argos 164 was probably extended in LH IIA and a more elaborate stomion was constructed, but the dromos seems to have been an original feature (Papadimitriou Reference Papadimitriou2001b, 47–8). Mitrou 73* was extended and rearranged in LH IIB but the dromos was an original construction (Moortel Reference Moortel, Dakouri-Hild and Boydforthcoming).

Some access devices in BCTs cannot be technically classified as dromoi, yet they reflect a similar concern for some sort of passageway. The central structure at Marathon tumulus II* [84] (Fig. 12) consisted of three compartments, the outermost of which was very low, empty of finds, had a monolithic threshold at its entrance and a built one at its rear side, and may have functioned as an open courtyard or an early form of dromos (Marinatos Reference Marinatos1970 15 fig. 4 and pl. 19b; Pelon Reference Pelon1976, 84; Cavanagh Reference Cavanagh1977, 105; Papadimitriou Reference Papadimitriou2001a, 103–4); another threshold was attached to the outer peribolos of the mound along the main axis of the tomb. The presence of two peribolos lines and of concentric lines of stones around the tomb may suggest successive stages of construction. Tomb E.III.7 at Eleusis [76] was divided by a partition wall, and one part was probably used as a ‘vestibule’ as it has yielded only pottery and animal bones; the entrance was not excavated, as it lay under the Peisistratian wall, but the arrangement is reminiscent of two LH II tombs from the West Cemetery (Λπ4 and Τ4), which had partition walls and covered vestibules with entrance in one long side (Cosmopoulos Reference Cosmopoulos2014, 139–40).

Fig. 12. Tumulus II at Marathon, Attica (plan also showing tumulus III and the ‘altar’) (courtesy of the Archaeological Society at Athens; source: Marinatos Reference Marinatos1970, 15 fig. 4).

IE. Pre-LH I tombs with lateral access devices (Appendix II: Table A2)

A number of MH tombs were also provided with lateral entrances, some of which were arranged in a fashion resembling later stomia. The most safely dated example is the MH II tomb 2 at Marathon tumulus I [G], whose entrance was defined by carefully built antae creating a small paved stomion and vestibule (Marinatos Reference Marinatos, Crossland and Birchall1973, pl. 15; Pantelidou Gofa et al. Reference Boyd, Dakouri-Hild and Boydforthcoming). The disturbed tomb 3 [H] at the same tumulus had upright slabs flanking the entrance (Marinatos Reference Marinatos1970, pl. 15a). The entrance of the central apsidal tomb at Papoulia [A], most probably of MH II–III date, was partly enclosed by a low curving wall which may have been meant to create a kind of vestibule or forecourt (Korres Reference Korres1978, 328 fig.1; Reference Korres1980, pl. 106a). A MH II(?) apsidal grave in the Kalogeropoulos mound, Routsi [B], and the apsidal tombs of Kato Samikon-Kleidi [C–E], reported to have contained MH III pottery, also had side entrances, although their precise form is unknown.

Two tombs which may predate the MH/LH I transition had short but well-defined passageways. A steep, earth-cut dromos c.2 m long is reported from a large cist grave at Argos [F], tentatively dated to the ‘final MH’ period (Morou Reference Morou1981, 107–9). The Xeropolis tomb [I] had a kind of roofed ‘vestibule’ or very short dromos obliquely set to the chamber and separated from it by a blocking wall; the tomb is dated to late MH on the basis of published sherd material, which however could also be early LH (Sapouna-Sakellaraki Reference Sapouna-Sakellaraki1995).

IIA. Floor depressions and grooves

A number of LH I tholoi have yielded evidence of special arrangements in the floor of the dromos. Usually, this is in the form of a wide depression (almost as wide as the stomion), starting at, or a little before, the entrance and continuing into the chamber (Fig. 13a ). This is attested in Peristeria 3, Tragana 2* and Gouvalari 1. At Peristeria 3 the depression (90 cm wide, c.80 cm max. depth) reached the centre of the chamber and has yielded spectacular finds (gold vases and ornaments, and a silver vessel), although no bone material was clearly associated with it (Marinatos Reference Marinatos1965, 116–19, pl. 133–144a; Korres Reference Korres1976b, 515–17, 533–5, pl. 272c; Reference Korres1977b, pl. 174). The Tragana 2* depression (1.35–1.5 m wide, c.40–50 cm deep)Footnote ²⁶ started at the entrance and reached the middle of the chamber; it contained intense traces of fire and a LH III piriform jar (Marinatos Reference Marinatos1955, 255, pl. 97a; Korres Reference Korres1976a, 270; Reference Korres1977a, pl. 142b; Iakovidis Reference Iakovidis2014, 126 plan 28). The Gouvalari 1 depression (1.4–1.9 m wide, 50 cm deep) carried on for more than 4 m inside the chamber, reaching beyond its centre, but no finds are reported (Marinatos Reference Marinatos1959, 176; Iakovidis Reference Iakovidis2014, 151 plan 33).

Fig. 13. Typology of depressions and grooves found in Early Mycenaean tombs (drawn by the author, graphically edited by Ms C. Stamati).

Some tombs had depressions which started in the dromos and split into two narrow grooves or furrows in the chamber (Fig. 13b ). This is attested at Routsi 2 and reported (without being illustrated) from Gouvalari 2. At Routsi 2, the depression was very deep (1.3 m) and continued for c.1 m inside the chamber before splitting into two furrows 1–1.8 m. long (Marinatos Reference Marinatos1957a, pl. 52a; Reference Marinatos1957b; Boyd Reference Boyd2002, 154; Reference Boyd2014a; Iakovidis Reference Iakovidis2014, 109 plan 26). At Gouvalari 2, the furrows are said to have been 60–70 cm long and 10–15 cm deep but no detailed description is available, nor are finds recorded.Footnote ²⁷

The character and function of these features – which are attested only in tholoi in this periodFootnote ²⁸ – are uncertain. The depressions created an obvious gap between the level of the entrance and the floor of the chamber. In Routsi 2, it is clear that this was not part of the original planning: the lower 1.3 m of the stomion walls was earthen (not reinforced with masonry), suggesting a secondary adaptation, when the level of the dromos was also lowered (Marinatos Reference Marinatos1957a). The same is true for Peristeria 3, where the earthen lower part of the stomion (Fig. 6) suggests also a later deepening in the level of the entrance and part of the chamber. The unsystematic recording of such depressions (there are no detailed descriptions or information about their extent in the dromos) and associated finds does not allow definite conclusions about their function,Footnote ²⁹ although the seeming lack of skeletal material suggests a non-funerary use. We should stress here that no example of the neat, parallel grooves of the type encountered in LH IIA and later tholoi and chamber tombs in the Argolid, Messenia and Boeotia (Fig. 13c ), is known from LH I tombs.Footnote ³⁰ The latter grooves have been convincingly interpreted as ceremonial features aimed at the symbolic communication with the dead through libations and offerings (Åkeström Reference Åkeström, French and Wardle1988, 202–5; Gallou Reference Gallou2005, 70–1). The LH I depressions may have been forerunners of the LH IIA grooves. As some of them are evidently secondary adaptations, dating to an advanced stage of LH I (Korres Reference Korres, Oliva and Frolikova1982, 91–3), they may mark the time during which the dromos and the entrance started acquiring ritual and symbolic significance. The LH IIA tholoi A and C at Kakovatos, which had similar depressions spanning the stomion and part of the chamber (although tholos C had no dromos; see Pelon Reference Pelon1976, 220–1, 277, 325), reinforce the possibility that threshold rituals became important in funerals and started acquiring standardised form during the LH I/IIA transition.Footnote ³¹ A much larger rectangular depression (5.46 × 1.80 m, and 15–70 cm deep), stretching from the entrance to the back wall of the chamber, has been found in the tholos of Corinth*, but it was obviously used for burials and does not seem to have continued in the stomion or the dromos (Kasimi Reference Kasimi, Kissas and Niemeier2013, 47–8).

IIB. ‘Benches’

Two LH I tombs have yielded evidence of some kind of ‘bench’ in the area of the dromos. An earthen ledge c.20 cm wide and 1–1.3 m high was found along the left side of the dromos at Routsi 2 (Fig. 14); it was probably created when the original floor of the dromos was dug to a lower level at some stage of the tomb's history (Marinatos Reference Marinatos1957a 118 and pl. 52b; Reference Marinatos1957b), which Korres places within the LH I period (Korres Reference Korres, Oliva and Frolikova1982, 91–3). A ‘hearth’ was discovered on this ledge (see section IIC).

Fig. 14. The dromos of tholos 2 at Routsi, Messenia (courtesy of the Archaeological Society at Athens; source: Marinatos Reference Marinatos1957a, pl. 52b).

In the BCT Argos 164 an earthen platform was found on the left side of the dromos (Fig. 11). It ran for most of the dromos’ length and was c.70 cm wide and c.70 cm high close to the entrance. Although it was not associated with finds, the excavator believes that it was an intentional feature perhaps used for the prothesis of the dead (Papadimitriou Reference Papadimitriou2001b, 43 and pls. 4a,b). Despite a possible rearrangement of the tomb's entrance in LH IIA, the dromos and the earthen platform seem to have been original features (Papadimitriou Reference Papadimitriou2001b, 47–8).

Only two more examples of (built) benches are known from dromoi of later date: one has been found on the left side of the dromos of the LH IIA tomb S2 at Medeon, and another one on the left side of the dromos of the LH IIIB(?) tomb Bπ1 at Eleusis (Papadimitriou Reference Papadimitriou2001a, 185 with references). None of them was associated with finds.

IIC. ‘Altars’, ‘hearths’, and possible ritual deposits in pits

Routsi 2 is the only LH I tomb where a permanent feature of possible ritual use has been reported from within the dromos (Fig. 14). It is described as an elliptical ‘hearth’ (60 × 30 cm) of local soft stone, placed at the outer end of the ledge formed along the left side of the dromos. The feature bore intense traces of fire and was associated with fragments of coarse pottery, some from cooking vessels (Marinatos Reference Marinatos1957a, 118; Reference Marinatos1957b; Korres Reference Korres, Oliva and Frolikova1982, 94).

A circular stone feature excavated at a distance of c.3 m from the entrance of Marathon tumulus II* (Fig. 12) has also been interpreted as an ‘altar’, although no finds were associated with it (Marinatos Reference Marinatos1970, 10 and pl. 8).

Permanent constructions of possible ritual use have also been found in association with the entrances of pre-LH I collective tombs. At Kato Samikon-Kleidi, two small circular ‘altars’ are reported to have been found at the outer ends of the long walls of tomb XI and another one in tomb IV (Papakonstantinou Reference Papakonstantinou1981, 149; Reference Papakonstantinou1982, 133). Although in the past I had doubted the character of these features (Papadimitriou Reference Papadimitriou2001a, 43–4, 183), the excavator Mrs Eleni Papakonstantinou has kindly shown me photographs which make clear that they were independent circular constructions placed slightly higher than the walls.Footnote ³² Close to the ‘altars’ of tomb XI a ‘Keftiu cup’ with ripple decoration (and, thus, possibly LH I) and a LH II scoop were found (Papakonstantinou Reference Papakonstantinou1981, 149). Mrs Papakonstantinou has also informed me that ashes and bones were recovered from the interior of the ‘altars’.Footnote ³³

At Papoulia, the small enclosed space in front of the central apsidal tomb has yielded abundant remains of fire, small animal bones and a large one possibly from a boar; similar remains are reported from an ‘altar’ found ‘at a small distance’ from the eastern end of the northern long side, which however has not been described in detail or illustrated (Marinatos Reference Marinatos1956, 254; Korres Reference Korres1978, 328 fig. 1). A sample of the ashes has been dated through C14 to 1850–1770+/−70 bc, thus probably falling within the late MH period or the MH/LH transition (Korres Reference Korres1978, 331; Boyd Reference Boyd2002, 122–3).

Next to the MH II tomb 2 in Marathon tumulus I, Marinatos unearthed an unusual feature consisting of one horizontal and two upright slabs leaning against the inner peribolos of the mound, together with the necks of two amphorae (Marinatos Reference Marinatos1970, 13 and pl. 12b). Kilian-Dirlmeier (Reference Kilian-Dirlmeier1997, 91) suggests a ritual function. The chronological relation of the feature to tomb 2, however, is problematic: tomb 2 seems to belong to the second phase of the tumulus use, when the ‘altar’ may have been covered by earth (Kilian-Dirlmeier Reference Kilian-Dirlmeier1997, 91–7).

Deposits of possible ritual character have been found in two cases. In the oblong tholos Thorikos IV*, a pit containing yellow earth together with MH/LH I – LH I/II sherds from open vases (cups, bowls) and coarse pottery was found in front of the western (left) jamb of the low blocking wall at the outer end of the dromos; traces of fire were found a little further, i.e. just behind the low blocking wall, suggesting to the excavators a rite associated with the closing of the dromos after a funeral (Servais and Servais-Soyez Reference Servais and Servais-Soyez1984, 26–9, 56). A shallow pit with broken bones and a few ‘Mycenaean’ sherds has been found in the dromos of the chamber tomb Volimidia-Angelopoulos 11*, close to the entrance (Marinatos Reference Marinatos1960, 199). Although it is not clarified whether the bones belonged to human or animal skeletons, I presume that if the former were the case it would have been noted by the excavator.

Finally, we should mention the ‘ritual deposits’ found in proximity to the BCTs 164 and 29 at Argos, in the area of the so-called ‘tumulus Γ’.Footnote ³⁴ Protonotariou-Deilaki reports three such deposits (two ‘bothroi’ and a horseshoe structure) containing much ash, animal bones, ‘portable tables of offerings’ and pottery dating from LH I/II to LH IIIA1, and thus probably in use from the beginning of the LBA (Protonotariou-Deilaki Reference Protonotariou-Deilaki and Morou-Kapokaki2009, 57–9, 204–5).

IID. Drinking and feasting remains, traces of fire, ceremonial vessels and other objects

Evidence of drinking or feasting activities in LH I tombs is rather limited, but this may be partly due to poor recording. In Pylos IV, 25 trays of pottery dating from MH to LH IIIB were recovered from the dromos, but no information is available about sherds associated with the earliest levels (Blegen et al. Reference Blegen, Rawson, Taylour and Donovan1973, 111); a patch of red earth found 15 cm above the ground, 1.7 m from the entrance, has been tentatively attributed to a purificatory fire (Blegen et al. Reference Blegen, Rawson, Taylour and Donovan1973, 100–1). At the Veves tholos, traces of fire mixed with bones of a bird, possibly the remains of a funerary meal, were found in the area of the built ‘vestibule’ (Choremis Reference Choremis1973, 51). Traces of fire have also been found at the beginning of the dromos of Thorikos IV* (see section IIC). At Volimidia-Kephalovryso 6, a pile of burnt stones mingled with ashes, animal bones and sherds from drinking cups of possible LH I date was found in front of the entrance and has been interpreted by the excavator as remains of sacrifice (Marinatos Reference Marinatos1965, 107; Lolos Reference Lolos1987, 206). Animal bones together with some sherds were also found in the ‘vestibule’ of the BCT E.III.7 at Eleusis, which apparently had not been used for burial (Cosmopoulos Reference Cosmopoulos2014, 139–41).

The dromoi of Mycenae 518, Prosymna 25 and 26 have yielded much pottery, including broken kylikes, which however date primarily to LH III (Wace Reference Wace1932, 75; Blegen Reference Blegen1937, 86, 93). In general, the custom of smashing drinking cups in front of the entrance as part of a final toasting ceremony is considered a LH III practice associated mostly with chamber tombs, at least in the Argolid (Cavanagh and Mee Reference Cavanagh and Mee1998, 55, 115; Reference Cavanagh, Mee, Galanakis, Wilkinson and Bennet2014; Gallou Reference Gallou2005, 25–6, 88–91). However, the stratigraphically excavated dromos of the BCT Argos 164 has yielded an intact layer (lot 8), which contained large numbers of LH IIB undecorated goblets together with animal bones and sea-shells, suggesting that the custom of drinking and feasting in dromoi may have started earlier (Shelton Reference Shelton2001, 77); an earlier layer (lot 9) is reported also to have contained pottery, sea-shells and animal bones, but the finds were not located during the study of the tomb (Shelton Reference Shelton2001, 77–8 and table I). At Messenia, dozens of intact and broken LH I cups have been found in tombs (Lolos Reference Lolos1987, 233–71 and tables 1–4; see also Chatzi Reference Chatzi1983, 111), although their exact findspots are rarely recorded; cups found in dromoi are explicitly mentioned from Routsi 2, Tragana 2* (Lolos Reference Lolos1987, 182, 210) and the aforementioned chamber tomb Volimidia-Kephalovryso 6. Boyd suggests an early start for the custom of ritual drinking in funerals in this part of the Peloponnese (Boyd Reference Boyd2002, 87–8).

Fragments of Palace Style jars are also occasionally found in dromoi (e.g. Prosymna 25, 26, Pylos IV), although they may have belonged to offerings taken out of the chamber in later cleaning operations (as was obviously the case in Pylos IV, Blegen et al. Reference Blegen, Rawson, Taylour and Donovan1973, 105). At Routsi 2, four LH I oval-mouthed amphorae were found near the entrance, whereas six large ‘Keftiu cups’ and a bronze double axe were recovered from other parts of the dromos (Marinatos Reference Marinatos1957a, 119; Lolos Reference Lolos1987, 210). Marinatos believed that they may have been remains of ritual feasts, but given the lack of stratigraphic details this can only be taken as a hypothesis. Other items discovered in the dromoi of some tombs (pieces of gold-leaf ornaments, bronze and stone tools, etc.) were probably leftovers from cleaning operations (e.g. Blegen et al. Reference Blegen, Rawson, Taylour and Donovan1973, 100–1; Marinatos Reference Marinatos1957a; Papadimitriou Reference Papadimitriou2001b, 51).

IIE. Burials in dromoi

Early Mycenaean tombs have yielded minimum evidence for the use of dromoi for burials.Footnote ³⁵ In chamber tomb Tiryns XIX*, a skeleton has been found in a pit dug in the dromos floor close to the entrance, but the accompanying vases date to LH IIB–III (Rudolph Reference Rudolph1973, pl. 50–1, nos. 4, 5, 6). A niche on the right side of the dromos of the chamber tomb Volimidia-Kephalovryso 2 contained two skeletons, one with Hellenistic, the other with LH IIIA vases, and thus was opened at an advanced stage of the tomb's history (Marinatos Reference Marinatos1964, 84–5). At Agia Anna 2*, a skeleton was found in the dromos but the accompanying stirrup jar was late in date (Keramopoulos Reference Keramopoulos1910, 214–15). The aforementioned shallow pit in the dromos of chamber tomb Volimidia-Angelopoulos 11* contained unspecified ‘broken bones’, but it is improbable that they were human. Otherwise, burials found in the upper levels of chamber tomb Mycenae 518 and Routsi tholos 1 were post-BA (most probably Byzantine).

We should note, however, that some chamber tombs had secondary chambers, accessible through the dromos, which might have been used for primary or secondary burials. They occur in Volimidia (Koronios 1* and 6, Angelopoulos 7 and 9) and Epidaurus Limera (Agia Triada A* and B*, Vamvakia*), but unfortunately no information is available about their contents and date. The side chamber at Volimidia-Koronios 1* had evidently been used for extended burials (Marinatos Reference Marinatos1953, 238; Iakovidis Reference Iakovidis2014, 5 plan 1), and the same is probable for some of the Epidaurus Limera tombs (Gallou Reference Gallou, Cavanagh, Gallou and Georgiadis2009, 89); given the lack of detailed data, however, we do not know whether side chambers were original features of the tombs or later additions (e.g. when the main chambers became overcrowded).

IIF. Liminal features

Features separating the dromos from the area outside the tomb have been identified in four cases. The tholoi of Psari* and Thorikos IV* had low blocking walls at the outer ends of their dromoi, carefully aligned with the stone periboloi of the covering mounds (Figs. 8, 9). At Thorikos IV*, in particular, the wall was flanked by two antae built of flat stones, which gave the whole ensemble an entrance-like appearance (Servais and Servais-Soyez Reference Servais and Servais-Soyez1984, 24–9). This arrangement (which is also seen in later tholoi; Pelon Reference Pelon1976, 296–7) and the ritual deposits found in association with the low blocking wall (see section IIC) suggest that such features were not designed to retain the earth filling of the dromos, as thought by some scholars (e.g. Tsountas and Manatt Reference Tsountas and Manatt1897, 140–1; Mylonas Reference Mylonas1966, 118–19; Kontorli-Papadopoulou Reference Kontorli-Papadopoulou and Laffineur1987, 154); instead, they must have served a strictly non-practical function. That such walls were symbolic in nature was long ago implied by Wace (Reference Wace1921–3, 358), who stressed that, had they been intended ‘to prevent free access to the dromos’, they ‘could not be very effective’. Moreover, the fact that they were aligned with the peribolos – which according to Pelon (Reference Pelon1976, 274–5) was an ‘enceinte sacrée’ – may suggest that they were meant as conceptual boundaries marking the passage to a space of special symbolic/ritual value.Footnote ³⁶

A symbolic function is also possible for the large stone threshold (flanked by two upright slabs) which abutted the outer peribolos of Marathon tumulus II* along the axis of the central tomb but at a distance from it (Marinatos Reference Marinatos1970, 14–16) (Fig. 12). That this feature did not serve practical needs is suggested both by the small height of the peribolos at that point (c.20 cm) and by the fact that no thresholds are known from tumuli outside Marathon. Its construction may have been meant to create a symbolic access-point, which gave direction to an otherwise unoriented monument and drove the funeral procession to the actual entrance of the tomb. The presence of a possible ‘altar’ at a distance of c.3 m (see section IIC) may suggest that the area in front of the tumulus was the focus of ritual activities.Footnote ³⁷

Finally, the entrance to the dromos of the intramural BCT Mitrou 73* was marked by a low clay wall and a kind of a porch with support bases on either side; according to the excavator, the wall may have been a symbolic boundary between the public street and the private tomb (Moortel Reference Moortel, Dakouri-Hild and Boydforthcoming).

Since three of the above examples come from LH I/IIA burial mounds with a single central tomb and a stone peribolos, it is possible that the practice of marking the end of the dromos with a liminal feature arose in the architectural environment of tumuli. From LH IIA onwards, blocking the dromos with a cross-wall became more common in good-quality tholoi and high-status chamber tombs, especially in the Argolid and occasionally in other regions.Footnote ³⁸

As for the inner end of the dromos, its liminal character has been stressed by many scholars, who have discussed the frequent discovery of smashed drinking vessels in front of the doorway (see section IID) and also the decoration of facades with plaster, painted motifs and added walls of poros blocks from LH IIA onwards, or even with wooden doors in later times.Footnote ³⁹ The sense of liminality was further emphasised by the corridor-like stomion. As we have seen, however, some early tholoi (Veves, Peristeria South, Peristeria 3, Gouvalari Tβ1, Corinth*) may have had their blocking walls at the inner (instead of the outer) end of this corridor-like feature, which therefore must have functioned as an open vestibule rather than a proper stomion; in these cases, a dromos is usually lacking.